The Dew Point Response of the Annulus Gas System of Bruce NGS A, Units 1 and 2: Derivation of a Model

Abstract

The dew point response of the Annulus Gas System in Bruce A, Units 1 and 2 has been modelled in order to alert the operator of the presence of heavy water and to estimate the leak rate into the annulus. The computer model can be easily adapted to determine the Annulus Gas System dew point response in any station. It models the complex arrangement of the system and the transportation of moisture through the annuli by a combination of plug flow and mixing of CO₂ and D₂O vapor. It predicts the response of the dew point monitor for a range of specified leak rates and positions of a leaking channel in a string of channels.

This model has been used to calculate the variation of dew point and rate of change of dew point with respect to time (t). It shows that there in a maximum in the rate of dew point change (dT/dt) with respect to the corresponding dew point (T). This maximum is unique for a given leak rate and channel position. It is independent of the starting time for the leak.

The computer programme has been verified by an analytical solution for the model.

The dew point response analysis is for the case of the D₂O leak remaining in the vapor phase throughout the annulus gas system. The analysis shows that the dew point in the leaking channel very quickly increases to the end shield temperature (60°C) for leaks > 15 g/h long before the dew point indicator even starts to respond. Condensation may occur in the lattice tubes and the dew point response at the dew point indicators will be as that of a of a 15 g/h leak. Condensation can be minimized by increasing the system CO₂ flow rate. Increasing the flow of CO₂ from 1.42 L/s (3 scfm) to 5.7 L/s (12 scfm) increases the leak size for which condensation occurs from 15 g/h to 60 g/h. In this case the analysis results are applicable for leaks up to 60 g/h.