Alejandro Castillo

BWR fuel reloads design using a Tabu search technique

Abstract We have developed a system to design optimized boiling water reactor fuel reloads. This system is based on the Tabu Search technique along with the heuristic rules of Control Cell Core and Low Leakage. These heuristic rules are a common practice in fuel management to maximize fuel assembly utilization and minimize core vessel damage, respectively. The system uses the 3-D simulator code CM-PRESTO and it has as objective function to maximize the cycle length while satisfying the operational thermal limits and cold shutdown constraints. In the system tabu search ideas such as random dynamic tabu tenure, and frequency-based memory are used. To test this system an optimized boiling water reactor cycle was designed and compared against an actual operating cycle. Numerical experiments show an improved energy cycle compared with the loading patterns generated by engineer expertise and genetic algorithms.

A new system to fuel loading and control rod pattern optimization in boiling water reactors

Abstract A new system to optimize both control rod pattern and fuel-loading design in boiling water reactors is shown. The system is named OCOTH, and it is based on heuristic techniques such as genetic algorithms, neural networks, and ant colonies. Each heuristic technique is used to design a part of the optimization process. So, the neural network finds an initial fuel loading with a Haling burnup calculation. The ant colony system optimizes full-power control rod pattern of the initial fuel loading. Finally, the genetic algorithm optimizes fuel loading with the optimized control rod patterns. The ant colony system and the genetic algorithm perform an iterative loop until a stop criterion is fulfilled; for example, control rod pattern and fuel-loading convergence. The OCOTH system was tested in an equilibrium cycle of Mexico’s Laguna Verde Nuclear Power Plant. We found very good results in control rod pattern and fuel-loading coupled optimization.

Nuclear Fuel Lattice Optimization Using Neural Networks and a Fuzzy Logic System

Abstract RENO-CC, a system to optimize nuclear fuel lattices for boiling water reactors using a multistate recurrent neural network, is shown. This kind of neural network is formed by only one layer of neurons. Each neuron is associated with a pin of the fuel lattice array. RENO-CC was tested through the fuel lattice design of 10 × 10 arrays with two water channels. Thus, the neural network has a total of 51 neurons; four neurons are associated with the channels (they correspond to a half fuel lattice). The neuron’s outputs are known as the neural states. The RENO-CC’s neural network works by changing the neural states in order to decrease or increase the value of an objective function. Neural states are chosen from an inventory of pins with different 235U enrichment and gadolinia concentrations. The objective function includes both the local power peaking factor and the infinite multiplication factor. These parameters are calculated with the HELIOS code. A fuzzy logic system is applied in order to decide if the designed fuel lattice is suitable to be evaluated by a three-dimensional reactor core simulator. To carry out the assessment, the fuel lattices with the best fuzzy qualification are placed at the bottom zone of a predesigned fuel assembly and predesigned fuel loading and control rod patterns. Fuel lattice performance is verified with the Core Master PRESTO core simulator. According to the obtained results, RENO-CC could be considered as a powerful tool to design fuel lattices. The system was programmed with Fortran 77 using a UNIX interface in an Alpha workstation.