1-Dimensional simulation of thermal annealing in a commercial nuclear power plant reactor pressure vessel wall section

Abstract

The objective of this work was to provide experimental heat transfer boundary condition and reactor pressure vessel (RPV) section thermal response data that can be used to benchmark computer codes that simulate thermal annealing of RPVs. This specific project was designed to provide the Electric Power Research Institute (EPRI) with experimental data that could be used to support the development of a thermal annealing model. A secondary benefit is to provide additional experimental data (e.g., thermal response of concrete reactor cavity wall) that could be of use in an annealing demonstration project. The setup comprised a heater assembly, a 1.2 m x 1.2 m x 17.1 cm thick [4 ft x 4 ft x 6.75 in] section of an RPV (A533B ferritic steel with stainless steel cladding), a mockup of the "mirror" insulation between the RPV and the concrete reactor cavity wall, and a 25.4 cm [lo in] thick concrete wall, 2.1 m x 2.1 m [lo ft x 10 ft] square. Experiments were performed at temperature heat-upkooldown rates of 7, 14, and 2 8 " C h C12.5, 25, and SO'Fh] as measured on the heated face. A peak temperature of 454°C [850°F] was maintained on the heated face until the concrete wall temperature reached equilibrium. Results are most representative of those RPV locations where the heat transfer would be 1-dimensional. Temperature was measured at multiple locations on the heated and unheated faces of the RPV section and the concrete wall. Incident heat flux was measured on the heated face, and absorbed heat flux estimates were generated from temperature measurements and an inverse heat conduction code developed at Sandia National Laboratories called "Sandia One Dimensional Direct and Inverse Thermal" (SODDIT). Through-wall temperature differences, concrete wall temperature response, heat flux absorbed into the RPV surface and incident on the surface are presented. All of these data are useful to modelers developing codes to simulate RPV annealing. Additionally, incident heat flux measurements can be used to design the heater system required to anneal a full-scale RPV. ~\ I Results compare favorably with those reported in NUREGKR-4212.

An investigation of temperature measurement methods in nuclear power plant reactor pressure vessel annealing

“…During the heat-up, the bottom required the greatest power, followed by the middle and the top; the top and bottom zones required increasing power while the middle zone was almost constant. This plot is consistent with those presented previously (Nakos et al, 1994), and indicates that the mounting fiame near the bottom heater bank absorbs a considerable amount of energy, thus requiring more power on the bottom'zone than either the middle or top zones. Convection heating also contributed to a lower power requirement for the middle and upper zones than for the lower zone.…”

“…It was demonstrated in an earlier investigation (Nakos et al, 1994) that when RFV temperature was measured, spot-mounted and intrinsic thermocouples gave temperature readings that were within 5°C [9'F] of each other over the entire temperature range used in this investigation; the spot-mounted thermocouples gave the hotter reading during the heat-up. At 454"C[85O0F] this difference is about 1%.…”

“…from the right side of the RPV edge. Based on temperature contours presented in Nakos et al (1994), it is reasonable to conclude that there are no significant edge effects present where PR2 is located.…”

“…The cladding was applied by a welding process and, consequently, this surface has a somewhat rough texture (see Figure 8). The RPV wall section was previously used in a one-dimensional heat transfer experiment (Nakos et al, 1994) and the instrumentation shown in Figure 1 was reconfigured for this study. The back side of the RPV wall, instrumented from a previous test series, can be seen in Figure 1.…”

Uncertainty of Heat Transfer Measurements in an Engulfing Pool Fire

“…This is due to the fact that SODDIT uses the slope of the temperature history data (change in temperature with time) to calculate heat flux, not the temperature magnitude. This method of uncertainty estimation for inverse heat flux calculation was used in reference [5]. The temperature magnitude is used to evaluate material properties, which for the most part change slowly with temperature.…”