- Email: [email protected]
Direct reuse of spent nuclear fuel
Nuclear Engineering and Design 278 (2014) 182–189
Contents lists available at ScienceDirect
Nuclear Engineering and Design journal homepage: www.elsevier.com/locate/nucengdes
Direct reuse of spent nuclear fuel Nader M.A. Mohamed ∗ Atomic Energy Authority, ETRR-2, Cairo, Egypt
h i g h l i g h t s • • • • •
A new design for the PWR assemblies for direct use of spent fuel was proposed. The PWR spent fuel will be transferred directly (after a certain cooling time) to CANDU reactors. The proposed assembly has four zircaloy-4 tubes contains a number of CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end. MCNPX is used for the calculations that showed that the burnup can be increased by about 25%. Acceptable linear heat generation rate in hot rods and improved Pu proliferation resistance.
a r t i c l e
i n f o
Article history: Received 20 April 2014 Received in revised form 6 July 2014 Accepted 10 July 2014
a b s t r a c t In this paper we proposed a new design for the PWR fuel assembly for direct use of the PWR spent fuel without processing. The PWR spent fuel will be transferred directly (after a certain cooling time) to CANDU reactors which preferably built in the same site to avoid the problem of transportations. The proposed assembly has four zircaloy-4 tubes contains a number of CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end. Each tube has the same inner diameter of that of CANDU pressure tube. The spaces between the tubes contain low enriched UO2 fuel rods and guide tubes. MCNPX code is used for the simulation and calculation of the burnup of the proposed assembly. The bundles after the discharge from the PWR with their materials inventories are burned in a CANDU cell after a certain decay time. The results were compared with reference results and the impact of this new design on the uranium utilization improvement and on the proliferation resistance of plutonium is discussed. The effect of this new design on the power peaking, moderator temperature coefficient of reactivity and CANDU coolant void reactivity are discussed as well. © 2014 Elsevier B.V. All rights reserved.
1. Introduction The development of the fuel cycle technology for recycling the spent fuel of nuclear reactors has a particular importance to improve the uranium utilization, to reduce the high level nuclear waste and to reduce the amount of plutonium in spent fuel per unit energy. Pressurized water reactors (PWRs) that represent about 62% (International Atomic Energy Agency, 2006) of the operational reactors generate about 68% of the electrical power produced from nuclear reactors. PWRs uses fuel enriched up to 5 wt% and discharge the fuel with a significant amount of fissile isotopes (U-235, Pu-239 and Pu-241), often about twice that of natural uranium (Chad and Bollmann, 1998). The average discharge fuel burnup in PWR has increased from about 30 MWd/kgU in the 1970s to about 50 MWd/kg today with increase of the fuel average enrichment
∗ Tel.: +20 1220196502. E-mail address: [email protected] http://dx.doi.org/10.1016/j.nucengdes.2014.07.017 0029-5493/© 2014 Elsevier B.V. All rights reserved.
from about 3 w/o to about 4.5 w/o (Xu, 2003). The burnup licensing is mainly related to the corrosion and hydrogen pickup of the clad and the high burnup properties of the fuel and the dimensional changes of the fuel assembly/bundle structure. The licensing limit of the fuel burnup is averaged on the fuel assembly, and it is from 60 to 70 MWd/kgU for light water reactors (LWRs) (IAEA-TECDOC1299, 2002). There are two methods to recycling the fissile isotopes in the PWR spent fuel. First method is by extracting the fissile isotopes from the spent fuel through the chemical processing (IAEA-TECDOC-1587, 2008) and recycling them in the reactors. Second method is the direct use of spent PWR fuel in CANDU reactors (DUPIC) which is originally proposed in Korea (Chad and Bollmann, 1998; Myung et al., 2006; Jeong and Choi, 2000). In DUPIC the PWR fuel is stored for a period of cooling time. Then the spent fuel is shipped to a special plant to be re-fabricated into CANDU fuel bundles using a dry processing such as AIROX plant (Zhao et al., 1999). In this plant, the cladding is punctured and fission gases are captured. The cladding is removed and the fuel pellets are transferred
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
183
Fig. 1. MCNP simulation of: (a) Westinghouse PWR assembly, (b) lattice cell of CANDU-6 and (c) lattice cell of ACR-700.
to a furnace at peak temperature of 1600 C where the volatile fission products are removed while the pellets are reduced to a fine powder. This powder is milled, shaped, and sintered into CANDU fuel pellets used to build the fuel bundles. The high neutron economy on-power refueling CANDU reactor is originally designed for natural uranium fuel or slightly enriched uranium fuel. Therefore, considerable amount of fuel burnup can be obtained from DUPIC. Another example of direct use of PWR spent fuel in CANDU reactors is cutting the PWR fuel elements into CANDU length (∼50 cm), straighten them, and then weld new end-caps to the ends. The smaller diameter of PWR elements would enable the use of a 48- or 61-element fuel bundle which would significantly reduce the linear element ratings compared with those of a 37-element bundle and enhance fuel performance, and would help to accommodate the variation in fissile content between elements (Zhonsheng and Boczar, 2014). The core of PWR is built from long fuel bundles (assemblies) running the length of the core and arranged in a near-cylindrical array, the entire reactor core is a single large pressure vessel containing the light water, which acts as moderator and coolant. The CANDU reactor is built from modular horizontal fuel channels surrounded by a heavy water moderator. CANDU fuel bundle contains 27, 37 or 43 (or more) half meter long fuel tubes with 12 such bundles lying end to end in a pressure tube. Fig. 1 shows the differences between a typical PWR and CANDU fuel bundles. The power distribution throughout the PWR assembly is more uniform than the CANDU fuel bundle since, in CANDU reactor the fission neutrons are thermalized in the moderator outside the fuel channel and the fission rate is decreased from the outer pins to the central pin. Many parameters affect the power distribution across the CANDU bundle. These parameters include fuel enrichments, lattice pitch size, fuel pin diameter and number of fuel pins. Increasing the fuel enrichment increases the pin power peaking factor due to the self-shielding of the fuel pins and therefore CANDU reactors only use slightly enriched uranium fuel. CANDU-6 burns natural uranium to about 7.5 MWd/kg while the 700 MWe advanced CANDU reactor (ACR-700) uses 2.1 w/o enriched uranium and has an average burnup of 20.5 MWd/kg (Lam, 2009). Another important parameter affecting the design of CANDU reactors is the coolant void reactivity. The coolant void coefficient of reactivity is negative in light water reactors, since the coolant and the moderator are the same fluid in the reactor pressure vessel. In pressure tube reactors the moderator remains when the coolant is lost, and consequently the void reactivity may be positive as in the case of CANDU reactors. Many methods are proposed for the reduction of coolant void coefficient of reactivity. The main two methods are decreasing the moderator volume and use of a burnable poison like dysprosium in the central pin (Marczak, 1990; Whitlock, 1995).
In this paper we proposed a new design for the PWR assemblies for direct reuse of the PWR spent fuel without processing. The PWR spent fuel will be transferred directly (after a certain cooling time) to CANDU reactors which preferably built in the same site to avoid the problem of transportations. Safety parameters such as power peaking, moderator temperature coefficient of reactivity and CANDU coolant void reactivity for the new design will be discussed. 2. Proposed design A Westinghouse PWR fuel assembly has been selected as reference for this study. The fuel assembly consists of UO2 fuel rods bundled in 17 × 17 pins fuel assembly (21.5 cm pitch) as shown in Fig. 1a. The design proposed is shown in Fig. 2. Four zircaloy-4 tubes contains a number of CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end. Each tube has the same inner diameter of that of CANDU pressure tube (5.17 cm) and with wall thickness of 1.3 mm. The spaces between the tubes contain UO2 fuel rods with enrichment lower than that in the four zircaloy-4 tubes and guide tubes as shown in the figure. The PWR with the new design fuel assembly can be considered as a vertical tubes reactor
Fig. 2. MCNP model of the proposed PWR assembly for direct spent fuel reuse.
184
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
(IAEA-TECDOC-1391, 2004) but with a number of UO2 fuel rods between the tubes that under the same pressure. The assembly has the same pitch of 21.5 cm. Many designs concerning the number and diameters of fuel elements in the CANDU bundle were conducted (Technical Reports Series No. 407, 2002). The 61-elements design uses the same pin diameter used in PWR and it has the advantage of lower linear element ratings (Zhonsheng and Boczar, 2014). The CANDU bundle design of the present research contains 61 elements arranged in four rings and a central element as shown in Fig. 2. The outer ring contains 24 elements, the outer-middle ring contains 18 elements, the inner-middle ring contains 12 elements and the inner ring contains 6 elements. 44 low enriched UO2 fuel rods are arranged between the zircaloy-4 tubes as shown in Fig. 2. This means that the proposed assembly contains 288 UO2 fuel rods while the reference assembly contains 264 UO2 fuel rods. Almost, the same pitch to rod diameter ratio inside the zircaloy-4 tubes and between them is considered. All elements have the diameter of the reference PWR pin. The diameter of UO2 fuel pellets is 8.19 mm and the outer diameter of fuel rod is 9.5 mm with zircaloy-4 cladding thickness of 0.57 mm. Therefore, the license burnup limit of the reference PWR (60 or 70 MWd/kgU) can be applied for the 61-elements CANDU bundle. The Westinghouse PWR uses the Integrated Fuel Burnable Absorber (IFBA) as a burnable poison. It is applied as a thin coating of zirconium diborite (ZrB2 ) coating on fuel pellet surfaces of about 100 fuel rods in the fuel assembly and limited to 0.618 mgB10 /cm (Xu, 2003). The new design applies the IFBA on the outer ring of each fuel bundle. The number of guide tubes are adapted by taking the 14 × 14 annular fuel assembly proposed for the new generation PWRs (Kazimi and Hejzlar, 2006) as a reference where it uses 12 guide tubes for the control rods and to hold the assembly together. 12 guide tubes with the four large zircaloy-4 tubes are proposed to hold the proposed assembly together. These 12 guide tubes will be used for the control rods. The flow into the regions between the zircaloy-4 tubes (that have not fuel rods) should be restricted by inserting flow restrictors. A typical current operating Westinghouse PWR is built from 193 fuel assemblies with active length of 3.66 m and generates 3411 MWth (average power per assembly is 17.67 MWth). Advanced PWR (APWR) is built from 257 fuel assemblies with the same active length and generates 4451 MWth (average power per assembly is 17.32 MWth), while the European APWR (EU-APWR) has the same number of assemblies and thermal power but with active length of 4.27 m (Saito et al., 2011; Suzuki et al., 2009). The length of CANDU bundle is 49.53 cm which means that inserting 7 or 8 bundles in zircaloy-4 tubes make the core active length 3.467 m or 3.962 m, respectively. Therefore, modification of the fuel assembly active length is required to match the new length or a small bundle is required to compensate the difference in the length. The design modification changes the fuel rod configuration and pitch to rod diameter ratio which will influence the assembly power distribution, coolant mass flux and the pressure drop. The new design changes the fuel rod configuration from square lattice to be approximately hexagonal lattice. Changing the fuel rod diameters in light water reactors (LWRs) and use of tight lattice are considered in many studies for high conversion ratios. In a conventional PWR the mass flux is about 4 Mg/m2 s, whereas in a high conversion ratio PWR a much higher mass flux, about 6 Mg/m2 s, is required (IAEA-TECDOC-1299, 2002; Oldekop et al., 1982). A comparison between the reference and proposed PWR assemblies is presented in Table 1. The effect of this new design on the safety parameters such as the power peaking, moderator temperature coefficient of reactivity and CANDU coolant void reactivity should be verified while the main
Table 1 Comparison between the reference and proposed PWR assemblies. Parameter
Reference PWR assembly
Proposed PWR assembly
Construction
17 × 17 UO2 fuel rods bundle: - 264 fuel rods - 25 guide tubes
Fuel rod configuration
Square lattice
Dimensions
- Fuel rod inner/outer diameter: 8.19/9.5 mm - Fuel cladding thickness: 0.57 mm - Assembly pitch: 21.5 cm - Fuel active length: 3.66 m
Operating power Burnable poisons
∼17.5 MW IFBA coating on fuel pellet surfaces of about 100 fuel rods
- Four zircaloy-4 tubes contains a number of 61-elements CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end - 44 low enriched UO2 rods between the tubes - Guide tubes Approximately hexagonal lattice - Zircaloy-4 tube inner/outer diameters: 5.17/5.3 cm - Fuel rod inner/outer diameter: 8.19/9.5 mm - Fuel cladding thickness: 0.57 mm - Assembly pitch: 21.5 cm - CANDU fuel bundle length: 49.53 cm ∼17.5 MW IFBA coating on fuel pellet surfaces of outer ring of each fuel bundle
challenge will be the fabrication which should fix the fuel bundles for easy extraction after irradiation. 3. MCNPX simulation The Monte Carlo N-Particle (MCNP) is a general-purpose, continuous-energy, generalized-geometry, time-dependent, coupled neutron/photon/electron Monte Carlo transport code. The advanced version, MCNPX2.7.0 (Pelowitz, 2011) is used in this research for simulation. The code is compatible with MCNP5 (X5, 2003) with many additional capabilities including material depletion/burnup calculations. All standard evaluated nuclear data libraries used by MCNP can be used by MCNPX 2.7.0. Data libraries containing particle-interaction can be replaced by physics models if the libraries are not available. The program also includes cross-section measurements, benchmark experiments, deterministic code development, and improvements in transmutation code and library tools through the CINDER90 project (Wilson et al., 1995). MCNP user creates an input file that is subsequently read by MCNP. This file contains information about the problem in areas such as: the geometry specification, the description of materials and selection of cross-section evaluations, the location and characteristics of the neutron, photon, or electron source, the type of answers or tallies desired, and any variance reduction techniques used to improve efficiency. The MCNP calculation results always have associated errors due to the statistical uncertainty, discrepancies in material composition and geometry and errors in nuclear data libraries and theoretical models. Therefore, it is better to compare the MCNP calculation results with reference results (for example from measurements) or taking a reference case as a “calibrator”. The 17 × 17 pins Westinghouse PWR fuel assembly and CANDU6 and ACR-700 lattice cells were simulated as reference cases. Input files for these cases and for the proposed PWR fuel assembly were created for the burnup calculations with the criticality calculations (KCODE calculations). The boundaries of the lattice cells were
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
185
Table 2 Simulation data and results. Data
Reference e cases
Studied case
PWR assembly
CANDU-6 bundle
ACR-700 bundle
Proposed PWR assembly
RPSFB
Description
17 × 17 pins fuel assembly, 264 enriched UO2 rods and 25 guide tubes
37 fuel elements (UO2 rods) arranged in 3 rings and a central element
43 fuel elements (UO2 rods) arranged in 3 rings and a central element
61 fuel elements (UO2 rods) arranged in 4 rings and a central element
Lattice pitch (cm) Active fuel bundle length (cm) Enrichment
21.5 366
28.6 49.53
22 49.53
Four zircaloy-4 tubes sort, end to end, eight fuel bundles, 44 UO2 rods between the tubes and 12 guide tubes 21.5 396
4.5 w/o 235 U
0.71 w/o 235 U
0.309 mgB10 /cm IFBA applied on all rods
Poisons free
Power per assembly/channel Average burnup (MWd/kg U) EFPD Maximum rod burnup (MWd/kg U) Fissile materials at discharge (g/kg U)
17.67 MW
5.3 MW
6.8 MW
4.5 w/o 235 U in the bundles and 2.4 w/o U-235 in the rods between the bundles 0.533 mgB10 /cm IFBA applied on elements in the outer ring 17.32 MW
Spent PWR fuel of the proposed design
Burnable poisons
2.1 w/o 235 U in 42 elements and the central element has natural: UO2 and Dy2 O3 Dy2 O3 with 7.5 wt% Dy in natural U
51.7
7.5
20.5
21
1350 51.7
330 8.5
640 24.8
Bundles: 43.6 and rods between bundles: 39.2 1350 47.4
235 U: 8.4 239 Pu: 6.22 241 Pu: 1.8 0.155
235 U: 2.16 239 Pu: 2.52 241 Pu: 0.21 0.364
235 U: 5.62 239 Pu: 3.35 241 Pu: 0.57 0.191
235 U: 11 239 Pu: 7.12 241 Pu: 1.72
235 U: 4 239 Pu: 3.5 241 Pu: 1.04
Average 239,241 Pu produced per MWd (g/MWd)
simulated with reflecting surfaces. Any particle hitting a reflecting surface is specularly (mirror) reflected. Therefore, the calculations were carried out for infinite reactors constructed from these unit cells. In the reference PWR fuel assembly, the 264 UO2 fuel rods have the same enrichment of 4.5 w/o and with applying the same IFBA of 0.309 mg/cm. CANDU-6 uses natural 37-UO2 elements bundle arranged in three rings and central pin with 13.1 mm outer diameter of each pin. The fuel materials are represented as a material card for each fuel ring. The ACR uses 2.1 w/o enriched uranium in the three rings while the center pin contains burnable poison (UO2 + Dy2 O3 ) pellet with 7.5 wt% dysprosium in natural uranium. The bundle has two elements size: center pin and inner ring of seven elements with a diameter of 13.5 mm while the outer two rings consist of 35 elements with 11.5 mm diameter. The fuel materials are represented as a material card for each ring. In the proposed design each zircaloy-4 tube contains eight 61element fuel bundle (with dimensions and arrangement mentioned above). The active fuel length will be 3.96 m. The UO2 fuel elements contain 4.5 w/o enriched uranium. Also, the fuel materials are represented as a material card for each ring. 0.533 mgB10 /cm IFBA is applied on each element in the outer ring of the fuel bundles. The 44 UO2 fuel rods between the zircaloy-4 tubes contain 2.4 w/o enriched UO2 fuel rods. All of these rods are represented by one material card. The neutron cross section data for the fuel, the coolant and CANDU moderator were recalled from ENDF/B-VII.0 library which is publicly available. The best matching temperature in the library for the fuel is 900 K, for the coolant is 600 K and for the D2 O moderator is 294 K. S(˛,ˇ) (X-5, 2003) data were recalled for hydrogen and deuterium with best matching temperature in the library. A criticality calculation (KCODE calculations) with the BURN cards is used to calculate the system criticality and the burnup of the fuel and fuel inventory after each time interval (defined in the BURN cards). The power was given as 5.3 MW for the CANDU-6 channel, it was given as 6.8 MW for the ACR-700 channel and it
28.6 49.53
Residual of the IFBA
5.3 MW
660 23.2
0.094
was 17.67 MW for the reference PWR. For the proposed designed assembly, it is assumed that it will be operated in the APWR. Therefore the average power per assembly will be 17.32 MW. The time intervals were input as 0.5, 1, 2, 4, 10, 20, and then 30 s and at the end of the cycles it were 60 s day in the BURN cards. A total of 150 cycles per burnup step of which 30 were skipped and 2000 histories per cycle were used to reach standard deviation of the k∞ values around 0.1%. In BOPT card the Tier 3 fission products (Pelowitz, 2011), which comprise fission products in ENDF/B-VII.0 was selected with using cross-section models for nuclides not containing tabular data and then allowing CINDER90 to calculate the 1-group cross section for these nuclides by importing a 63-group flux and matching to a 63-group cross-section set. The 61-element bundles after the discharge from the PWR with their materials inventories are burned in a CANDU cell after a certain decay time (assumed to be 90 days in this research). This was carried out in the simulation by entering 90 days as the first time interval in the BURN cards with zero power fraction in the PFRAC card. Table 2 summarizes data used in the simulation. The MCNP simulated unit cells of the reference cases are shown in Fig. 1 and the MCNP model for the proposed design is shown in Fig. 2. 4. Results and discussions One method to calculate the discharge burnup is by calculating the single-batch core burnup and the results are generalized for multi-batch fuel management using the Linear Reactivity Model (LRM) (Driscoll et al., 1990). The single batch discharge burnup can be approximated by taking the burnup value from a reactivity vs. burnup curve (depletion history) where the infinite medium eigenvalue, k∞ decreases to 1.03 (a 3% leakage effect is assumed for PWRs) (Xu, 2003). This leakage will be decreased for the PWR built from 257 assemblies (∼4 m length) and reflected by stainless steel (Tahara et al., 2001; Tahara and Sekimoto, 2002). From
186
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
Fig. 3. The infinite medium eigenvalue, k∞ as function of EFPD for reference cases and studied case.
literature (Xu, 2003; Saglam et al., 2004), Current Westinghouse 193 assemblies 3-batch cores PWR fueled with 4.5 w/o enriched UO2 has a cycle of 450 Effective Full Power Days (EFPD) where each batch spends 1350 EFPD in the reactor with average discharge burnup of 51.7 MWd/kgU. The average bundle discharge burnup of CANDU-6 fueled with natural uranium is 7.5 MWd/kgU while it is 20.5 MWd/kgU for ACR-700 which fueled with 2.1 w/o enriched UO2 (Lam, 2009). The average irradiation time of CANDU-6 bundle is 330 EFPD and it is 640 EFPD for ACR-700 as given in Table 2. Fig. 3 gives the infinite medium eigenvalue, k∞ vs. the EFPD curves of the three reference cases, for the proposed PWR assembly and for the Recycled PWR Spent Fuel Bundles (RPSFB) when they were recycled in CANDU-6. Comparing the proposed design with the reference PWR assembly, we find that the proposed design decreases the reactivity at the beginning of the cycle while after about 1000 EFPD the reactivity of the proposed assembly becomes larger than the reference case as shown in the figure. Therefore, more than 1350 EFPDs can be achieved by the proposed assembly. The proposed design uses larger number of fuel rods (288 rods) with effective length of 396 cm and average enrichment of 4.18 w/o U235 as given in Table 2. This means that the initial amount of U-235 is higher than the reference case by 9.6% but taking into account the residual amount of U-235 at 1350 EFPDs burnup, the proposed design consumes less amount of U-235 by about 0.7%. Also, the amount of Plutonium fissile nuclides is higher than in the references case by about 30% at 1350 EFPDs. This of course due to the higher conversion ratio in the proposed design. The fuel inventory with their concentrations (average on the 61-element bundle) from the proposed PWR assembly after 1350 EFPDs followed by 90 days cooling time that simulated in the CANDU-6 pressure channel is given in Appendix A. Assuming that the same ACR-700 fuel management will be applied in CANDU-6 when fueled with the RPSFB, the RPSFB can achieve about 660 EFPD in CANDU-6 as shown in Fig. 3. Table 2 gives the burnup at the end of irradiation for all the cases. The average burnup of the 61-element bundles which contains 4.5 w/o enriched UO2 fuel is 43.6 MWd/kgU after the discharge from the PWR. After the discharge of the RPSFB from CANDU-6 it gives additional 21 MWd/kg. Therefore after the two cycles the average discharge of the 4.5 w/o enriched UO2 will be 64.6 MWd/kgU. The average burnup discharge of the 44 UO2 fuel rods that enriched to 2.4 w/o is 39.2 MWd/kgU as given in Table 2. This means that using this method, the average enrichment of 4.18 w/o UO2 fuel achieves average burnup of
60.7 MWd/kgU. The maximum burnup appears in the outer ring of the reference CANDU and proposed bundles as given in Table 2. The burnup of the outer ring of the proposed bundle after the discharge from the PWR is 47.4 MWd/kgU and after the discharge from CANDU-6 is 23.2 MWd/kgU with a total of 70.6 MWd/kgU. Three-batch fuel managements discharge one-third of the fuel assemblies per core cycle. Assuming 18 calendar months core cycle for the APWR: Number of assemblies discharged per year =
257 (assemblies) Assemblies ∼57 y 3 (batch) × 1.5 (year)
But, there are 4 zircaloy-4 tubes per assembly and 8 fuel bundles per tube, therefore: Number of bundles discharged per year = 57 × 4 × 8 = 1824
bundles y
The class 600+ MWe natural uranium CANDU reactor has 380 fuel channels while the class 900+ MWe has 480 channels with 12 fuel bundles per channel. Since, the EFPD of RPSFB is 660 days, Number of bundles loaded per year = 380 or 480 (channels) 365 days/y ×12 (bundles) × 660 EFPD Bundles ∼2500 or 3185 (for the two classes, respectively). y Assuming one CANDU reactor per APWR, 698 or 1361 slightly enriched fresh fuel bundles per year are needed with the RPSFBs for the 380 or 480 channels CANDU, respectively. Poisoning the central element of these fresh fuel bundles will control the full core coolant void reactivity. Increasing the burnup has another advantage of improving the plutonium proliferation resistance. Calculations resulted in much less plutonium produced per unit energy when burning the spent PWR fuel in CANDU reactors as given in Table 2. 4.1. Power peaking One important safety parameter when considering a new core design is the power peaking. The fuel rod power is limited such that
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
187
Fig. 4. The hot rod linear heat generation rate, q’ as function of the burnup for the reference and studied (for non-leakage reactors).
the fuel melting will not occur. The average axial power peaking of the proposed design will be slightly less than that of the reference case, since the active length in the proposed design (∼4 m) is longer than that of the reference case (3.66 m). However, the power distribution throughout the reference PWR assembly is more uniform than the proposed fuel assembly, since the fission rate is decreased from the outer pins to the central pin in the zircaloy-4 tubes. It follows that, higher fuel rod power peaking in the proposed design is expected, but considering that the number of fuel rods per assembly in the proposed design is 288 while it is 264 in the reference case, the average fuel rod power is less than that in the reference case. The hot rod power can by calculated as: Pmax = FA pmax ,
(1)
where Pmax is the hot rod power, FA is the assembly power peeking defined as the hot assembly average power density to the core average power density and pmax is the hot rod power for nonleakage reactor (infinite reactor) that it is each assembly has the same reactivity (same power). In the proposed design, if the same core pattern and fuel management as the reference case is considered, the same fuel assembly power peaking, FA can be achieved. The burnup was calculated as average on the 264 UO2 fuel rods in the reference assembly while in the proposed assembly it was averaged on each ring in reference CANDU bundles and in the 61element fuel bundle and average on the 44 UO2 fuel rods between the bundles. This means that the calculations assumed that the power is uniformly distributed in the reference assembly and in each ring of the fuel bundle of the reference CANDU reactors and in the proposed 61-element fuel bundle. From the calculation, the hot fuel ring during all burnup remains the outer ring of the reference and proposed fuel bundles. The linear heat generation rate, q’ of the hot rod as function of the burnup of the reference assembly, proposed assembly, CANDU-6 bundle, ACR-700 bundle and the RPSFB (for non-leakage reactor) is given in Fig. 4. It is calculated by multiplying the power fraction generated in the hot rod by the fuel lattice cell power divided by effective fuel length (assuming infinite reactor). In the simulation it is assumed that the proposed assembly will be operated in the APWRs with active fuel length of about 4 m. Therefore, from Fig. 4, the hot rod in the reference design for the non-leakage reactor will be operated at 67.5 and 65 kW at the beginning and end of burning, respectively, while the hot rod in the proposed design for the non-leakage reactor will be operated at 67 kW. For the same assembly power peaking, the hot rods in the
proposed and references designs will operate almost at the same power. The hot rods in RPSFB operates with much lower q’ than the 37-element CANDU-6 bundle as shown in Fig. 4. This is due to the higher number of fuel elements (61) in the RPSFB. This can help in increasing the power of CANDU-6. 4.2. Moderator temperature coefficient The moderator temperature coefficient (MTC) of reactivity is defined as the change of reactivity per degree change of the core averaged moderator temperature. MTC is an important safety parameter. Reactors are designed such that MTC has a negative value for a negative reactivity feedback. The value of MTC, a is calculated by dividing the change in the reactor reactivity (ıp) due to the change of the moderator temperature (ıT) (Mourtzanos et al., 2001): a=
ıp ıT
(2)
In PWRs, the dominant effect of moderator temperature on reactivity is due to the effect of temperature on moderator density (Xu, 2003; Mourtzanos et al., 2001). Increasing the water temperature leads to a decrease in its density which introduces a negative reactivity and vice versa. Decreasing the moderator to fuel ratio as in the proposed design makes MTC more negative. However, PWRs uses boric acid (H2 BO2 ) dissolved in the primary loop water as a soluble poisons (chemical shim) as a mean of reactivity control. This soluble poison has undesirable effect on the MTC because the decrease in the moderator density leads to decrease in the amount of soluble poison which introduces positive reactivity. Therefore, the soluble poison is limited to about 2000 ppm natural boron in the coolant for Westinghouse PWRs. As a verification the simulation was carried out at the beginning of the core cycle for calculating the change in the non leakage-reactor reactivity when the coolant density changed from 0.69 g/cm3 to 0.72 g/cm3 . That means 0.03 g/cm3 change in the coolant density around 0.705 g/cm3 (0.705 g/cm3 assumed for the coolant density for the full power normal operation). 900 ppm soluble boron is assumed in the reference and proposed designs. A total of 150 cycles of which 10 were skipped and 40,000 histories per cycle were used to reach standard deviation of about 20 pcm. The simulation gave a change in the reactor reactivity ıp, of 170 and 260 pcm for the reference and proposed designs, respectively. This change in the coolant density can be caused by a change in the
188
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
Fig. 5. Effect of RPSFB on CANDU-6 coolant void reactivity.
moderator temperature (ıT) by about −13 K around 583 K at 15.5 MPa coolant pressure (from compressed water tables). This means that for the above conditions, the MTC value of the reference design is approximately −13 pcm/K while it is approximately −20 pcm/K for the proposed design. 4.3. CANDU coolant void reactivity Marczak (1990) listed a number of reasons that make the positive coolant void reactivity with a value greater than the delayed neutron fraction, ˇ in CANDU-6 is acceptable. These reasons includes: (1) the coolant voids during a LOCA at a slow rate ensuring effective shutdown system action, (2) the reactor is divided into two separate cooling loops and a LOCA can generally expected to leave the cooling system, (3) the negative fuel temperature coefficient of reactivity, (4) the long prompt neutron lifetime (compared to that in the LWRs) limits the power increase rate following a LOCA and (5) two separate and independent shutdown system in CANDU600 increasing confidence in the ability of shutdown systems to operate effectively. However, the reduction of CANDU coolant void reactivity is a main target in the design of advanced CANDU reactors. The simulation was carried out to calculate the CANDU-6 coolant void reactivity as function of the fuel burnup when it is fueled by natural fuel bundles and when it is fueled by the RPSFBs. The material inventories at different fuel burnup were inputted in the CANDU-6 lattice cell MCNP files to run 150 cycles of which 10 were skipped and 20,000 histories per cycle to reach standard deviation of about 30 pcm for normal and voided coolant. The calculations assumed non-leakage reactor. The change in non-leakage probability during LOCA in CANDU-6 is about −0.7 mk (Marczak, 1990). This value is added to the results, since the calculations assumed nonleakage reactor. The calculation results are presented in Fig. 5. The equilibrium CANDU core contains fuel at a range of burnups, from 0 to exit-burnup values. Use the mid-burnup coolant void reactivity for the comparison, we find that in the case of fueling CANDU-6 by the RPSFB, coolant void reactivity is increased by about 4.5 mk. This increase is due to the increase of the moderator to fuel ratio and the increase in the amount of zircaloy-4 fuel cladding. As given above, the CANDU reactor will be fueled by about 27% or 43% (depending on the number of fuel channels, 380 or 480) slightly enriched fresh fuel bundles with the RPSFBs. Poisoning the central element of these fresh fuel bundles by dysprosium can reduce the full core coolant void reactivity.
5. Conclusion The proposed design of the PWR assembly enables reuse of the spent fuel without processing. The proposed assembly has four zircaloy-4 tubes contains a number of CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end. Each tube has the same inner diameter of that of CANDU pressure tube. The spaces between the tubes are filled by low enriched UO2 rods and guide tubes. 61-elements fuel bundles were considered for this assembly where each element has the same diameter of a reference PWR UO2 fuel rods. 44 low enriched UO2 fuel rods and 12 guide tubes were designed between the zircaloy-4 tubes. MCNPX code is used for the simulation and calculation of the burnup of the proposed assembly with 4.5 w/o enriched UO2 and 2.4 w/o enriched UO2 rods in the spaces between the tubes. The bundles after the discharge from the PWR with their materials inventories are burned in a CANDU cell after 90 days cooling time. The calculations showed that the burnup can be increased by about 25%. Much less plutonium produced per unit energy than reference case which improves the plutonium proliferation resistance. The calculations showed acceptable linear heat generation rate, q’ in the hot rods and higher negative MTC in the proposed assembly design than the reference design. Fueling CANDU600 by the spent 61-element fuel bundle increase the coolant void reactivity by about 4.5 mk. The number of discharged fuel bundles per year from the APWRs is less than that required for fueling the 380 or 480 fuel channels CANDU reactor by about 27% or 43%, respectively. Therefore, CANDU reactor will be fueled by about 27% or 43% (depending on the number of fuel channels) slightly enriched fresh fuel bundles with the RPSFBs. Poisoning the central element of these fresh fuel bundles by dysprosium can reduce the full core coolant void reactivity. The main challenge of this proposed fuel assembly will be the fabrication which should fix the fuel bundles for easy extraction after irradiation. Appendix A. The fuel inventory with their concentrations (average on the 61elemnt bundle) from the proposed PWR assembly after 1350 EFPDs followed by 90 days cooling time.
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
189
Actinide inventory Nuclidea 90,232 92,234 92,235
Mass fraction
Nuclide
Mass fraction
Nuclide
Mass fraction
Nuclide
Mass fraction
Nuclide
Mass fraction
5.18E−10 4.05E−06 1.08E−02
92,236 92,237 92,238
5.16E−03 9.01E−10 8.15E−01
93,236 93,237 94,236
2.94E−09 5.41E−04 3.32E−10
94,238 94,239 94,240
2.20E−04 6.49E−03 2.49E−03
94,241 94,242 95,241
1.49E−03 5.07E−04 1.78E−05
Non-actinide inventory 4.31E−07 6012 9.41E−06 6013 6.52E−11 6014 7015 2.40E−08 8016 1.19E−01 8017 4.43E−08 6.35E−10 31,069 31,071 2.48E−09 32,072 6.68E−09 32,073 1.73E−08 32,074 5.08E−08 3.48E−07 32,076 1.28E−07 33,075 5.50E−09 34,076 34,077 8.58E−07 2.52E−06 34,078 5.09E−06 34,079 1.45E−05 34,080 3.80E−05 34,082 35,079 2.31E−10 2.27E−05 35,081 36,080 2.88E−10 8.08E−07 36,082 4.71E−05 36,083 1.26E−04 36,084 36,085 2.59E−05 36,086 2.27E−04 1.06E−04 37,085 1.05E−09 37,086 2.84E−04 37,087 38,086 7.25E−07 38,087 4.80E−09 3.92E−04 38,088 380S9 6.54E−06 6.13E−04 38,090 39,089 5.21E−04
39,090 39,091 40,090 40,091 40,092 40,093 40,094 40,095 40,096 41,094 41,095 42,094 42,095 42,096 42,097 42,098 42,100 43,099 44,099 44,100 44,101 44,102 44,103 44,104 44,106 45,103 45,106 46,102 46,104 46,105 46,106 46,107 46,108 46,110 47,109 48,108
1.59E−07 1.20E−05 3.54E−05 6.68E−04 7.37E−04 8.08E−04 8.79E−04 2.13E−05 9.16E−04 2.51E−09 1.94E−05 1.29E−08 8.18E−04 5.42E−05 8.94E−04 9.28E−04 1.06E−03 8.62E−04 3.58E−08 1.31E−04 8.64E−04 8.99E−04 7.64E−06 6.06E−04 1.26E−04 4.76E−04 5.09E−11 2.14E−10 2.59E−04 4.06E−04 2.60E−04 2.29E−04 1.48E−04 4.81E−05 7.19E−05 4.10E−10
48,110 48,111 48,112 48,113 48,114 48,116 49,113 49,115 50,115 50,116 50,117 50,118 50,119 50,120 50,122 50,123 50,124 50,126 51,121 51,123 51,124 51,125 52,122 52,123 52,124 52,125 52,126 52,128 52,130 53,127 53,129 53,131 54,126 54,128 54,129 54,130
3.66E−05 2.31E−05 1.09E−05 1.19E−07 1.32E−05 5.03E−06 3.01E−10 8.18E−07 2.09E−07 2.57E−06 4.59E−06 3.93E−06 4.12E−06 3.99E−06 5.24E−06 9.43E−08 8.88E−06 2.08E−05 3.76E−06 5.07E−06 1.14E−08 7.44E−06 3.12E−07 3.14E−09 2.30E−07 4.10E−06 9.52E−07 9.70E−05 4.33E−04 4.91E−05 1.68E−04 2.65E−09 5.83E−11 3.83E−06 2.63E−08 6.74E−06
54,131 54,132 54,134 54,136 55,133 55,134 55,135 55,136 55,137 56,132 56,133 56,134 56,135 56,136 56,137 56,138 56,140 57,138 57,139 57,140 58,139 58,140 58,141 58,142 58,144 59,141 59,143 60,142 60,143 60,144 60,145 60,146 60,147 60,148 60,150 61,147
4.68E−04 1.27E−03 1.75E−03 2.49E−03 1.25E−03 1.44E−04 5.63E−04 6.48E−09 1.38E−03 1.75E−10 4.89E−10 8.43E−05 5.09E−07 2.76E−05 6.83E−05 1.50E−03 1.33E−07 1.81E−08 1.39E−03 2.02E−08 3.70E−09 1.40E−03 6.21E−06 1.29E−03 2.56E−04 1.25E−03 1.82E−07 2.37E−05 9.11E−04 1.22E−03 7.42E−04 8.18E−04 2.06E−08 4.23E−04 2.00E−04 1.38E−04
62,147 62,148 62,149 62,150 62,151 62,152 62,154 63,151 63,152 63,153 63,154 63,155 63,156 64,152 64,153 64,154 64,155 64,156 64,157 64,158 64,160 65,159 65,160 66,160 66,161 66,162 66,163 66,164 67,165 68,166 6S167 68,168 68,170
8.41E−05 1.28E−04 3.29E−06 3.06E−04 1.41E−05 9.99E−05 4.04E−05 4.18E−08 1.23E−08 1.19E−04 3.03E−05 8.39E−06 6.02E−08 3.15E−08 1.98E−09 3.74E−06 4.21E−07 9.31E−05 1.59E−07 1.99E−05 1.09E−06 2.57E−06 3.06E−08 2.79E−07 3.87E−07 2.80E−07 1.80E−07 3.60E−08 4.95E−08 1.63E−08 7.92E−10 3.42E−09 2.76E−10
a Nuclides are defined by four or five digits in which the last three digits are the mass number and the rest digit(s) is/are the atomic number. For example O-16 is given by 8016 and U-235 is given by 92,235.
References Chad, A., Bollmann, 1998. Optimization of DUPIC Cycle Environmental and Economic Performance (M.Sc thesis). Massachusetts Institute of Technology, USA. Driscoll, M.J., Downar, T.J., Pilat, E.E., 1990. The Linear Reactivity Model for Nuclear Fuel Management. American Nuclear Society, La Grange Park, IL. IAEA-TECDOC-1299, 2002. Technical and Economic Limits to Fuel Burnup Extension. IAEA, Austria. IAEA-TECDOC-1391, 2004. Status of Advanced Light Water Reactor Designs. IAEA, Vienna, ISBN: 92-0 104804-1 1011-4289. IAEA-TECDOC-1587, 2008. Spent Fuel Reprocessing Options. IAEA, Vienna, ISBN: 978-92-0-103808-1, ISSN: 1011-4289. International Atomic Energy Agency, 2006. Nuclear Power Reactors in the World. IAEA, Vienna, IAEA-RDS 2/26, ISBN: 92-0-106906-5, ISSN: 1011-2642. Jeong, Chang-Joon, Choi, Hangbok, 2000. Compatibility analysis on existing reactivity devices in CANDU 6 reactors for DUPIC fuel cycle. Nucl. Sci. Eng. 134, 265–280. Kazimi, M.S., Hejzlar, P., 2006. High Performance Fuel Design for Next Generation PWRs: Final Report. Nuclear Energy Research Initiative, USA. Lam, W.K., 2009. Advanced CANDU Reactor (ACR-700) Simulator, User Manual. Cassiopeia Technologies Inc., Canada. Marczak, J.V., (Ph.D thesis) 1990. Low Enriched Uranium Fuel Cycle in Heavy Water Reactors. McMaster University. Mourtzanos, K., Housiadas, C., Antonopoulos-Domis, M., 2001. Calculation of the moderator temperature coefficient of reactivity for water moderated reactors. Ann. Nucl. Energy 28, 1773–1782. Myung, Seung Yang, Choi, Hangbok, Jeong, Chang Joon, Song, Kee Chan, Lee, Jung Won, Park, Geun Il, Kim, Ho Dong, Ko, Won Il, Park, Jang Jin, Kim, Ki Ho, Lee, Ho Hee, Park, Joo Hwan, 2006. The status and prospect of DUPIC fuel technology. Nucl. Eng. Technol. 38, 254–259. Oldekop, W., Bereg, H.D., Zeggel, W., 1982. General features of advanced pressurized water reactors with improved fuel utilization. Nucl. Technol. 59, 212–227.
Pelowitz, D.B. (Ed.), 2011. MCNPX User’s Manual, Version 2.7.0. Los Alamos National Laboratory, USA, LA-CP-11-00438. Saglam, M., Sapyta, J.J., Spetz, S.W., Hassler, L.A., 2004. Core designs and economic analyses of homogeneous thoria–urania fuel in light water reactors. Nucl. Technol. 147, 8–19. Saito, Takehiko, Yamashita, Junichi, Ishiwatari, Yuki, Oka, Yoshiaki, 2011. Advances in Light Water Reactor Technologies. Springer Science + Business Media LLC, New York, NY, USA. Suzuki, Shigemitsu, Kanda, Makoto, Shimizu, Shunsuke, Tanaka, Kazumasa, Terada, Seiji, Kawahara, Hiroto, 2009. Safe, reliable, economical advanced pressurized water reactor for European market. Mitsubishi Heavy Ind. Tech. Rev. 46 (March (1)). Tahara, Yoshihisa, Sekimoto, Hiroshi, 2002. Transport equivalent diffusion constants for reflector region in PWRs. J. Nucl. Sci. Technol. 39, 716–728. Tahara, Yoshihisa, Sekimoto, Hiroshi, Miyoshi, Yoshinori, 2001. Reactivity effect of iron reflector in LWR cores. J. Nucl. Sci. Technol. 38, 102–111. Technical Reports Series No. 407, 2002. Heavy Water Reactors: Status and Projected Development. IAEA, Vienna, ISBN 92-0–111502-4. Whitlock, J.J., 1995. Reduction of the Coolant Void Reactivity Effect in a CANDU Lattice Cell. McMaster University, Canda. Wilson, W.B., England, T.R., George, D.C., Young, P.G., 1995. Recent Development of the CINDER90 Transmutation Code and Data Library for Actinide Transmutation Studies. Los Alamos National Laboratory, New Mexico, USA. X-5 MONTE CARLO TEAM, 2003. MCNP – A General Monte Carlo: N-Particle Transport Code, Version 5. Los Alamos National Laboratory, USA, LA-UR-03-1987. Xu, Zhiwan, 2003. Design Strategies for Optimizing High Burnup Fuel in Pressurized Water Reactor (Ph.D thesis). MIT, USA. Zhao, Xianfeng, Driscoll, M.J., Kazimi, M.S., 1999. Results of spent fuel in PWRs via dry processing. In: 7th International Conference on Nuclear Engineering, Tokyo, Japan, April 19–23, JSME. Zhonsheng, Xie, Boczar, P.G., 2014. Candu Fuel-Cycle Vision, http://www.nuceng.ca/ canteachmirror/library/20054415.pdf.
Contents lists available at ScienceDirect
Nuclear Engineering and Design journal homepage: www.elsevier.com/locate/nucengdes
Direct reuse of spent nuclear fuel Nader M.A. Mohamed ∗ Atomic Energy Authority, ETRR-2, Cairo, Egypt
h i g h l i g h t s • • • • •
A new design for the PWR assemblies for direct use of spent fuel was proposed. The PWR spent fuel will be transferred directly (after a certain cooling time) to CANDU reactors. The proposed assembly has four zircaloy-4 tubes contains a number of CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end. MCNPX is used for the calculations that showed that the burnup can be increased by about 25%. Acceptable linear heat generation rate in hot rods and improved Pu proliferation resistance.
a r t i c l e
i n f o
Article history: Received 20 April 2014 Received in revised form 6 July 2014 Accepted 10 July 2014
a b s t r a c t In this paper we proposed a new design for the PWR fuel assembly for direct use of the PWR spent fuel without processing. The PWR spent fuel will be transferred directly (after a certain cooling time) to CANDU reactors which preferably built in the same site to avoid the problem of transportations. The proposed assembly has four zircaloy-4 tubes contains a number of CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end. Each tube has the same inner diameter of that of CANDU pressure tube. The spaces between the tubes contain low enriched UO2 fuel rods and guide tubes. MCNPX code is used for the simulation and calculation of the burnup of the proposed assembly. The bundles after the discharge from the PWR with their materials inventories are burned in a CANDU cell after a certain decay time. The results were compared with reference results and the impact of this new design on the uranium utilization improvement and on the proliferation resistance of plutonium is discussed. The effect of this new design on the power peaking, moderator temperature coefficient of reactivity and CANDU coolant void reactivity are discussed as well. © 2014 Elsevier B.V. All rights reserved.
1. Introduction The development of the fuel cycle technology for recycling the spent fuel of nuclear reactors has a particular importance to improve the uranium utilization, to reduce the high level nuclear waste and to reduce the amount of plutonium in spent fuel per unit energy. Pressurized water reactors (PWRs) that represent about 62% (International Atomic Energy Agency, 2006) of the operational reactors generate about 68% of the electrical power produced from nuclear reactors. PWRs uses fuel enriched up to 5 wt% and discharge the fuel with a significant amount of fissile isotopes (U-235, Pu-239 and Pu-241), often about twice that of natural uranium (Chad and Bollmann, 1998). The average discharge fuel burnup in PWR has increased from about 30 MWd/kgU in the 1970s to about 50 MWd/kg today with increase of the fuel average enrichment
∗ Tel.: +20 1220196502. E-mail address: [email protected] http://dx.doi.org/10.1016/j.nucengdes.2014.07.017 0029-5493/© 2014 Elsevier B.V. All rights reserved.
from about 3 w/o to about 4.5 w/o (Xu, 2003). The burnup licensing is mainly related to the corrosion and hydrogen pickup of the clad and the high burnup properties of the fuel and the dimensional changes of the fuel assembly/bundle structure. The licensing limit of the fuel burnup is averaged on the fuel assembly, and it is from 60 to 70 MWd/kgU for light water reactors (LWRs) (IAEA-TECDOC1299, 2002). There are two methods to recycling the fissile isotopes in the PWR spent fuel. First method is by extracting the fissile isotopes from the spent fuel through the chemical processing (IAEA-TECDOC-1587, 2008) and recycling them in the reactors. Second method is the direct use of spent PWR fuel in CANDU reactors (DUPIC) which is originally proposed in Korea (Chad and Bollmann, 1998; Myung et al., 2006; Jeong and Choi, 2000). In DUPIC the PWR fuel is stored for a period of cooling time. Then the spent fuel is shipped to a special plant to be re-fabricated into CANDU fuel bundles using a dry processing such as AIROX plant (Zhao et al., 1999). In this plant, the cladding is punctured and fission gases are captured. The cladding is removed and the fuel pellets are transferred
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
183
Fig. 1. MCNP simulation of: (a) Westinghouse PWR assembly, (b) lattice cell of CANDU-6 and (c) lattice cell of ACR-700.
to a furnace at peak temperature of 1600 C where the volatile fission products are removed while the pellets are reduced to a fine powder. This powder is milled, shaped, and sintered into CANDU fuel pellets used to build the fuel bundles. The high neutron economy on-power refueling CANDU reactor is originally designed for natural uranium fuel or slightly enriched uranium fuel. Therefore, considerable amount of fuel burnup can be obtained from DUPIC. Another example of direct use of PWR spent fuel in CANDU reactors is cutting the PWR fuel elements into CANDU length (∼50 cm), straighten them, and then weld new end-caps to the ends. The smaller diameter of PWR elements would enable the use of a 48- or 61-element fuel bundle which would significantly reduce the linear element ratings compared with those of a 37-element bundle and enhance fuel performance, and would help to accommodate the variation in fissile content between elements (Zhonsheng and Boczar, 2014). The core of PWR is built from long fuel bundles (assemblies) running the length of the core and arranged in a near-cylindrical array, the entire reactor core is a single large pressure vessel containing the light water, which acts as moderator and coolant. The CANDU reactor is built from modular horizontal fuel channels surrounded by a heavy water moderator. CANDU fuel bundle contains 27, 37 or 43 (or more) half meter long fuel tubes with 12 such bundles lying end to end in a pressure tube. Fig. 1 shows the differences between a typical PWR and CANDU fuel bundles. The power distribution throughout the PWR assembly is more uniform than the CANDU fuel bundle since, in CANDU reactor the fission neutrons are thermalized in the moderator outside the fuel channel and the fission rate is decreased from the outer pins to the central pin. Many parameters affect the power distribution across the CANDU bundle. These parameters include fuel enrichments, lattice pitch size, fuel pin diameter and number of fuel pins. Increasing the fuel enrichment increases the pin power peaking factor due to the self-shielding of the fuel pins and therefore CANDU reactors only use slightly enriched uranium fuel. CANDU-6 burns natural uranium to about 7.5 MWd/kg while the 700 MWe advanced CANDU reactor (ACR-700) uses 2.1 w/o enriched uranium and has an average burnup of 20.5 MWd/kg (Lam, 2009). Another important parameter affecting the design of CANDU reactors is the coolant void reactivity. The coolant void coefficient of reactivity is negative in light water reactors, since the coolant and the moderator are the same fluid in the reactor pressure vessel. In pressure tube reactors the moderator remains when the coolant is lost, and consequently the void reactivity may be positive as in the case of CANDU reactors. Many methods are proposed for the reduction of coolant void coefficient of reactivity. The main two methods are decreasing the moderator volume and use of a burnable poison like dysprosium in the central pin (Marczak, 1990; Whitlock, 1995).
In this paper we proposed a new design for the PWR assemblies for direct reuse of the PWR spent fuel without processing. The PWR spent fuel will be transferred directly (after a certain cooling time) to CANDU reactors which preferably built in the same site to avoid the problem of transportations. Safety parameters such as power peaking, moderator temperature coefficient of reactivity and CANDU coolant void reactivity for the new design will be discussed. 2. Proposed design A Westinghouse PWR fuel assembly has been selected as reference for this study. The fuel assembly consists of UO2 fuel rods bundled in 17 × 17 pins fuel assembly (21.5 cm pitch) as shown in Fig. 1a. The design proposed is shown in Fig. 2. Four zircaloy-4 tubes contains a number of CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end. Each tube has the same inner diameter of that of CANDU pressure tube (5.17 cm) and with wall thickness of 1.3 mm. The spaces between the tubes contain UO2 fuel rods with enrichment lower than that in the four zircaloy-4 tubes and guide tubes as shown in the figure. The PWR with the new design fuel assembly can be considered as a vertical tubes reactor
Fig. 2. MCNP model of the proposed PWR assembly for direct spent fuel reuse.
184
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
(IAEA-TECDOC-1391, 2004) but with a number of UO2 fuel rods between the tubes that under the same pressure. The assembly has the same pitch of 21.5 cm. Many designs concerning the number and diameters of fuel elements in the CANDU bundle were conducted (Technical Reports Series No. 407, 2002). The 61-elements design uses the same pin diameter used in PWR and it has the advantage of lower linear element ratings (Zhonsheng and Boczar, 2014). The CANDU bundle design of the present research contains 61 elements arranged in four rings and a central element as shown in Fig. 2. The outer ring contains 24 elements, the outer-middle ring contains 18 elements, the inner-middle ring contains 12 elements and the inner ring contains 6 elements. 44 low enriched UO2 fuel rods are arranged between the zircaloy-4 tubes as shown in Fig. 2. This means that the proposed assembly contains 288 UO2 fuel rods while the reference assembly contains 264 UO2 fuel rods. Almost, the same pitch to rod diameter ratio inside the zircaloy-4 tubes and between them is considered. All elements have the diameter of the reference PWR pin. The diameter of UO2 fuel pellets is 8.19 mm and the outer diameter of fuel rod is 9.5 mm with zircaloy-4 cladding thickness of 0.57 mm. Therefore, the license burnup limit of the reference PWR (60 or 70 MWd/kgU) can be applied for the 61-elements CANDU bundle. The Westinghouse PWR uses the Integrated Fuel Burnable Absorber (IFBA) as a burnable poison. It is applied as a thin coating of zirconium diborite (ZrB2 ) coating on fuel pellet surfaces of about 100 fuel rods in the fuel assembly and limited to 0.618 mgB10 /cm (Xu, 2003). The new design applies the IFBA on the outer ring of each fuel bundle. The number of guide tubes are adapted by taking the 14 × 14 annular fuel assembly proposed for the new generation PWRs (Kazimi and Hejzlar, 2006) as a reference where it uses 12 guide tubes for the control rods and to hold the assembly together. 12 guide tubes with the four large zircaloy-4 tubes are proposed to hold the proposed assembly together. These 12 guide tubes will be used for the control rods. The flow into the regions between the zircaloy-4 tubes (that have not fuel rods) should be restricted by inserting flow restrictors. A typical current operating Westinghouse PWR is built from 193 fuel assemblies with active length of 3.66 m and generates 3411 MWth (average power per assembly is 17.67 MWth). Advanced PWR (APWR) is built from 257 fuel assemblies with the same active length and generates 4451 MWth (average power per assembly is 17.32 MWth), while the European APWR (EU-APWR) has the same number of assemblies and thermal power but with active length of 4.27 m (Saito et al., 2011; Suzuki et al., 2009). The length of CANDU bundle is 49.53 cm which means that inserting 7 or 8 bundles in zircaloy-4 tubes make the core active length 3.467 m or 3.962 m, respectively. Therefore, modification of the fuel assembly active length is required to match the new length or a small bundle is required to compensate the difference in the length. The design modification changes the fuel rod configuration and pitch to rod diameter ratio which will influence the assembly power distribution, coolant mass flux and the pressure drop. The new design changes the fuel rod configuration from square lattice to be approximately hexagonal lattice. Changing the fuel rod diameters in light water reactors (LWRs) and use of tight lattice are considered in many studies for high conversion ratios. In a conventional PWR the mass flux is about 4 Mg/m2 s, whereas in a high conversion ratio PWR a much higher mass flux, about 6 Mg/m2 s, is required (IAEA-TECDOC-1299, 2002; Oldekop et al., 1982). A comparison between the reference and proposed PWR assemblies is presented in Table 1. The effect of this new design on the safety parameters such as the power peaking, moderator temperature coefficient of reactivity and CANDU coolant void reactivity should be verified while the main
Table 1 Comparison between the reference and proposed PWR assemblies. Parameter
Reference PWR assembly
Proposed PWR assembly
Construction
17 × 17 UO2 fuel rods bundle: - 264 fuel rods - 25 guide tubes
Fuel rod configuration
Square lattice
Dimensions
- Fuel rod inner/outer diameter: 8.19/9.5 mm - Fuel cladding thickness: 0.57 mm - Assembly pitch: 21.5 cm - Fuel active length: 3.66 m
Operating power Burnable poisons
∼17.5 MW IFBA coating on fuel pellet surfaces of about 100 fuel rods
- Four zircaloy-4 tubes contains a number of 61-elements CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end - 44 low enriched UO2 rods between the tubes - Guide tubes Approximately hexagonal lattice - Zircaloy-4 tube inner/outer diameters: 5.17/5.3 cm - Fuel rod inner/outer diameter: 8.19/9.5 mm - Fuel cladding thickness: 0.57 mm - Assembly pitch: 21.5 cm - CANDU fuel bundle length: 49.53 cm ∼17.5 MW IFBA coating on fuel pellet surfaces of outer ring of each fuel bundle
challenge will be the fabrication which should fix the fuel bundles for easy extraction after irradiation. 3. MCNPX simulation The Monte Carlo N-Particle (MCNP) is a general-purpose, continuous-energy, generalized-geometry, time-dependent, coupled neutron/photon/electron Monte Carlo transport code. The advanced version, MCNPX2.7.0 (Pelowitz, 2011) is used in this research for simulation. The code is compatible with MCNP5 (X5, 2003) with many additional capabilities including material depletion/burnup calculations. All standard evaluated nuclear data libraries used by MCNP can be used by MCNPX 2.7.0. Data libraries containing particle-interaction can be replaced by physics models if the libraries are not available. The program also includes cross-section measurements, benchmark experiments, deterministic code development, and improvements in transmutation code and library tools through the CINDER90 project (Wilson et al., 1995). MCNP user creates an input file that is subsequently read by MCNP. This file contains information about the problem in areas such as: the geometry specification, the description of materials and selection of cross-section evaluations, the location and characteristics of the neutron, photon, or electron source, the type of answers or tallies desired, and any variance reduction techniques used to improve efficiency. The MCNP calculation results always have associated errors due to the statistical uncertainty, discrepancies in material composition and geometry and errors in nuclear data libraries and theoretical models. Therefore, it is better to compare the MCNP calculation results with reference results (for example from measurements) or taking a reference case as a “calibrator”. The 17 × 17 pins Westinghouse PWR fuel assembly and CANDU6 and ACR-700 lattice cells were simulated as reference cases. Input files for these cases and for the proposed PWR fuel assembly were created for the burnup calculations with the criticality calculations (KCODE calculations). The boundaries of the lattice cells were
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
185
Table 2 Simulation data and results. Data
Reference e cases
Studied case
PWR assembly
CANDU-6 bundle
ACR-700 bundle
Proposed PWR assembly
RPSFB
Description
17 × 17 pins fuel assembly, 264 enriched UO2 rods and 25 guide tubes
37 fuel elements (UO2 rods) arranged in 3 rings and a central element
43 fuel elements (UO2 rods) arranged in 3 rings and a central element
61 fuel elements (UO2 rods) arranged in 4 rings and a central element
Lattice pitch (cm) Active fuel bundle length (cm) Enrichment
21.5 366
28.6 49.53
22 49.53
Four zircaloy-4 tubes sort, end to end, eight fuel bundles, 44 UO2 rods between the tubes and 12 guide tubes 21.5 396
4.5 w/o 235 U
0.71 w/o 235 U
0.309 mgB10 /cm IFBA applied on all rods
Poisons free
Power per assembly/channel Average burnup (MWd/kg U) EFPD Maximum rod burnup (MWd/kg U) Fissile materials at discharge (g/kg U)
17.67 MW
5.3 MW
6.8 MW
4.5 w/o 235 U in the bundles and 2.4 w/o U-235 in the rods between the bundles 0.533 mgB10 /cm IFBA applied on elements in the outer ring 17.32 MW
Spent PWR fuel of the proposed design
Burnable poisons
2.1 w/o 235 U in 42 elements and the central element has natural: UO2 and Dy2 O3 Dy2 O3 with 7.5 wt% Dy in natural U
51.7
7.5
20.5
21
1350 51.7
330 8.5
640 24.8
Bundles: 43.6 and rods between bundles: 39.2 1350 47.4
235 U: 8.4 239 Pu: 6.22 241 Pu: 1.8 0.155
235 U: 2.16 239 Pu: 2.52 241 Pu: 0.21 0.364
235 U: 5.62 239 Pu: 3.35 241 Pu: 0.57 0.191
235 U: 11 239 Pu: 7.12 241 Pu: 1.72
235 U: 4 239 Pu: 3.5 241 Pu: 1.04
Average 239,241 Pu produced per MWd (g/MWd)
simulated with reflecting surfaces. Any particle hitting a reflecting surface is specularly (mirror) reflected. Therefore, the calculations were carried out for infinite reactors constructed from these unit cells. In the reference PWR fuel assembly, the 264 UO2 fuel rods have the same enrichment of 4.5 w/o and with applying the same IFBA of 0.309 mg/cm. CANDU-6 uses natural 37-UO2 elements bundle arranged in three rings and central pin with 13.1 mm outer diameter of each pin. The fuel materials are represented as a material card for each fuel ring. The ACR uses 2.1 w/o enriched uranium in the three rings while the center pin contains burnable poison (UO2 + Dy2 O3 ) pellet with 7.5 wt% dysprosium in natural uranium. The bundle has two elements size: center pin and inner ring of seven elements with a diameter of 13.5 mm while the outer two rings consist of 35 elements with 11.5 mm diameter. The fuel materials are represented as a material card for each ring. In the proposed design each zircaloy-4 tube contains eight 61element fuel bundle (with dimensions and arrangement mentioned above). The active fuel length will be 3.96 m. The UO2 fuel elements contain 4.5 w/o enriched uranium. Also, the fuel materials are represented as a material card for each ring. 0.533 mgB10 /cm IFBA is applied on each element in the outer ring of the fuel bundles. The 44 UO2 fuel rods between the zircaloy-4 tubes contain 2.4 w/o enriched UO2 fuel rods. All of these rods are represented by one material card. The neutron cross section data for the fuel, the coolant and CANDU moderator were recalled from ENDF/B-VII.0 library which is publicly available. The best matching temperature in the library for the fuel is 900 K, for the coolant is 600 K and for the D2 O moderator is 294 K. S(˛,ˇ) (X-5, 2003) data were recalled for hydrogen and deuterium with best matching temperature in the library. A criticality calculation (KCODE calculations) with the BURN cards is used to calculate the system criticality and the burnup of the fuel and fuel inventory after each time interval (defined in the BURN cards). The power was given as 5.3 MW for the CANDU-6 channel, it was given as 6.8 MW for the ACR-700 channel and it
28.6 49.53
Residual of the IFBA
5.3 MW
660 23.2
0.094
was 17.67 MW for the reference PWR. For the proposed designed assembly, it is assumed that it will be operated in the APWR. Therefore the average power per assembly will be 17.32 MW. The time intervals were input as 0.5, 1, 2, 4, 10, 20, and then 30 s and at the end of the cycles it were 60 s day in the BURN cards. A total of 150 cycles per burnup step of which 30 were skipped and 2000 histories per cycle were used to reach standard deviation of the k∞ values around 0.1%. In BOPT card the Tier 3 fission products (Pelowitz, 2011), which comprise fission products in ENDF/B-VII.0 was selected with using cross-section models for nuclides not containing tabular data and then allowing CINDER90 to calculate the 1-group cross section for these nuclides by importing a 63-group flux and matching to a 63-group cross-section set. The 61-element bundles after the discharge from the PWR with their materials inventories are burned in a CANDU cell after a certain decay time (assumed to be 90 days in this research). This was carried out in the simulation by entering 90 days as the first time interval in the BURN cards with zero power fraction in the PFRAC card. Table 2 summarizes data used in the simulation. The MCNP simulated unit cells of the reference cases are shown in Fig. 1 and the MCNP model for the proposed design is shown in Fig. 2. 4. Results and discussions One method to calculate the discharge burnup is by calculating the single-batch core burnup and the results are generalized for multi-batch fuel management using the Linear Reactivity Model (LRM) (Driscoll et al., 1990). The single batch discharge burnup can be approximated by taking the burnup value from a reactivity vs. burnup curve (depletion history) where the infinite medium eigenvalue, k∞ decreases to 1.03 (a 3% leakage effect is assumed for PWRs) (Xu, 2003). This leakage will be decreased for the PWR built from 257 assemblies (∼4 m length) and reflected by stainless steel (Tahara et al., 2001; Tahara and Sekimoto, 2002). From
186
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
Fig. 3. The infinite medium eigenvalue, k∞ as function of EFPD for reference cases and studied case.
literature (Xu, 2003; Saglam et al., 2004), Current Westinghouse 193 assemblies 3-batch cores PWR fueled with 4.5 w/o enriched UO2 has a cycle of 450 Effective Full Power Days (EFPD) where each batch spends 1350 EFPD in the reactor with average discharge burnup of 51.7 MWd/kgU. The average bundle discharge burnup of CANDU-6 fueled with natural uranium is 7.5 MWd/kgU while it is 20.5 MWd/kgU for ACR-700 which fueled with 2.1 w/o enriched UO2 (Lam, 2009). The average irradiation time of CANDU-6 bundle is 330 EFPD and it is 640 EFPD for ACR-700 as given in Table 2. Fig. 3 gives the infinite medium eigenvalue, k∞ vs. the EFPD curves of the three reference cases, for the proposed PWR assembly and for the Recycled PWR Spent Fuel Bundles (RPSFB) when they were recycled in CANDU-6. Comparing the proposed design with the reference PWR assembly, we find that the proposed design decreases the reactivity at the beginning of the cycle while after about 1000 EFPD the reactivity of the proposed assembly becomes larger than the reference case as shown in the figure. Therefore, more than 1350 EFPDs can be achieved by the proposed assembly. The proposed design uses larger number of fuel rods (288 rods) with effective length of 396 cm and average enrichment of 4.18 w/o U235 as given in Table 2. This means that the initial amount of U-235 is higher than the reference case by 9.6% but taking into account the residual amount of U-235 at 1350 EFPDs burnup, the proposed design consumes less amount of U-235 by about 0.7%. Also, the amount of Plutonium fissile nuclides is higher than in the references case by about 30% at 1350 EFPDs. This of course due to the higher conversion ratio in the proposed design. The fuel inventory with their concentrations (average on the 61-element bundle) from the proposed PWR assembly after 1350 EFPDs followed by 90 days cooling time that simulated in the CANDU-6 pressure channel is given in Appendix A. Assuming that the same ACR-700 fuel management will be applied in CANDU-6 when fueled with the RPSFB, the RPSFB can achieve about 660 EFPD in CANDU-6 as shown in Fig. 3. Table 2 gives the burnup at the end of irradiation for all the cases. The average burnup of the 61-element bundles which contains 4.5 w/o enriched UO2 fuel is 43.6 MWd/kgU after the discharge from the PWR. After the discharge of the RPSFB from CANDU-6 it gives additional 21 MWd/kg. Therefore after the two cycles the average discharge of the 4.5 w/o enriched UO2 will be 64.6 MWd/kgU. The average burnup discharge of the 44 UO2 fuel rods that enriched to 2.4 w/o is 39.2 MWd/kgU as given in Table 2. This means that using this method, the average enrichment of 4.18 w/o UO2 fuel achieves average burnup of
60.7 MWd/kgU. The maximum burnup appears in the outer ring of the reference CANDU and proposed bundles as given in Table 2. The burnup of the outer ring of the proposed bundle after the discharge from the PWR is 47.4 MWd/kgU and after the discharge from CANDU-6 is 23.2 MWd/kgU with a total of 70.6 MWd/kgU. Three-batch fuel managements discharge one-third of the fuel assemblies per core cycle. Assuming 18 calendar months core cycle for the APWR: Number of assemblies discharged per year =
257 (assemblies) Assemblies ∼57 y 3 (batch) × 1.5 (year)
But, there are 4 zircaloy-4 tubes per assembly and 8 fuel bundles per tube, therefore: Number of bundles discharged per year = 57 × 4 × 8 = 1824
bundles y
The class 600+ MWe natural uranium CANDU reactor has 380 fuel channels while the class 900+ MWe has 480 channels with 12 fuel bundles per channel. Since, the EFPD of RPSFB is 660 days, Number of bundles loaded per year = 380 or 480 (channels) 365 days/y ×12 (bundles) × 660 EFPD Bundles ∼2500 or 3185 (for the two classes, respectively). y Assuming one CANDU reactor per APWR, 698 or 1361 slightly enriched fresh fuel bundles per year are needed with the RPSFBs for the 380 or 480 channels CANDU, respectively. Poisoning the central element of these fresh fuel bundles will control the full core coolant void reactivity. Increasing the burnup has another advantage of improving the plutonium proliferation resistance. Calculations resulted in much less plutonium produced per unit energy when burning the spent PWR fuel in CANDU reactors as given in Table 2. 4.1. Power peaking One important safety parameter when considering a new core design is the power peaking. The fuel rod power is limited such that
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
187
Fig. 4. The hot rod linear heat generation rate, q’ as function of the burnup for the reference and studied (for non-leakage reactors).
the fuel melting will not occur. The average axial power peaking of the proposed design will be slightly less than that of the reference case, since the active length in the proposed design (∼4 m) is longer than that of the reference case (3.66 m). However, the power distribution throughout the reference PWR assembly is more uniform than the proposed fuel assembly, since the fission rate is decreased from the outer pins to the central pin in the zircaloy-4 tubes. It follows that, higher fuel rod power peaking in the proposed design is expected, but considering that the number of fuel rods per assembly in the proposed design is 288 while it is 264 in the reference case, the average fuel rod power is less than that in the reference case. The hot rod power can by calculated as: Pmax = FA pmax ,
(1)
where Pmax is the hot rod power, FA is the assembly power peeking defined as the hot assembly average power density to the core average power density and pmax is the hot rod power for nonleakage reactor (infinite reactor) that it is each assembly has the same reactivity (same power). In the proposed design, if the same core pattern and fuel management as the reference case is considered, the same fuel assembly power peaking, FA can be achieved. The burnup was calculated as average on the 264 UO2 fuel rods in the reference assembly while in the proposed assembly it was averaged on each ring in reference CANDU bundles and in the 61element fuel bundle and average on the 44 UO2 fuel rods between the bundles. This means that the calculations assumed that the power is uniformly distributed in the reference assembly and in each ring of the fuel bundle of the reference CANDU reactors and in the proposed 61-element fuel bundle. From the calculation, the hot fuel ring during all burnup remains the outer ring of the reference and proposed fuel bundles. The linear heat generation rate, q’ of the hot rod as function of the burnup of the reference assembly, proposed assembly, CANDU-6 bundle, ACR-700 bundle and the RPSFB (for non-leakage reactor) is given in Fig. 4. It is calculated by multiplying the power fraction generated in the hot rod by the fuel lattice cell power divided by effective fuel length (assuming infinite reactor). In the simulation it is assumed that the proposed assembly will be operated in the APWRs with active fuel length of about 4 m. Therefore, from Fig. 4, the hot rod in the reference design for the non-leakage reactor will be operated at 67.5 and 65 kW at the beginning and end of burning, respectively, while the hot rod in the proposed design for the non-leakage reactor will be operated at 67 kW. For the same assembly power peaking, the hot rods in the
proposed and references designs will operate almost at the same power. The hot rods in RPSFB operates with much lower q’ than the 37-element CANDU-6 bundle as shown in Fig. 4. This is due to the higher number of fuel elements (61) in the RPSFB. This can help in increasing the power of CANDU-6. 4.2. Moderator temperature coefficient The moderator temperature coefficient (MTC) of reactivity is defined as the change of reactivity per degree change of the core averaged moderator temperature. MTC is an important safety parameter. Reactors are designed such that MTC has a negative value for a negative reactivity feedback. The value of MTC, a is calculated by dividing the change in the reactor reactivity (ıp) due to the change of the moderator temperature (ıT) (Mourtzanos et al., 2001): a=
ıp ıT
(2)
In PWRs, the dominant effect of moderator temperature on reactivity is due to the effect of temperature on moderator density (Xu, 2003; Mourtzanos et al., 2001). Increasing the water temperature leads to a decrease in its density which introduces a negative reactivity and vice versa. Decreasing the moderator to fuel ratio as in the proposed design makes MTC more negative. However, PWRs uses boric acid (H2 BO2 ) dissolved in the primary loop water as a soluble poisons (chemical shim) as a mean of reactivity control. This soluble poison has undesirable effect on the MTC because the decrease in the moderator density leads to decrease in the amount of soluble poison which introduces positive reactivity. Therefore, the soluble poison is limited to about 2000 ppm natural boron in the coolant for Westinghouse PWRs. As a verification the simulation was carried out at the beginning of the core cycle for calculating the change in the non leakage-reactor reactivity when the coolant density changed from 0.69 g/cm3 to 0.72 g/cm3 . That means 0.03 g/cm3 change in the coolant density around 0.705 g/cm3 (0.705 g/cm3 assumed for the coolant density for the full power normal operation). 900 ppm soluble boron is assumed in the reference and proposed designs. A total of 150 cycles of which 10 were skipped and 40,000 histories per cycle were used to reach standard deviation of about 20 pcm. The simulation gave a change in the reactor reactivity ıp, of 170 and 260 pcm for the reference and proposed designs, respectively. This change in the coolant density can be caused by a change in the
188
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
Fig. 5. Effect of RPSFB on CANDU-6 coolant void reactivity.
moderator temperature (ıT) by about −13 K around 583 K at 15.5 MPa coolant pressure (from compressed water tables). This means that for the above conditions, the MTC value of the reference design is approximately −13 pcm/K while it is approximately −20 pcm/K for the proposed design. 4.3. CANDU coolant void reactivity Marczak (1990) listed a number of reasons that make the positive coolant void reactivity with a value greater than the delayed neutron fraction, ˇ in CANDU-6 is acceptable. These reasons includes: (1) the coolant voids during a LOCA at a slow rate ensuring effective shutdown system action, (2) the reactor is divided into two separate cooling loops and a LOCA can generally expected to leave the cooling system, (3) the negative fuel temperature coefficient of reactivity, (4) the long prompt neutron lifetime (compared to that in the LWRs) limits the power increase rate following a LOCA and (5) two separate and independent shutdown system in CANDU600 increasing confidence in the ability of shutdown systems to operate effectively. However, the reduction of CANDU coolant void reactivity is a main target in the design of advanced CANDU reactors. The simulation was carried out to calculate the CANDU-6 coolant void reactivity as function of the fuel burnup when it is fueled by natural fuel bundles and when it is fueled by the RPSFBs. The material inventories at different fuel burnup were inputted in the CANDU-6 lattice cell MCNP files to run 150 cycles of which 10 were skipped and 20,000 histories per cycle to reach standard deviation of about 30 pcm for normal and voided coolant. The calculations assumed non-leakage reactor. The change in non-leakage probability during LOCA in CANDU-6 is about −0.7 mk (Marczak, 1990). This value is added to the results, since the calculations assumed nonleakage reactor. The calculation results are presented in Fig. 5. The equilibrium CANDU core contains fuel at a range of burnups, from 0 to exit-burnup values. Use the mid-burnup coolant void reactivity for the comparison, we find that in the case of fueling CANDU-6 by the RPSFB, coolant void reactivity is increased by about 4.5 mk. This increase is due to the increase of the moderator to fuel ratio and the increase in the amount of zircaloy-4 fuel cladding. As given above, the CANDU reactor will be fueled by about 27% or 43% (depending on the number of fuel channels, 380 or 480) slightly enriched fresh fuel bundles with the RPSFBs. Poisoning the central element of these fresh fuel bundles by dysprosium can reduce the full core coolant void reactivity.
5. Conclusion The proposed design of the PWR assembly enables reuse of the spent fuel without processing. The proposed assembly has four zircaloy-4 tubes contains a number of CANDU fuel bundles (7 or 8 bundles per tube) stacked end to end. Each tube has the same inner diameter of that of CANDU pressure tube. The spaces between the tubes are filled by low enriched UO2 rods and guide tubes. 61-elements fuel bundles were considered for this assembly where each element has the same diameter of a reference PWR UO2 fuel rods. 44 low enriched UO2 fuel rods and 12 guide tubes were designed between the zircaloy-4 tubes. MCNPX code is used for the simulation and calculation of the burnup of the proposed assembly with 4.5 w/o enriched UO2 and 2.4 w/o enriched UO2 rods in the spaces between the tubes. The bundles after the discharge from the PWR with their materials inventories are burned in a CANDU cell after 90 days cooling time. The calculations showed that the burnup can be increased by about 25%. Much less plutonium produced per unit energy than reference case which improves the plutonium proliferation resistance. The calculations showed acceptable linear heat generation rate, q’ in the hot rods and higher negative MTC in the proposed assembly design than the reference design. Fueling CANDU600 by the spent 61-element fuel bundle increase the coolant void reactivity by about 4.5 mk. The number of discharged fuel bundles per year from the APWRs is less than that required for fueling the 380 or 480 fuel channels CANDU reactor by about 27% or 43%, respectively. Therefore, CANDU reactor will be fueled by about 27% or 43% (depending on the number of fuel channels) slightly enriched fresh fuel bundles with the RPSFBs. Poisoning the central element of these fresh fuel bundles by dysprosium can reduce the full core coolant void reactivity. The main challenge of this proposed fuel assembly will be the fabrication which should fix the fuel bundles for easy extraction after irradiation. Appendix A. The fuel inventory with their concentrations (average on the 61elemnt bundle) from the proposed PWR assembly after 1350 EFPDs followed by 90 days cooling time.
N.M.A. Mohamed / Nuclear Engineering and Design 278 (2014) 182–189
189
Actinide inventory Nuclidea 90,232 92,234 92,235
Mass fraction
Nuclide
Mass fraction
Nuclide
Mass fraction
Nuclide
Mass fraction
Nuclide
Mass fraction
5.18E−10 4.05E−06 1.08E−02
92,236 92,237 92,238
5.16E−03 9.01E−10 8.15E−01
93,236 93,237 94,236
2.94E−09 5.41E−04 3.32E−10
94,238 94,239 94,240
2.20E−04 6.49E−03 2.49E−03
94,241 94,242 95,241
1.49E−03 5.07E−04 1.78E−05
Non-actinide inventory 4.31E−07 6012 9.41E−06 6013 6.52E−11 6014 7015 2.40E−08 8016 1.19E−01 8017 4.43E−08 6.35E−10 31,069 31,071 2.48E−09 32,072 6.68E−09 32,073 1.73E−08 32,074 5.08E−08 3.48E−07 32,076 1.28E−07 33,075 5.50E−09 34,076 34,077 8.58E−07 2.52E−06 34,078 5.09E−06 34,079 1.45E−05 34,080 3.80E−05 34,082 35,079 2.31E−10 2.27E−05 35,081 36,080 2.88E−10 8.08E−07 36,082 4.71E−05 36,083 1.26E−04 36,084 36,085 2.59E−05 36,086 2.27E−04 1.06E−04 37,085 1.05E−09 37,086 2.84E−04 37,087 38,086 7.25E−07 38,087 4.80E−09 3.92E−04 38,088 380S9 6.54E−06 6.13E−04 38,090 39,089 5.21E−04
39,090 39,091 40,090 40,091 40,092 40,093 40,094 40,095 40,096 41,094 41,095 42,094 42,095 42,096 42,097 42,098 42,100 43,099 44,099 44,100 44,101 44,102 44,103 44,104 44,106 45,103 45,106 46,102 46,104 46,105 46,106 46,107 46,108 46,110 47,109 48,108
1.59E−07 1.20E−05 3.54E−05 6.68E−04 7.37E−04 8.08E−04 8.79E−04 2.13E−05 9.16E−04 2.51E−09 1.94E−05 1.29E−08 8.18E−04 5.42E−05 8.94E−04 9.28E−04 1.06E−03 8.62E−04 3.58E−08 1.31E−04 8.64E−04 8.99E−04 7.64E−06 6.06E−04 1.26E−04 4.76E−04 5.09E−11 2.14E−10 2.59E−04 4.06E−04 2.60E−04 2.29E−04 1.48E−04 4.81E−05 7.19E−05 4.10E−10
48,110 48,111 48,112 48,113 48,114 48,116 49,113 49,115 50,115 50,116 50,117 50,118 50,119 50,120 50,122 50,123 50,124 50,126 51,121 51,123 51,124 51,125 52,122 52,123 52,124 52,125 52,126 52,128 52,130 53,127 53,129 53,131 54,126 54,128 54,129 54,130
3.66E−05 2.31E−05 1.09E−05 1.19E−07 1.32E−05 5.03E−06 3.01E−10 8.18E−07 2.09E−07 2.57E−06 4.59E−06 3.93E−06 4.12E−06 3.99E−06 5.24E−06 9.43E−08 8.88E−06 2.08E−05 3.76E−06 5.07E−06 1.14E−08 7.44E−06 3.12E−07 3.14E−09 2.30E−07 4.10E−06 9.52E−07 9.70E−05 4.33E−04 4.91E−05 1.68E−04 2.65E−09 5.83E−11 3.83E−06 2.63E−08 6.74E−06
54,131 54,132 54,134 54,136 55,133 55,134 55,135 55,136 55,137 56,132 56,133 56,134 56,135 56,136 56,137 56,138 56,140 57,138 57,139 57,140 58,139 58,140 58,141 58,142 58,144 59,141 59,143 60,142 60,143 60,144 60,145 60,146 60,147 60,148 60,150 61,147
4.68E−04 1.27E−03 1.75E−03 2.49E−03 1.25E−03 1.44E−04 5.63E−04 6.48E−09 1.38E−03 1.75E−10 4.89E−10 8.43E−05 5.09E−07 2.76E−05 6.83E−05 1.50E−03 1.33E−07 1.81E−08 1.39E−03 2.02E−08 3.70E−09 1.40E−03 6.21E−06 1.29E−03 2.56E−04 1.25E−03 1.82E−07 2.37E−05 9.11E−04 1.22E−03 7.42E−04 8.18E−04 2.06E−08 4.23E−04 2.00E−04 1.38E−04
62,147 62,148 62,149 62,150 62,151 62,152 62,154 63,151 63,152 63,153 63,154 63,155 63,156 64,152 64,153 64,154 64,155 64,156 64,157 64,158 64,160 65,159 65,160 66,160 66,161 66,162 66,163 66,164 67,165 68,166 6S167 68,168 68,170
8.41E−05 1.28E−04 3.29E−06 3.06E−04 1.41E−05 9.99E−05 4.04E−05 4.18E−08 1.23E−08 1.19E−04 3.03E−05 8.39E−06 6.02E−08 3.15E−08 1.98E−09 3.74E−06 4.21E−07 9.31E−05 1.59E−07 1.99E−05 1.09E−06 2.57E−06 3.06E−08 2.79E−07 3.87E−07 2.80E−07 1.80E−07 3.60E−08 4.95E−08 1.63E−08 7.92E−10 3.42E−09 2.76E−10
a Nuclides are defined by four or five digits in which the last three digits are the mass number and the rest digit(s) is/are the atomic number. For example O-16 is given by 8016 and U-235 is given by 92,235.
References Chad, A., Bollmann, 1998. Optimization of DUPIC Cycle Environmental and Economic Performance (M.Sc thesis). Massachusetts Institute of Technology, USA. Driscoll, M.J., Downar, T.J., Pilat, E.E., 1990. The Linear Reactivity Model for Nuclear Fuel Management. American Nuclear Society, La Grange Park, IL. IAEA-TECDOC-1299, 2002. Technical and Economic Limits to Fuel Burnup Extension. IAEA, Austria. IAEA-TECDOC-1391, 2004. Status of Advanced Light Water Reactor Designs. IAEA, Vienna, ISBN: 92-0 104804-1 1011-4289. IAEA-TECDOC-1587, 2008. Spent Fuel Reprocessing Options. IAEA, Vienna, ISBN: 978-92-0-103808-1, ISSN: 1011-4289. International Atomic Energy Agency, 2006. Nuclear Power Reactors in the World. IAEA, Vienna, IAEA-RDS 2/26, ISBN: 92-0-106906-5, ISSN: 1011-2642. Jeong, Chang-Joon, Choi, Hangbok, 2000. Compatibility analysis on existing reactivity devices in CANDU 6 reactors for DUPIC fuel cycle. Nucl. Sci. Eng. 134, 265–280. Kazimi, M.S., Hejzlar, P., 2006. High Performance Fuel Design for Next Generation PWRs: Final Report. Nuclear Energy Research Initiative, USA. Lam, W.K., 2009. Advanced CANDU Reactor (ACR-700) Simulator, User Manual. Cassiopeia Technologies Inc., Canada. Marczak, J.V., (Ph.D thesis) 1990. Low Enriched Uranium Fuel Cycle in Heavy Water Reactors. McMaster University. Mourtzanos, K., Housiadas, C., Antonopoulos-Domis, M., 2001. Calculation of the moderator temperature coefficient of reactivity for water moderated reactors. Ann. Nucl. Energy 28, 1773–1782. Myung, Seung Yang, Choi, Hangbok, Jeong, Chang Joon, Song, Kee Chan, Lee, Jung Won, Park, Geun Il, Kim, Ho Dong, Ko, Won Il, Park, Jang Jin, Kim, Ki Ho, Lee, Ho Hee, Park, Joo Hwan, 2006. The status and prospect of DUPIC fuel technology. Nucl. Eng. Technol. 38, 254–259. Oldekop, W., Bereg, H.D., Zeggel, W., 1982. General features of advanced pressurized water reactors with improved fuel utilization. Nucl. Technol. 59, 212–227.
Pelowitz, D.B. (Ed.), 2011. MCNPX User’s Manual, Version 2.7.0. Los Alamos National Laboratory, USA, LA-CP-11-00438. Saglam, M., Sapyta, J.J., Spetz, S.W., Hassler, L.A., 2004. Core designs and economic analyses of homogeneous thoria–urania fuel in light water reactors. Nucl. Technol. 147, 8–19. Saito, Takehiko, Yamashita, Junichi, Ishiwatari, Yuki, Oka, Yoshiaki, 2011. Advances in Light Water Reactor Technologies. Springer Science + Business Media LLC, New York, NY, USA. Suzuki, Shigemitsu, Kanda, Makoto, Shimizu, Shunsuke, Tanaka, Kazumasa, Terada, Seiji, Kawahara, Hiroto, 2009. Safe, reliable, economical advanced pressurized water reactor for European market. Mitsubishi Heavy Ind. Tech. Rev. 46 (March (1)). Tahara, Yoshihisa, Sekimoto, Hiroshi, 2002. Transport equivalent diffusion constants for reflector region in PWRs. J. Nucl. Sci. Technol. 39, 716–728. Tahara, Yoshihisa, Sekimoto, Hiroshi, Miyoshi, Yoshinori, 2001. Reactivity effect of iron reflector in LWR cores. J. Nucl. Sci. Technol. 38, 102–111. Technical Reports Series No. 407, 2002. Heavy Water Reactors: Status and Projected Development. IAEA, Vienna, ISBN 92-0–111502-4. Whitlock, J.J., 1995. Reduction of the Coolant Void Reactivity Effect in a CANDU Lattice Cell. McMaster University, Canda. Wilson, W.B., England, T.R., George, D.C., Young, P.G., 1995. Recent Development of the CINDER90 Transmutation Code and Data Library for Actinide Transmutation Studies. Los Alamos National Laboratory, New Mexico, USA. X-5 MONTE CARLO TEAM, 2003. MCNP – A General Monte Carlo: N-Particle Transport Code, Version 5. Los Alamos National Laboratory, USA, LA-UR-03-1987. Xu, Zhiwan, 2003. Design Strategies for Optimizing High Burnup Fuel in Pressurized Water Reactor (Ph.D thesis). MIT, USA. Zhao, Xianfeng, Driscoll, M.J., Kazimi, M.S., 1999. Results of spent fuel in PWRs via dry processing. In: 7th International Conference on Nuclear Engineering, Tokyo, Japan, April 19–23, JSME. Zhonsheng, Xie, Boczar, P.G., 2014. Candu Fuel-Cycle Vision, http://www.nuceng.ca/ canteachmirror/library/20054415.pdf.