Abstract:
One of the passive safety systems employed in nuclear reactors is a large water pool positioned at an elevation, which acts as a heat sink to remove decay heat from the reactor core in the event of an accident using the natural circulation phenomena. A large mass of water at an elevated position has concerns regarding the safety of support structure, and therefore an alternative methodology where the heat sink is positioned just below the heater section is investigated in this study. Results of an experimental and numerical investigation for the thermal-hydraulic behavior of single-phase natural circulation in a rectangular loop of height 4300 mm and width 1900 mm with the heat sink located at an elevation just below the heater section is reported here. The heat sink was a finite one, a constant volume insulated tank filled with water, whereas the heat source was a constant heat flux type. The loop was constructed with circular pipe elements with a nominal inner diameter equal to 40 mm. The heater section was 1m in height and three input power values equal to 1.15 kW, 1.55 kW, and 2 kW were used in the study. The flow behavior in a single and twin parallel heater system with a single riser was studied. The loop fluid is initially in a single-phase and becomes two-phase as the heat sink temperature rises, but only the results for the loop fluid in the single-phase regime are presented in the current study. The measured mass flow rate, after an initial transient, settles down to values which slowly decreases with time. The average temperature of the heat sink is shown to rise continuously with increasing time, demonstrating the effectiveness of the heat sink in removing the energy input to the working fluid in the heater section. However, the heat sink removes only 60-70% of the input heat while the rest is either lost to the ambient or utilized to raise the enthalpy of the working fluid. The Nusselt numbers in the heater section are also presented which steadily rise after an initial transient. In addition, the values are much higher than those for the corresponding Reynolds number values in forced convection due to the influence of buoyancy-induced convection. The Nusselt number is well correlated by the expression (Nu) over bar = 1.35((Ra*) over bar)(0.3) for 2 x 10(5) <= (Ra*) over bar <= 6.5 x 10(5) and 600 <= Re <= 1000. Numerical predictions for loop flow rate, temperature variation, and sink temperature using the RELAP5/MOD3.2 code for the experimental configuration of the current study are presented which compare well with the measurements the maximum deviation being less than 10%. Numerical results for pressurizing the working fluid and permitting boiling in the heat sink to drive the system towards steady-state operation are also presented.