D. Ridikas - Pr 03 Fusion–fission hybrid system for nuclear waste transmutation

0
0
12
2 months ago
Preview
Full text

  Progress in Nuclear Energy 48 (2006) 235–246

  

Fusion–fission hybrid system for nuclear waste transmutation (I):

Characterization of the system and burn-up calculations

a, b b c

D. Ridikas *, R. Plukiene , A. Plukis , E.T. Cheng

  a

DSM/DAPNIA/SPhN, CEA Saclay, F-91191 Gif-sur-Yvette, France

b

  

Institute of Physics, Savanoriu pr. 231, LT-2053 Vilnius, Lithuania

c

  

TSI Research, 312 S. Cedros Avenu, Solana Beach, CA 92075, USA

Abstract

Generally, a very intensive neutron source is needed for the efficient transmutation process of nuclear waste and minor actinides

in particular. A plasma based fusion device is naturally a candidate neutron source. In this paper, we analyze a molten salt (flibe)

blanket, driven by a plasma based fusion device, dedicated to burn actinides extracted from the LWR spent fuel. Performance

parameters such as k , burn-up, fluxes and equilibrium conditions were calculated to assess the characteristics of a flibe based

scr

transmutation plant. It seems that nearly 1.1metric ton of spent fuel TRU could be burned annually with an output of 3 GW fission

th

power. Detailed burn-up calculations show that equilibrium conditions of the fuel concentration in the flibe transmutation blanket

could be achieved, provided that fission products were removed at least in part and fresh spent fuel TRU was added during burn-up

on-line. q 2005 Elsevier Ltd. All rights reserved.

  Keywords:

  Hybrid system; Nuclear waste transmutation

  1. Introduction The main goal of our present work is to investigate the performance of the dedicated molten salt (flibe) blanket for

nuclear waste transmutation as a function of different treatment of fission products as well as different refueling of

already destroyed actinides. This work is a continuation of the related studies ( ;

on a flibe based actinide transmutation hybrid system driven by a fusion device.

  

Below, we present only a brief summary of the concept sufficient for understanding this follow-up study. A detailed

description of the fusion–fission hybrid system including some burn-up calculations can be found in Refs.

).

  In order to transmute efficiently the nuclear waste, a very high intensity neutron source is needed. It seems that an

inertial confinement fusion (ICF) device could provide with a neutron source as intense as a high power accelerator,

  18

  19

i.e. of the order of 10 –10 n/s A molten salt (LiF–BeF –(HN)F ) blanket,

  2

  4

surrounding this neutron source, then serves as a medium for transuranic actinides (TRU) to be burned. Flibe also has

  6

a function of both coolant and carrier of tritium breeding material ( Li in this case). A well-known advantage of

* Corresponding author. Tel.: C33 1 690 878 47; fax: C33 1 690 875 84.

  E-mail address: ridikas@cea.fr (D. Ridikas).

  0149-1970/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.

  236

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246

  Table 1 Main parameters of the fusion–fission hybrid system used in this study Fusion power

  250 MW Fission power

  3000 MW

  th k at equilibrium 0.743 eff

  M at equilibrium

  19 E TBR at equilibrium

  1.4 Initial TRU composition LWR spent fuel Initial TRU inventory (kg) 3040

  M E stands for the energy multiplication of the system, while TBR for the tritium breeding ratio (

the molten salt is a possibility of both refueling of burned TRU and extraction of fission products (FP) on-line.

Namely, these two parameters we are going to investigate in more detail in this work.

  The main characteristics of the fusion–fission hybrid system are presented in . In order to produce

3000 MW fission power, a 250 MW power fusion device operating with a deuterium–tritium fuel cycle is sufficient

th

  19 n/s. The initial load of the fuel composition and to provide with an external 14 MeV neutron source of w9!10

subsequent feeding material is a typical TRU vector defined by the LWR spent fuel to be given below. A simplified

fusion–fission hybrid system was modelled with the main geometry and material parameters given in

presents a geometry setup used in our calculations.

  6 As presented in flibe (a) contains 0.1% of Li compared to the total amount of Li. Flibe (b) density is 2 g/

  3

  3

  6

cm and corresponding TRU density is 0.074 g/cm of flibe or 0.5 M% in flibe. 0.6% of Li enrichment for flibe (b)

  6

was taken as a reference case ). In both cases, Li serves as a breeding material for

tritium production. The initial loading and subsequent feeding TRU compositions were those from LWR spent fuel

when fission products and uranium are separated (to be provided in detail later in this paper in Section 3).

  2. Modelling tools and procedure All calculations on flibe based actinide transmutation blanket were made employing the Monteburns code system

( codes. Monteburns was already successfully benchmarked by simulating the uranium based fuel cycle

of the high flux reactor at ILL Grenoble ). MCNP was also used to obtain

k in the mode of a criticality eigenvalue problem (kcode card in MCNP input), and also to estimate neutron flux as

eff well as k in its external source mode (sdef card in the MCNP input). In this case, k is defined as: scr scr

  K 1Þ=ðM K 1=nÞ (1) k scr Z ðM n n

where n—is an average number of neutrons per fission and M is a total neutron multiplication factor of the system.

n

  

The effect of 14 MeV neutrons is included in the k calculations—the multiplication factor will be different for

scr

  Table 2 Neutronics model: zone dimension and material compositions for a molten salt transmutation blanket as from (Cheng and Wong, 2000)

  KR Zone name Radius R i iC

  1 (cm) Thickness (cm) Material composition

  Cavity 0–200 200 Void Liquid wall 200–201

  1 Flibe (a) Metallic wall 201–201.3

  0.3 SS316, 50% Graphite 201.3–202.3

  1 Graphite Blanket Region 1 202.3–217.3

  15 Flibe (b) Region 2 217.3–232.3

  15 Flibe (b) Region 3 232.3–247.3

  15 Flibe (b) Region 4 247.3–262.3

  15 Flibe (b) Reflector 262.3–282.3

  20 Graphite Envelope 282.3–287.3

  5 SS316, 50%

  D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246 237 Fig. 1. A simplified geometry model of the fusion–fission hybrid used in the calculations. Also see . FD stands for a fusion device.

  

14 MeV and for fission source neutrons. The average value of n in a problem is calculated by dividing the fission

neutrons gained per fission by the fission neutrons lost to fission.

  All burn-up calculations were done employing this external source mode. The advantage of the Monteburns code is

that it can use different nuclear data libraries as long as they are compatible with MCNP format (see for example Ref.

  

being most frequently employed, while for the fission products JENDL data library ) was taken

due to the biggest number of fission products available (w200).

  The main goal of the modelling was to find the equilibrium conditions of the fusion–fission hybrid system and to

investigate how these conditions could be influenced due to the different modes of refueling (continuous or discrete

feeding) and also due to the different treatment of fission products (partial/full removal and/or continuous/discrete

removal). For this purpose, we had to:

  † stabilize an effective neutron multiplication coefficient and the total TRU mass in the system, (E) as a function of burn-up, † stabilize the neutron spectrum F n ) to be fueled on-line. † define the actual amount of fresh TRU material (DM TRU In all cases, the fission power is kept constant (3000 MW ), what corresponds to a variable fusion power of the th

fusion device (Section 3.2). In our simulations, this implies a renormalization of the absolute neutron flux if an

effective neutron multiplication coefficient of the system is changed.

  3. Results

3.1. Neutron flux and energy spectra

   presents the main neutron flux characteristics in different regions of the molten salt transmutation blanket

In the 239-energy group, neutron spectrum the

  239 240 242

  19

  6 resonance absorption by transuranium ( Pu, Pu and Pu) and molten salt components (

  F, Li) is clearly visible. We note that the averaged neutron flux is very high and corresponds to the flux typical only for high flux reactors.

In addition, the amount of thermal neutrons is very small (!1%). Most of the neutrons are epithermal or fast, what, in

principle, could favor the transmutation of actinides in the resonance region. Indeed, the fast neutron flux (E n

  O

  22

  2 0.1 MeV) is about 10 n/cm per full power year.

   gives a variation of averaged neutron spectrum as a function of burn-up. The neutron spectrum in the

  238

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246

  Table 3 Neutron flux characteristics in different regions of molten salt transmutation blanket at the beginning of the irradiation

  2 Zone name F (total), n/s cm F (E n n n O0.1 MeV), %

  15 Region 1 2.47!10

  26.2

  15 Region 2 1.84!10

  22.4

  15 Region 3 1.22!10

  20.4

  15 Region 4 0.75!10

  20.6

  15 Regions 1–4 1.48!10

  22.9 Also see

  19

  14

  6

10 F

  Li 242

  13 Pu 240

  10 Pu

  • s* lethargy)

  2

  12

  10 239

  Pu region 1 region 2 region 3 region 4

  11

  10 all regions

  10

  10 Neutron flux, n/(cm -8 -7 -6 -5 -4 -3 -2 -1

  1

  10

  10

  10

  10

  

10

  10

  10

  10

  10

  10 E, MeV Fig. 2. Neutron energy spectra in different flibe regions. A solid line indicates the average neutron flux over all regions with fuel.

the spectrum established by the fission neutrons. We note that the final spectrum due to 14 MeV source neutrons in the

blanket is almost constant. Equally, the flux level is constant too if the fusion power is not changed. The spectrum due

to fission neutrons is also very stable, except the flux level, which depends on fission reaction rates. This is because the

majority of the fission neutrons are characterized by the same fission neutron spectrum and only then moderated in the

  239 241

blanket. Indeed, most of the fission reactions still occur in Pu and Pu, even though at a long burn-up stage

significant amount of Am and Cm accumulates. From it is clearly seen that the neutron spectrum remains

rather stable as a function of burn-up. In addition, this is also true if either continuous or discrete feeding of the fresh

fuel is simulated. This might justify a less time consuming approach for the fuel burn-up calculations, i.e. as a good

approximation one could calculate the neutron flux (and consequently flux weighted neutron cross-sections) only

once and perform burn-up calculations with a constant neutron spectrum.

  Finally, gives one group flux weighted neutron capture (s ) and fission (s ) cross-sections including their c f

  ) for transuranium actinides and comparison with other transmutation systems being either critical =s ratios a (aZ s c f

or accelerator driven as reported by It is clear that epithermal flux is more favored for the most of

  Table 4 Averaged neutron spectrum over all flibe regions, presented in different energy groups (%) as a function of burn-up (time)

  F F

  Time, n, % (Discrete) n, % (Continuous) days [0.,1.] [1.,100.] [.1,100.] [.1,1.] [1.,20] [0.,1.] [1.,100.] [.1,100.] [.1,1.] [1.,20] (eV) (eV) (keV) (MeV) (MeV) (eV) (eV) (keV) (MeV) (MeV)

  0.16

  14.58

  62.49

  12.28

  10.45

  0.04

  14.58

  62.49

  12.28

  10.45 400

  0.16

  15.37

  62.07

  12.02

  10.33

  0.03

  14.84

  62.55

  12.24

  10.19 800

  0.16

  15.52

  61.87

  12.07

  10.35

  0.04

  15.09

  62.33

  12.18

  10.22 1200

  0.14

  15.43

  61.87

  12.11

  10.42

  0.04

  15.12

  62.12

  12.2

  10.38 1600

  0.15

  15.30

  61.98

  12.03

  10.51

  0.04

  15.25

  62.03

  12.08

  10.47 2000

  0.16

  15.33

  61.72

  12.14

  10.62

  0.03

  14.93

  62.22

  12.22

  10.47

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246 239

  Table 5 Average fission, capture cross-sections, and their ratios for a number of selected actinides

  ; ; Isotope s b s b a (f this work) a (f thermal) a (f fast)

  c f

  ( )

  237

  Np

  18.20

  0.26

  70.0

  63

  5.3

  238

  Pu

  6.79

  1.90

  3.6

  12

  0.53

  239

  Pu

  7.70

  10.80

  0.7

  0.58

  0.3

  240

  Pu

  14.60

  0.31 47.1 396.6

  1.6

  241

  Pu

  6.48

  20.70

  0.3

  0.4

  0.19

  242

  Pu

  14.20

  0.22

  64.5

  65.5

  1.8

  241

  Am

  20.60

  0.36 57.2 100

  7.4

  242m

  Am

  7.20

  44.00

  0.2

  0.23

  0.18

  243

  Am

  27.50

  0.21 57.2 111

  8.6

  243

  Cm

  3.25

  21.10

  0.2

  0.16

  0.14

  244

  Cm

  16.30

  0.65

  25.1

  16

  1.4

  245

  Cm

  3.25

  22.70

  0.1

  0.15

  0.18 Ð Ð sð EÞ4ðEÞ dE= 4ð EÞ dE. Average cross-section (barn) is defined as sZ

  239 237

actinides compared with thermal flux (only Pu and Np is transmuted more efficiently in this case). However,

also shows that a values are the smallest in fast neutron spectra—direct incineration by fission being maximal

in this case.

3.2. Equilibrium conditions

  A typical evolution of the TRU isotopic composition is presented in This is the case when the system is refueled continuously. In addition, the fission products are also continuously removed on-line at 100% rate. In order to define the equilibrium, we employ the following procedure. Since not all actinides are stabilized even

after 2000 days of operation (Cm isotopes in particular), we assume that the system reaches equilibrium (or system is

approaching equilibrium) when the variation of Pu isotope mass is less than 4% with respect to the actual plutonium

mass in the blanket. This is represented in detail in , where the time dependence of the ratio DM /M of Pu

i i

isotope masses is illustrated. According to this analysis, we conclude that the equilibrium is reached after about

1000G50 days of irradiation.

  A corresponding behavior of k and required fusion power P at a constant fission power of 3000 MW are scr fusion th 240 241

presented in . The excursion of k occurs mainly due to the mass conversion of Pu into Pu during early

scr 241

stage of irradiation ( for detailed mass dependence of Pu irradiation). The values of k were

scr

calculated in the case when all fission products are continuously removed from the blanket. The variation of fusion

power is between 220 and 242 MW. We remind that 1 MW fusion power corresponds to the external neutron source

  17 n/s of 14 MeV neutrons. intensity of w4!10 237 3 238 Np Pu

  10 239 2 240 Pu Pu 10 241 1 242 Pu Pu

  10 241

  Am M, kg 242

  Am

  10 243

  • –1 244

  Am Cm

  10 245

  • –2 246

  Cm

  10 Cm 500 1000 1500 2000

  Time, d

  240

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246

  20

  10

  • –10

  , % i

  238 Pu

  • –20

  / M i

  239 Pu

  M ∆ 240

  • –30 Pu

  241 Pu

  • –40

  242 Pu

  • –50 500 1000 1500 2000 Time, d

  Fig. 4. Variation in mass (%) of Pu isotopes as a function of the irradiation time in the case of continuous feeding of TRU material and total removal of FP; i stands for a different isotope of Pu.

  Finally, we note that w3.1 kg of TRU is burned per day when the equilibrium situation is reached, what

corresponds to nearly 1.1 metric ton of TRU burned annually. gives detailed isotopic composition of the

initial TRU vector and also at equilibrium, i.e. after w1000 days of irradiation. Different options DD, DC, CD, and

CC correspond to different refueling conditions and also to different treatment of the fission products (see the caption

of We note

that in all cases equilibrium is reached and a reasonable agreement between Pu, Np and Cm isotopes is obtained when

compared to the calculations reported in Ref. On the other hand, in our case we

  242 243 obtained much less Am and Am. This difference is mainly due to the different data libraries used.

3.3. Effect of different feeding options

  In general, it is preferred that an effective neutron multiplication factor is as stable as possible for any nuclear

reactor being either critical or sub-critical. , illustrates the effect of different feeding options on k . As

scr

expected, the discrete feeding results in sharp variations of k , consequently, possible sharp variations of the reactor

scr

power. Of course, this effect could be diminished with additional control rods. On the other hand, the situation is

improved considerably if continuous feeding option is chosen (see the same ).

  We also note separately that in both cases as above a slightly different isotopic composition at equilibrium stage is

obtained , the difference appears mainly in the first step of burn-up calculation,

  0.84 400 380

  0.82 360 340

  0.80 320

  W

  300

  k

scr scr , M

k

  0.78 280 fusion

  ~(1-k P )/k P

scr

fusion scr

  260 0.76 240

  220 0.74 200 500 1000 1500 2000

  

Time, d

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246 241

  Table 6 TRU composition (%) in the molten salt at initial and equilibrium stages Stage Initial Equilibrium Cases DD DC CD CC (Cheng et al.,

  2000) Time, days 1000 1000 1000 1000 w980

  237

  Np (%)

  3.78

  1.50

  1.48

  1.63

  1.62

  1.51 Total Np (%)

  3.78

  1.51

  1.50

  1.64

  1.64

  1.51

  238

  Pu (%)

  1.81

  6.92

  6.98

  6.47

  6.57

  4.83

  239

  Pu (%)

  51.64

  21.23

  21.05

  22.95

  22.75

  20.90

  240

  Pu (%)

  24.61

  29.49

  29.16

  30.42

  30.58

  30.10

  241

  Pu (%)

  8.22

  16.09

  16.18

  15.35

  15.31

  15.70

  242

  Pu (%)

  4.61

  10.14

  9.96

  9.90

  9.71

  10.70 Total Pu (%)

  90.89

  83.88

  83.35

  85.10

  84.92

  82.20

  241

  Am (%)

  4.21

  1.34

  1.32

  1.49

  1.48

  1.49

  242m

  Am (%)

  0.00

  0.29

  0.28

  0.29

  0.28

  0.59

  243

  Am (%)

  0.83

  1.82

  1.88

  1.60

  1.67

  4.80 Total Am (%)

  5.04

  3.45

  3.48

  3.38

  3.43

  6.89

  244

  Cm (%)

  0.19

  6.74

  6.87

  5.83

  6.05

  6.31

  245

  Cm (%)

  0.00

  2.91

  2.99

  2.45

  2.52

  2.68

  246

  Cm (%)

  0.00

  0.52

  0.54

  0.41

  0.43

  0.49 Total Cm (%)

  0.19

  10.17

  10.40

  8.70

  9.01

  9.49 Others (%)

  1.01

  1.31

  1.21

  1.03 DD, discrete feeding and discrete FP removal; DC, discrete feeding and continuous FP removal; CD, continuous feeding and discrete FP removal; CC, continuous feeding and continuous FP removal; in all cases 100% of FP were removed.

  

after which the behavior of the curves is rather similar. TRU mass at equilibrium with discrete feeding is 2770 kg and

with continuous feeding is w3100 kg. The reason for such a difference is that in the discrete feeding case a fresh TRU

material was added at the beginning of each step (100 days long), while in continuous feeding case—it was

replenished during each step. Therefore, in the first case less initial TRU material remains in the molten salt from the

very beginning. Smaller equilibrium mass of TRU material could be important for system safety requirements. On the

other hand, discrete feeding results in sharp fluctuations of k , as already discussed above, what would be difficult to

scr tolerate in the realistic system.

  One could also ask if different feeding options could change the equilibrium conditions. , clearly shows 239

that in both cases (with discrete and continuous feeding), Pu mass equilibrates with the same speed. In other

words, less than 4% variation is achieved in both cases at the same time. The same is also true for other isotopes of

plutonium.

  0.90

  0.85

  0.80 scr k

  0.75

discrete feeding

continuous feeding

  0.70

100% discrete removal of FP

  0.65 500 1000 1500 2000 Time, d

  242

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246

  237 Np 1600

  238 Pu

  239 1400

  Pu 240

  Pu 1200 241

  Pu 242

  1000 Pu

  241 Am 800

  M, kg 243

  Am 244

  600 Cm

  245 Cm 400 solid symbol line: 200 continuous TRU feeding open simbol line:

  500 1000 1500 2000 discrete TRU feeding Time, d Fig. 7. TRU isotopic composition profile in molten salt as a function of irradiation in the cases of discrete and continuous feeding.

3.4. Effect of fission products

  The importance of fission products in the molten salt blanket for fusion–fission hybrid system was analyzed

partially in Ref. ( It was concluded that on-line removal and disposal of some neutron poisonous

fission products may be needed in order to reach deep burn-up level and, at the same time, to keep the effective

neutron multiplication coefficient of the system (equally blanket fission power) as close to the initial value as possible.

In the same study, it was found that fission products are responsible at least for K3150 pcm decrease in the effective

neutron multiplication coefficient at 15% burn-up. It was also pointed out that the tolerance of the fission products in

molten salt may allow a batch process, and the allowable burn-up tolerance level could be about 10% or slightly more.

  In we identify the most important fission products contributing more than 0.1% to the total macroscopic K

  

3

neutron absorption of the entire blanket, being 6.3!10 (1/cm) for 0% burn-up at the beginning of the irradiation.

  

Typically, these fission products are long lived or stable nuclei with an averaged (flux weighted) neutron capture

cross-sections of the order of 5–80 barns. All these fission products, due to their constant accumulation and also

because of their important neutron capturing, will degrade the performance of the entire system as we already

  135

mentioned above. The only exception is Xe, know as the most ‘poisonous’ fission product, which is short lived

(T

  1/2 Z9.14 h only) but is equally important in a long run due to its very high neutron capture cross-section (about

  135

1200 b in this particular neutron flux). As it can be seen from the same the concentration of Xe gets

saturated and it will not perturb the system more as it has done by now (so called Xe-effect in reactor physics). One

  24

  20

  

239

Pu continuous feeding

  

239

  16 Pu discrete feeding

  , % i M

  12

  / i M ∆

  8

  4 500 1000 1500 2000 Time, d

  239

  Fig. 8. Variation in mass (%) of Pu as a function of the irradiation time in the case of continuous and discrete feeding and with 100% removal of

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246 243

  Table 7 Fission products contributing more than 0.1% to the total macroscopic neutron absorption in a self-cooled molten salt transmutation blanket

  ; Fission products s b 6% burn-up, atoms/b-cm 10% burn-up, atoms/b-cm 15% burn-up, atoms/b-cm

  c

  99 K

  7 K

  7 K

  6 Tc

  10 4.18!10 9.89!10 1.38!10 K K K

  101

  7

  6

  6 Ru

  6 6.21!10 1.06!10 1.48!10 K K K

  103

  7

  7

  6 Rh

  9 2.54!10 6.17!10 1.02!10 K K K

  105

  7

  7

  6 Pd

  5 5.44!10 9.46!10 1.33!10 K K K

  107

  7

  7

  7 Pd

  7 3.80!10 6.52!10 9.12!10 K K K

  108

  7

  7

  7 Pd

  13 2.68!10 4.66!10 6.62!10 K K K

  109

  7

  7

  7 Ag

  34 1.61!10 2.69!10 3.98!10 K K K

  131

  7

  7

  7 Xe

  30 2.53!10 4.22!10 5.67!10 K K K

  135

  9

  9

  9 Xe 1200 6.89!10 6.67!10 6.44!10 K K K 133

  7

  6

  6 Cs

  12 5.77!10 1.04!10 1.46!10 K K K

  145

  7

  7

  7 Nd

  8 3.02!10 5.11!10 7.06!10 K K K

  147

  7

  7

  7 Pm

  58 1.17!10 1.90!10 2.32!10 K K K

  149

  8

  7

  7 Sm

  86 9.88!10 1.55!10 1.98!10 K K K

  151

  8

  8

  7 Sm

  67 6.41!10 9.92!10 1.27!10 K K K

  152

  8

  7

  7 Sm

  71 6.36!10 1.07!10 1.45!10 K K K

  153

  8

  8

  7 Eu

  43 4.88!10 9.73!10 1.50!10 Note that the macroscopic neutron absorption due to a particular fission product in (1/cm) could be easily obtained by multiplying s c in (barns) with a corresponding atomic density in (atoms/barn cm) as a function of burn-up.

  

has to note that the actual effect of non-soluble or gaseous fission products for the performance of the entire system

will be somewhat lower than presented in this work; simply because these fission products will have a chance to leave

the blanket before neutron interaction takes place.

   presents the behavior of k when all fission products are removed in two different modes, i.e. continuously scr

and discretely. It seems that the way fission products are removed (as long as they are removed 100%) has little

influence on the performance of the system. On the other hand, the percentage of removed FP relative to the total mass

of fission products present in the blanket becomes crucial as shown in

  Four different scenarios were considered with the following FP removal options: 100, 50, 20 and 0%, respectively.

0% corresponds to the case when all fission products are left in the molten salt flibe, while in the case of 100%—all

fission products are removed.

  One can ask why there is such a big difference between removal of 50, 20 and 0% of fission products. This can be

explained by analyzing the creation/destruction and removal ratios of the fission products. We will take, for example,

132

a stable fission product Xe, which has a small neutron capture cross-section. In , we present its mass as a

function of burn-up when either 100% or part of the fission products is left. When all fission products are left, the

  132 amount of Xe increases linearly (triangles).

  0.90

  0.85

  0.80

  scr k

  0.75 Discrete removal of FP 100%

  0.70 Continuous removal of FP 100%

  0.65 500 1000 1500 2000

  Time, d

  244

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246

  0.90

  0.85

  0.80

  0.75 scr k

  FP removal:

  0.70 100% 50%

  0.65 20% 0%

  0.60 500 1000 1500 2000 Time, d Fig. 10. k scr as a function of different percentage of fission products removed from the flibe.

  132 When 20% of FP are taken away, the Xe mass reaches the equilibrium after w1200 days and later its mass

remains stable; in the case of 50% removal the equilibrium is reached already after w400 days. It should be pointed

out separately that as long as fission products reach equilibrium, k also varies very little (compare the curves in

scr

  

Rather similar situation will be with other (stable or

long-lived and with similar capture cross-sections) fission products—they will reach equilibrium at different times

and at different quantities. For the rest of fission products (short lived or with high capture cross-sections), the

equilibrium may be reached much faster and is some cases even without removal.

  Finally, we add that a mechanism to remove the fission products is definitely needed in the molten salt

transmutation system. This has to be done not only due to the better neutron balance of the system as discussed above

but also due the following reason. The burn-up of the TRU materials is about 1.1 metric ton per year, equivalent to

about 30% of the TRU inventory in the blanket. Solidification of fission products will occur in the molten salt quite

soon after the operation, if some fission products are not removed. The next study of this area would be to look into

strategy and reality how these fission products can be removed.

  3

  3

  3.5. Effect of the He H/ The neutron multiplication coefficient k in the blanket depends not only on accumulation of fission products as it scr

  3

  3

  3

was described in the previous chapter, but also on accumulation of volatile materials— H and He. H is produced

  6

  6

  3 mainly from Li ( Li(n,a)

  H), which should be continuously replenished with new TRU in the molten salt blanket in

  3

  3

  3

order to provide the continuous supply of H for fusion device. At the same time, due to the H b-decay the He is

  250 Continuous FP removal: 50% 200 20% g 0% k

  150 ass, e m

  100

  X 132

  50 500 1000 1500 2000 Time, d

  132

  Fig. 11. Mass variation of the stable fission product Xe during the irradiation time in the case of continuous feeding of TRU and different removal

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246 245

  0.84

  0.83

  0.82

  0.81

  0.80

  0.79

  scr k

  0.78

  3

  3

  0.77 k without H and He

  scr

  

3

  3

  0.76 k with H and He

  scr

  0.75

  0.74 200 400 600 800 1000 1200 1400 1600 1800 2000

  Time, d

  

3

  3

  3

  3 Fig. 12. k behaviour in the hybrid system blanket with H and He and in the system with H and He removal. scr

  3

  3

produced. The microscopic neutron capture cross-section of He(n,p) H is about 1000 b. The efficient neutron

  3

absorption in He disturbs neutron balance in the hybrid system blanket as it is shown in . Due to the presence

  3

  3

of H and He in the molten salt the k of the system decreases all the time starting from 400 days and in such case the

scr

equilibrium stage in the system cannot be achieved. After 2000 days 5% difference in k is obtained for the two

scr

  3

  3

  3

  3

cases: blanket with H and He and blanket without these nuclides. In more realistic scenario volatile H and He

  3 nuclides either will be removed (separated for fusion process as in the case of

  H) or will escape by leakage from the

molten salt blanket. Indeed, the solubility of hydrogen in the flibe molten salt is extremely low such that it is very

often the total inventory of hydrogen in the flibe blanket is only a few grams, compared to about 80 metric ton of flibe.

  3

3 On the other hand, the certain part of H and He will be present in the blanket during the system operation and this should be taken into account when making modelling calculations for fusion–fission hybrid system.

  4. Conclusions The performance a fusion–fission hybrid system with the molten salt (flibe) blanket, driven by a plasma based

fusion device and dedicated to burn actinides extracted from the LWR spent fuel TRU, was analyzed. Detailed burn-

up calculations with different modes of replenishing the destroyed TRU material and also with different treatment of

fission products were done using Monteburns code system. Major system performance parameters of the system (e.g.

k , F (E ), equilibrium conditions, etc.) were investigated. We conclude that equilibrium concentration for most of

scr n n

the TRU materials in the blanket can be achieved after about 3 years of exposure. Such a plant could annually burn

about 1.1 metric ton of spent fuel actinides when the fission power output is 3000 MW . Different replenishing of the

th

destroyed TRU material was tested (continuous and discrete feeding). A stable neutron energy spectrum was obtained

in both cases. The behavior of k in the case of a discrete feeding has sharp fluctuations if compared with continuous

scr

feeding. On the other hand, a different TRU mass in the blanket at equilibrium was obtained (3100 and 2770 kg with

continuous and discrete feeding, respectively). Although a smaller TRU mass at equilibrium is preferred due to safety

reasons, a more stable performance of the system obtained in the case of a continuous feeding is recommended.

Accumulation of fission products in the blanket was investigated including their influence on the performance of the

system. We found that k is rather stable (0.81–0.79) during the calculation time when all or at least 50% of fission

scr

products are taken away. The removal of a smaller percentage of fission products is not recommended since, k

scr

would fall down considerably (e.g. 0.81(0.76 and 0.81(0.65 with only 20 or 0% FP removal, respectively, within

2000 days of burn-up). The most poisonous fission products were identified in terms of their macroscopic capture

cross-sections. We found that fission products are responsible at least for K3150 pcm decrease in k at 15% burn-up.

scr

  3

  3 The presence of volatile H and He has significant influence on hybrid system k stability and should be taken into scr

account when performing hybrid system model calculations. In spite of very promising results on the efficient TRU

destruction, we found that at equilibrium, significant amount of curium isotopes is accumulated, while for the rest of

actinides fully equilibrated concentrations are reached. In this respect an optimization procedure should be developed

  246

D. Ridikas et al. / Progress in Nuclear Energy 48 (2006) 235–246

  

including all minor actinides. Another important issue is directly related to the benchmark of different sets of nuclear

data libraries in the case of minor actinide incineration scenarios. These problems are addressed in our companion

paper by . Acknowledgements This work was accomplished in the frame of the French–Lithuanian cooperation program ‘Gilibert’ (Ref. No

06452PG) and partly was supported by a Lithuanian State Science and Studies Foundation within C-03049 project.

References

  Briesmeister J. for Group X-6 (1993), MCNP-A, A General Monte Carlo Code for Neutron and Photon Transport, Version 4A, LANL, preprint LA- 12625-M. Cheng E.T. and Wong Clement P.C. (2000), Transmutation of Actinide in Fusion Reactors and Related Systems Analysis, Proc. of 10th Int. Conf. on Emerging Nuclear Energy Systems, ICENES’2000, Petten, The Netherlands, 24–28 September (and Refs. therein). Cheng, E.T., Cheng, E.T., 2001. Characteristics of Promising Transmutation Blanket Concepts Proc. of the International Workshop on Blanket and Fusion Concepts for the Transmutation of Actinides, General Atomics, San Diego, California, USA 2001. ENDF/B-VI.The US Evaluated Neutrons Nuclear Data Library for Neutron Reactions, IAEA-Vienna, IAEA-NDS-100 (1995), Rev 6. JENDL-3.2, Japanese Evaluated Nuclear Data Library, IAEA-Vienna, IAEA-NDS-110, (1994), Rev 5. Plukiene R., Plukis A., Ridikas D., Cheng E.T., Fusion - Fission Hybrid System for Nuclear Waste Transmutation (II): From the Burn-up Optimization to the Tests of Different Data, accepted for publication in Progress in Nuclear Energy (June 2005). Ridikas D., Fioni G., Goberis P., Marie F., A (2000), On the Fuel Cycle and Neutron Fluxes of the High Flux reactor at ILL Grenoble, Proc. of the 5th Int. Specialists Meeting SATIF-5, OECD/NEA Paris, France, 17–21 July. Ridikas D., Bletzacker L., Goberis P., Fioni G., Marie F., B (2000), “Critical and Sub-critical GT-MHRs for Waste Disposal and Proliferation-

  Resistant Plutonium Fuel Cycles”, paper 7 in the Proc. of the 6th OECD/NEA Information Exchange Meeting on Actinide and Fission Product Partitioning and Transmutation, p. 71. Madrid, Spain, 11–13 December. Ridikas, D., Fioni, G., Cheng, E.T.A., 2001a. A Importance of Fission Products in the Molten Salt Blanket for Fusion - Fission Hybrid Systems

  Proc. of the International Workshop on Blanket and Fusion Concepts for the Transmutation of Actinides, General Atomics, San Diego, California, USA 2001. Ridikas, D., Bletzacker, L., Goberis, P., Fioni, G., Marie, F., 2001b. Comparative Analysis of ENDF, JEF, and JENDL Data Libraries by Modelling

  High Temperature Reactors (GT-MHR) and Plutonium Based Fuel Cycles Proc. of the Int. Conference on Nuclear Data for Science and Technology, ND2001, Tsukuba, Ibaraki, Japan 2001. RSICC Computer Code Collection (1991), ORIGEN 2.1 -Isotope Generation and Depletion Code Matrix Exponential Method, Radiation shielding Information Center, report CCC-371. Salvatores, M., 2002. The physics of transmutation in critical or sub-critical reactors. C.R. Physique 3, 999–1012. Trellue H.R. and Poston D.I. (1999), User’s manual, Version 2.0 for Monteburns, Version5B, LANL, preprint LA-UR-99-4999.