logo

Analysis of fixation method of fuel assembly for lead-alloy cooled reactor

NUCLEAR ENERGY SCIENCE AND ENGINEERING

Analysis of fixation method of fuel assembly for lead-alloy cooled reactor

HAN Qian
WU Qing-Sheng
CHEN Jian-Wei
MEI Hua-Ping
HUANG Qun-Ying
Nuclear Science and TechniquesVol.26, No.5Article number 050601Published in print 20 Oct 2015Available online 20 Oct 2015
38600

As a potential candidate for generation IV reactors, lead-alloy cooled reactor has attracted much attentions in recent years. The China LEAd-based research Reactor (CLEAR) is proposed as the primary choice for the accelerator driven subcritical system project launched by Chinese Academy of Sciences. Lead-bismuth eutectic (LBE) is selected as the coolant of CLEAR owing to its efficient heat conductivity properties and high production rate of neutrons. In order to compensate the buoyancy due to the high density of lead-alloy, fixation methods of fuel assembly (FA) have become a research hotspot worldwide. In this paper, we report an integrated system of ballast and fuel element for CLEAR FA. It guarantees the correct positioning of each FA in normal and refueling operations. Force calculation and temperature analysis prove that the FA will be stable and safe under CLEAR operation conditions.

Lead-alloy cooled reactorFuel assembly (FA)Fixation methodsBallast

I. INTRODUCTION

The Accelerator Driven subcritical System (ADS) is an advanced nuclear energy system for transmutation of long-lived radioactive nuclear wastes and fission fuel breeding [1, 2]. The lead-alloy of lead-bismuth eutectic (LBE) is advantageous in its high production rate of neutrons, efficient heat removal properties, thermal hydraulics and safety characteristics etc. Therefore, LBE-cooled reactor is designed as a main candidate for the subcritical reactor of ADS [3]. In 2011, CAS launched the ADS project and the China LEAd-based research Reactor (CLEAR) proposed by the Institute of Nuclear Energy Safety Technology (INEST) was selected as the reference reactor [4].

Fuel assembly (FA) is a key component in lead-alloy cooled reactor. A conventional FA would float in the liquid lead-alloy due to the high density, and this problem can be solved by a FA-fixation method to compensate the buoyancy force. Typical fixation methods worldwide include ballast and locking device [5-9]. With a locking device, it is not necessary to increase the FA and core volume, but this causes difficulties in positioning the FA accurately to the lower grid plate in refueling, hence the needs of structure modification of transfer machine gripper and lower grid plate. Considering that the service life of FA in lead-alloy cooled reactor is longer than commercial reactors, no sufficient experimental and operation data can be used to prove its structural stability and reliability.

The density of an FA added with ballast is larger than LBE, so the FA can be easily installed into the correct position and will not float in abnormal operating conditions. The ballast is of inherent safety. It requests no changes in structures of transfer machine gripper and other coupling components, and is simple to be implemented. However, its disadvantage is that volume of the ballast-added FA increases dramatically.

II. TYPICAL STRUCTURE DESIGN AND FIXATION METHOD OF FAS

A. Buoyancy problems

The force analysis of the CLEAR FA (without ballast) is shown in Fig. 1. The direction of total force of the FA in water is the same as gravity, whereas the direction of total force of the FA in LBE is the same as buoyancy, hence the need of a specific design to solve the problem of floating FA in the coolant.

Fig. 1.
(Color online) Force analysis of the FA (without ballast).
pic
B. Fixation methods of typical reactors

In the European project of the Advanced Lead-cooled Fast Reactor European Demonstrator (ALFRED), the core consists of wrapped hexagonal FAs with pins, arranged in a triangular lattice. The 171 FAs are subdivided into two radial zones with different plutonium enrichment to ensure an effective power flattening, and surrounded by two rows of dummy elements serving as reflector. The ALFRED design uses upper tungsten ballast to guarantee the correct position of each FA in refueling operations. The high density of lead alloy causes a great buoyancy force and leads to the FA floatation. In normal operation, FAs are connected to the upper diagrid plate by preloaded springs; while in refueling operation, the upper diagrid plate is removed in order to reach the upper head of each FA. Thus, the FAs cannot change their position because of the upper ballast counterbalanced the buoyancy force.

In the Russian project of experimental accelerator driven system (XADS), the core includes 120 FAs, 162 dummy assemblies and 12 absorber assemblies. Each FA, consisting of hexagonal wrapper tube, operation head and position spike, is anchored to the lower diagrid through a locking device. The operation head can be connected to the transfer machine gripper, which has been improved referring to the sodium reactor technology and possesses the additional possibility to push the FA inside the LBE and keep its vertical trim. The transfer machine gripper locks/unlocks the FA to/from the lower diagrid, and unlocks/locks the FA from/to itself.

In the Belgian project of Multipurpose Hybrid Research Reactor for High-tech Applications (MYRRHA), the fresh core contains 68 FAs, which are kept to the core support plate by inserting them in holes of the core support plate from underneath. Each FA includes 91 pins and retains at their location by the buoyancy force. The sub-critical core is surrounded by the core barrel to prevent the FAs from moving away from each other. The refueling machine can move up and down for about 2 m, to extract the FAs from the core. The fuel handling system is performed underneath the core.

After weighing advantages and disadvantages of the three core designs (Table 1), an integrated system of ballast and fuel element guaranteed the correct FA-positioning in normal and refueling operations. In this paper, structural design of the FA, especially the fixation methods, will be described.

TABLE 1.
Advantages and disadvantages of the fixation methods
Reactors ALFRED/ CLEAR-I XADS MYRRHA
Fixation methods Ballast Locking device FA-inserted from underneath
Advantages Inherent safety, Hydraulic autofrettage No change in FA weight No change in FA weight
Disadvantages Weight increase of FA Relatively short service time Big change in reactor structure
Decrease of one fuel rod per FA
Special system for installation
Show more

III. STRUCTURE DESIGN OF CLEAR FA

A. The FA structure

In view of reducing technical risk and R&D cost of FA, UO2 was chosen as the reactor fuel due to its adequate chemical compatibility with the coolant lead-alloy. Based on the premise of the nuclear nonproliferation treaty and present situation in China, the enrichment of UO2 was designed as 19.75%. The maximum temperature limit of fuel pellet was 2300 ℃, and the temperature gradient limit was 4×104 ℃/m. Each FA includes 61 fuel pins, featuring a stainless steel cladding and holding UO2 pellets [10, 11]. Considering service conditions and design load cases in CLEAR, the preliminary structure design parameters of FA are listed in Table 2.

TABLE 2.
The design parameters of CLEAR FA [12, 13]
Items Values
Fuel elements 61
FA length (mm) 2110
FA diameter (mm) 117.2
Fuel material UO2 (19.75%)
Show more
B. The fixation method

Ballast was chosen as the fixation method for CLEAR FA. An integrated system of ballast and fuel element ensures correct positioning of FAs in normal and refueling operations. The design of ballast was analyzed with a detailed assessment of the service condition and accident risk.

1. Material of the ballast

Density of the ballast material must be much larger than coolant LBE. Performance of the ballast should be suitable for the lead-alloy cooled reactor environment. Thus, depleted uranium (DU) and tungsten are the candidate material for ballast. Their densities are almost twice as large as LBE. Their physical properties, such as high thermal conductivity and low thermal expansion, were suitable for the reactor environment. Most importantly, they are good at resisting LBE corrosion and irradiation swelling. DU and tungsten (W) are of different effective multiplication factors (keff), calculated at keff(W)=1470 pcm and keff(DU)=1638 pcm, considering 0.7% of the 235U enrichment in DU. Therefore, DU was chosen as the ballast material of CLEAR-I FA.

2. Force analysis of the ballast

The force analysis of FA without and with ballast added in coolant LBE is shown in Fig. 2.

Fig. 2.
(Color online) Force analysis of FA without and with ballast in coolant LBE.
pic

From Fig. 2(a), the total force of FA without ballast in coolant LBE is

Ftotal=|Fb+FsfG|, (1)

where, Fb = ρLBE·g·VFA is the buoyancy, ρLBE=11 g/cm-3 is the density of LBE, g is the acceleration of gravity, VFA is the volume of FA; Fsf is the coolant scouring force estimated by thermal hydraulics calculations; and G is the gravity of FA.

With the ballast of a DU cylinder in radius of rba = 5.45 cm, height of hba, and gravity of Fba, we have (Fig. 2(b))

hba=Fba/(SbaρDUg)=(Fb'+Fsf'G)/(πrba2ρDUg), (2)

where, Sba is the section of ballast; Fb' is the buoyancy of FA; Fsf' is the scouring force estimated by thermal hydraulics calculations; and ρDU=19.1 g/cm-3 is the density of DU. The thermal hydraulic calculation was performed, and the ballast height was estimate at hba=510 cm (Ftotal =0).

DU ballast in two versions was added into the FA to improve comprehensive density and to balance the upward force. Version A ballast was outside the FA, while Version B ballast was inside FA on every fuel element.

3. Comparison of the two ballast versions

Force analysis of the two versions were presented in Table 3.

TABLE 3.
Parameters of force analysis
Designs V (cm3) G (N) Fba (N) Fb (N) Fsf_max (N)
No ballast 10292.5 846.83 0 1109.4 26.25
Ballast A 15657.7 1851.1 1004.3 1687.8 26.25
Ballast B 14512.3 1482.3 635.5 1422.2 26.25
Show more

The Version A ballast (Fig. 3(a)) was made of several DU plates, connected with FA on the entrance of coolant. The pressure at entrance of coolant, PEC, can be calculated by

Fig. 3.
(Color online) Schematics of the fuel assembly with Version A (a) and Version B (b) ballast.
pic
PEC=Fcw/SE, (3)

where, Fcw = Fb + FsfG = 600 N and SE = 2.34 cm2 is the sectional area of entrance of coolant. So, PEC = 2.56 MPa. The refueling cycle was designed at ten years. Corrosion at corner will occur due to the small section area at the entrance of coolant. In the ten years, performance of stainless steel will decrease significantly, and deformation of coolant entrance was possible in normal and refueling operations. In addition, the crack grew much frequently owing to holes at the entrance of coolant.

Figure 3(b) shows the FA structure with the Version B ballast, which is divided to every fuel element. Design specifications of the fuel element are given in Table 4.

TABLE 4.
Design specifications (in mm) of CLEAR fuel element
Items Values
Pin length 1675
Pin diameter 12
Pellet diameter 10.9
Active zone length 800
Ballast length 510
Cladding material Stainless steel
Show more

The pressure of one cladding tube, PCT, can be calculated by

PCT=fcw/SCT, (4)

where fcw=10.4 N is the gravity of ballast in one fuel element, and SCT=0.1457 cm2> is the sectional area of one cladding tube. Thus, PCT=0.714 MPa, being much smaller than PEC. Therefore, under the same service environment, the safety margin of Version B is much larger than that of Version A. After weighting the pros and cons (Table 5), Version B was proposed as the most potential design for CLEAR FA, with fewer technological risks.

TABLE 5.
The pros and cons of two versions
Features Version A Version B
Force analysis –, Deformation of coolant entrance occurs. +, Smaller deformation
Inherent safety –, FA may float in reactor accidents. +, FA always sinks under coolant.
Safety margin –, Crack grows much easier. +, Service life is longer.
Show more

In normal operations, sufficient safety margin of ballast design should be given for service environment [14]. Therefore, safety analysis of the ballast should be carried out.

IV. TEMPERATURE FIELD ANALYSIS OF BALLAST

A. Model of ballast

As cross section of the ballast is central symmetry, 1/6 of the structure was chosen to establish the calculation model [15, 16] (Fig. 4). The DU pellet has a 15-15Ti cladding, with a helium layer between the pellet and cladding. Properties of the materials are given in Table 6.

Fig. 4.
(Color online) The cross section of the ballast.
pic
TABLE 6.
Properties of the ballast materials
Material DU 15-15Ti Helium
Density (g/cm3) 19.1 8 0.1347×10-3
Thermal conductivity 29.721 19.3 0.144
(400 ℃) (W/(m ℃))
Specific heat capacity (J/(kg ℃)) 140.73 561 3116 (constant volume)
Show more
B. Boundary conditions

The heat of ballast is transferred from DU pellet to cladding by helium, and the heat of cladding is transferred by the LBE coolant [17, 18]. When the reactor is in normal operation, the electromagnetic load is too small to be considered. Service parameters of the ballast in the lead alloy-cooled reactor are given in Table 7. Temperatures of the cladding and DU were obtained using empirical formulas. To ensure a sufficient safety, they were calculated conservatively.

TABLE 7.
Service parameters of the ballast
Parameters Value
Thermal power of the reactor (MW) 10
Ballast material DU
Coolant temperature (℃) 400
Heat transfer supervisor of cladding-pellet (W/(m2 ℃)) 1000
Maximum linear power density (kW/m) 4.06
Thermal conductivity of 15-15Ti (400 ℃) (W/(m ℃)) 19.3
Average velocity of the coolant (m/s) 0.5
Show more
C. The simulation results of steady-state operation

The finite element code ANSYS was used to simulate the ballast calculation model. The temperature field is shown in Fig. 5(a), on the basis of steady-state operation. Figure 5(b) shows the radial temperature of ballast pellet. After an amount of theoretical and experimental verification, the operation temperature limit of 15-15Ti was set to 550 ℃ [19, 20]. The simulation results show that the ballast temperature did not increase dramatically when the reactor was in steady-state operation. Thus, DU does not introduce damage to the cladding and other structural materials. In summary, the ballast design meets the requirements of design criteria.

Fig. 5.
(Color online) Temperature field of ballast (a) and radial temperature variation of ballast pellet (b).
pic

V. CONCLUSION

In this paper, the structure of fixation ballast for FA in the lead-alloy cooled reactor of CLEAR-I is investigated and designed. The main conclusion can be summarized as follows:

1) Comparing to locking devices, ballast is of attractive characteristics, including inherent safety, small change to core structure and reactor internals, hence the choice of ballast as the fixation method for CLEAR-I FA.

2) Calculation results of keff show that DU is better than tungsten in terms of the effective multiplication factor and initial loading capacity. So, DU is chosen for the ballast material.

3) The force calculation results of coolant entrance show that the ballast divided into every fuel element has sufficient safety margin under the service environment. The simulation results of temperature field prove that the temperatures of DU and cladding are stable and safe under normal operation condition.

The sufficient safety margin should be given for anticipated accidents, therefore, the safety analysis of ballast in abnormal operating conditions will be given in the near future.

References
[1] M Wang, C Lian, Y Li, et al.

Preliminary conceptual design of a lead-bismuth cooled small reactor (CLEAR-SR)

. Int J Hydrogen Energ, 2015, in press. DOI: 10.1016/j.ijhydene.2015.03.097
Baidu ScholarGoogle Scholar
[2] M H Zeng, Y Song, X L Ding, et al.

Kinematics analysis and verification of In-Vessel Fuel Handling System for China Lead-based Research Reactor

. Ann Nucl Energy, 2015, 85: 301-305. DOI: 10.1016/j.anucene.2015.05.031
Baidu ScholarGoogle Scholar
[3] M K Rowinski, T J White and J Y Zhao.

Innovation model of annular fuel design for lead-cooled fast reactors

. Prog Nucl Energ, 2015, 83: 270-282. DOI: 10.1016/j.pnucene.2015.04.002
Baidu ScholarGoogle Scholar
[4] Y C Wu, J P Qian and J N Yu.

The fusion-driven hybrid system and its material selection

. J Nucl Mater, 2002, 307-311: 1629-1636. DOI: 10.1016/S0022-3115(02)01272-2
Baidu ScholarGoogle Scholar
[5] S Qvist.

Optimization method for the design of hexagonal fuel assemblies

. Ann Nucl Energy, 2015, 75: 498-506. DOI: 10.1016/j.anucene.2014.08.060
Baidu ScholarGoogle Scholar
[6] H Hernandez-Lopez.

Thermo-mechanic behavior simulation for fuel assemblies manufactured in Mexico with FEMAXI-VI code

. Ann Nucl Energy, 2014, 64: 100-106. DOI: 10.1016/j.anucene.2013.09.040
Baidu ScholarGoogle Scholar
[7] J K Suh, J W Kim, S G Kwon, et al.

Experimental study of press drops through LOCA-generated debris deposited on a fuel assembly

. Nucl Eng Des, 2015, 289: 49-59. DOI: 10.1016/j.nucengdes.2015.04.009
Baidu ScholarGoogle Scholar
[8] C Fazio, A Alamo, A Almazouzi, et al.

European cross-cutting research on structural materials for Generation IV and transmutation systems

. J Nucl Mater, 2009, 392: 316-323. DOI: 10.1016/j.jnucmat.2009.03.020
Baidu ScholarGoogle Scholar
[9] D Maes.

Mechanical design of the small-scale experimental ADS: MYRRHA

. Energ Convers Manage, 2006, 47: 2710-2723. DOI: 10.1016/j.enconman.2006.02.009
Baidu ScholarGoogle Scholar
[10] L Damiani, M Monttcucco and A P Prato.

Conceptual design of a bayonet-tube steam generator for the ALFRED lead-cooled reactor

. Nucl Eng Des, 2013, 265: 154-163. DOI: 10.1016/j.nucengdes.2013.06.021
Baidu ScholarGoogle Scholar
[11] G Grasso, C Petrovich, D Mattiol, et al.

The core design of ALFRED, a demonstrator for the European lead-cooled reactors

. Nucl Eng Des, 2014, 278: 287-301. DOI: 10.1016/j.nucengdes.2014.07.032
Baidu ScholarGoogle Scholar
[12] L Cinotti, B Giraud and H A Abderrahim.

The experimental accelerator driven system (XADS) designs in the EURATOM 5th framework programme

. J Nucl Mater, 2004, 335: 148-155. DOI: 10.1016/j.jnucmat.2004.07.006
Baidu ScholarGoogle Scholar
[13] S Ishida and H Sekimoto.

Finding the best fuel assemblies shuffling scheme of ADS for MA transmutation using dynamic programming

. Nucl Eng Des, 2010, 240: 3645-3653. DOI: 10.1016/j.nucengdes.2010.07.008
Baidu ScholarGoogle Scholar
[14] C R F Azevedo.

Selection of fuel cladding material for nuclear fission reactors

. Eng Fail Anal, 2011, 8: 1943-1962. DOI: 10.1016/j.engfailanal.2011.06.010
Baidu ScholarGoogle Scholar
[15] P Roberto, B Andrea, C Antonio, et al.

Object-oriented modelling and simulation for the ALFRED dynamics

. Prog Nucl Energ, 2014, 71: 15-29. DOI: 10.1016/j.pnucene.2013.10.013
Baidu ScholarGoogle Scholar
[16] Y W Wu, X Li, X L Yu, et al.

Subchannel thermal-hydraulic analysis of the fuel assembly for liquid sodium cooled fast reactor

. Prog Nucl Energ, 2013, 68: 65-78. DOI: 10.1016/j.pnucene.2013.05.001
Baidu ScholarGoogle Scholar
[17] D V Di Maio, L Cretara, F Giannetti, et al.

An alternative solution for heavy liquid metal cooled reactors fuel assemblies

. Nucl Eng Des, 2014, 278: 503-514.DOI: 10.1016/j.nucengdes.2014.08.011
Baidu ScholarGoogle Scholar
[18] X J Liu and N Scarpelli.

Development of a sub-channel code for liquid metal cooled fuel assembly

. Ann Nucl Energy, 2015, 77: 425-435. DOI: 10.1016/j.anucene.2014.10.030
Baidu ScholarGoogle Scholar
[19] K D Hamman and R A Berry.

A CFD simulation process for fast reactor fuel assemblies

. Nucl Eng Des, 2010, 240: 2304-2312. DOI: 10.1016/j.nucengdes.2009.11.007
Baidu ScholarGoogle Scholar
[20] N Li.

Lead-alloy coolant technology and materials-technology readiness level evaluation

. Prog Nucl Energ, 2008, 50: 140-151. DOI: 10.1016/j.pnucene.2007.10.016
Baidu ScholarGoogle Scholar