Next Article in Journal
A Novel Approach to Enhance the Generalization Capability of the Hourly Solar Diffuse Horizontal Irradiance Models on Diverse Climates
Next Article in Special Issue
Multitask Support Vector Regression for Solar and Wind Energy Prediction
Previous Article in Journal / Special Issue
Machine Learning-Based Approach to Predict Energy Consumption of Renewable and Nonrenewable Power Sources
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Minimizing the Standard Deviation of the Thermal Load in the Spent Nuclear Fuel Cask Loading Problem

ETSEIB, Universitat Politècnica de Catalunya, Avda. Diagonal 647, 08028 Barcelona, Spain
*
Author to whom correspondence should be addressed.
Energies 2020, 13(18), 4869; https://doi.org/10.3390/en13184869
Submission received: 4 August 2020 / Revised: 13 September 2020 / Accepted: 15 September 2020 / Published: 17 September 2020
(This article belongs to the Special Issue Soft Computing Techniques in Energy System)

Abstract

:
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).

Graphical Abstract

1. Introduction

Optimization problems are found at different levels in electric power systems, from both technological and operations management standpoints. In transport and distribution, not intending to make an exhaustive list, there are studies related to classical modeling and problem solving for electric energy distribution static planning [1,2,3]; other works, also on planning, propose a flexible distribution attending to the emerging variability of demand [4]. There are also many examples of optimization problems at the source point, from renewable generation (see for instance [5] for wind) to nuclear power.
A nuclear power plant is a complex system in which many optimization processes take place. The balance between safety requirements, competition in a liberalized electricity market, and access to limited human and material resources is resolved in nuclear power plants through an approach that prioritizes safety (see, for instance, INSAG-4 [6] and IAEA-TECDOC-1329 [7]).
Nuclear Power Plants (NPP) produce electricity from the heat released in a fission chain reaction; the so-called “nuclear fuel” is actually fissionable material. In Light Water Reactors (LWR), the type of reactor that constitutes the majority of the world’s nuclear fleet and the total of the Spanish fleet [8], nuclear fuel is made of uranium (slightly enriched in the 235U isotope), in the form of ceramic pellets of UO2 protected and waterproofed by a metallic cladding (zirconium alloy) and grouped together, forming the so-called fuel assemblies (FA) [9]. The fuel used in the two units of Ascó NPP (the plant that has inspired the present study) is manufactured by ENUSA under a license agreement with Westinghouse, the owner of the technology [10].
Presently, the two Ascó units have fuel burn-up cycles of one and a half year. Each 18 months, some 60 FA (out of the 157 contained in the core of the reactor) are replaced (each one containing about 500 kg of uranium). Up to the end of its service, each Spent Fuel Assembly (SFA) has produced around 50 GWd/t of heat.
In the fission process, uranium atoms are transformed into radioactive fission products. Roughly speaking, in an SFA, some 5% of the initial uranium mass has undergone fission. A smaller part of the uranium is also transformed, through neutron capture nuclear reactions, into transuranic elements, which are also radioactive. The hazard of spent fuel is due to the potentially harmful effects of ionizing radiation. Thus, it is necessary to isolate the spent fuel for a long time to avoid the dispersion of the radioactive substances. An ultimate storage solution is proposed in repositories excavated in deep geological formations [11], in a way similar to that followed with chemical contaminants such as mercury [12].
The decay of radioactive substances in the spent fuel causes the so-called “decay heat”, which decreases as the activity of those substances declines, but which requires careful management of SFAs to prevent them from being damaged by excess temperature. After being removed from the core of the reactor, spent fuel is temporarily stored in the fuel pool of each power unit. Pool water provides shielding and cooling. Only when the power of the SFA has been sufficiently reduced can it be considered for dry storage in appropriate containers (either temporarily or permanently in a deep geological repository). The lack of political consensus for a definitive solution in Spain made it necessary to propose a Central Interim Storage Facility (CISF) [13], which was never built (now, it is proposed for 2028 [14]).
Over time, the spent fuel pools of the Spanish NPP have been reaching the limit of their capacity. In the case of the two units of Ascó, the pool in each of them can house 1421 FA, which is a figure that includes the 157 positions needed to discharge the entire reactor core if necessary. At the end of 2018, there were 1160 FAs in the Ascó 1 pool and 1104 in the Ascó 2 pool. The plant has 608 SFAs in an Independent Spent Fuel Storage Installation (ISFSI) in 19 HI-STORM containers [15], each hosting 32 SFAs [16].
In Spain, ENRESA (a state-owned company) is responsible for the decommissioning of the nuclear installations, as well as for the ultimate disposal of nuclear and radioactive wastes [17]. At the end of the exploitation period, it is in the interest of the plant operator to be able to transfer ownership of the installation to ENRESA in the shortest possible time. In order to do it, previously, all the fuel must be completely removed from the spent fuel pool [18]. At this moment, the ISFSI can be used as a logistics storage, housing spent fuel containers prior to transportation to the CISF. If the CISF is not yet available, the ISFSI should be usable as a full-scope storage (perhaps even in the long-term).
Thus, an optimization problem arises: it is necessary to remove all of the SFA from the pool within the minimum time and at the minimum cost [19]? Indeed, optimization problems can be found in a wide variety of operations in NPPs. For instance, in the spent fuel pool, where the SFAs are stored in racks, problems arise in optimizing the reliability of the design of the racks and the arrangement of the assemblies in the event of their possible displacement [20], and, as the capacity of the pool is limited to a specific number of assemblies, other management problems of spent nuclear fuel storage arise linked to this limitation [21]. At the right time, SFAs are removed from the pool and relocated in casks, often for long-term storage. The optimization here ranges from determining the SNF removal allocation strategy for a fleet of plants [22], to the design of the containers (e.g., what capacity, how much cooling time needed, etc.) [23], to the strategies to be followed in the loading of the casks.
As for the latter, Zerovnik et al. [24] propose combinatorial methods to optimize the cask loading, aiming at final geological repository. Considering as well the final deposition, Ranta and Cameron [25] used a two-phase heuristic method in order to minimize the thermal load of the canister with the largest load. Gomes et al. [26] propose a methodology, using genetic algorithms, to determine the optimal time to load the casks for dry interim storage, taking into account several cost factors. Spencer et al. [27] have develop a method (embedding greedy randomized adaptive search procedures in a memetic algorithm) to optimize the cask loading in a multi-objective search of configurations that minimize the number of casks, the thermal load, and the time when the transportation requirements are met.
In the present work, the cask loading problem is solved with two hierarchical optimization criteria: (i) minimize the cost of the regionalized casks necessary to relocate the spent fuel located in the pool at a given moment [28] using mixed integer linear programming [29]; and (ii) minimize the standard deviation of the thermal load in the set of casks using a multi-start metaheuristic with local search procedures [30].
In Section 2, we describe some dry storage systems for spent nuclear fuel. Section 3 is devoted to the description, modeling, and resolution of the spent nuclear fuel cask loading problem. Section 4 illustrates the procedure proposed in a case study inspired in Ascó NPP. Finally, Section 5 provides the conclusions derived from this work and a brief description of future research lines.

2. Dry Storage Systems for Spent Nuclear Fuel

Different solutions exist for the interim dry storage of spent nuclear fuel; as a comparison, the 14 closed sites in the USA use 11 different storage systems from four different suppliers [31].
In Spain, ENRESA has adopted different solutions. Some plants are using dual-purpose containers (transport and storage) whose structure provides shielding, confinement, and heat dissipation. Ascó’s two units are using the multi-purpose HI-STORM system, which is described below [15,16].
The spent fuel storage and transport system in Ascó is made up of different elements: the multi-purpose canister (MPC-32), the storage module (HI-STORM), the transport overpack (HI-STAR), and the transfer cask (HI-TRAC), which allows the MPC to be transferred to HI-STORM and HI-STAR [15] (see Figure 1).
The time required to completely remove the spent fuel from the pool depends on the minimum cooling time necessary for the last SFA to be loaded into a cask. This cooling period (that is, the time elapsed since the unload of the SFA from the reactor core) depends, in turn, on the type of container used and the technical specifications approved by the regulator for it.
To guarantee the integrity of the SFAs, the temperature of the metal cladding that surrounds the UO2 pellets must be limited. This fact imposes requirements of sufficient heat dissipation in the cask design. However, also, it imposes a limit to the thermal power that the SFAs can generate inside the cask. For instance, in the HI-STORM system, under uniform charging conditions, each of the positions of the MPC-32 can house an SFA with a maximum power of 0.9375 kW at the time of loading (i.e., the total thermal power dissipated by the canister will be limited to 30 kW) [32]. The system allows regionalized loads; so, SFAs with higher decay heat can be located in the central region of the canister (region R a , formed by 12 positions), at the cost of limiting the maximum power per position in the peripheral region (region R b , 20 positions) and also limiting the total thermal load to a value less than 30 kW [33]. Figure 2 represents the positions assigned to each region.
In fact, the loading of a cask is constrained by the technical specifications in a more complex way, since they take into account, in addition to the cooling time, multiple fuel characteristics such as decay heat, burn-up, initial enrichment, the possibility that the assembly has some damage, or the possible presence of elements such as neutron sources, control rods, burnable poisons, etc. These conditions vary according to the type of cask, its use (storage or transport), and other specific limitations.
In what follows, it is assumed that at the end of the operational life of the plant, an ISFSI will exist that allows transfering all of the SFAs to it using the HI-STORM system (without imposing the present transport requirements, which are more restrictive than storage requirements). It is considered here that the storage requirements apply only to the decay heat of each SFA (calculated in turn from the burn-up, cooling time, and initial enrichment). The minimum time for the complete removal of all of the SFA from the pool is bounded by the moment when the hottest SFA fits into the requirements. The cost will depend on the number of containers to be used. So, the number of containers needed to remove all the SFA from the spent fuel pool needs to be minimized: this is the optimization problem addressed in the present study.

3. The Spent Nuclear Fuel Cask Loading Problem

In this section, we describe and formalize the spent nuclear fuel cask loading problem, and we propose a hybrid metaheuristic to solve it. Our method works in three phases using mixed integer linear programming (MILP), a deterministic algorithm, and local search heuristics.

3.1. Description of the Problem

At a given instant T , corresponding to the planned date for transferring in one go the contents of the spent fuel pool to the ISFSI, there are n SFAs, forming a set I ( i = 1 , , n ( n | I | )).
At the time of removal, T, each fuel assembly i I is characterized by a thermal power (decay heat), q i ( T ) .
On the other end, there is an ISFSI ready to host all of the casks needed. Casks are regionalized, i.e., they are compartmentalized into regions with attributes of capacity and decay heat. Thus, there is a set of virtual regions J ( j = 0 , 1 , , m ( | J | m + 1 ) ) , in which each type of region ( j J ) is characterized by the upper and lower bounds on the decay heat a single SFA may have [ q j , q j + ] , and the number of SFAs in that region that can be held by a single cask, h j . j = 0 symbolizes the type of universal region capable of housing any fuel assembly without restrictions.
There is a set K ( k = 0 , 1 , , l ( | K | l + 1 ) ) of cask classes; k = 0 symbolizes the universal class, which is capable of accommodating any fuel assembly without restrictions at an arbitrary high cost γ 0 . Each class is defined in relation to the maximum thermal load allowed in each region of the cask.
An adversity coefficient γ k is associated to each class of cask ( k K ) , corresponding to the arbitrary cost of selecting a cask of class k K to house at least one SFA. We also assume for the class k K a maximum number of units available, d k   ( k ) . The total number of available casks is D = k d k .
The relationship between a virtual region type ( j J ) and a cask class ( k K ) is given by matrix B , whose coefficients b j , k   ( j k ) take the value 1 when the region of type j J is associated to the cask of class k K , and 0 otherwise.
On the other hand, the relationship between the fuel assembly i I and the type of region j J is given by matrix A ( T ) , whose coefficients, a i , j ( T )   ( i j ) , are equal to 1 when the fuel assembly i I can be located in the region j J at time T , and 0 in any other case; i.e., a i , j ( T ) = 1 q i ( T ) [ q j , q j + ] . It follows from it that the relationship between the fuel assembly i I and the class of cask k K is given by matrix C ( T ) resulting from the boolean product of matrices A ( T ) and B ; that is, C ( T ) = A ( T ) B , thus fulfilling the following property: ( a i , j ( T ) = 1 b j , k = 1 ) c i , k ( T ) = 1 .
Under such conditions [28], the spent nuclear fuel cask loading problem consists of transferring the maximum number of the fuel assemblies, which are located in the pool of a nuclear power plant on a date T and expressed as time after the end of operation, into a set of regionalized casks of different classes, meeting the restrictions and minimizing costs (i.e., minimizing the total adversity).

3.2. Solving the Problem

As discussed above, our proposal solves the problem in three phases. The first phase uses MILP to allocate fuel assemblies to virtual regions with minimum cost (i.e., a minimum number of casks or minimum total adversity) [28]. In the second phase, a family of deterministic algorithms has been designed to assign each SFA to a specific cask actual region. Finally, in the third phase, a local search heuristic is proposed to minimize the standard deviation of the thermal load of the casks set.

3.2.1. Phase 1: MILP Model to Minimize the Cask Loading Cost

To formalize the proposed resolution methods, we use the following nomenclature.
Parameters:
T Time or expected date to remove in one go the SFA from the pool in the nuclear power plant, expressed as time after the end of operation.
I Set of SFA located in the pool of the nuclear power plant at time T . Let the elements be i = 1 , , n ( n | I | ).
J Set of virtual region types according to the upper and lower bounds on the decay heat of the SFAs. Let the regions be j = 0 , 1 , , m with | J | m + 1 . The universal region, j = 0 , does not impose any limit; for practical purposes, the assigned elements to the universal region will remain in the pool.
K Set of classes of casks: k = 0 , 1 , , l with | K | l + 1 . The type of cask considered here corresponds to MPC-32 described above. We will assume that there is a type of universal container, k = 0 , associated with the universal region j = 0 ; for practical purpose, the elements assigned to the universal container will remain in the pool.
q i ( T ) Decay heat of the fuel assembly i I at the time T .
q j + Upper limit of a single SFA’s decay heat in the virtual region j J .
q j Lower limit of a single SFA’s decay heat in the virtual region type j J . Here, we will assume
q j = 0   j J .
A ( T ) , a i , j ( T ) Element–region compatibility matrix; technological coefficients a i , j equal to 1 if the fuel assembly i I can be located in the virtual region j J at time T , and 0 otherwise. Formally:
a i , j a i , j ( T ) = { 0 , i f   q i ( T ) > q j + 1 , otherwise } .
B , b j , k Region-cask compatibility matrix; technological coefficients b j , k equal to 1 if the region of type j J is associated with the cask class k K , and 0 otherwise.
C ( T ) , c i , k ( T ) Element–cask compatibility matrix, which is defined as C ( T ) = A ( T ) B ; technological coefficients c i , k ( T ) are equal to 1 when the element i I can be placed in a cask of class k K at time T , and 0 otherwise. Therefore, it is true that:
c i , k c i , k ( T ) = { 1 , i f j J a i , j ( T ) b j , k = 1 0 , otherwise } .
h j Storage capacity of a region of type j J measured by the number of SFAs a single cask can hold in that region at most.
H k Maximum storage capacity of a cask of class k K measured by the number of fuel assemblies. The relationship H k = j b j , k × h j is fulfilled.
d k , D Number of available casks of the class k K at time T . The expected total cask availability is D = k d k .
γ k Adversity coefficient for a cask of class k K . The value γ k symbolizes the arbitrary cost if a cask of class k K is selected to hold SFAs.
Variables:
x i , j Binary variable that equals 1 if the fuel assembly i I is assigned to the virtual region j J , and 0 otherwise.
y j Number of fuel assemblies assigned to the virtual region j J .
z k Number of casks of class k K .
Γ ( T ) Total adversity function. It represents the overall cost of the casks used to hold the spent nuclear fuel on the scheduled date, T , after the pool of the nuclear power plant is emptied.
Using these parameters and variables, we formulate the following mathematical model.
MILP-1: Model for the assignment of fuel assemblies to types of virtual region
min Γ ( T ) = k = 0 l γ k z k
Subject to:
i = 1 n j = 0 m x i , j = n
j = 0 m a i , j x i , j = 1                                                         i = 1 , , n
y j = i = 1 n a i , j x i , j                                                                     j = 0 , 1 , , m
z k b j , k h j y j j = 0 , 1 , , m ;                                             k = 0 , 1 , , l
z k d k                                                               k = 0 , 1 , , l
x i , j { 0 , 1 } i = 1 , , n               j = 0 , 1 , , m
y j + { 0 }                                                                                     j = 0 , 1 , , m
z k + { 0 }                                                                     k = 0 , 1 , , l
In the MILP-1 model, the objective Function (1) represents the minimization of the cask loading cost or total adversity, whose value coincides with the total number of casks when all the fuel assemblies are removed from the pool ( z 0 = 0 ) and the adversity coefficients are unitary for any class of cask ( γ k = 1   k = 1 , , l ) . Equalities (2) and (3) guarantee the assignment of all the fuel assemblies to any region, including the universal region j = 0 . Equality (4) determines the number of fuel assemblies that must be assigned to each type of region j J . Constraint (5) lower bounds the number of casks of each class k K , while Constraint (6) upper bounds that value according to their availability. Condition (7) imposes that the decision variables ( x i , j ) are binary and, finally, the non-negativity and integrity of the auxiliary variables y j and z k are given by (8) and (9), respectively.
Running the MILP-1 model offers three types of results: (i) an optimal assignment of the fuel assemblies to the various virtual regions ( x i , j * { 0 , 1 } i j ) , (ii) the optimal number of SFA assigned to each region ( y j * j ) , and (iii) the optimal number of casks of each class ( z k * k ) corresponding to the minimum cost.
Although the previous results are interesting, the location of the fuel assemblies in the real regions of the casks remains to be determined.

3.2.2. Phase 2: Assignment of the Fuel Assemblies to the Real Cask Regions

Any sequence of objects corresponding to the triplet of sets ( I , J , K ) can lead to a different solution. For example, if the sequence is built on the set I of fuel assemblies, then the number of different ways to assign is n ! This can lead to an astronomical number of potentially different solutions for instances of realistic dimensions (e.g., 1164 ! in our case study). Based on this idea, we propose a family of constructive algorithms, A 1 [ π ( n ) ], which depends on the regulatory sequence π ( n ) of the assignment of fuel assemblies to virtual regions.
Starting from an optimal solution given by MILP-1 ( x i , j *   i j ,   y j *   j ,   z k *   k ) , the procedure A 1 [ π ( n ) ] creates a sequence π ( n ) of fuel assemblies (Statement 5 in A 1 [ π ( n ) ] ) to establish the sequence of assignment of SFAs to casks. Once the procedure determines the class of the cask k t * that corresponds to the current element π t , the element is recorded in the matrix M P C 32 [ · ] (Statement 13 in A 1 [ π ( n ) ] ), whose rows correspond to casks and whose columns correspond to positions within the cask. Subsequently, the heat loads Q m p c 32 [ · ]   in the casks are determined (Statement 16 in A 1 [ π ( n ) ] ), and the elements are sequenced in the casks according to such loads (Statement 17 in A 1 [ π ( n ) ] ). Finally, the results of the algorithm are returned (Statement 18 in A 1 [ π ( n ) ] ).
Algorithm 1. A 1 [ π ( n ) ] : Algorithms to assign fuel assemblies to regions according to π ( n )
  • // Initialization
  • Input I , ( n , m , l ) , B [ b j , k ] , ( q i   i ) ,   ( x i , j *   i j ) ,   ( y j * j ) ,   ( z k * k ) , Ρ κ
  • Initialize t = 0 , M P C 32 [ z k * , 32 ] = { }  
  • // Create assignment sequence of SFAs
  • π ( n ) = ( π 1 , π 2 , , π n ) G e n e r a t e _ s e q u e n c e ( I , n , Ρ κ )
  • While ( t n ) do
  • Set t = t + 1
  • // Find the virtual region assigned to element π t . Let j t *
  • Set j t * = a r g m a x j ( x π t , j * )
  • // Find cask class associated with region j t * . Let k t *
  • Set k t * = a r g m a x k ( b j t * , k )
  • // Record the SFA in the matrix MPC32
  • M P C 32 [ · ]   R e c o r d _ e l e m e n t ( I , π t , j t * , k t * )
  • End while
  • // Calculate residual heat loads and order elements in matrix MPC32
  • Q m p c 32 [ · ] C a l c u l a t e _ l o a d s ( I , q i , M P C 32 [ · ] )
  • M P C 32 [ · ]   O r d e r _ e l e m e n t s ( I , Q m p c 32 [ · ] , M P C 32 [ · ] )
  • Return Q m p c 32 [ · ] ,   M P C 32 [ · ]
  • // End Algorithm A 1 [ π ( n ) ]
Note that Algorithm 1 A 1 [ π ( n ) ] can generate many solutions thanks to the fact that the starting point is a permutation π ( n )   of the fuel assemblies of the set I (Statement 5 in A 1 [ π ( n ) ] ). Such permutations can be created using a family of rules Ρ κ applied on the set I , which is initially ordered according to the SFA’s identifier ( i I ). The family of rules Ρ κ can be of diverse nature; they may consist of the same initial sequence (lexicographic order) or the simple sequence of the elements according to the decay heat or in more complex rules used for the generation of neighborhoods (exchange and insertion) in algorithms based on trajectories through the search domain.
Obviously, all the solutions generated by A 1 [ π ( n ) ] are optimal when the function objective is the minimization of the cost of the casks [28], as happens in Phases 1 and 2 of this work; however, when the objective function corresponds to another optimization criterion, such as the balancing of the heat load in the set of casks, A 1 [ π ( n ) ] can be used in, at least, two ways:
(1)
In the construction phase (or phases) of multi-start algorithms, v.gr. greedy randomized adaptive search procedure (GRASP) algorithms, to build initial solutions to be used as seeds for the next phase of trajectory-based improvement in the search domain.
(2)
In the systematic generation of solutions, to assist metaheuristics based on populations (v.gr. genetic and memetic algorithms).
In this work, Algorithm 1 A 1 [ π ( n ) ] constitutes one of the fundamental pillars for the construction of initial solutions, and the chosen way is (1).

3.2.3. Phase 3: Reduction of the Dispersion of the Thermal Load of the Casks

Phase 3 of the proposed procedure aims to reduce, by a predetermined number of iterations, the dispersion of the thermal load (measured in kW) in the set of casks. The dispersion metric to minimize chosen here is the standard deviation σ ( Q , T ) ; this is:
min σ ( Q , T ) = 1 z t o t * k = 1 l δ k = 1 z k * ( Q δ k Q a v g ) 2
where:
z t o t * Total optimal number of casks: z t o t * = k z k * . It is obtained by applying the MILP-1 procedure by setting values to the adversity coefficient γ k ( k K ).
Q δ k Total thermal load due to the SFA loaded on date T in the δ k -th cask of class k : δ k = 1 , , z k *   ( k K ) . These values are obtained after applying the Algorithm 1 A 1 [ π ( n ) ] by fixing a permutation π ( n ) of the elements of set I .
Q a v g Average value of the thermal load in the set of casks on date T :
Q a v g = i I q i ( T ) z t o t *
σ ( Q , T ) Standard deviation of the thermal load in the set of casks on date T .
In order to carry out Phase 3, the non-exhaustive descent Algorithm 2 A 2 [ π ( n ) ] is applied, based on local search optimization (see Algorithm 2 A 2 [ π ( n ) ] below). Its main characteristics are as follows.
Phase 3.1: Initialization of Algorithm 2 A 2 [ π ( n ) ]
  • An initial solution X 0 is obtained, applying the mathematical program MILP-1, after prefixing the values of the vector γ = ( γ 1 , , γ l ) of adversity coefficients; this solution is characterized by the triple of sets of values ( x i j *   i j ,   y j *   j ,   z k *   k ) .
  • The Algorithm 1 A 1 [ π ( n ) ] is applied to the solution X 0 , considering the arbitrary sequence π ( n ) , which initially corresponds to π 0 ( n ) : the lexicographic order of the SFA’s identifiers ( i I ). Thus, the extended solution X m p c 0 = { X 0 , π ( n ) , M P C 32 [ z k * , 32 ] } is obtained, which contains detailed information on the location of each element inside the casks z k *   ( k ) in the matrix M P C 32 [ z k * , 32 ] .
  • Solution X m p c 0 is fixed as the incumbent solution of the local search iterative procedure ( X m p c c X m p c 0 ) . In addition, X m p c 0 is fixed as the best solution ( X m p c * X m p c 0 ) , and the standard deviation of this solution σ ( X m p c * ) is determined.
Phase 3.2: Improvement of the initial solution with A 2 [ π ( n ) ]
  • A tentative solution X ^ m p c c belonging to the restricted neighborhood of the incumbent solution X m p c c is generated—that is, X ^ m p c c N ( X m p c c , Ρ L S , S F ) —where Ρ L S is the rule for generating neighboring solutions and S F is the set of possible solutions. Specifically, given the incumbent solution X m p c c , the generation rule Ρ L S operates as follows: for all pairs of different casks ( δ k , δ k ) and all pairs of positions in these casks ( p o s ( δ k ) , p o s ( δ k ) ) , the tentative neighboring solutions ( X ^ m p c c ) exchange the elements between these positions, within the set of possible solutions S F that is regulated by the SFA–cask compatibility matrix C ( T ) .
  • If the tentative solution X ^ m p c c is a possible solution ( X ^ m p c c S F ) and σ ( X ^ m p c c ) < σ ( X m p c * ) , then the solution X ^ m p c c becomes the best solution ( X m p c * X ^ m p c c ) and also the new incumbent solution ( X m p c c X ^ m p c c ) .
  • The above steps are repeated within Phase 3.2 as long as there is an improvement to the incumbent solution or the number of tentative solutions generated does not reach the preset value i t ^ (v.gr. i t ^ = 1000 tentative solutions).
Algorithm 2. A 2 [ π ( n ) ] : Non-exhaustive local search descent algorithm based on π ( n )
  • // Initialization
  • Input I , ( n , m , l ) , C [ c i , k ] , ( q i   i ) ,   X 0 , π ( n ) , Ρ L S , i t ^  
  • X m p c 0 G e n e r a t e _ s o l u t i o n ( X 0 , A 1 [ π ( n ) ]
  • X m p c c X m p c 0 ;   X m p c * X m p c 0 ;   i t e r = 0  
  • // Solution improvement by local search
  • Repeat
  • X ^ m p c c X m p c c ; e x i t = t r u e ;   i t e r = i t e r + 1
  • N ( X c , Ρ κ , S F ) O p e r a t o r _ N e i g h b o u r h o o d ( X m p c c , Ρ L S , S F )
  • Repeat for all X N ( X m p c c , Ρ L S , S F )
  • If [ σ ( X ) < σ ( X ^ m p c c ) ] = t r u e then X ^ m p c c X
  • Until ( [ σ ( X ^ m p c c ) < σ ( X m p c c ) ] = t r u e )
  • If [ σ ( X ^ m p c c ) < σ ( X m p c c ) ] = t r u e then X m p c * X ^ m p c c X m p c c X ^ m p c c e x i t = f a l s e
  • Until ( e x i t = t r u e     i t e r i t ^ )
  • Return { X m p c * ,   σ ( X m p c * ) }
  • // End Algorithm A 2 [ π ( n ) ]

3.2.4. Multi-Start Metaheuristics Assisted by Assignment Sequences

The proposed metaheuristic integrates the procedures of the three phases described above (see Schema M S [ X 0 , π ( n ) ] ) . This metaheuristic can be classified within the category of multi-start algorithms, since it uses a multitude of initial solutions generated by the algorithm A 1 [ π ( n ) ] of Phase 2. In the generation of solutions, A 1 [ π ( n ) ] uses a current sequence string ( π c ( n ) ) ; this chain is progressively constructed by moderately altering the current sequence giving rise to the next ongoing permutation. The main characteristics of M S [ · ] are the following:
  • The Algorithm 3 metaheuristic M S [ X 0 , π ( n ) ] ) is initialized by executing Phase 3.1, starting with an initial solution X 0 offered by MILP-1 (Phase 1) and registering the sequence π ( n ) as an assignment current sequence; that is: π c ( n ) π ( n ) .
  • In a specific iteration, given an arbitrary current sequence, which we will denote as π c ( n ) , the algorithms A 1 [ π ( n ) ] and A 2 [ π ( n ) ] are cascaded, making π c ( n )   π ( n ) , and thus obtaining a solution X m p c * ( π c ( n ) ) that is locally optimal.
  • In an iteration, to go from an ongoing assignment current sequence ( π c ( n ) ) to the next current sequence ( π   c ( n ) ) , there are many alternatives. Specifically, in this work, we have chosen to apply a slight mutation on the first sequence π c ( n ) . This mutation, characterized by the P m rule, consists of exchanging, in pairs, the fuel elements located in four positions of the current sequence chosen at random.
  • Thus, for each X 0 , the number of initial solutions locally optimized by M S [ X 0 , π ( n ) ] is equal to the number of sequences generated with the P m mutation rule. The M S [ · ] procedure ends when the number of mutated sequences reaches a maximum value equal to i t π (v.gr. i t π = 1000 mutate sequences).
Algorithm 3.Metaheuristic M S [ X 0 , π ( n ) ] : Three-Phase Hybrid Metaheuristics Based on X 0 and π ( n )
  • // Initialization from a solution X 0 found by MILP-1 // (Phase 1)
  • Input   X 0 , π ( n ) , Ρ L S ,   Ρ m , i t π
  • π c ( n )   π ( n ) ; i t e r = 1
  • X m p c c G e n e r a t e _ s o l u t i o n ( X 0 , A 1 [ π ( n ) ] )           // Initial solution from A 1 [ π ( n ) ] .
  • X m p c * I m p r o v e _ s o l u t i o n ( X m p c c , A 2 [ π ( n ) ] , Ρ L S ) // Improved solution from A 2 [ π ( n ) ] .
  • Repeat
  • π ( n ) M u t a t e _ s e q u e n c e ( π c ( n ) , Ρ m )
  • π c ( n ) π ( n )
  • X m p c c G e n e r a t e _ s o l u t i o n ( X 0 , A 1 [ π c ( n ) ] )        // Phase 2: Initial solution (iter).
  • X I m p r o v e _ s o l u t i o n ( X m p c c , A 2 [ π c ( n ) ] , Ρ L S )   // Phase 3: Improved solution (iter).
  • If [ σ ( X ) < σ ( X m p c * ) ] = t r u e then X m p c * X
  • i t e r = i t e r + 1
  • Until ( i t e r i t π )
  • Return { X m p c * ,   σ ( X m p c * ) }
  • // End Metaheuristic M S [ X 0 , π ( n ) ]

4. Case Study

This section is dedicated to a case study inspired in the spent fuel pools of Ascó NPP (Tarragona, Spain). It contains the following points: (i) the model assumptions, (ii) the requirements for spent fuel cask loading, (iii) the procedures used, (iv) the data used, and (v) the results achieved.

4.1. Hypothesis for the Case

The following hypotheses are assumed for the case at hand:
Hypothesis 1 (H1). 
Fuel assemblies are characterized by their type, burn-up ( M W d / t ), initial enrichment (% by mass), and the time elapsed since they are unloaded from the core (in years). From these data, the decay heat of each element can be determined at the time it is transferred from the pool to the cask.
Hypothesis 2 (H2). 
There are no fuel assemblies damaged or affected by other additional constraints.
Hypothesis 3 (H3). 
The elements associated with the fuel, such as neutron sources, control rods, consumable poisons, and others, are not considered in the model.
Hypothesis 4 (H4). 
It is assumed that only one type of cask is available in terms of physical dimensions and number of positions for fuel assemblies, the MPC-32.
Hypothesis 5 (H5). 
The fuel assemblies are loaded into the casks taking into account the requirements for storage in module-type HI-STORM storage containers. It is assumed that there is no capacity limit in the ISFSI.
Hypothesis 6 (H6). 
The contents of the spent fuel pool at the end of the operation of the plant (T = 0) are assumed to be known. To carry out the case study, the real content of the spent fuel pool of NPP Ascó 1 is used here. The decay heat of each of the SFA at any given time T can be computed from their cooling time (this is T plus the time each SFA has previously spent in the pool) and the SFA’s burn-up and initial enrichment.
Hypothesis 7 (H7). 
The pool should be emptied using the minimum possible number of casks on the earliest possible date. The date has been chosen as T = 6 years in order to guarantee that even the hottest SFA meets the requirements.
Hypothesis 8 (H8). 
It may be interesting that the thermal load of the casks is balanced. Therefore, we aim at minimizing the standard deviation of the thermal load of the set of casks.

4.2. Requirements for the Storage of Fuel Assemblies

The requirements to take into account for emptying the pool are as follows:
R1.
Casks are of the MPC-32 type, which are compartmentalized in 32 positions, each one containing a single fuel assembly (see Figure 2).
R2.
The positions of a MPC-32 are grouped into two regions: ( R a ) a permissive region with a high limit for the decay heat formed by h a = 12 positions and ( R b ) a restrictive region with a low limit for the decay heat formed by h b = 20 positions.
R3.
A fuel encapsulation position (concept for which we will use the Latin noun locus–loci) can house an SFA whose cooling time is greater than 5 years (in our case, with T = 6 years, this does not constitute any limitation).
R4.
The contents of each locus in any cask must have a decay heat less than or equal to the maximum allowed in the locus: q a + in region R a and q b + in region R b .
R5.
The maximum heat load allowed in a cask is equal to Q M P C + = 30   kW .
R6.
The maximum values for decay heat at loci are determined according to the case:
-
Case of uniform loading: the maximum decay heat allowed per locus in the MPC-32 equals q a + = q b + = 0.9375   kW (30 kW / 32 ). Therefore, if all SFAs have a decay heat q i ( T ) 0.9375   kW   i I , then only one class of container is needed, and the problem has a trivial solution (in what concerns the minimization of the number of casks)
-
Case of regionalized loading: if any SFA has a decay heat q i ( T ) > 0.9375   kW , then some casks allowing regionalized load are needed. In this case, the maximum values of decay heat per locus ( q a + or q b + ) according to the region to which it belongs ( R a o R b ) are determined as follows:
(1)
Set a value of the ratio of the maximum values of decay heat per locus ρ = q a + / q b + , such that   1 ρ 3 ( ρ = 1 corresponts to the uniform loading, ρ = 3 is the maximum value allowed by the regulator for this ratio).
(2)
Determine the maximum allowed thermal load per locus,   q b + = q b + ( ρ , h a , h b ) for the restrictive region R b , from the expression:
( h a × ρ + h b ) × q b + ( ρ , h a , h b ) = f ( ρ , h a , h b ) × Q M P C +
where f ( ρ , h a , h b ) is a function bounded to 1 for ρ = 1 and decreasing with ρ . The function f ( ρ , h a , h b ) for the MPC-32 is:
f ( ρ , h a , h b ) = 2 ( 1 + ρ ( 0.23 / ρ 0.1 ) ) .
(3)
Determine the maximum allowed thermal load per locus, q a + ( ρ , h a , h b ) , for the permissive region R a :
q a + ( ρ , h a , h b ) = q b + ( ρ , h a , h b ) × ρ

4.3. Data and Other Characteristics of the Case

Obviously, the dimensions of the MILP-1 model are a function of the instances used. In this work, we use the ASCÓ#01_6.0 instance (see Table A1 in Appendix A) as data for decay heat. The main characteristics of the experiment are as follows:
(1)
The number of spent fuel assemblies is equal to n = 1164 . The elements are identified with the indices i I corresponding to the n first natural numbers.
(2)
The expected date for the removal of elements from the pool is T = 6 years after the end of operation. The decay heat values q i ( T )   i I listed in Table A1 of Appendix A correspond to T = 6 years.
(3)
The number of virtual regions of decay heat is equal to m = 6 . Let r j = { q : q [ q j ,   q j + ] }   ( j = 0 , 1 , , 5 ) be those regions that are defined with closed intervals of decay heat per locus. The regions are specified from three values of the ratio of maximum thermal loads: ρ = 1 , 2 , 3 . Let:
r 0 = { q : q [ 0.00 ,   30   kW ] }
r 1 = { q : q [ 0.00 ,   1.4250   kW ] }
r 3 = { q : q [ 0.00 ,   0.9375   kW ] }
r 4 = { q : q [ 0.00 ,   0.6310   kW ] }
r 5 = { q : q [ 0.00 ,   0.4750   kW ] } .
For convenience, virtual regions have been numbered in descending order of the upper limit q j + j of each interval.
(1)
The definition of virtual regions ( r 0 to r 5 ) allows establishing the element–region compatibility matrix A ( T ) , whose technological coefficients adopt the value 1 if the fuel assembly i I can be assigned to a virtual region of type j J at the time T , and 0 otherwise. Formally:
a i , j a i , j ( T = 6 ) = { 0 , i f   q i ( T = 6 ) > q j + 1 , otherwise }   i = 1 , , 1164 ;   j = 0 , 1 , , 5
(2)
The number of cask classes is l = 3 if we omit the neutral virtual region r0, which corresponds to the pool. The correspondence between the virtual regions ( r 1 to r 5 ) and the real classes of the regionalized cask ( C R 1 a C R 3 ) is established through the region-cask compatibility matrix B , whose coefficients b j , k are equal to 1 when the region r j ( j = 1 , , 5 ) is associated with cask C R k ( k = 1 , , 3 ) , and they are 0 otherwise. In our case, matrix B is as follows:
B m × l = B 5 × 3 = (   1   0   0     0   1   0     0   0   1     0   1   0     1   0   0   )
Thus, a cask C R 1 is compartmentalized in a permissive region with the characteristics of r 1 and a restrictive region associated with r 5 ; a cask C R 2 is associated with a permissive region r 2 and another restrictive r 4 ; and finally, a cask C R 3 has only one region with the qualities of the intermediate region r3.
(1)
From a technological point of view, all the containers in this study correspond to MPC-32, with 12 positions in the permissive region and 20 positions in the restrictive regions. Obviously, an MPC-32 of class C R 3 has 32 positions in the unique region.
(2)
Taking into account that the number of cask classes is l = 3 , it remains to define the unit cost per each class. In order to measure the impact of the adversity coefficients γ k   ( k ) on the solutions offered by MILP-1, different values are given to the arbitrary costs γ = ( γ 1 , γ 2 , γ 3 ).
(3)
Taking into account the values and characteristics exposed in the previous points, the dimensions of the MILP-1 model for the Ascó # 01_6.0 instance are the following: (i) 6984 binary decision variables ( x i , j ) corresponding to the assignment of fuel assemblies to virtual regions; (ii) 6 integer variables ( y j ) associated with the number of fuel assemblies hosted by each virtual region; (iii) 4 representative integer variables ( z k ) for the number of casks of each class; and (iv) 1195 constraints (1199 if the availability of casks is bounded).

4.4. Procedures

All compiled codes corresponding to the procedures used have been executed on a DELL Inspiron-13 (Intel(R) Core(TM) i7-7500U @ 2.70 GHz CPU 2.90 GHz, 16 GB de RAM, x64 Windows 10 Pro). The models and tools used to find the optimal solutions for this case study are the following:
(1)
Model MILP-1 [28] based on MILP that allows minimizing the cost ( Γ * ( T ) ) of the containers required ( z k * k ) to transfer all the SNF from the NPP pool. MILP-1 also offers the number of SFAs assigned to each region ( y j * j ) , as well as an optimal assignment ( x i , j *   i j ) of each SFA to a virtual region. For the exploitation of MILP-1, we use the IBM ILOG CPLEX solver (Optimization Studio v.12.2, win-x86-64).
(2)
Deterministic method A 1 [ π ( n ) ] that starts from the result of MILP-1 and offers an assignment of SFAs to the permissive and restrictive regions of the casks based on the sequence of assignment of SFAs to casks π ( n ) . Here, 1000 consecutive allocation sequences are considered, which are generated by slightly mutating one of them to obtain the next, following the neighborhood transition rule P L S . MILP-1 plus A 1 [ π ( n ) ] (with lexicographic order of the fuel assembly codes) constitute the deterministic procedure D P [ π ( n ) ] proposed in [28].
(3)
Local search heuristic A 2 [ π ( n ) ] that starts from a solution generated by A 1 [ π ( n ) ] and explores neighborhoods non-exhaustively until reaching a local optimum. The procedures MILP-1 plus A 1 [ π ( n ) ] plus A 2 [ π ( n ) ] (with the mutation rule P m applied to the π ( n ) sequence in i t π iterations) constitute the metaheuristic M S [ X 0 , π ( n ) ] proposed in this work.
In short, metaheuristic M S [ X 0 , π ( n ) ] works in this manner: from an optimal solution found by MILP-1, which is characterized by the triple of sets of values [ ( x i , j *   i j ) ,   ( y j * j ) ,   ( z k * k ) ] , the algorithm A 1 [ π ( n ) ] is applied i t π times, transferring one by one and following i t π sequences π ( n ) , the SFAs assigned to the virtual regions r j ( j ) toward the physical regions (permissive and restrictive) of the casks C R k ( k ) , using the region-cask compatibility matrix B . Afterwards, the local optimization corresponding to heuristics A 2 [ π ( n ) ] is applied to all ( i t π ) solutions generated by algorithm A 1 [ π ( n ) ] .

4.5. Results

In addition to considering the solution from MILP-1 with the triplet γ 0 = ( 1 , 1 , 1 ) , which corresponds to minimizing the total number of casks, and in order to assess the impact of the adversity coefficients γ k   ( k ) on the solutions offered by MILP-1, 24 other optimal solutions are also analyzed, fixing the list of arbitrary costs γ = ( γ 1 , γ 2 , γ 3 ) to the following values:
γ 1 = ( 3 , 2 , 1 ) ,   γ 2 = ( 3 , 1 , 2 ) ,   γ 3 = ( 2 , 3 , 1 ) , γ 4 = ( 2 , 1 , 3 ) ,   γ 5 = ( 1 , 3 , 2 ) , γ 6 = ( 1 , 2 , 3 )
γ 7 = ( 3 , 3 , 2 ) , γ 8 = ( 3 , 3 , 1 ) , γ 9 = ( 3 , 2 , 3 ) , γ 10 = ( 3 , 2 , 2 ) , γ 11 = ( 3 , 1 , 3 ) , γ 12 = ( 3 , 1 , 1 )
γ 13 = ( 2 , 3 , 3 ) , γ 14 = ( 2 , 3 , 2 ) , γ 15 = ( 2 , 2 , 3 ) , γ 16 = ( 2 , 2 , 1 ) , γ 17 = ( 2 , 1 , 2 ) , γ 18 = ( 2 , 1 , 1 )
γ 19 = ( 1 , 3 , 3 ) , γ 20 = ( 1 , 3 , 1 ) , γ 21 = ( 1 , 2 , 2 ) , γ 22 = ( 1 , 2 , 1 ) , γ 23 = ( 1 , 1 , 3 ) , γ 24 = ( 1 , 1 , 2 ) .
Using the deterministic procedure D P [ π ( n ) ] proposed in [28], after obtaining an optimal solution with MILP-1 [ ( x i , j *   i j ) ,   ( y j * j ) ,   ( z k * k ) ] , the algorithm A 1 [ π ( n ) ] is applied following the sequence π ( n ) corresponding to the lexicographic order of the fuel assembly identifiers. The global results of the 25 solutions are shown in Table 1.
Specifically, in Table 1, the following data are collected for each solution (#0 to #24):
(1)
The cost γ = ( γ 1 , γ 2 , γ 3 ) imputed to each cask class.
(2)
The optimal number of casks of each class ( z k * k ) and the total number ( z t o t * ) of casks required by each solution.
(3)
The optimal values of the objective cost function ( Γ * ( T ) for the MILP-1 model.
(4)
The average thermal load corresponding to the configuration of each solution ( Q a v g ) .
(5)
The standard deviation of the thermal load in the set of casks: σ ( Q ) .
Table 1 shows 16 optimal solutions that require 37 MPC-32, which is a number that also coincides with the minimum number of casks, according to the lower bound ( 1164 / 32 ). However, these solutions show different values of the standard deviation of the thermal load, depending on the configuration of the 37 casks.
As an example, initial solution #9, one with the lowest thermal load dispersions between the casks: σ ( Q ) = 2.90 , corresponds to the triplet of costs γ 9 = ( 3 , 2 , 3 ) , in which it is considered that the adversity to the use of casks of classes C R 1 and C R 3 is 1.5 times higher that of class C R 2 casks.
Based on solution #9, the FAs column in Figure 3 shows the number of SFAs assigned to each virtual region, i.e., y 1 * = 24 ,   y 2 * = 274 ,   y 3 * = 369 , y 4 * = 457 , y 5 * = 40 . This is translated into a total requirement of 37 casks (see the MPC[zk, hj] column in Figure 3), z 1 * = 2 ,   z 2 * = 23 and z 3 * = 12 regionalized casks of classes C R 1 , C R 2 , and C R 3 , respectively.
On the other hand, Figure 4 shows the detailed loads of the containers MPC-32 number n.1 (class C R 1 ), number n.3 (class C R 2 ) and number n.26 (class C R 3 ) of the initial solution #9. These loads are characterized by the identifiers and decay heat of the fuel assemblies located in each MPC-32.
In view of Figure 4, the following can be stated:
  • Cask n.1 of initial solution #9 is class C R 1 with a permissive region R a r 1 of 12 positions and a restrictive region R b r 5 of 20. The 12 hottest assemblies in this cask are located in region R a , so locus-1 is occupied by the fuel assembly 751 with a decay heat of 1.3740 kW, while the fuel assembly 753 (the last in R a ) is located in locus-12 with 1.3234 kW. Complementarily, 20 “cold” elements are located in positions 13 to 32 of the restrictive region R b of MPC-32 n.1, assembly 23 in locus-13, and assembly 21 in locus-32, whose decay heats are 0.4189 kW, coinciding with the decay heat of all the elements located in this region.
  • Casks n.3 (initial solution #9) is class C R 2 , with its permissive region R a r 2 and its restrictive region R b r 4 . In region R a , the 12 hottest elements of this cask are located with element 255 occupying locus-1 and element 281 located in locus-12 with residual heats equal to 0.7813 kW and 0.6320 kW, respectively. In addition, all the SFAs located in the restrictive region R b are considered to have a decay heat of 0.4189 kW.
  • Finally, cask n.26 of initial solution #9 is class 3 with a uniform region consisting of 32 positions: ( R a , R b ) r 3 . The loci-{1,12,13,32} are occupied by the elements 244, 126, 144, and 122, their respective decay heats being as follows: 0.7673, 0.6917, 0.6898, and 0.6332 kW.
Other interesting results shown in Figure 4 are as follows: (i) the total thermal load casks n.1, n.3, and n.26, which are 24.55, 16.39, and 21.59 kW, respectively; (ii) the thermal loads in the permissive and restrictive ( R a , R b ) regions of these casks (v.gr. in MPC-32 n.26: 8.57 and 13.01 kW); (iii) the average thermal load, equal to 22.77 kW, corresponding to the load of the 1164 SFAs in the 37 casks that constitute initial solution #9; and, finally, (iv) the standard deviation of the thermal load associated with initial solution # 9, σ ( Q ) = 2.90   kW .
On the other end, to illustrate the performance of the M S [ X 0 , π ( n ) ] metaheuristic in terms of improving the standard deviation of the thermal load in the set of casks, nine solutions have been selected from the 25 solutions presented in Table 1. The solution selection criterion meets the following conditions:
(1)
Have a total number of z t o t * containers that is minimal; that is: z t o t * = 37 .
(2)
If two or more solutions with z t o t * = 37 , have the same configuration of containers by type ( z 1 * , z 2 * , z 3 * ) , the one that presents smaller standard deviation σ ( Q ) is chosen; for example, between solutions #2 and #9, we select solution #9.
The final solutions selected to apply Phase 3 to improve σ ( Q ) using metaheuristic M S [ X 0 , π ( n ) ] are shown in Table 2. The indicators selected to compare these solutions are as follows: (i) the standard deviations σ A 1 ( Q ) and σ A 2 ( Q ) corresponding to the initial solutions provided by A 1 [ π ( n ) ] and the improved solutions provided by A 2 [ π ( n ) ] , (ii) the iteration i t * corresponding to the best solution with M S [ X 0 , π ( n ) ] ; (iii) the computing time in seconds, C P U i t * , corresponding to the iteration i t * ; and (iv) the computing time in seconds, C P U 1000 , given to M S [ X 0 , π ( n ) ] to improve 1000 initial solutions provided by A 1 [ π ( n ) ] from a solution X 0 provided by MILP-1.
In view of Table 2, we can draw the following conclusions:
(1)
Metaheuristic M S [ X 0 , π ( n ) ] uses on average 632 s of CPU time to reach the best local optimum of 1000 initial solutions from algorithm A 1 [ π ( n ) ] . This average is included in the range of values [511, 729] in seconds.
(2)
Metaheuristic M S [ X 0 , π ( n ) ] takes on average 254.4 s of CPU time to obtain the best solution (locally optimized) from 1000 initial solutions generated by A 1 [ π ( n ) ] . This average is included in the range of values [7.9, 652.3] in seconds.
(3)
On average, the iteration i t * (best solution using M S [ X 0 , π ( n ) ] ) is i t * 361 (over 1000 iterations). This average is included in the range of values [10, 998] in iterations.
(4)
The procedure D P [ π ( n ) ] proposed in [28] obtains, on average, solutions with a standard deviation of the thermal load equal to σ A 1 ( Q ) = 3.21 . The range of values for σ A 1 ( Q ) is [2.90, 3.52].
(5)
On average, the metaheuristic M S [ X 0 , π ( n ) ] proposed in this work obtains solutions with a standard deviation of the thermal load of the set of casks equal to σ A 2 ( Q ) = 0.1658 , from the solutions generated by A 1 [ π ( n ) ] and the rule Ρ L S . The range of values for σ A 2 ( Q ) is [0.0001, 0.5464].
(6)
On average, the standard deviation σ ( Q ) is reduced approximately 20 times by applying the metaheuristic M S [ X 0 , π ( n ) ] versus procedure D P [ π ( n ) ] proposed in [28].
To illustrate the reduction in the dispersion of the thermal load in the set of casks, Figure 5 shows the detailed loads of the containers MPC-32 number n.1 (type C R 1 ), number n.3 (type C R 2 ), and number n.22 (type C R 3 ) from solution #10 provided by metaheuristic M S [ X 0 , π ( n ) ] .
Figure 6 and Figure 7 show graphically the thermal loads corresponding to the solutions provided by procedure D P [ π ( n ) ] proposed in [28] (see Figure 6) versus metaheuristic M S [ X 0 , π ( n ) ] proposed in this work (see Figure 7) for each region (permissive and restrictive) and for each of the 37 casks. Note the significant differences that exist between the thermal loads of the 37 used MPC-32 in both figures, despite the fact that the containers present the same thermal limitation characteristics number by number (1 to 37).
In addition, Appendix B contains the detailed configurations of the 37 containers corresponding to solution #10 provided by metaheuristic M S [ X 0 , π ( n ) ] .

5. Conclusions

A hybrid multi-start procedure assisted by mixed integer linear programming and local search optimization algorithms, that we have called here metaheuristic M S [ X 0 , π ( n ) ] , has proven to be competitive in solving the spent fuel cask loading problem with minimal storage cost and a minimum standard deviation of thermal load. This is confirmed through the experiments carried out with the database of Ascó NPP. It allows generating instances with industrial dimensions (e.g., 1200 fuel assemblies and from 1 to 30 classes of regionalized casks, which translates into a number of decay heat virtual regions between 2 and 60).
The method proposed here, M S [ X 0 , π ( n ) ] , to solve the single-stage cask loading of spent nuclear fuel is computationally competitive in its three phases: MILP-1 mathematical program, deterministic algorithm A 1 [ π ( n ) ] , and local search heuristic algorithm A 2 [ π ( n ) ] .
(i) The MILP-1 model is used to minimize the cost of casks required to host the nuclear fuel; it offers, as a solution, an optimal assignment of fuel assemblies to virtual thermal regions, in less than 0.5 CPU seconds, using instances with 1200 fuel assemblies, 6 types of regions, and 4 classes of casks. (ii) Algorithm A 1 [ π ( n ) ] , which starts from an MILP-1 solution, can assign the SFAs to the real regions of the casks in less than 0.25 s for the tested instances. (iii) The local search algorithm A 2 [ π ( n ) ] , based on 1000 initial solutions generated by A 1 [ π ( n ) ] by slight mutations of the π ( n ) sequences, is capable of reducing the initial values of the standard deviation σ ( Q ) 20 times compared to the solutions provided by procedure   D P [ π ( n ) ]   proposed   in   [28], in an average CPU time equal to 4.24 min.
In other words, metaheuristic M S [ X 0 , π ( n ) ] (made up MILP-1 plus A 1 [ π ( n ) ] plus A 2 [ π ( n ) ] ) is capable of obtaining, on average, a solution with a Pearson’s coefficient of variation lower than 0.75% in less than 260s CPU (1000 iterations). In the slowest approximation, M S [ X 0 , π ( n ) ] obtains a solution with a Pearson’s coefficient of variation lower than 2.4%, starting from around 15.5% in little more than 12 min CPU (1000 iterations).
Future working lines taking advantage of these results include the following: (1) the formulation and exploitation of other models based on MILP for the regionalized cask loading according to several optimization criteria, and (2) the extension of the above models and procedures to the multi-stage load case.

Author Contributions

Conceptualization, J.B.-V. and L.B.; methodology, J.B.-V.; software, J.B.-V.; validation, J.B.-V., L.B. and M.M.; formal analysis, J.B.-V., L.B. and M.M.; investigation, J.B.-V., L.B. and M.M.; resources, J.B.-V., L.B. and M.M.; data curation, J.B.-V. and L.B.; writing—original draft preparation, J.B.-V.; writing—review and editing, J.B.-V., L.B. and M.M.; visualization, J.B.-V., L.B. and M.M.; supervision, J.B.-V., L.B. and M.M.; project administration, J.B.-V. and L.B.; funding acquisition, J.B.-V. and L.B. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Acknowledgments

This work has been subsidized by the Ministry of Science, Innovation and Universities of the Government of Spain with the OPTHEUS project (ref. PGC2018-095080-B-I00), including Funds for European regional development, and the UPC project “Tool development for the simulation and optimization of the emptying of spent fuel pools in nuclear power plants” financed by ENDESA Generation. The authors also thank the Asociación Nuclear Ascó–Vandellós II (ANAV) the support provided and, especially, to Jordi Estrampes, head of Reactor Engineering and Nuclear Safeguards at NPP Ascó, for his tips and clarifications.

Conflicts of Interest

The authors declare no conflict of interest.

Nomenclature

CISF Central Interim Storage Facility
IAEAInternational Atomic Energy Agency
INSAGInternational Nuclear Safety Group
ISFSIIndependent Spent Fuel Storage Installation
LWRLight Water Reactor
MILPMixed Integer Linear Programming
NPPNuclear Power Plant
SFASpent Fuel Assembly

Appendix A. Instance Ascó#01_06

Table A1. Set of SFAs corresponding to the Ascó instance #01_6.0. The characteristics are as follows: identifier i   ( i = x + y ) of the element, corresponding to row x and column y. The values in the table correspond to the decay heat q i ( T ) (in kW) for T = 6 years.
Table A1. Set of SFAs corresponding to the Ascó instance #01_6.0. The characteristics are as follows: identifier i   ( i = x + y ) of the element, corresponding to row x and column y. The values in the table correspond to the decay heat q i ( T ) (in kW) for T = 6 years.
x\y12345678910
00.41890.41890.41890.41890.41890.41890.41890.41890.41890.4189
100.41890.41890.41890.41890.41890.41890.41890.41890.41890.4189
200.41890.41890.41890.41890.41890.41890.41890.41890.41890.4189
300.41890.41890.41890.41890.41890.41890.41890.41890.41890.4189
400.41890.41890.41890.41890.41890.41890.41890.41890.41890.4189
500.41890.41890.41890.59790.58410.57890.57890.59620.58180.6080
600.53840.60840.53100.57660.57320.57840.60140.58010.55620.5721
700.57660.54770.57040.54770.58350.53020.54840.53180.57490.6098
800.60980.60980.57890.60100.57720.52360.53020.53020.57720.6049
900.59620.61010.53020.47720.48200.47560.48360.47400.48360.4756
1000.47560.53350.63500.59790.62960.59620.63850.59970.59620.5979
1100.53350.59970.59790.59620.68600.53510.59620.59400.63470.6010
1200.62960.63320.62960.59270.53350.69170.68980.61650.61470.7068
1300.49420.61830.69170.49250.63630.49420.49090.49580.49740.7105
1400.49250.74420.61120.68980.49740.58750.54510.58530.53430.5401
1500.53350.54510.58580.54440.53180.54180.54340.59100.59050.5841
1600.57890.53020.58580.53350.53510.53270.54340.54180.53020.6955
1700.70520.50550.50390.58130.50880.52700.58440.50640.52530.5031
1800.50720.52700.69170.58310.69730.50230.52530.58480.59900.5467
1900.59730.55330.59730.55000.65710.65710.54170.54170.65350.6094
2000.55330.59730.65530.59390.55000.60250.55330.55500.60070.5384
2100.61720.53510.61540.53600.61540.53940.54770.52360.61010.7112
2200.54840.61190.53600.61540.52360.54940.53510.54770.53680.5410
2300.53510.61370.61010.64420.63030.65120.61650.63550.64070.6233
2400.64600.64420.64420.76730.60280.75930.62850.60790.61990.6407
2500.59770.62330.62160.64070.78130.61990.60620.64770.64950.6079
2600.64250.63900.64770.60450.60450.64070.60450.62850.63900.6320
2700.63200.63550.63200.63550.63370.63720.63550.64070.62510.6372
2800.63200.63370.63720.64420.65120.64600.71990.71800.65430.7238
2900.65070.64900.65250.65250.64720.64900.65070.64540.64540.6578
3000.65430.65250.77530.77930.71170.71170.77330.77130.70970.7203
3100.71840.71650.72410.69230.71840.69230.69040.72600.71650.7203
3200.71840.72030.69040.71840.71460.70710.71280.70710.71090.7090
3300.70710.71090.70340.72220.71650.71460.71280.67570.67940.6775
3400.67940.67570.67750.67750.67570.62180.45350.62610.62610.4566
3500.62440.62790.45200.62890.62440.44890.63070.67830.69150.6839
3600.68020.69720.68390.69910.69150.83150.82740.83360.83150.8212
3700.83360.82330.84820.97840.97100.97590.98580.97350.97840.9759
3800.98830.72390.71820.72580.72390.83570.83570.75990.76380.7677
3900.83150.76580.76380.76380.76380.72840.77170.76380.73430.7323
4000.84410.73040.76770.76770.76580.76580.74830.73860.76380.7463
4100.74830.74830.75210.75020.75600.74630.74630.75020.75020.7463
4200.74060.77760.77360.77950.76770.76800.77200.76200.75600.7640
4300.75400.76400.76400.86660.85580.86230.86230.85580.87100.8688
4400.88200.89520.88640.88860.89080.88420.89080.88860.90630.7704
4500.76440.77240.77040.78820.77000.77200.77810.77440.77810.7680
4600.79430.79230.78820.78420.79030.78420.79030.78820.79030.8579
4700.76800.77200.86010.85140.77000.77610.84920.79030.79230.7943
4800.79030.81280.78010.80860.80660.80460.82560.81280.81070.7882
4900.82130.80460.79030.81490.80870.81920.78821.00270.92030.9955
5000.99550.92491.00270.91800.91110.91800.92950.99310.99070.9835
5100.91570.91570.98351.13061.12231.12781.13890.80300.89970.8907
5200.80300.79680.91340.89740.81340.83030.83450.83670.83450.8303
5300.81130.81340.81970.80930.81550.82180.82180.83030.80930.8155
5400.82390.82610.81970.81340.82180.84390.82180.85270.83510.8197
5500.83950.81970.81550.82180.82180.84830.84600.83730.81970.8218
5600.84601.02010.79011.01761.01260.79891.02260.78140.78360.7923
5700.78791.01261.01511.00760.79010.77491.02011.00761.00261.0026
5801.00260.84830.83290.83290.84170.83950.85050.83290.85490.8615
5900.84830.85270.85710.84600.85050.85710.85710.89290.88160.8839
6000.88390.82850.89290.88610.84600.83510.83510.88390.89070.8771
6100.85490.87270.87270.99160.87710.85490.85490.86150.98900.8749
6200.87940.87040.99940.98900.87491.11961.11681.13061.11411.0868
6301.04391.08141.06521.05191.07331.02811.04121.04391.03331.0733
6401.06791.06521.04121.02541.07331.11961.12781.11681.12780.8965
6500.88720.88950.89420.88950.89180.89650.89180.90360.89650.8918
6601.06791.25431.07061.06791.25750.91061.23810.90831.05991.0760
6700.90831.07061.05721.06520.89651.25430.90590.90360.90120.9059
6800.91301.06030.90360.91770.91300.90590.92010.90120.89891.2510
6901.25430.90121.23801.16141.25021.23361.16141.23671.15261.1496
7001.15551.24281.23671.23061.23061.15851.14851.23670.94330.9408
7100.93830.94080.95841.12331.11750.95340.94581.11461.12040.9508
7200.97620.96090.96600.97621.07160.97880.95591.08591.08021.0830
7300.96350.98131.06310.98391.06311.05750.97881.05470.98130.9762
7400.97620.98131.05471.06601.05470.98900.97881.06030.96401.3368
7501.37401.36381.32340.97751.33351.34361.33011.35031.36720.9721
7601.37061.34361.33680.96671.06311.05471.04911.05751.08301.0688
7701.06601.07451.07161.07451.06031.07161.08591.25431.25101.0773
7801.06881.25751.25101.07451.09731.08871.09731.09441.25431.0916
7901.09161.26081.26081.26081.09441.08870.87491.10930.87240.8698
8001.10930.88261.11241.11851.11541.11850.86730.87490.86731.1124
8101.10930.87491.15561.14941.15561.15870.52071.21241.22840.5186
8200.51231.22521.22200.52071.23491.25101.25751.23171.05191.2349
8301.26411.06091.06391.06391.25751.23491.25431.20921.24780.5375
8401.21241.24781.20601.19971.21561.20291.21241.24461.25101.2060
8501.20921.23810.59230.59470.59470.59230.40690.39130.51590.3975
8600.42920.39750.39750.39750.40380.47850.40220.46510.39750.5133
8700.52340.43720.48360.43250.54770.46350.52230.45240.54260.5375
8800.46510.46030.52230.45710.46510.51560.54260.52730.52400.4492
8900.41960.52060.45550.52230.46670.46030.45710.41350.54770.5140
9000.54090.41960.51730.54600.52060.46190.51400.51900.45390.4587
9100.46030.40890.52400.51400.45940.48200.48200.47870.45620.5455
9200.49020.49340.55580.55070.48850.51670.45460.47710.47550.5576
9300.51670.41170.46560.50150.43120.63660.61140.61200.46510.5190
9400.47310.52570.46350.44760.54770.46830.52100.50120.43540.4700
9500.50120.46560.47440.45540.42980.45540.46560.47290.43680.4714
9600.51790.43120.51480.51790.44390.43970.42130.52560.43680.4326
9700.46560.51480.46410.51330.50270.50120.43540.45830.43540.4598
9800.52400.43680.43260.49970.52560.52870.50120.49970.53180.5302
9900.50420.50190.59250.50330.49450.59570.49310.50190.49010.4887
10000.49600.50330.49010.50480.49890.60050.49600.60210.59890.4887
10100.59890.49750.59730.60530.54700.50780.51220.50630.53480.5424
10200.53940.46130.46280.45850.53940.46280.45990.50780.51520.5078
10300.51220.45990.45990.54090.45850.51520.53630.50480.54090.5237
10400.57130.57130.47060.46910.45760.56810.45470.45900.57130.4706
10500.47500.47350.47200.45320.47940.47500.53820.53050.46620.4504
10600.45180.44900.53820.45760.46330.46190.46190.44900.55390.4561
10700.46480.56020.53360.53510.44900.55540.56180.53360.44900.4677
10800.45180.53360.45610.53210.46040.52900.52900.47640.47500.5649
10900.55390.53980.55540.46040.47060.47200.45470.56020.46770.4461
11000.46770.45180.53360.46770.65120.64600.64070.63720.64070.6390
11100.64770.63900.63370.65300.65650.66010.65480.64950.65480.6601
11200.69050.69240.69240.69240.66900.66540.66720.66900.67080.6672
11300.66900.67080.70970.64070.71460.71840.72220.71460.72030.6794
11400.67390.67200.71460.67890.72030.71650.71650.67390.67200.7053
11500.72600.66840.72030.71840.67340.71460.70900.72600.75930.7494
11600.74750.72790.75340.7713

Appendix B. Cask Configurations of the Improved Solution #10 Provided by M S [ X 0 , π ( n ) ]

Table A2. Solution #10 M S [ X 0 , π ( n ) ] ): List of the fuel assemblies loaded (1 … 1164) in each locus (1 … 32) of the casks (n.1 … n.37) of classes C1 to C3.
Table A2. Solution #10 M S [ X 0 , π ( n ) ] ): List of the fuel assemblies loaded (1 … 1164) in each locus (1 … 32) of the casks (n.1 … n.37) of classes C1 to C3.
n.12345678910111213141516171819
Locus\ClassC1C1C2C2C2C2C2C2C2C2C2C2C2C2C2C2C2C2C2
1761751848827794835792665793782691662789849779690837702839
2752759852667783703842826693698708818838845823828696836830
3762758841819704851822705844814694515701700816815846697707
4750756628514706517813699648714649803516715629718804646719
5755763805801626811806627798785632777786796630795790791788
6753757730729728645784769772780670774640635663672776770661
7829831682669781771641664674633743833733832834744748775735
8574573738768736634673745638767631643567564577581637562636
9498379503565644579578572623501508737500624742377619509746
10374749724732381726513734754510741727375723740721376764747
11575722760566486480469522461462570711712688709720731717716
12397425255568569423261274115271476576457464463468466105458
1310561051346355349256252348237253938248222200211937232215132
14953982671291282332192131088162113265902458020692257
15105295811026082841014264101367541778551891008191109911011
169601053117193120202100611257251158566468124160204159188
179501095993581141481189966915385104610691049708310775973
189461044651638531042184146923161104110157719410989012262051093
1911041080945109055109179662072088991256116816710251039904210
201059895102192489879729301522011471021551691641037165214880
21971957182107422819715610761501572309137617688931084151985
2210718688831628872292161921063101911038208779428941878710871031
23943881102982410342279862211057149989914817947926981104088999
249118859741028840108287174223107394099490896497221890010001055
251032106586617811678903217166111103013993188687082114595928
2610351094108895110781036101687510866310188761819071731809951101897
2710479159809529631099104392096817510041022172103810029759221027909
289659441083105497695610238889051001971068997100710129989181024979
2996738729831064966109710059921388739551371411089917347935891
30114149129022697010105292991632939925109691011023851
3131372440244046510601010459731645815860
32862000000000086500000863869
n.202122232425262728293031323334353637
Locus\ClassC2C2C3C3C3C3C3C3C3C3C3C3C3C3C3C3C3C3
1676778677666505511507502684499687506504671685523681512
2825695650660656689598679680668686449658678683692524659
3843847448443652654445653675442519609603651520657446447
4647850615604625444441655599802608601600621610620613612
5787810800590439799812797808809622807440437436618470593
6725773596438617434473616435589611597591595477561605546
7765642556373557587474592548582594386606558549387584391
8639766602527528585551401586529538368583607369542555560
9614580545537366547487588371530367372536554543533553559
10380739491550488540525526541370552535544532489485534531
11713378492484467481465478563496494482493539495521518479
122737103074834594524273034221164456304571472490424454497
13249279426406390455460450308404475453244388403413471392
14602244324153934304334054093983944514293991160414417428
15100920940742024611634124184193953893964023182904114161161
16854104115140840041042114238411594311162313115811373853821145
177585611561154315310113933432211532873203063193351136383321
187117433332711461138311140324114731228832832911353323051150
191072190112422032533636411333161703371143112211233263621157361
20105810201105331133330344185360365359309114012613018311211141
2186198111834332331411311155112734111443173393401711271441116
229612312411152112536311203001119203338358342345114211481149293
23948990262112619511151993022913012851128113211302362961111258
2410261792832782891114292286110711062952971129196243298299250
25896859277276294259234254269266113426328411172722421031108
2610339913522811113239110993611122352822801351110354121270275
27107094114310921017999238934122987186892119107154247357123
28959882106988131105035310678931369331085106694134225212268
29139699219849610481079919356906106188492735010031001081351
30364010621495423398789629771100890982107597821949240
318670301251948874293447422250861337101
3285808579128641720689835289322718539430
Table A3. Solution #10 ( M S [ X 0 , π ( n ) ] ): Thermal load in each of the casks (n.1…n.37) of classes 1 to 3, as obtained from metaheuristic M S [ X 0 , π ( n ) ] : (MILP-1, A 1 [ π ( n ) ] , A 2 [ π ( n ) ] ), with 1000 initial solutions from A 1 [ π ( n ) ] for each cask. Qavg = 22.7683 kW and SD-Q = 1.04e-4 kW.
Table A3. Solution #10 ( M S [ X 0 , π ( n ) ] ): Thermal load in each of the casks (n.1…n.37) of classes 1 to 3, as obtained from metaheuristic M S [ X 0 , π ( n ) ] : (MILP-1, A 1 [ π ( n ) ] , A 2 [ π ( n ) ] ), with 1000 initial solutions from A 1 [ π ( n ) ] for each cask. Qavg = 22.7683 kW and SD-Q = 1.04e-4 kW.
MPC32 n.1 (Class 1) Qa = 13.6741 kWQb = 9.0943 kW Qab = 22.7684 kW
MPC32 n.2 (Class 1) Qa = 14.0522 kWQb = 8.7163 kW Qab = 22.7685 kW
MPC32 n.3 (Class 2)Qa = 12.8714 kWQb = 9.8968 kWQab = 22.7682 kW
MPC32 n.4(Class 2) Qa = 12.7357 kW Qb = 10.0325 kW Qab = 22.7682 kW
MPC32 n.5 (Class 2) Qa = 12.8346 kWQb = 9.9338 kWQab = 22.7684 kW
MPC32 n.6 (Class 2) Qa = 12.6921 kW Qb = 10.0762 kW Qab = 22.7683 kW
MPC32 n.7 (Class 2) Qa = 12.6314 kW Qb = 10.1369 kW Qab = 22.7683 kW
MPC32 n.8 (Class 2) Qa = 12.6429 kW Qb = 10.1255 kWQab = 22.7684 kW
MPC32 n.9 (Class 2) Qa = 12.5654 kWQb = 10.2029 kWQab = 22.7683 kW
MPC32 n.10 (Class 2) Qa = 12.4591 kW Qb = 10.3091 kW Qab = 22.7682 kW
MPC32 n.11 (Class 2) Qa = 12.5739 kWQb = 10.1945 kWQab = 22.7684 kW
MPC32 n.12 (Class 2) Qa = 12.6148 kWQb = 10.1537 kWQab = 22.7685 kW
MPC32 n.13 (Class 2) Qa = 12.6799 kWQb = 10.0885 kWQab = 22.7684 kW
MPC32 n.14 (Class 2) Qa = 12.6146 kWQb = 10.1538 kWQab = 22.7684 kW
MPC32 n.15 (Class 2) Qa = 12.6762 kWQb = 10.0919 kWQab = 22.7681 kW
MPC32 n.16 (Class 2)Qa = 12.6875 kWQb = 10.0808 kWQab = 22.7683 kW
MPC32 n.17 (Class 2) Qa = 12.7866 kWQb = 9.9817 kW Qab = 22.7683 kW
MPC32 n.18 (Class 2) Qa = 12.5323 kWQb = 10.2358 kWQab = 22.7681 kW
MPC32 n.19 (Class 2) Qa = 12.7007 kWQb = 10.0677 kWQab = 22.7684 kW
MPC32 n.20 (Class 2) Qa = 12.6462 kWQb = 10.1222 kWQab = 22.7684 kW
MPC32 n.21 (Class 2) Qa = 13.1250 kWQb = 9.6433 kWQab = 22.7683 kW
MPC32 n.22 (Class 3) Qa = 10.1928 kWQb = 12.5754 kWQab = 22.7682 kW
MPC32 n.23 (Class 3) Qa = 10.2051 kWQb = 12.5633 kWQab = 22.7684 kW
MPC32 n.24 (Class 3)Qa = 10.1933 kWQb = 12.5751 kWQab = 22.7684 kW
MPC32 n.25 (Class 3) Qa = 10.2239 kWQb = 12.5445 kWQab = 22.7684 kW
MPC32 n.26 (Class 3) Qa = 10.2224 kWQb = 12.5459 kWQab = 22.7683 kW
MPC32 n.27 (Class 3) Qa = 10.2675 kWQb = 12.5007 kWQab = 22.7682 kW
MPC32 n.28 (Class 3) Qa = 10.2498 kWQb = 12.5186 kWQab = 22.7684 kW
MPC32 n.29 (Class 3) Qa = 10.2534 kWQb = 12.5149 kWQab = 22.7683 kW
MPC32 n.30 (Class 3) Qa = 10.2452 kWQb = 12.5233 kWQab = 22.7685 kW
MPC32 n.31 (Class 3) Qa = 10.2235 kWQb = 12.5449 kWQab = 22.7684 kW
MPC32 n.32 (Class 3) Qa = 10.1969 kWQb = 12.5714 kWQab = 22.7683 kW
MPC32 n.33 (Class 3) Qa = 10.1802 kWQb = 12.5881 kWQab = 22.7683 kW
MPC32 n.34 (Class 3) Qa = 10.1898 kWQb = 12.5785 kWQab = 22.7683 kW
MPC32 n.35 (Class 3) Qa = 10.1594 kWQb = 12.6090 kWQab = 22.7684 kW
MPC32 n.36 (Class 3) Qa = 10.1419 kWQb = 12.6266 kWQab = 22.7685 kW
MPC32 n.37 (Class 3) Qa = 10.1415 kWQb = 12.6267 kWQab = 22.7682 kW

References

  1. Sempértegui, R.; Bautista, J.; Griño, R.; Pereira, J. Models and Procedures for Electric Energy Distribution Planning. A review. IFAC Proc. Vol. 2002, 35, 395–400. [Google Scholar] [CrossRef] [Green Version]
  2. Paiva, P.C.; Khodr, H.M.; Dominguez-Navarro, J.A.; Yusta, J.M.; Urdaneta, A.J. Integral planning of primary-secondary distribution systems using mixed integer linear programming. IEEE Trans. Power Syst. 2005, 20, 1134–1143. [Google Scholar] [CrossRef]
  3. Georgilakis, P.S.; Hatziargyriou, N.D. A review of power distribution planning in the modern power systems era: Models, methods and future research. Electr. Power Syst. Res. 2015, 121, 89–100. [Google Scholar] [CrossRef]
  4. Klyapovskiy, S.; You, S.; Cai, H.; Bindner, H.W. Incorporate flexibility in distribution grid planning through a framework solution. Int. J. Electr. Power Energy Syst. 2019, 111, 66–78. [Google Scholar] [CrossRef]
  5. Chehouri, A.; Younes, R.; Ilinca, A.; Perron, J. Wind Turbine Design: Multi-Objective Optimization. In Wind Turbines. Design, Control and Applications; Aissaoui, A.G., Tahour, A., Eds.; InTech: Rijeka, Croatia, 2016; pp. 121–148. [Google Scholar] [CrossRef] [Green Version]
  6. International Atomic Energy Agency. Safety Culture. A Report by the International Nuclear Safety Advisory Group; Safety Series No. 75-INSAG-4; IAEA: Vienna, Austria, 1991; Available online: https://www-pub.iaea.org/MTCD/publications/PDF/Pub882_web.pdf (accessed on 31 August 2020).
  7. International Atomic Energy Agency. Safety Culture in Nuclear Installations. Guidance for Use in the Enhancement of Safety Culture; IAEA-TECDOC-1329; IAEA: Vienna, Austria, 2002; Available online: https://www-pub.iaea.org/MTCD/Publications/PDF/te_1329_web.pdf (accessed on 31 August 2020).
  8. International Atomic Energy Agency. Nuclear Power Reactors in the World; Reference Data Series No. 2; IAEA: Vienna, Austria, 2019; Available online: https://www-pub.iaea.org/MTCD/Publications/PDF/RDS-2-39_web.pdf (accessed on 31 August 2020).
  9. Nuclear Fuel and Its Fabrication—World Nuclear Association. Available online: https://www.world-nuclear.org/information-library/nuclear-fuel-cycle/conversion-enrichment-and-fabrication/fuel-fabrication.aspx (accessed on 31 August 2020).
  10. ENUSA Industrias Avanzadas. Memoria Anual 2018; ENUSA Industrias Avanzadas, S.A.: Madrid, Spain, 2019; Available online: http://www.enusa.es/wp-content/uploads/2019/06/enusa-2018-ESP_Programada_3.pdf (accessed on 31 August 2020).
  11. International Atomic Energy Agency. Scientific and Technical Basis for the Geological Disposal of Radioactive Wastes; Technical Report Series No. 413; IAEA: Vienna, Austria, 2003; Available online: https://www-pub.iaea.org/MTCD/Publications/PDF/TRS413_web.pdf (accessed on 31 August 2020).
  12. Beratungsgesellschaft für integrierte Problemlösungen. Requirements for Facilities and Acceptance Criteria for the Disposal of Metallic Mercuri; European Commission: Brussels, Belgium, 2010; Available online: https://ec.europa.eu/environment/chemicals/mercury/pdf/bipro_study20100416.pdf (accessed on 31 August 2020).
  13. Turismo y Comercio, Ministerio de Industria, Gobierno de España. Sexto Plan. General de Residuos Radiactivos; Turismo y Comercio, Ministerio de Industria: Madrid, Spain, 2006; Available online: http://www.enresa.es/documentos/6PGRR_Espa_ol_Libro_versi_n_indexada.pdf (accessed on 31 August 2020).
  14. Ministerio para la Transición Ecológica y el Reto Demográfico, Gobierno de España. Borrador de 7º Plan. General de Residuos Radiactivos; Ministerio para la Transición Ecológica y el Reto Demográfico: Madrid, Spain, 2020; Available online: https://energia.gob.es/es-es/Novedades/Documents/borrador_7_PGRR.pdf (accessed on 31 August 2020).
  15. López, M.C.R.; Hernández, G.O.; Martín, F.Z.; Leiva, M.G. Contenedores para el almacenamiento temporal y el transporte del combustible gastado. Alfa Revista de Seguridad Nuclear y Protección Radiológica CSN 2015, 27, 8–15. Available online: https://www.csn.es/documents/10182/13557/Alfa+27 (accessed on 31 August 2020).
  16. Consejo de Seguridad Nuclear. Informe del Consejo de Seguridad Nuclear al Congreso de los Diputados y al Senado 2018; Colección: Informes del CSN. Referencia: INF-01.18; Consejo de Seguridad Nuclear: Madrid, Spain, 2019; Available online: https://www.csn.es/documents/10182/13529/Informe+anual+2018/a7a2e69e-9b17-09fa-de54-aa046987be65 (accessed on 31 August 2020).
  17. Gobierno de España. Real Decreto 1522/1984, de 4 de julio, por el que se autoriza la constitución de la «Empresa Nacional de Residuos Radiactivos, S.A.» (ENRESA). Boletín Oficial del Estado 1984, 201, 24186–24187. Available online: https://www.boe.es/boe/dias/1984/08/22/pdfs/A24186-24187.pdf (accessed on 31 August 2020).
  18. Gobierno de España. Real Decreto 1836/1999, de 3 de diciembre, por el que se aprueba el Reglamento sobre instalaciones nucleares y radiactivas. Boletín Oficial del Estado 1999, 313. Available online: https://www.boe.es/eli/es/rd/1999/12/03/1836/con (accessed on 31 August 2020).
  19. Batet, L.; León, P.; Serra, E.; Ciruelos, J.; Estrampes, J.; Culebras, F. Desarrollo de una herramienta de ayuda a la gestión y optimización de carga de contenedores de combustible gastado. In Proceedings of the SNE 44. Reunión anual de la Sociedad Nuclear Española, Ávila, Spain, 26–28 September 2018. [Google Scholar]
  20. Gonzalez-Merino, A.; Gonzalez, A. Reliability methods in the design point of free-standing spent fuel racks under seismic conditions. Progr. Nucl. Energy 2019, 115, 208–220. [Google Scholar] [CrossRef]
  21. Chen, Y.S. Thermal analysis for the integrated spent fuel pool of the Chinshan plant in the decommissioning process. Ann. Nucl. Energy 2018, 119, 163–174. [Google Scholar] [CrossRef]
  22. Petersen, G.M.; Skutnik, S.E.; Ostrowski, J.; Joseph, R.A., III. Determining Optimal Used Fuel Allocation Strategies. Nucl. Technol. 2017, 200, 208–224. [Google Scholar] [CrossRef]
  23. International Atomic Energy Agency. Optimization Strategies for Cask Design and Container Loading in Long Term Spent Fuel Storage; IAEA-TECDOC-1523; IAEA: Vienna, Austria, 2006; Available online: https://www-pub.iaea.org/MTCD/Publications/PDF/te_1523_web.pdf (accessed on 31 August 2020).
  24. Žerovnik, G.; Snoj, L.; Ravnik, M. Optimization of Spent Nuclear Fuel Filling in Canisters for Deep Repository. Nucl. Sci. Eng. 2009, 163, 183–190. [Google Scholar] [CrossRef]
  25. Ranta, T.; Cameron, F. Heuristic methods for assigning spent nuclear fuel assemblies to canisters for final disposal. Nucl. Sci. Eng. 2012, 171, 41–51. [Google Scholar] [CrossRef]
  26. Gomes, I.B.; Saldanha, P.L.C.; Alvim, A.C.M. A Methodology for Optimizing the Management of Spent Fuel of Nuclear Power Plants Using Dry Storage Casks. Sci. Technol. Nucl. Install. 2019. [Google Scholar] [CrossRef]
  27. Spencer, K.Y.; Tsvetkov, P.V.; Jarrell, J.J. Optimization of dry cask loadings for used nuclear fuel management strategies. Progr. Nucl. Energy 2018, 108, 11–25. [Google Scholar] [CrossRef]
  28. Bautista, J.; Batet, L.; Mateo, M. Modelado y resolución del problema del encapsulamiento de combustible nuclear gastado. Dirección y Organización 2020, 71, 46–70. [Google Scholar] [CrossRef]
  29. Wolsey, L.A.; Nemhauser, G.L. Integer and Combinatorial Optimization; John Wiley & Sons Inc: Danvers, MA, USA, 1999. [Google Scholar]
  30. Aarts, E.; Lenstra, J.K. Local Search in Combinatorial Optimization; Princeton University Press: Princeton, NJ, USA, 2000. [Google Scholar]
  31. U.S. Department of Energy Office of Integrated Waste Management. Preliminary Evaluation of Removing Used Nuclear Fuel from Shutdown Sites. Spent Fuel and Waste Disposition; SFWD-IWM-2017-000024. PNNL-22676 (Rev. 10); U.S. Department of Energy Office of Integrated Waste Management, 2017. Available online: https://www.energy.gov/sites/prod/files/2018/06/f53/ne-Shutdown_Sites_Report_Sept2017.pdf (accessed on 31 August 2020).
  32. Consejo de Seguridad Nuclear. Contenedor de combustible HI-STORM 100 de CN Ascó. Available online: https://www.csn.es/documents/10182/989190/Ficha%20HI-STORM (accessed on 31 August 2020).
  33. Holtec International. Final Safety Analysis Report for the HI-STORM 100 Cask System; Report No. HI-2002444 (Rev. 7); Holtec International: Jupiter, FL, USA, 2008. Available online: https://www.nrc.gov/docs/ML0824/ML082401629.pdf (accessed on 31 August 2020).
Figure 1. Components of the HI-STORM system for the storage and transportation of spent nuclear fuel. The HI-TRAC transfer cask allows the MPC to be transferred to the HI-STORM and HI-STAR [15].
Figure 1. Components of the HI-STORM system for the storage and transportation of spent nuclear fuel. The HI-TRAC transfer cask allows the MPC to be transferred to the HI-STORM and HI-STAR [15].
Energies 13 04869 g001
Figure 2. Cross-section of a regionalized MPC-32, showing the positions of the permissive region R a and the restrictive region R b (inspired from Holtec).
Figure 2. Cross-section of a regionalized MPC-32, showing the positions of the permissive region R a and the restrictive region R b (inspired from Holtec).
Energies 13 04869 g002
Figure 3. Solution #9 (MILP-1): number of fuel assemblies and casks in each virtual region.
Figure 3. Solution #9 (MILP-1): number of fuel assemblies and casks in each virtual region.
Energies 13 04869 g003
Figure 4. Initial solution #9 ( P r o c e d u r e   D P [ π ( n ) ] proposed in [28]): identifier and decay heat of the SFAs contained in the MPC-32 with identifying numbers n.1, n.3, and n.26.
Figure 4. Initial solution #9 ( P r o c e d u r e   D P [ π ( n ) ] proposed in [28]): identifier and decay heat of the SFAs contained in the MPC-32 with identifying numbers n.1, n.3, and n.26.
Energies 13 04869 g004
Figure 5. Improved solution #10 ( M S [ X 0 , π ( n ) ] , i t π = 1000 ) : identifier and decay heat of the SFAs contained in MPC-32 with identifying numbers n.1, n.3, and n.22.
Figure 5. Improved solution #10 ( M S [ X 0 , π ( n ) ] , i t π = 1000 ) : identifier and decay heat of the SFAs contained in MPC-32 with identifying numbers n.1, n.3, and n.22.
Energies 13 04869 g005
Figure 6. Solution #10 ( procedure   D P [ π ( n ) ]   proposed   in [28]): thermal loads for each of the 37 used MPC-32 as an addition of the ones in the permissive (Qa) and restrictive (Qb) regions.
Figure 6. Solution #10 ( procedure   D P [ π ( n ) ]   proposed   in [28]): thermal loads for each of the 37 used MPC-32 as an addition of the ones in the permissive (Qa) and restrictive (Qb) regions.
Energies 13 04869 g006
Figure 7. Solution #10 (metaheuristic M S [ X 0 , π ( n ) ] ): thermal loads for each of the 37 used MPC-32 in addition to the ones in the permissive (Qa) and restrictive (Qb) regions.
Figure 7. Solution #10 (metaheuristic M S [ X 0 , π ( n ) ] ): thermal loads for each of the 37 used MPC-32 in addition to the ones in the permissive (Qa) and restrictive (Qb) regions.
Energies 13 04869 g007
Table 1. Summary of results of the 25 optimal initial solutions (#0 to #24) provided by procedure D P [ π ( n ) ] proposed in [28] for the Ascó #01_06 instance with the following technical characteristics: (i) 1164 Spent Fuel Assemblies (SFAs), (ii) T = 6 years, (iii) 6 virtual regions (including the neutral region) defined by three values of the ratio ρ (1, 2, and 3), and (iv) 4 cask classes (including the neutral container).
Table 1. Summary of results of the 25 optimal initial solutions (#0 to #24) provided by procedure D P [ π ( n ) ] proposed in [28] for the Ascó #01_06 instance with the following technical characteristics: (i) 1164 Spent Fuel Assemblies (SFAs), (ii) T = 6 years, (iii) 6 virtual regions (including the neutral region) defined by three values of the ratio ρ (1, 2, and 3), and (iv) 4 cask classes (including the neutral container).
Solution # γ 1 γ 2 γ 3 z 1 * z 2 * z 3 * z t o t * Γ *   ( T ) Q a v g   [ kW ] σ ( Q )   [ k W ]
#011141716373722.77 3.35
#132121718375822.773.29
#231222312375322.773.00
#323119024436219.594.42
#42132540565815.044.55
#513219024436719.594.61
#612340160567215.044.10
#733261318379322.773.20
#833161318377522.773.20
#932322312378822.772.90
#1032221916377622.773.52
#113132540566015.044.49
#1231121916374122.773.52
#13233913153710222.773.11
#1423291018378422.772.91
#1522342112378622.773.37
#1622171218375622.773.27
#1721222312375122.772.93
#1821121916373922.773.52
#191338300838310.152.97
#2013119024434319.594.62
#2112291315376522.773.13
#2212119024434319.594.45
#2311333230565615.044.30
#2411242112374922.773.51
Table 2. Summary of results of the solutions selected to apply Phase 2 plus Phase 3 of the proposed procedure M S [ X 0 , π ( n ) ] . The table shows the values: σ A 1 ( Q ) , σ A 2 ( Q ) , i t * , C P U i t * , and C P U 100 .
Table 2. Summary of results of the solutions selected to apply Phase 2 plus Phase 3 of the proposed procedure M S [ X 0 , π ( n ) ] . The table shows the values: σ A 1 ( Q ) , σ A 2 ( Q ) , i t * , C P U i t * , and C P U 100 .
Solution # z 1 * z 2 * z 3 * z t o t * Γ * ( T ) σ A 1 ( Q )   [ kW ] σ A 2 ( Q )   [ k W ] i t * C P U i t * [s] C P U 1000   [ s ]
#04171637373.350.0001998652.3654
#12171837583.290.105912797.5608
#76131837933.200.000810780.7583
#92231237882.900.22407356.2635
#102191637763.520.0001535422.8729
#1391315371023.110.5464647421.7605
#149101837842.910.3364107.9511
#154211237863.370.2772662487.5717
#167121837563.270.00088362.8643
Avg.-----3.210.1658360.2254.4632
Max-----3.520.5464998652.3729
Min-----2.900.0001107.9511

Share and Cite

MDPI and ACS Style

Bautista-Valhondo, J.; Batet, L.; Mateo, M. Minimizing the Standard Deviation of the Thermal Load in the Spent Nuclear Fuel Cask Loading Problem. Energies 2020, 13, 4869. https://doi.org/10.3390/en13184869

AMA Style

Bautista-Valhondo J, Batet L, Mateo M. Minimizing the Standard Deviation of the Thermal Load in the Spent Nuclear Fuel Cask Loading Problem. Energies. 2020; 13(18):4869. https://doi.org/10.3390/en13184869

Chicago/Turabian Style

Bautista-Valhondo, Joaquín, Lluís Batet, and Manuel Mateo. 2020. "Minimizing the Standard Deviation of the Thermal Load in the Spent Nuclear Fuel Cask Loading Problem" Energies 13, no. 18: 4869. https://doi.org/10.3390/en13184869

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop