A STUDY ON THE CHANNEL HEAT-UP DURING POSTULATED LOCA

Abstract

The nuclear power industry adopts the defense in depth (DiD) model, which encompasses a number of stringent operating practices and response systems in the form of multiple redundant protective barriers. The goals of the DiD model are: (a) to ensure the safe operation of the reactors by avoiding any evolution of severe accident conditions, (b) to form an effective barrier to avert the consequences of any malfunction/accident, and (c) to protect the public community and environment from any harm in case the barriers are not fully effective. A recent accident at the Fukushima Daiichi nuclear power plant in 2011 questioned the effectiveness of the DiD model. In light of this accident, almost all countries, including India, took measures to strengthen the effectiveness of the nuclear accident regulations and reduce the impact on the public mind. In India, the government is committed to scaling up the capacity of nuclear power production; therefore, Pressurized Heavy Water Reactors (PHWRs) of large capacities (540 and 700 MWe) are being developed. These large capacity PHWRs consist of assemblies of 392 coolant channels which are mounted horizontally through the calandria vessel. Each coolant channel has two concentric tubes referred to as a pressure tube (PT) made of Zr-2.5% Nb alloy and a calandria tube (CT) made of Zr-2 alloy. A series of fuel bundles having a configuration of 37 elements are axially loaded into the PT. Under normal operating conditions, the pressurized heavy-water coolant of the primary heat transport system (PHTS) flows through the PT (at around 10 MPa and 266 °C) and extracts the heat generated from the fission chain reaction. This heavy-water coolant later exchanges heat with the light-water circulating in the secondary heat transport system (SHTS), thereby generating steam to drive the turbines. During a postulated accident situation, such as a loss of coolant accident (LOCA) coincident with the loss of emergency core cooling system (LOECCS), the cooling environment of the channel is degraded because steam is the only coolant available to remove decay heat of the fuel. This causes the channel to undergo severe heating, which may deteriorate its structural integrity (such as sagging, ballooning, or both of the PT, slumping of fuel elements, bursting of clad tubes). Furthermore, the temperature of the fuel may escalate due to secondary heat liberated during the auto-catalytic chemical reaction (oxidation) between steam and Zr-4 cladding. In the worst scenario, the explosion of hydrogen gas generated during the steam-Zr-4 oxidation reaction could breach the containment, posing a greater threat to the community due to the release of volatile and semi-volatile fission products.Considering the severity of such postulated accidents, it is very important to understand the thermal response of the fuel channel. In light of this experimental facility was developed to investigate the behavior of simulated fuel channel housing 37 elements under heat-up condition. Steady-state experiments were conducted with two different configurations of fuel bundle simulator (FBS): (a) assembled and (b) disassembled at different decay heats. The second configuration of the FBS simulated the slumping of fuel pins onto PT. For this, a conceptualized disassembled FBS was fabricated, assuming that the 37 fuel elements of a bundle are collapsed and collected in a tightly packed stack at the bottom of the PT of the fuel channel. The influence of sagging on the temperatures of the fuel channel was also investigated by varying the position of PT with respect to CT. The experiments were performed under non-oxidizing and oxidizing environments. The experiments under a non-oxidizing environment were conducted at different steady-state temperatures of 550 °C, 700 °C, 850 °C and 1000 °C in center FES, and for the oxidizing tests at three temperatures of 700 °C, 850 °C and 1000 °C. The purpose of conducting experiments under a non-oxidizing environment (using argon gas) was to avoid the effects of possible oxidation and emissivity changes on the analysis of radiation heat transfer in the channel. The numerical simulations of the non-oxidizing experiments were also performed using ANSYS Fluent 19.0. The radiation heat transfer was modeled using the Discrete Ordinates (DO) model. The experiments under oxidizing environment intend to simulate the late phase of the event scenario; hence a weak steam flow was passed through the channel. In fact, a mixture of steam and argon gas was passed through the test section. The purpose of the argon gas is to drive out the generated hydrogen gas as the weak steam flow is not sufficient for the said purpose. The investigations with assembled FBS have shown that a significant radial temperature gradient gets established in FBS. The radial temperature distribution was found to be opposite to the relative power distribution, i.e., the temperature decreased from the center ring to the outer ring. The maximum and minimum temperatures for each FES were obtained at the circumferential segment, which corresponds to the inward and outward orientations from PT, respectively. The maximum and minimum circumferential temperature variation was found in the FESs of the outer ring and center ring, respectively. Similarly, for the disassembled FBS, minimal circumferential temperature distribution was found in the center FES and a significant radial temperature variation was seen in the FES of the outer ring. The nodes of FES in contact with the PT had a relatively lower temperature as compared to its diametric opposite node. The slumping of fuel pins resulted in local heating of the PT in the contact area, which in turn developed a circumferential temperature gradient in the PT. For the concentric configuration of the PT with disassembled FBS, a higher temperature was obtained at the bottom sector of the PT as compared to the top sector. The thermal stratification effect of the gas between the concentric tubes (PT and CT) and coolant caused circumferential temperature variations in the PT, with the bottom nodes having lower temperatures than the corresponding top nodes with an assembled configuration of FBS. Significant circumferential temperature gradients were established in the channel during a PT–CT contact configuration. The PT–CT contact caused localized heating of CT and localized cooling of PT due to enhancement in heat transfer through conduction from the contact points. The temperature distribution of FBS in both assembled and disassembled configurations was also affected by the PT–CT contact. The temperature was found to decrease from top to bottom symmetric FES. However, contact had minimal effect on the circumferential temperature distribution of the center FES. Under an oxidizing environment, a significant temperature gradient gets established in radial, circumferential and axial directions in the fuel channel compared to the non-oxidizing environment. Furthermore, as temperature increased, the values of these temperature gradients increased significantly. For each oxidizing test, hydrogen gas was generated at test temperatures of 850 ℃ and 1000 ℃. The downstream side of the FBS was found to be more oxidized from the physical inspection. An estimation based on the heat balance showed that during an accidental situation like LOCA coincident with ECCS failure, the moderator acts as an effective heat sink because 76-88% of the total heat supplied to the bundle was transferred to the moderator.