Please use this identifier to cite or link to this item:
|Title:||Optimization of In-Core Fuel Management in CANDU Nuclear Reactors|
|Keywords:||Nuclear Engineering;Nuclear Engineering|
|Abstract:||<p>A numerical approach was taken to optimize in-core fuel management in CANDU reactors at equilibrium refuelling. A computer program called, OPTEX was written to obtain an optimum distribution of fuelling rates and initial fuel enrichment.</p> <p>In an on-power refuelled reactor such as CANDU, depletion effects are quite localized and the global power shape results from the distribution of fuelling rates and fixed absorbers in the core. An important group of fixed absorbers, called adjuster rods, can be replaced (and modified) a number of times during the life of the reactor. The optimization process in OPTEX can be applied to the adjuster grading and the fuel management problems simultaneously, yielding the desired reactivity worth for the adjusters as well as a nominal power distribution which satisfies the design limits on the fuel.</p> <p>The OPTEX approach consists in linearizing the system characteristics about points in the feasible domain of the state variables, and then using mathematical programing techniques to direct the search towards an improved state. The originality of the approach is that variational techniques, found in Generalized Perturbation Theory (GPT) or in sensitivity theory, are used to define the true gradients of the system characteristics.</p> <p>The computational advantages of this approach were illustrated with two applications of the code. Using a new correlation to properly account for the effects of enrichment and axial fuelling schemes on the channel power peaking, fresh fuel enrichment was optimized in a 850 MWe CANDU reactor using Lightly Enriched Uranium (LEU) fuel. The optimum uniform enrichment was of the order of 0.91%, but was found to be sensitive to the axial fuelling schemes used. In all cases, a significant reduction of 25-30% in fuelling costs was observed.</p> <p>The second problem dealt with adjuster grading in an existing 600 MW reactor. Starting with a current (nonuniform) distribution of adjuster weights, a new distribution of weights was found along with an optimized burnup distribution to yield a 3% reduction in fuelling costs while preserving the global adjuster reactivity worth.</p> <p>A number of other applications were identified. In particular, it was shown how, by introducing a time-dependent GPT, the UPTEX approach could be extended to deal with the optimization of the initial core and other fuelling transients.</p>|
|Appears in Collections:||Open Access Dissertations and Theses|
Items in MacSphere are protected by copyright, with all rights reserved, unless otherwise indicated.