A new technique for estimating the bulk fluid temperature of stratified flow in pressurized water reactors

In some Westinghouse reactors, the computed hot-leg bulk temperature is being biased by thermal stratification of the core exit flow. The stratification results from large temperature differences between the inner and outer regions of the core exit flow and limited fluid mixing between the core exit and the hot-leg temperature sensors. Previously, the bulk temperature obtained from hot-leg temperature measurements was considered to be an accurate estimate, but recent studies suggest that the computed bulk temperature is changing without corresponding flow changes. The objectives of this research are to develop a method for estimating the hot-leg bulk fluid temperature from exterior pipe surface temperatures and to investigate the relationship between the core exit temperatures and the fluid temperatures measured in the hot-leg(s). The basis for the proposed estimating method is that as the flow mixes along the pipe, the surface temperature of the pipe will approach the bulk fluid temperature within some small temperature difference. The process of determining the bulk fluid temperature begins with discrete measurements of the pipe surface temperature. These measurements are processed into axial temperature profile(s) that is extrapolated to their asymptotic value(s). The bulk temperature is computed by averaging the final asymptotic temperature(s) and adjusting this value for the temperature drop across the pipe wall. An experimental facility was constructed to simulate stratified flow and verify the process of estimating the bulk fluid temperature from pipe surface temperatures. The facility was used to perform experiments with various degrees of flow stratification. The experimental results are bulk temperature estimates that on an average are within 0.15°F (0.13%) of the actual bulk temperature. One of the primary agents believed to promote stratification in the reactor coolant system, RCS, is the temperature variations present in the core exit flow. Therefore, the influence of the core exit temperature on the bulk fluid temperature is analyzed. The analysis uses artificial neural systems technology to develop this relationship using core exit temperature measurements and RSC hot-leg temperature measurements made during reactor operations. The results of the analysis are promising but are restricted by the availability of measured data; only one set of data was available. Therefore, the results are not conclusive, but they illustrate the potential benefits of this procedure if more information becomes available.