Modelling Inlet or Outlet Feeder Voiding and Fuel Simulator Cooling Under Natural Circulation Conditions in the RD-14M Loop

Abstract

To better understand thermosyphoning behaviour in CANDU, a number of two-phase thermosyphoning experiments were conducted in the multiple-channel RD-14M test facility. Above some loop void fraction in these experiments, the flow permanently reverse direction in some of the simulated fuel channel assemblies. Following flow reversal in these assemblies, the assembly outlet feeders became water-filled. Subsequently, void appeared in some of these outlet feeders causing the fuel element simulators (FES) to heat up. At higher loop void fraction in some of the tests, similar voiding occurred in the forward-flow inlet feeders. This paper presents models, dubbed BENDORY and TALSUALL, which were developed to explain and predict feeder voiding and FES temperature behaviour observed in the RD-14M tests. BENDORY models partial voiding of an initially water-filled reverse-flow outlet feeder under oscillatory low void loop. BENDORY assumes that, for certain conditions, reverse outlet feeder flow entrains steam bubbles of sufficiently small size from the connected outlet header into the feeder. The paper presents some of the RD-14M test results to justify this and other model assumptions. TALSMALL was developed previously to model water draining in a forward-flow inlet feeder following steam -water phase separation in the connected inlet header. This phase separation exposed the inlet feeder nozzles to steam causing the feeder water to drain. The channel conditions predicted by BENDORY and TALSMLL were used in the model AMPTRACT developed previously to predict FES temperature. The comparison of BENORY/TALSMALL/AMPTRACT/CATHENA predictions and of the RD- 14M test results presented in this paper forms part of a validation basis for these models which were used to predict fuel and fuel channel cooling under natural circulation conditions in the intact loop for accident scenarios in the Gentilly 2 plant.