STUDY OF SAGGING AND BALLOONING BEHAVIOUR OF INDIAN PHWR PRESSURE TUBE DURING LOCA

Abstract

Consideration of severe accidents at a nuclear power plant (NPP) is an essential component in nuclear safety analysis. The extent and nature of deformation is important to study from reactor safety point of view, Under normal operating condition, the heat generated by the fission reaction in the fuel bundle is removed by pressurized heavy water coolant. During postulated Loss of Coolant Accidents (LOCA), the decay heating is removed by the Emergency Core Cooling System (ECCS). Even in case of complete failure of the ECCS, the moderator system will prevent gross damage and maintain the integrity of the fuel and fuel channels. Under postulated low frequency (< 10-6 per year) accidents, heat up of Pressure Tube (PT) along with internal pressure may lead to its deformation. There are two modes of deformation of PT: sagging and ballooning. The deformation mode will depend on internal pressure in PT. If the internal pressure is high (more than 1.0 MPa) then the PT will deform radialy outward (balloons) at high temperature and may contact the Calandria Tube (CT). Due to localized nature of deformation, the whole PT is prone to balloon except the zone near the garter springs. If the pressure is low, the PT will sag under the self weight and the weight of fuel bundles. The sagging of PT would always lead to an initial contact at the bottom of tube. Bhabha Atomic Research Centre (BARC), the nodal agency in India, is actively involved in the design and development of nuclear reactor for power generation. Leading technical Institutions/Universities in India are collaboratively supporting this national activity. In this program, Indian Institute of Technology, Roorkee, as a collaborative partner had agreed to conduct channel heat up experiments: (a) PT sagging due to its own weight and weight of fuel bundles and (b) PT ballooning due to high internal pressure. lii These investigations will help in understanding the behaviour of PT, used in Indian Pressurised Heavy Water Reactor (IPHWR), in case of LOCA. Moreover, experimental data will help BARC, Mumbai to validate their code "PTCREEP", which predicts the creep deformation of PT. The present thesis describes the experimental and numerical investigations carried out at IIT Roorkee for channel heat up experiments. To simulate LOCA with complete failure of ECCS, experimental set-ups were designed and fabricated to carry out investigations to study sagging and ballooning deformation of PT used in Indian PHWR. Both sagging and ballooning deformation experiments were carried out while the CT was wrapped by ceramic fibers and subsequently the experiments were repeated while CT was submerged in water. In the sagging experiment heat up rates were varied whereas in ballooning experiment heat up rate and internal pressure were varied. In the first stage of sagging experiment, when the PT was covered by ceramic wools. three set of experiments were conduced at heat up rates of 32.5, 27.8 and l5 kW respectively. The downward deformation of PT was measured at five axial locations and the temperatures of the PT and CT were measured at different axial locations. At each axial location, six thermocouples were placed on the circumference. In the second stage, the set-up was modified to carry out the measurements while the CT was submerged in water. Three set of experiments were conduced at heat up rates of 32.5, 27.8 and 26.7 kW. "The temperatures of PT and CT were recorded and the boiling at the outer surface of CT was visually observed. However, in these experiments sagging was not measured. Commercially available FEM package ANSYS 11.0 is also used for the estimation of contact conductance, In this process, ANSYS uses coupled field (structural-thermal) iv analysis, radiation heat transfer, and contact element concept to estimate the contact conductance between PT and CT. The ballooning experiments were also conducted in two stages. In the first stage, the ballooning and the temperature of the PT and CT were measured. In this stage, the CT was wrapped with ceramic wools and the experiment was conducted maintaining 2 MPa internal pressure in the PT. In the second stage, the set up was modified to carry out the measurements while CT was submerged in water. Three set of experiments were conduced at internal pressure of 2, 4 and 6 MPa with different heat up rates and initial conditions. From the experimental results, it is found that sagging of the PT initiates at around 400°C and it does not depend on the initial heat up rate. Sagging of PT is very fast after 450T irrespective of heat up rate and initial PT temperature. PT-CT contact takes place at PT temperature in the range of 540°C to 640°C near the contact point. The average PT temperature is 580 to 700°C near the contact location when the CT is wrapped by ceramic wool. When CT is submerged in water, contact takes place at PT temperature in the range of 550°C to 610°C with average PT temperature in the range of 600 to 620 °C near—the contact location. The contact location in all the experiments was near the centre of the tube. Nucleate boiling was visually observed at the outer surface of CT throughout the length, but bubble density was high at the central location. Circumferential temperature difference was set up in the PT with the top being at higher temperature than the bottom. Although the temperature at the top and bottom of PT were not measured, the temperature at nodal positions at 30° and 150° from PT top clearly shows this trend. It was observed that the temperature difference between these two nodal points was from 180°C to 220°C. The maximum temperature attained by PT, while CT was wrapped by ceramic fiber and submerged in water, was 840°C and 740°C respectively. The contact conductance estimated using ANSYS is 3000 W/m2 °C. In the ballooning deformation experiment, it was found that the sagging of PT takes place at 500°C before ballooning initiation. Maximum sagging is less than 2 mm. It was observed that once PT touches the CT, there is no further rise in PT temperature and the maximum PT temperature was below 740°C. PT ballooning initiation took place within temperature range of 550°C-640°C for pressure range of 2-6 MPa. The PT-CT contact temperature was in the range of 620°C-725°C. The maximum percentage strain rate is found to be 0.250%, 0.285% and 0.143 % per second in ballooning deformation at 2, 4 and 6 MPa pressure respectively. The transverse creep correlation for CANDU material, adopted in the "PTCREEP" code, is found to hold good for Indian PT material [or the temperature range expected during LOCA. the arrest of temperature rise of PT, mechanical integrity of PT (no breach) and non-occurrence of Critical Heat Flux on CT shows the channel integrity under efficient cooling of the simulated moderator. This demonstrates that the functioning of "moderator as a heat sink" as an inherent safety feature of