INVESTIGATION OF DPA IN THE REACTOR PRESSURE VESSEL OF

VVER-1000/V320
NGUYEN HUU TIEP1*, PHAM NHU VIET HA1, NGUYEN MINH TUAN2
1
Institute for Nuclear Science and Technology, Vietnam Atomic Energy Institute
179 Hoang Quoc Viet Street, CauGiay, Ha Noi, Viet Nam
2
Dalat Nuclear Research Institute, Vietnam Atomic Energy Institute
01 Nguyen Tu Luc, Da Lat, Lam Dong, Viet Nam
*E-mail: [email protected]

Abstract: The most important ageing effect on the reactor pressure vessel (RPV) is radiation
embrittlement, which is mainly caused by fast neutrons during operation lifetime of nuclear
reactors. The aim of this study was to investigate the DPA (displacement per atom) rate, an
important parameter describing radiation damage to the RPV, and identify the position of the
maximum DPA rate in the RPV of the VVER-1000/V320 reactor using the Monte Carlo code
MCNP5. To reduce statistical errors in the MCNP5 simulation, the weight window technique
was applied to non-repeated structures outside the reactor core. The results showed the
distribution of the DPA rate in the RPV and the maximum DPA rate was found to be at the first
millimeters of the RPV. Consequently, these calculations could be useful for assessment of
radiation damage to the RPV of VVER reactors.

Keywords: VVER, reactor pressure vessel embrittlement, DPA rate, weight window technique

I. INTRODUCTION

During the operation of nuclear power plants (NPPs), assessment of radiation embrittlement
of the structure materials and reactor pressure vessels (RPVs) by neutron and gamma is one of the
most important issues to ensure their integrity. In particular, it is widely recognized that the service
lifetime of an RPV is limited by neutron irradiation embrittlement [1].
As of 2014, there have been more than 100 serious nuclear accidents and incidents from the
use of NPPs, including the Three Mile Island (1979), Chernobyl (1986), and Fukushima Daiichi
(2011) accidents. The RPV acts as a barrier that keeps radioactive fuel contained and out of the
environment, and therefore ensuring the integrity of the RPV during normal operation of NPPs or
under accident conditions is indispensable. To this end, investigating the displacement per atom
(DPA) rate in the RPV, which is a key parameter describing radiation embrittlement of the RPV,
has received much attention so far [2]-[4].
As published by the OECD/NEA state-of-the-art report in 1996 [2], the introduction of DPA
to represent the metal damaging effects of neutrons at all neutron energy levels was presented.
Besides, the reconsideration of the computation techniques for calculating neutron/gamma radiation
damage to RPV and the methods used in the NEA member countries for computing long-term
cumulative dose rates were also reported. The report disclosed that the results of neutron/gamma
fluence and radiation doses were within 20 percent difference when compared between calculations
and measurements or calculations with different computer codes. Another report of Boehmer et al
[3] showed the results such as the neutron/gamma spectra, several fluence integrals, and the DPA
and freely migrating defect (FMD) rates of ex-core components of Russian (VVER-1000) and
German light water reactors (1300 MW PWR and 900 MW BWR). Nonetheless, the neutron
fluence and DPA ditributions at the RPV have not been shown. Recently, the calculation of DPA in
the RPV of the Argentinian Atucha II reactor (PHWR type) [4] was performed using the Monte
1
Carlo code MCNP, determining the areas at the RPV where the neutron fluence and DPA rate are
maximum. However, application of variance reduction techniques (VRTs) to reduce statistical
errors and computational time for such neutron deep penetration calculation with MCNP has not
been mentioned.
In this paper, we investigate the DPA distributions on RPV of a Russian pressurized water
reactor, the VVER-1000/V320 [5], using the Monte Carlo code MCNP5 [6], thereby identifying the
maximum radiation exposure areas in the RPV. The purpose is to establish the method for DPA
calculations for evaluating radiation damage to the RPV of the VVER-1000/V392. In the MCNP5
simulation, the weight window VRT was applied to non-repeated structures outside the reactor core,
leading to a significant decrease of statistical errors in the neutron fluence and DPA calculations. As
a result,the maximum neutron fluence and DPA rate were found at the first millimeters of the RPV
areas that are nearest to the peripheral fuel assemblies.

II. CALCULATION METHODOLOGY

The VVER-1000 reactor core consists of 163 fuel assemblies (FA). Each FA has 312 fuel
rods and 18 guiding channels. The main characteristics of VVER reactor core and FA parameters
are described in Table I and Table II, respectively. Detailed description of the reactor core materials
can be found in [5].

Table I: A briefinformation of VVER-1000/V320

Parameter Value
Reactor type VVER-1000
Version V320
Nominal power, MWt 3000
Nominal electric power, MWe 1000
Coolant inlet temperature, 0C 288
Number of fuel assemblies, pcs 163
Effective core radius, mm 1580
Pressure vessel inner radius, mm (without 7mm of 2075
cladding thickness)
Pressure vessel outer radius, mm 2267.5

Table II: Fuel assembly (FA) description

Parameter Value
FA pitch, mm 236
FA wrench size, mm 234
FA gap, mm 2
Number of fuel rods, pcs 312
Fuel pin pitch, mm 12.75
Fuel pin grid triangular
Fuel pin
Cladding:

2
 Material Zirconium alloy (Zr+1%Nb)
Density, g/cm3 6.52
Outer diameter, mm 9.1
Wall thickness, mm 0.65
Pellet:
Material UO2
Density, g/cm3 10.22
Outer diameter, mm 7.55
Center hole diameter, mm 2.4
Height of UO2, mm 3550
Mass of UO2, g 1460

The MCNP5 input file for VVER-1000/V320 reactor core modelled the fuel assemblies as
repeated structures up to the steel baffle, while the regions outside the core from the baffle to the
RPV (see Fig. 1) were simulated as non-repeated structures. The full core model in MCNP5 for
VVER-1000/V320 was described in Fig. 2.
The nuclear data for this calculation were taken from the ENDF/B-VII.1 library. To calculate
the neutron fluence on the RPV of the VVER-1000/V320 reactor, the FMESH tally card was
utilized in the MCNP5 calculation. The FMESH card calculates the track length estimate of particle
flux, averaged over a mesh cell, in units of particles/cm2. This card can be used for the calculation
of flux distributions, power peaking factor and power distributions. The neutron fluences calculated
by the MCNP5 code were plotted using the "pcolor" graphics module of the Matlab-like open-
source Scilab [7]. The formulae for calculating the neutron flux and DPA rate from the FMESH
tally results are described as follows.
The neutron flux can be determined using the following equation.

Pcore (W) . ν 1 particle

Φ (E ) = . . ϕF , (1)
1.6022. 10 .Q k cm

where Q is the energy release in one fission, Pcore the thermal power of the reactor,  the
average number of neutrons emitted in one fission, and ϕF is the fluence obtained by FMESH
in neutron energy group i.
To calculate DPA (displacement per atom), which is the number of times an atom is displaced
from the normal lattice by interaction with neutrons, the DPA cross-section for iron was used [8]
and the following formula was applied.

R ≅ σ Φ(E )dE = σ .ϕ , (2)

3
where σ is the DPA microscopic cross-section, ϕ is the neutron flux in the i group (obtained from
Eq. (1)), and N the number of neutron energy groups (N= 213 in this case).
Finally, the DPA rate can be calculated as follows.

R
DPA = , (3)
n

where n is the number of atoms.

The statistical errors for the FMESH tally results were found as high as 0.1 without applying
any VRT (with a huge number of neutron history of 109). To reduce the statistical errors and
computational time in the MCNP5 calculation, the weight window generator, which outputs the
reciprocal of the average score (importance) generated by particles entering a given phase-space
region and helps correct poor track distributions [9], was applied in this study for the regions
outside the reactor core (non-repeated structures).
The areas on the RPV inner surface where the neutron fluence is highest were identified at the
core mid plane. Then the average DPA rate in the RPV thickness at the core mid plane was
calculated to determine the position at which the DPA rate reaches maximum. The DPA spectrum
was also evaluated to figure out contributions to the DPA rate from each neutron energy group. The
calculation results are presented in the following Section.

Fig. 1 VVER-1000/V320 core in 600 symmetry

4
 Fig. 2 The VVER-1000/V320 full core model in MCNP5

III. CALCULATION RESULTS

To identify the maximum neutron fluence in the RPV, the neutron fluence at the inner
surface of the RPV was calculated and investigated depending on the azimuthal angle (θ) and the
reactor core axial position (z). The long distance from the core center to the RPV outer surface of
226.75 cm (the thickness of RPV is 19.25 cm) requires application of advanced VRTs to reduce
statistical errors in the neutron fluence calculation; otherwise, analog calculations without any
VRTs for such a neutron deep penetration problem could lead to unreliable results even with a huge
number of neutron history.
Specifically, the weight window generator was not produced for repeated structures, because
the geometry splitting uses the product of the importance at different levels [6]. However, in our
MCNP5 simulation, we used both repeated structures (reactor core) and non-repeated structures
(regions outside the reactor core). Thus, it is possible to apply the weight window technique for the
regions outside the reactor core in this study. First, we performed the analog calculation to produce
the average score generated by particles entering a given phase-space region for all regions
including fuel assemblies and the regions outside the reactor core. Second, the weight window
lower bounds of the RPV cladding were observed and the weight window factors for the F4 tally
region (the whole RPV) were determined. Table. 3 illustrated the neutron fluence calculation results
for the whole RPV region in which using the weight windows significantly reduced the statistical
error from 0.0682 to 0.0028.

5
 Table III. The F4 tally results for the whole RPV region with and without weight windows technique
(nps: total number of neutron histories, FOM: figure of merit)
No weight windows Weight windows
nps mean error FOM nps mean error FOM
1024000 1.3140E-10 0.6321 3.6E-01 1024000 1.1405E-10 0.0540 9.0E-01
2048000 1.2170E-10 0.2186 6.4E-02 2048000 1.3746E-10 0.0088 7.1E-01
3072000 1.3742E-10 0.1400 8.2E-02 3072000 1.3931E-10 0.0062 7.2E-01
4096000 1.1784E-10 0.1207 7.4E-02 4096000 1.3954E-10 0.0051 7.1E-01
5120000 1.1846E-10 0.1057 7.3E-02 5120000 1.3755E-10 0.0044 7.2E-01
6144000 1.2638E-10 0.1003 6..5E-02 6144000 1.3782 E-10 0.0039 7.2E-01
7168000 1.3375E-10 0.0881 7.0E-02 7168000 1.3810 E-10 0.0036 7.2E-01
8192000 1.2626E-10 0.0826 6.9E-02 8192000 1.3779 E-10 0.0033 7.2E-01
9216000 1.2582E-10 0.0761 7.1E-02 9216000 1.3736 E-10 0.0031 7.2E-01
10240000 1.2432E-10 0.0712 7.2E-02 10240000 1.3734 E-10 0.0029 7.2E-01
10997019 1.2432E-10 0.0682 7.3E-02 10999762 1.3713 E-10 0.0028 7.2E-01

The FMESH tally was then applied to determine the neutron fluence and distribution of
DPA rate in the RPV using the weight window technique. In this case, the neutron number history
of 107 was chosen and the relative error of the FMESH tally results was found as low as less than
0.035. It is noted that we used a fine mesh for the FMESH tally (Δr, Δz, and Δθ = 0.5 cm, 35.3 cm,
and 10 respectively) to obtain the distribution of DPA rate in the RPV; while the case in Table III
used the F4 tally for the whole RPV region. As the FMESH tally was used, the relative error was as
high as 0.1 without using the weight window technique.
Fig. 3 showed the neutron fluence, ( , ), at the inner surface of the RPV (inner radius of
the RPV = 207.5 cm). As it was expected, the maxima of the neutron fluence were found at the
positions close to the azimuthal angles where the distance between the RPV and the peripheral fuel
assemblies was shortest. The peaks of the neutron fluence were found at z = 176.5cm (core mid-
plane) and =70, = 530, =670, =1130, =1270, =1730, = 1870, =2330,
=2470, =2930, =3070, =3530. It can be seen that each peak was repeated every 60°
due to the one-sixth symmetry of the core. Also, the neutron fluence were symmetric with respect to
the core mid-plane, mainly caused by the use of uniform coolant and fuel temperatures along the
core axial direction in the MCNP5 calculation.

6
Fig. 3 The neutron fluence at the inner surface of the RPV (1/cm2)

Fig. 4 The DPA rate at the RPV on the core mid-plane (s-1)
7
 Fig. 4 displayed the DPA rate at the RPV on the mid-plane of the core (outer radius of the
RPV = 226.75 cm). It was found that the maxima of the DPA rate appeared at the same azimuthal
positions with the peaks of the neutron fluence. In this case, the DPA was linearly dependent on the
neutron fluences, because only one neutron energy group was used for calculation of the DPA rate
(see Eq. (2)). In addition, the maximum neutron fluence and DPA rate were identified at the first
millimeters of the RPV. The contribution of each neutron energy group to the DPA rate will be
examined and presented below.

Fig. 5 The neutron flux spectra at the barrel and RPV

Fig.5 represented the neutron flux spectra at the steel barrel (r =181 cm), the inner surface of
the RPV (r =207.5 cm), the 1/4 thickness of the RPV (r =212.31 cm), and the outer surface of the
RPV (r =226.75 cm). It can be seen that the neutron spectrum was hardened as neutrons penetrated
from the steel barrel into the RPV. The highest spectrum was at the steel barrel (before the down-
comer region) and the lowest was identified at the outer surface of RPV. It can be explained by the
presence of the down-comer region where the neutrons were slowed down and partially absorbed by
the boric acid in the water.

Fig. 6 The DPA rate at the inner surface, the 1/4T thickness and the outer surface of the RPV

8
 Combining the neutron flux and DPA cross-section [9], the DPA rate distribution was
calculated following the Eqs. (2) - (3). As shown in Fig. 6, the DPA rate in each energy group is
plotted as a function of neutron energy at the inner surface of the RPV, 1/4 thickness of the RPV,
and the outer surface of the RPV. The contributions of thermal neutrons to the DPA rate at the inner
surface of the RPV and 1/4 thickness of the RPV were higher than that at the outer surface of the
RPV. This difference was reduced in the intermediate and fast energy ranges.

Table IV The neutron flux and DPA rate for inner surface and 1/4 thickness of the RPV
Energy Neutron fluence (1/cm2) DPA rate (s-1)
group
(MeV) Inner surface % 1/4Thickness % Inner surface % 1/4Thickness %
0 to 4e-7 1.00E-09 57.8 1.13E-10 17.4 1.08E-10 1.9 1.11E-11 0.2
4e-7 to 0.1 3.31E-10 19.1 2.06E-10 31.8 2.30E-10 4.2 1.94E-10 5.0
0.1 to 1 2.53E-10 14.6 2.37E-10 36.5 1.75E-09 31.4 1.63E-09 41.8
1 to 20 1.48E-10 8.5 9.23E-11 14.3 3.49E-09 62.5 2.06E-09 53.0
Total 1.73E-09 100 6.47E-10 100 5.57E-09 100 3.90E-09 100

The neutron fluence and DPA rate contributed from the four commonly used energy groups
(thermal, epithermal, intermediate and fast neutron energies) for the inner surface and 1/4 thickness
of the RPV were presented in Table IV. As shown in this Table, significant contributions to the
DPA rate on the inner surface of the RPV were from the fast neutrons (62.5% of the total DPA rate)
and the intermediate neutrons (31.4% of the total DPA rate). These contributions from fast and
intermediate neutrons correspond to their fraction of 23.1% of the total flux while the contribution
from thermal and epithermal neutron groups (76.9% of the total flux) is small (only 6.1% of the
total DPA rate). The same results were found at the 1/4 thickness of the RPV. However, the
contribution from the fast neutrons to the DPA rate was decreased about 10% while that of the
intermediate neutrons was increased about 10% as compared with the case at the inner surface.

IV. CONCLUSION
In this study, we performed the calculation of the neutron fluence and DPA rate on the RPV
of the VVER-1000/V392 with the Monte Carlo code MCNP5. The neutron fluence and DPA rate at
different positions in the RPV were investigated to figure out the position at which these quantities
are maximum. The main results were summarized as follows:
 The weight window technique was applied to reduce statistical errors in the MCNP5
calculations. By using this VRT, the relative error of the FMESH tally results was
reduced from 0.1 to an acceptable value of 0.035.
 The maxima of the neutron fluence and DPA rate were found at the same positions at the
core mid-plane, which are close to the peripheral fuel assemblies.
 These maxima were identified at the first millimeters of the RPV. The DPA rate versus
neutron energy was investigated in difference positions of the RPV including its inner
surface, 1/4 thickness and the outer surface. It was found that the rate of DPA decreased
when the neutron penetrated through the RPV. The results also showed that the main
contribution to the DPA rate came from intermediate and fast neutron energy groups
(93.9% at the inner surface of the RPV and 94.8% at 1/4 thickness of the RPV).

9
 In future work, several VRTs will be applied together to further reduce the above-mentioned
statistical error of the FMESH tally results. Additionally, verification calculation by using another
nuclear code is also being planned along with using different nuclear data libraries.

REFERENCE

1. ODETTE, G., R., LUCAS, G., E. Embrittlement of Nuclear Reactor Pressure Vessels: JOM journal,
No. 7, 2001, p. 18-22

2. OECD/NEA State-of-the-art Report, "Computing Radiation Dose to Reactor Pressure Vessel and
Internals," NEA/NSC/DOC (96)5, 1996.

3. B. Boehmer, J. Konheiser, K. Noack, A. Rogov, G.Borodkin, E. Polke, P. Vladimirov, “Neutron and

gamma fluence and radiation damage parameters of ex-corecomponents of Russian and German light
water reactors”. Proceedings of the 11th International Symposium on Reactor Dosimetry, 18-23
August 2002 in Brussels, Belgium. World Scientific Publishing Co. ISBN #9789812705563; 2003,
286-294.

4. J. A.Mascitti and M. Madariaga,"Method for the Calculation of DPA in the Reactor Pressure Vessel
of Atucha II," Science and Technology of Nuclear Installations, Volume 2011, Article ID 534689,
2011.

5. G.Borodkin, B.Boehmer, K.Noack, and N.Khrennikov. "Balakovo-3 VVER-1000 EX-vessel neutron

dosimetry benchmark experiment," Forschungszentrum Rossendorfe.V, Moscow - Dresden,
November 2002.

6. X-5 Monte Carlo Team, MCNP5 - A General Monte Carlo N-Particle Transport Code - Volume I, II,
III, Version 5,Los Alamos National Laboratory Report LA-UR-03-1987, Apirl 24, 2003.

7. S.L. Campbell, J.P. Chancelier, and R. Nikoukhah, Modeling and Simulation in Scilab/Scicos,
Springer, 2000.

8. https://www-nds.iaea.org/CRPdpa/

9. A Sample Problem for Variance Reduction in MCNP, Thomas Booth Los Alamos National Lab.
Report: LA-10363-MS, 1985.

10. https://www.nndc.bnl.gov/sigma/getPlot.jsp?evalid=15019&mf=3&mt=1&nsub=10

10
TÍNH TOÁN TỐC ĐỘ DPA LỚN NHẤT CỦA THÔNG LƯỢNG NƠTRON
VÀ GAMMA TRÊN VỎ THÙNG LÒ VVER

NGUYỄN HỮU TIỆP1, PHẠM NHƯ VIỆT HÀ1, NGUYỄN MINH TUÂN2
1
Viện Khoa học và Kỹ thuật Hạt nhân
179 - Hoàng Quốc Việt, Nghĩa Đô, Cầu Giấy, Hà Nội, Việt Nam
2
Viện Nghiên cứu hạt nhân Đà Lạt
01 – Nguyên Tử Lực, Đà Lạt, Lâm Đồng, Việt Nam
Email: [email protected]

Tóm tắt: Sự giòn hóa vật liệu kim loại khi chịu bức xạ năng lượng cao là một trong những hệ
số quan trọng liên quan tới giới hạn tuổi thọ của vỏ thùng lò trong nhà máy điện hạt nhân, đó
cũng là hiệu ứng ảnh hưởng quan trọng nhất trong quá trình lão hóa của RPV. Sự giòn hóa này
là do RPV chịu sự chiếu xạ của nơtron và gamma, đặc biệt là nơtron nhanh trong suốt quá trình
vận hành lò phản ứng hạt nhân. Trong bài báo này chúng tôi xác định được vùng mà bị chiếu xạ
nhiều nhất trên vỏ thùng lò phản ứng VVER-1000/V320, dựa vào đó chúng tôi xác định được
tốc độ DPA lớn nhất do thông lượng nơtron và gamma gây ra trên vỏ thùng lò phản ứng. Ngoài
ra, phổ nơtron và gamma cũng như giá trị thông lượng tích phân của nơtron và gamma tại nhiều
vị trí khác nhau cũng được khảo sát và đánh giá.

Từ khóa: DPA, VVER-1000/V320, MCNP5, RPV, Vỏ thùng lò, Sự giòn hóa, Phân bố thông
lượng nơtron.

11