During the normal operation of light water reactor, fuel defects can reside on fuel cladding from various reason such as fuel-pellet mechanical interaction, the hydriding of the Zr clad, the grid fritting from the assembly support, the warp wire fritting with the clad and also the stress corrosion cracking (SCC). The formation of defects on fuel cladding will result water ingression to the gap and react with nuclear fuel and the cladding inner surface. The interaction of water and nuclear fuel will affect the fuel thermal properties and deteriorate the cladding by hydriding and change of oxygen potential in the fuel. The change of fuel thermal properties will decrease the thermal conductivity, lead the decrease of heat transfer coefficient which may increase the fuel melting risk. The volatile fission products and fission gas will release to the coolant through cladding defects and increase the coolant activity and the defective nuclear fuel becomes a fission product source term when reactor is under normal operation. Experimental and modeling are applied to understand the behavior of a defective fuel pin. The experimental part focuses on the dissolution test of rare earth fission products in simulated LWR coolant chemistry and the diffusion coefficient measurement of cesium iodide in simulated LWR coolant chemistry using Nuclear Magnetic Resonance (NMR) technique. Rare earth fission products significantly contribute the residual heat and large quantity of radioactivity after the core shut down or in severe accident, therefore, their dissolution kinetic parameters in LWR are important to reactor safety and the understanding the source terms.

A general model was developed to estimate the activities of fission products in reactor coolant and hence to predict a value for the I-129/Cs-137 scaling factor; the latter can be applied along with measured Cs-137 activities to estimate I-129 levels in reactor waste. The model accounts for fission product release from both defective fuel rods and uranium contamination present on in-core reactor surfaces. For simplicity, only the key release mechanisms were modeled. A mass balance, considering the two fuel source terms and a loss term due to coolant cleanup was solved to estimate fission product activity in the primary heat transport system coolant. Steady state assumptions were made to solve for the activity of shortlived fission products. Solutions for long-lived fission products are time-dependent. Data for short-lived radioiodines I-131, I-132, I-133, I-134 and I-135 were analyzed to estimate model parameters for I-129. The estimated parameter values were then used to determine I-1 29 coolant activities. Because of the chemical affinity between iodine and cesium, estimates of Cs-137 coolant concentrations were also based on parameter values similar to those for the radioiodines; this assumption was tested by comparing measured and predicted Cs-137 coolant concentrations. Application of the derived model to Douglas Point and Darlington Nuclear Generating Station plant data yielded estimates for I-129/I-131 and I-129/Cs-137 which are consistent with values reported for pressurized water reactors (PWRs) and boiling water reactors (BWRs). The estimated magnitude for the I-129/Cs-137 ratio was 10-8 - 10-7.

An increasing desire to transition from traditional fossil fuels for energy production coupled with an aging nuclear power infrastructure presents a critical demand for new power production designs. Advanced Generation IV reactor designs offer potential solutions to this looming future demand. Of these Gen IV reactors, pool-type sodium fast reactors are a leading candidate with hundreds of reactor years of operation and many inherent benefits over traditional light water reactors. However, prior to licensing a commercial sodium fast reactor, a mechanistic source term assessment is likely required. An assessment of the current state of knowledge for sodium fast reactor source term assessment was conducted at Argonne National Laboratory. From this, several key gaps were identified within the knowledge base. Fission product scrubbing via bubble release was of particular importance due to its impact on the overall source term calculation and associated uncertainty. Gaseous fission products entering the coolant pool following fuel pin failure have the potential to be transported directly to cover gas region. Solid/liquid fission products may either be injected into the coolant pool fluid or be entrained within gaseous fission products. While the coolant pool itself is expected to scrub the majority of radionuclides, those becoming entrained within bubbles have the potential to be transported directly to the cover gas region. From the cover gas, there is a considerably greater risk of release to the environment. To better understand this phenomenon, recent efforts in sodium fast reactors have been towards developing computational tools to model fission product retention following accident scenarios. Researchers at Argonne National Laboratory have developed the Simplified Radionuclide Transport code to model fission product transport. The model is able to estimate fission product types and quantities based on failure criteria and then predict their transport from the fuel pin to the coolant pool, and eventually to the environment. The bubble scrubbing portion of this code utilizes classical aerosol scrubbing theories to predict an overall decontamination factor. To validate this code, experimental data is needed. Currently, there are limited bubble scrubbing data in a sodium coolant pool. This thesis aims to validate the Simplified Radionuclide Transport bubble scrubbing code in a sodium environment while conducting a parametric study to analyze the effects of aerosol size, bubble size, pool temperature, pool depth, aerosol density, and aerosol concentration. Through a series of experiments, it was determined that aerosol sizes ranging from 0.001 to 1 microns are of primary concern as aerosols in this range have a ratio of aerosol mass entering the sodium pool from the fuel pin to aerosol mass exiting the sodium pool to the cover gas region of less than 10. The experimental results were found to match the trends found in the scrubbing model closely, but significantly more scrubbing was seen experimentally. Decreasing bubble size and increasing pool depth and aerosol density were all found to increase scrubbing both experimentally and theoretically. Pool temperature was found to have a negligible effect on scrubbing amounts; however, this was largely due to a subsequent increase in bubble size corresponding to increasing temperatures. Varying aerosol concentration was found to have no effect on scrubbing ratios. A final series of tests was conducted for a more prototypic fuel pin failure with a heterogenous bubble swarm. From these tests, it was found that the experimental scrubbing quantities were larger than for the single bubble case. Overall, it was found that the simplified bubble transport scrubbing code accurately models the trends of the bubble scrubbing but provides a conservative estimate of scrubbing quantities.