Applying a nodal diffusion-micro depletion sequence for high-fidelity analysis of sodium cooled fast reactors
Pereira, Gustavo Sigwalt Horn
MetadataShow full item record
Recent developments in the global nuclear industry have led to the need of reactor designs that are not only safe, but also address the challenges involving nuclear waste while producing clean electricity at low costs. One of the designs proposed to fill these requirements is the Advanced Burner Reactor (ABR), a sodium cooled, metal fuel fast reactor system that uses spent fuel from current light water reactors as part of its energy source. Due to the complex nature of nuclear reactors, extensive modeling of a system must be performed in order to demonstrate the viability of such system. This thesis combines two established reactor modeling techniques in order to efficiently model the ABR core. The computational methods used in this work are Monte Carlo (MC) and nodal diffusion. The MC method is a well-established computational approach for modeling of nuclear systems, and is considered to be very accurate and versatile. However, the MC requires extensive time and computational resources, and its applicability becomes prohibitively expensive when performing analyses of accident scenarios. Meanwhile, the nodal diffusion method requires much fewer resources to perform such analyses, but theoretically the accuracy is compromised due to the simplifications applied to the model. The main focus of the work presented in this thesis revolves around expanding the capabilities of nodal diffusion codes to calculate local isotopic concentrations, activities and decay heat quantities, which is a first-of-a-kind demonstration of the applicability of nodal diffusion codes for such calculations. Establishing this approach allows for the capability of decay heat to be calculated rapidly and efficiently, allowing for the performance of transient analyses in accident scenarios. The work presented in this thesis uses the MC code Serpent as a macroscopic and microscopic cross-section generation tool, and the nodal diffusion code DYN3D for full core analysis of the ABR core. The Serpent-DYN3D code sequence is then applied for various scenarios, including decay heat analysis, and compared to reference MC solutions. It is found that the Serpent DYN3D sequence is an adequate tool for modeling of sodium cooled, metal fuel fast reactors, providing accurate solutions while saving on time and computational resources required.