American Nuclear Society

Home / Publications / Journals / Nuclear Science and Engineering / Volume 184 / Number 2

Experimental Investigation on Characteristics of Boiling Two-Phase-Flow Instability in a Parallel-Multichannel Natural-Circulation System

Sapna Singh, A. K. Nayak, J. Aparna

Nuclear Science and Engineering / Volume 184 / Number 2 / October 2016 / Pages 263-279

Technical Paper /

First Online Publication:September 12, 2016
Updated:September 30, 2016

Natural-circulation systems are gaining wide acceptance due to their inherent passive safety features, which makes them more reliable to use in nuclear power plants. In view of this, many Generation III and III+ nuclear reactors have been designed with natural circulation as a mode of core cooling. But, the major area of concern in these kinds of systems is still the complex phenomenon of boiling two-phase-flow instabilities, which is yet to be fully understood especially when power and pressure in the system are low (type I instability). A major factor in this regard is to know the sensitivity of the number of parallel channels to the characteristic behavior of these flow oscillations in systems like those of a boiling water reactor. Based on mathematical models, in the past, some authors reported that any number of channels behaves in the same way as a twin-channel system. There is no experimental study to validate this. This experimental investigation has been done to add insight. A parallel-multichannel closed loop filled with water and maintained at atmospheric pressure was used for the study. Power in the individual heated sections of the loop was increased from 0 to 2 kW in steps of 250 W. Each power level was maintained for ~30 min. After reaching 2 kW, power was decreased to 0 kW. Three cases of experiments were done by taking two, four, and six active channels at a time, respectively. Different flow oscillation parameters such as amplitude, frequency, phase difference, general characteristics, etc., were studied to see if they were affected when the number of parallel channels was changed. The present analysis showed that their behavior is sensitive toward changes in the number of parallel channels. We cannot extrapolate twin-channel data when there are more channels in a system. The present paper discusses the experiments performed and the detailed results in support of this argument.

Questions or comments about the site? Contact the ANS Webmaster.