Plasma Science and Technology, Vol.5, No.2, Feb. 20 Experimental Studies on the Self-Shielding Effect in Fissile Fuel Breeding Measurement in Thorium Oxide Pellets Irradiated with 4 MeV Neutrons Mitul ABHANGI, Nupur JAIN 2, Rajnikant MAKWANA, Sudhirsinh VALA, Shrichand JAKHAR, T. K. BASU, and C. V. S. RAO Fusion Neutronics Laboratory, Institute for Plasma Research, Bhat, Gandhinagar-82428, Gujarat, India 2 School of Nuclear Energy, Pandit Deendayal Petroleum University, Gandhinagar, Gujarat 82007, India Abstract The 4 MeV neutrons produced in the D-T fusion reactions have the potential of breeding Uranium-2 fissile fuel from fertile material Thorium-22. In order to estimate the amount of U-2 produced, experiments are carried out by irradiating thorium dioxide pellets with neutrons produced from a 4 MeV neutron generator. The objective of the present work is to measure the reaction rates of 22 Th + n 2 Th 2 Pa 2 U in different pellet thicknesses to study the self-shielding effects and adopt a procedure for correction. An appropriate assembly consisting of high-density polyethylene is designed and fabricated to slow down the high-energy neutrons, in which Thorium pellets are irradiated. The amount of fissile fuel ( 2 U) produced is estimated by measuring the 2 kev gammas emitted by Protactinium-2 (half-life of 27 days). A calibrated High Purity Germanium (HPGe) detector is used to measure the gamma ray spectrum. The amount of 2 U produced by Th 22 (n, γ) is calculated using MCNP code. The self-shielding effect is evaluated by calculating the reaction rates for different foil thickness. MCNP calculation results are compared with the experimental values and appropriate correction factors are estimated for self-shielding of neutrons and absorption of gamma rays. Keywords: 2 U breeding, fissile fuel, MCNP, 4 MeV neutron source PACS: 28.4.Bm DOI: 0.088/009-060/5/2/8 Introduction The strategy for nuclear power engineering development in the first half of the 2st century is based on the following principles: nuclear fuel breeding [,2], comprehensive safety, and competitiveness. Nuclear fuel breeding is a process of producing new fissile material from material which was previously non-fissile. One of the most important candidates in view of this mechanism is the Th-22 chain. Although not fissile itself, Th- 22 will absorb slow neutrons to produce uranium-2, which is fissile. The irradiated fuel can then be unloaded from the reactor, the U-2 separated from the thorium, and fed back into another reactor as part of a closed fuel cycle. Alternatively, U-2 can be bred from thorium in a blanket, the U-2 separated, and then fed into the core. In this experiment the thorium dioxide pellets are irradiated using 4 MeV neutrons from a neutron generator which has a yield of 0 0 n/s []. Upon irradiation Th-22 captures a neutron and forms radioactive nuclei Th-2. Thorium-2 then undergoes two consecutive beta-decays to form Uranium-2, the fissile isotope. The emitted 2 kev gamma, from the produced radioactive nuclei, are counted and analysed with a high resolution gamma-ray spectrometer. To eliminate the sample induced effects during the analysis of gamma spectrum, the sample used in neutron activation must be very thin. If the sample is made thin then the activity produced in the foil after irradiation may not be sufficient to produce significant counts. A thicker foil is thus preferred to undergo the irradiation, which in turn gives rise to sample induced effects. The sample induced effects discussed in this paper are neutron self-shielding and gamma ray attenuation within the pellet. Monte Carlo calculations are performed to analyze the neutron self-shielding and gamma ray attenuation on the counting of fissile fuel breeding. 2 Experiment particulars Thorium dioxide pellets are irradiated with 4 MeV neutrons. The cross-section for (n, γ) reaction of Thorium-22 is considerably high ( 7.7 barns) for
Mitul ABHANGI et al.: Experimental Studies on the Self-Shielding Effect in Fissile Fuel Breeding thermal neutrons (energy < ev) and is unsteady in epithermal and fast regions. To slow down the 4 MeV neutrons, a moderator assembly made up of high density poly-ethylene (HDPE) is designed and constructed shown in Fig.. The MCNP calculations are performed to estimate the thermal flux along the circumference of the neutron generator which is found to be 4.25E+06 ± 0.0 n/cm2 /s. Thermal flux at the four pellet locations is also calculated and tabulated in Table. Fig. locations (Table 2). The sample pellets and reference foils are irradiated for two hours in the irradiation facility. After the irradiation is complete, the foils and samples are counted using the gamma-ray spectroscopy system [4]. The HDPE assembly housing the neutron generator Table. The thermal flux at all the pellet locations calculated by MCNP code Thickness Thermal flux (n/cm2 /s) 5 mm 4.5078E+06 ± 0.0 4 mm 4.295E+06 ± 0.0 mm 4.2257E+06 ± 0.0 2 mm 4.255E+06 ± 0.0 Fig.2 The arrangement of sample pellets (a) and reference foils around the neutron generator (b) To study the effect of self-shielding, different thickness pellets are irradiated and analyzed. The sample pellets of different thickness viz. 2 mm, mm, 4 mm, 5 mm are arranged around the circumference of the neutron generator 90 degrees apart from each other. To make an estimate about the nature of flux at the pellet locations, reference foils are also kept beside the sample pellets as shown in Fig. 2. By using appropriate reference foils and selecting proper reaction, inference can be drawn about the nature of the thermal flux and fast flux at the sample Table 2. Before irradiating the samples, their gamma-ray spectrum is taken using HPGe. Gamma rays from the unirradiated sample are then used to identify the presence of other radionuclides in the sample. Most of the gamma ray energies were found to be originating from the radionuclide produced in the natural decay chain of Thorium-22. As the half life of Th-22 is.405e+0 years, the amount of other radionuclide is considered negligible in further calculation and analysis. The characteristics and reaction of interest of the reference foils Gold Reaction Threshold (MeV) Analysis 97 Zirconium 98 Au(n, γ) Au 90 89 Zr(n, 2n) Zr Copper 6 Cu(n, 2n)62 Cu 0 >0 σ (barn) 92. 0.56 0.49 Half-life (hour) 64.56 78.4 9.74 min 0.545 0.697 4.8 908.96 5 Abundance Eγ (kev) 67
Plasma Science and Technology, Vol.5, No.2, Feb. 20. Experimental estimation To estimate the number of atoms of fissile fuel produced, gamma rays of energy 2 kev are counted, which are emitted by Protactinium-2 during its decay to Uranium-2. The decay scheme of Protactinium2 is shown in Fig.. The gamma ray of energy 2 kev can be observed in the decay [5] of Protactinium-2. detector window, the reaction rate is proportional to the counts per unit mass of the sample. After irradiation, the thorium dioxide pellets are counted at a distance of 2 cm from the detector window for 800 s. The counts at the peak of 2 kev are recorded and tabulated in Table. As all the pellets are exposed to the same thermal flux, the counts/g or counts/target atom must be equal in all the pellets (ideal case). But it is found that the counts from the 5 mm thick pellet are only 52% of the counts from the 2 mm pellet. Similarly, counts from the 4 mm and mm pellets are 6% and 8% of the counts from the 2 mm pellet respectively. The sample pellets are also counted at a distance of 0 cm from the detector window to account for the gamma rays escaping because of the edge effect. The counts recorded are tabulated in Table 4..2 Fig. The decay scheme of Pa-2 Reaction rate and the peak counts hold the following relation [4]. Reaction rate (N σφ) = Peak count λ Molecular mass, NA maiγ εω(e λtcool e λtcount ) () Here, λ - Decay constant Ω - Solid angle subtended at the detector by the sample ε - Efficiency of detector at x cm from the detector m - Mass of sample a - Abundance of the target radionuclide NA - Avogadro s number At a distance x from the detector window, the solid angle and efficiency are constant. Avagadro s number, intensity of gamma ray, molecular mass, decay constant and isotopic abundance are also constant for a radionuclide. Hence, if the different samples of the same material are kept at the same distance x cm from the Table. Thickness (mm) 5 4 2 Table 4. Thickness (mm) 5 4 2 68 Study of the pellet-induced effects Thermal neutron self-shielding within large samples is studied using the Monte Carlo neutron transport code MCNP [6]. The code enabled a three-dimensional modelling of the actual source and geometry configuration including the neutron generator, HDPE assembly and thorium dioxide samples shown in Fig. 4. Fig.4 The HDPE assembly and arrangement of pellets around the neutron generator as modeled in MCNP The counts per second recorded from the sample pellet at a distance of 2 cm Mass (g) 6.057 4.774.8750 2.49 Counts (cps) of Eγ 298 562 246 9092 Counts/g 2027 2440 4 89 Ratio (normalized to 2 mm) 0.52 0.6 0.8 The counts per second recorded from the sample pellet at a distance of 0 cm Counts 88 ±.79% 84 ±.96% 70 ±.77% 279 ± 4.5% Counts/g 645 74 869 08 Ratio with respect to 2 mm 0.64 0.70 0.854
Mitul ABHANGI et al.: Experimental Studies on the Self-Shielding Effect in Fissile Fuel Breeding The same geometry is also modelled by keeping void material in place of the ThO 2 pellet. Table 5 shows the result of simulation of self-shielding. It is observed that in the 5 mm and 2 mm pellets, the ratio of reaction rate for (n, γ) reaction in ThO 2 to the reaction rate for (n, γ) reaction in void material is a mere 0.94 and 0.905 respectively. We can conclude the effect of self-shielding [7,8] of neutrons is not reflected in the MCNP calculation. Also the effect of ThO 2 material is not markedly significant. After irradiation the pellets are counted in the gamma ray spectrometer. The gamma rays of energy 2.7 kev emitted by Protactinium-2 are counted to determine the amount of Uranium-2 (fissile material) bred. Before reaching the detector gamma rays gets attenuated within the pellet and hence, the amount of fissile material bred is also underestimated. Since attenuation of emitted low-energy gamma radiation in voluminous bulk samples is an obstruction for determining Pa-2 (2.7 kev, I γ 8.6%) quantitatively by means of gamma-spectroscopy, self-attenuation [7,8] correction must be taken into account. Thorium dioxide pellets of different thicknesses viz. 5 mm, 4 mm, mm and 2 mm are modelled in the MCNP code. The gamma ray source is defined inside the pellet with energy 2.7 kev. The geometry shown in Fig. 5 is simulated with ThO 2 material being filled in the pellet and the intensity of gamma ray on the surface of pellet facing the detector (surface ) is calculated. This gives the intensity of gamma ray (I) which has seen a thickness x in the ThO 2 material. The geometry is then simulated with air being filled in the pellet and the intensity of gamma ray on the same surface of the pellet is calculated. This indicates the intensity of gamma ray (I 0 ) which has seen no thickness or zero thickness. Table 6 shows the gamma ray intensity calculated to demonstrate the effect of material and thickness. The results demonstrate that in a 5 mm pellet the intensity of gamma ray after going through the pellet thickness is 5% of the source intensity of gamma ray. In the 4 mm, mm and 2 mm pellets the intensity observed after going through the pellet thickness is 0.55, 0.6 and 0.68 respectively. As the thickness seen by gamma ray increases the ratio (I/I 0 ) decreases. The gamma ray intensities going through different pellet thickness are also compared in the table. The gamma ray intensity, I, in a 5 mm pellet is 70% of the gamma ray intensity in a 2 mm pellet. The remaining 0% is attenuated inside the pellet because of the 50% increase in the thickness. Similar attenuation in the 4 mm and mm pellets is observed as normalized to the 2 mm pellet. Fig.5 Thorium dioxide pellet and HPGe detector geometry modeled in MCNP The results obtained from MCNP calculations depict that there is considerable attenuation of gamma ray within the pellet. When the radionuclide is estimated by counting the number of gamma rays coming out of the sample, this effect of gamma ray attenuation within the pellet is very significant. To avoid underestimation of the activity of the radionuclide, a correction factor is mandatory. 4 Results In Table 7 the reaction rate for the reaction 22 Th (n, γ) 2 Th is calculated using MCNP and theoretical calculation. Table 5. The reaction rate calculated by MCNP code to demonstrate the effect of material and thickness Thickness Reaction rate Reaction rate Ratio Ratio (mm) (ThO 2 material) (void material) (ThO 2/void) (normalized to 2 mm) 5.2E+04 ± 0.02.07E+04 ± 0.08 0.94.00 4.89E+04 ± 0.02.08E+04 ± 0.09 0.909 0.974.25E+04 ± 0.02.7E+04 ± 0.022 0.854 0.922 2.22E+04 ± 0.02.49E+04 ± 0.020 0.905 Table 6. The gamma ray intensity calculated to demonstrate the effect of material and thickness Thickness I γ with ThO 2 I γ with air Ratio Ratio I γ- ThO 2 (mm) (I) (I 0) (I/I 0) (normalized to 2 mm) 5 0.85295 ± 0.0009 0.678 ± 0.0005 0.5 0.629 4 0.2226 ± 0.0004 0.8428 ± 0.000 0.55 0.80 0.2476 ± 0.0005 0.40644 ± 0.0004 0.6 0.892 2 0.29448 ± 0.0004 0.4884 ± 0.0002 0.68 69
Plasma Science and Technology, Vol.5, No.2, Feb. 20 Table 7. The comparison of the number of fissile atoms produced in each of the pellets calculated manually and using MCNP code Thickness (mm) MCNP atoms/g Manual calculation Nσφ/g % error 5.22089E+04.6927E+04 4.22 4.252E+04.7674E+04 6.0.8942E+04.5984E+04 2.48 2.22E+04.847E+04.94 The results from MCNP and theoretical calculation are matching at par with a maximum deviation of 6%. This represents that the experiment is modeled appropriately in MCNP and the theoretically calculated results match with MCNP results. Table 8 shows the ratio of gamma ray intensity of x mm thick pellet normalized to a 2 mm thick pellet. The MCNP calculation and the experiment results are matching closely as shown in Fig. 6. to the 2 mm pellet is found to be 42%, 7% and 29% respectively. Table 8. The comparison of the gamma ray attenuation effect results obtained by MCNP calculation and Experiment Distance MCNP (Normalized to 2 mm) (mm) (Normalized to 2 mm) Experiment 5.7 0.629 0.64 4.25 0.80 0.70.5 0.892 0.854 2.02 Fig.7 The variation of correction factor with respect to the thickness of the pellet This reduction in the number of counts can be attributed to self-shielding effects in the pellet. As found from MCNP calculation, the effect of neutron selfshielding is not prominent. The majority of thermal fluxes at all pellet locations make sure that the neutrons do not encounter the resonance peaks and hence the self-shielding effect is absent. The influence of gamma attenuation within the pellet is seen to be significant and thus needs correction. References Fig.6 Comparison of calculated & experimental results This shows that the MCNP model represents the experimental scenario. The correction factor for the gamma ray self-attenuation is calculated and presented in Table 9 and graphically shown in Fig. 7. In the real estimates of the radionuclide produced this correction factor has to be considered. Table 9. The correction factor for each of the pellets Thickness (mm) Correction factor 5.7.96 4.25.82.5.64 2.02.47 5 Conclusion The counts/g of the gamma ray recorded from each of the pellets must ideally be equal. But experiments show that they differ by as large as 52%. The attenuation in the 5 mm, 4 mm and mm pellets as compared 70 Bethe Hans A. 979, Physics Today, 2: 44 2 Ma X B, Chen Y X, Wang Y, et al. 200, Fusion Engineering and Design, 85: 2227 Lee J D and Moir R W. 98, Journal of Fusion Energy, : 299 4 Gerhard Erdtmann and Hermann petri. Neutron activation analysis: Fundamentals and techniques. Second edition, part I, volume 4. John Wiley & Sons 5 Firestone Richard B. 999, Table of Isotopes, Volume II. Eighth edition. Wiley-Interscience 6 Shultis J K and Faw R E. Mcnp Primer. 20, Dept. of Mechanical and Nuclear Engineering, Kansas State University, Manhattan 7 Knoll Glenn F. 2000, Radiation Detection and Measurement. Third edition. John Wiley & Sons, New York 8 Tsoulfanidis Nicholas. 995, Measurement and detection of radiation. Second edition. Taylor andfrancis, Dallas, TX, USA (Manuscript received 6 January 202) (Manuscript accepted 24 July 202) E-mail address of corresonding author S. VALA: sudhir@ipr.res.in