When a nuclear power plant shuts down, the heat from the reactor doesn't disappear right away. Even after the chain reaction stops, the fuel continues to give off what scientists call decay heat. In the event of an unplanned shutdown or accident scenario, the heat must be removed fully, safely and without fail.
Researchers at the U.S. Department of Energy's (DOE) Argonne National Laboratory study ways to remove decay heat without pumps or active controls. They recently looked more closely at how certain systems behave when the amount of decay heat changes over time, just as it does in the real world.
One concept, called the Reactor Cavity Cooling System (RCCS), is being considered for several advanced nuclear reactor designs. Argonne's Natural Convection Shutdown Heat Removal Test Facility (NSTF) is the largest experimental facility of its kind built to study how such systems work at realistic size and conditions.
"Most advanced reactor vendors rely on passive systems like RCCS to prove their designs can stay safe even during unlikely accident scenarios," said Argonne Nuclear Engineer Qiuping Lu. "Our experiments give them the high-quality data they need to show regulators that their safety systems will perform as expected."
The RCCS concept uses passive safety principles - relying on natural phenomena like gravity and buoyancy instead of mechanical pumps - to move water through metal tubes that run around the outside of a reactor vessel. When the reactor shuts down, decay heat travels from the vessel walls into these tubes. Warmer water rises, cooler water falls and the loop carries heat away to a large storage tank at a higher elevation. Because the water can eventually start to boil and form steam, engineers must understand not only how much decay heat the system can remove but also how it behaves when liquid water and steam coexist.
One key question is how the system responds when the amount of decay heat is higher or lower than the standard set point at which engineers calculate the performance of their designs. Decay heat is typically higher right after shutdown, and then it gradually decreases, which means a system might vary in performance.
Using the NSTF, the team ran controlled tests at power levels scaled based on the range of typical decay heat levels in a full-size reactor. They then changed the height of the main water tank inlet, which can change across different designs. The inlet's height affects how much water is available and how fluid in the loop circulates and flows.
"We were asking two questions," Lu said. "What happens when decay heat is higher or lower than designers expect- And does moving the tank connection up or down change how reliably the system runs-"
Using NSTF's 59-foot-tall loop, which is roughly half the size of a large-scale commercial nuclear reactor, Argonne researchers slowly increased power levels over several hours. They observed as the water was heated, produced steam and, in some cases, developed boiling conditions with anticipated surges in flow.
One key effect they observed was a phenomenon called flashing. Flashing happens when water in the lower portions of the loop rises in elevation and reaches a critical point where it turns suddenly into steam, even without extra heat. These bursts of steam - flashing - can affect surges in the flow of water and steam, reducing cooling efficiency and placing extra stress on reactor components. Argonne worked to identify where these surges and swings begin, how strong they become and whether they fade.
They found that decay heat at lower power levels slowed the spread of boiling and flashing from NSTF's tank into its chimney. At an equivalent full-scale power of 2.4 megawatts, steam from flashing reached the upper chimney about 20 minutes after boiling began in the tank. At 1.75 megawatts, it took about 90 minutes. At 1.4 megawatts, steam never reached the chimney during the test.
The height of the tank inlet also mattered. With a mid-level inlet, boiling and flashing reached the chimney rapidly, in less than 15 minutes. With a lower inlet, the researchers observed flashing-induced instability throughout the tests. They identified a clear boundary where metrics changed and concluded that mid-level inlets contributed to more stable operating conditions in certain scenarios.
"These findings have practical value for companies designing next-generation reactors," Lu said. "By answering questions about how passive safety performs at different power levels, we give designers more confidence in RCCS technology and help move advanced nuclear energy toward deployment."
This research was supported by the DOE Office of Nuclear Energy's Advanced Reactor Technologies program.