Onset of Channel Flow Reversal in RD- l4M Natural Circulation Tests

Abstract

This paper presents an analysis of the conditions leading to the onset of channel flow reversal in a series of natural circulation tests conducted in the RD-14M facility. This series of tests was carried out to investigate two-phase natural circulation under conditions of decreasing primary inventory, but with no break in the heat transport system. In these tests, the channel flows were found to be unidirectional in the early part of each experiment. Then, depending on the specific test conditions, the flow in some channels reversed, while the flow in the remaining channels continued in the same direction. A conceptually simple flow-reversal criterion is applied to these tests: the prevailing header-to-header pressure differential (APuH) must be sufficiently negative to overcome the forward driving force resulting from the density gradient between the inlet and outlet feeders. A comparison between the predictions using the above criterion and the experimental data is made in the following areas : (1) the magnitude of the APHH at the onset of channel flow reversal, and (2) which of the channels is the first to reverse. For those tests which exhibited relatively steady APHH, the APHH was observed to become increasingly negative as the loop liquid inventory was reduced. The mechanism responsible for the increasingly negative APHH is described. Using the above flow- reversal criterion, the header-to-header pressure differential required for the onset of channel flow reversal, APrev, is computed using experimental values as a function of time. Onset of channel flow reversal is observed to occur when the absolute magnitude of the experimentally measured APHH exceeds that of APrev. Furthermore, the criterion predicts the top channels to be the first to reverse in these tests, which is in agreement with the experimental results for the tests which exhibited steady flow. For those tests which exhibited oscillatory APHuf the amplitude of the oscillations was observed to increase as the loop liquid inventory was reduced. A qualitative explanation for the- oscillatory behaviour of the loop is given. Using the above flow-reversal criterion, the header-to-header pressure differential required for the onset of channel flow reversal, APrev, is computed as a function of time. When the experimentally measured APHH approaches APrev, temporary flow slowdowns are observed. When the absolute magnitude of the experimentally measured APm, significantly exceeds that of APrev, onset of sustained channel flow reversal is observed. Applying the above criterion to these tests, it is shown that the top channels are not necessarily the first to reverse. However, the above flow reversal criterion is unable to predict which of the channels is the first to reverse in these tests. Development of a dynamic flow-reversal criterion, accounting for-the transient and feedback processes occurring in these tests, is in progress. A model of the below-header components in the RD-14M loop is under development. The model has been used to predict the fluctuation in APHH required for channel flow reversal as a function of the average APHH. The model also predicts that smaller fluctuations in APHH are sufficient for reversal of the channel flow when the net channel power is decreased. It is planned to integrate the present "below-header" model with an "above-header" model to predict the primary inventory at the onset of channel flow reversal. When an integrated model is developed, the effect of heat losses on the primary inventory at the onset of channel flow reversal will be determined. The experiments described in this paper were performed by Whiteshell Laboratories of AECL Research, and funding for these experiments was provided by the CANDU Owners Group. The analysis described in this paper was performed by Ontario Hydro.