The Development of a Nodal Method for the Analysis of PWR Cores with Advanced Fuels

摘要

This thesis outlines the development of a nodal method with the purpose of addressing difficultiesudencountered in the modelling of advanced fuels. The standard calculational route used whenudmodelling a Uranium (U) fuelled Pressurised Water Reactor (PWR) is not accurate enough toudanalyse a PWR containing U and Plutonium (Pu). This is because the assumptions madeudwhen developing the standard route are not necessarily representative of situations involvingudadvanced fuels.udTo address some of these poor assumptions a nodal method has been developed which canudsolve the SPN equations in multiple energy groups. The SPN equations are an asymptoticudapproximation of the full neutron transport equation, and as such will include more physicaludeffects than the neutron diffusion equation. The theory behind the development of this nodaludmethod is outlined in this thesis along with an extensive set of benchmark tests for verificationudof the method. It is found that through a similarity transformation of the determining equations,udexisting nodal diffusion solvers can obtain solutions to the SPN equations without anyudapproximations.udPreviously EDF Energy have developed an embedded methodology to address the shortcomingsudof the standard calculational route. This procedure solves the diffusion equation in greater detailudon local sub-meshes in order to correct the standard 2 group nuclear data, and reduces the pinudpower errors by ≈ 50% by capturing spectral effects on the interface between two significantlyuddifferent fuel types. In this thesis the incorporation of the SPN nodal method into the embeddedudmethodology is described.udA small light water reactor benchmark is solved to test the accuracy of the embedded methodologyudcombined with the SPN nodal method. It is concluded that similar accuracy to diffusionudis attained with the SPN equations. This is because the homogenisation procedure producesudan error larger than the improvements due to the use of the SPN equations. To address theudlimitations discovered in this thesis future work is proposed based on the author’s experienceudof research in the area.