Minimizing the standard deviation of the thermal load in the spent nuclear fuel cask loading problem

The paper assumes that, at the end of the operational period of a Spanish nuclear power plant, an Independent Spent Fuel Storage Installation will be used for long-term storage. Spent fuel assemblies are selected and transferred to casks for dry storage, with a series of imposed restrictions (e.g.,...

Descripción completa

Detalles Bibliográficos
Autores: Bautista Valhondo, Joaquín|||0000-0002-2214-4991, Batet Miracle, Lluís|||0000-0003-1882-6313, Mateo Doll, Manuel|||0000-0002-3975-9116
Tipo de recurso: artículo
Fecha de publicación:2020
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/334748
Acceso en línea:https://hdl.handle.net/2117/334748
https://dx.doi.org/10.3390/en13184869
Access Level:acceso abierto
Palabra clave:Spent reactor fuels
Mixed integer linear programming (MILP)
Multi-start metaheuristics
Spent nuclear fuel
Nuclear power plant operations
Independent Spent Fuel Storage Installation (ISFSI)
Combustibles nuclears gastats
Àrees temàtiques de la UPC::Física
Descripción
Sumario:The paper assumes that, at the end of the operational period of a Spanish nuclear power plant, an Independent Spent Fuel Storage Installation will be used for long-term storage. Spent fuel assemblies are selected and transferred to casks for dry storage, with a series of imposed restrictions (e.g., limiting the thermal load). In this context, we present a variant of the problem of spent nuclear fuel cask loading in one stage (i.e., the fuel is completely transferred from the spent fuel pool to the casks at once), offering a multi-start metaheuristic of three phases. (1) A mixed integer linear programming (MILP-1) model is used to minimize the cost of the casks required. (2) A deterministic algorithm (A1) assigns the spent fuel assemblies to a specific region of a specific cask based on an MILP-1 solution. (3) Starting from the A1 solutions, a local search algorithm (A2) minimizes the standard deviation of the thermal load among casks. Instances with 1200 fuel assemblies (and six intervals for the decay heat) are optimally solved by MILP-1 plus A1 in less than one second. Additionally, A2 gets a Pearson’s coefficient of variation lower than 0.75% in less than 260s CPU (1000 iterations)