Radiation Protection Dosimetry (2014), Vol. 161, No. 1–4, pp. 176 –180 Advance Access publication 1 July 2014

doi:10.1093/rpd/ncu195

STUDY OF NEUTRON SCATTERING CONTRIBUTION ON Hp(10) AND H*(10) CALIBRATION IN THE BRAZILIAN NATIONAL LOW SCATTERING LABORATORY B. M. Freitas, W. W. Pereira, K. C. S. Patra˜o, E. S. Fonseca and C. L. P. Mauricio* Instituto de Radioprotec¸a˜o e Dosimetria, Av. Salvador Allende, s/n, Recreio, Rio de Janeiro, CEP: 22783-127 RJ, Brazil

The neutron scattering at the Low Scattering Laboratory of the Brazilian National Neutron Laboratory has been studied using three different methods. The measurements have been done with a traceable standard 241Am–Be from source-to-detector distances of 0.52 –3.00 m. The obtained results with the variation distance methods are in agreement. Measurements with a large shadow cone are not worth for larger distances due to overshadowing. As the quantity required in a calibration is the response of the device being calibrated to the scattered neutron component in order to subtract this from the total response, for these purposes, the distance variation method must be used for each device. To quantify absolutely the scattering contribution on the quantity rates of fluence, Hp(10) and H*(10) in irradiation procedures, a Bonner sphere spectrometer with the shadow cone was employed. The evaluated scattering correction factor value may be employed for a distance of 1.00 m.

INTRODUCTION The Neutron Laboratory (LN) of the Instituto de Radioprotec¸a˜o e Dosimetria was created in 1973 as one of the laboratories that form the Brazilian National Metrology Laboratory of Ionizing Radiation (LNMRI). LNMRI is part of the International Metrology System. Besides primary standardisation of neutron sources, LN also calibrates and irradiates dosemeters or any kind of sample. LN is the Brazilian reference laboratory in the field of neutron metrology, being responsible for dissemination of the following neutron quantities: fluence, emission rate and ambient and personal dose equivalent rate. Its traceability is obtained through a manganese sulphate bath that provides the source emission corrected for encapsulation and geometric factors. LN participates in the key comparisons sponsored by the Bureau International des Poids et Me´sures (BIPM). Since 13 September 2013, BIPM Calibration and Measurement Capabilities (CMC) database states, for LN, a relative expanded uncertainty (for k ¼ 2, level of confidence 95 %) of 8.5 % for calibration with a standard 241Am –Be radionuclide source at a distance of 1.00 m. This uncertainty is valid in the range of 1.1`  1026 to 4.4`  1024 Sv h21 for ambient dose equivalent H*(10) and 1.2`  1026 to 4.7 `  1024 Sv h21 for personal dose equivalent Hp(10). Calibrations are typically performed under lowscatter conditions. For this purpose the LN comprises a Low Scattering Laboratory (LSL). The LSL is a building of 7 m`  18 m and 6 m high. The floor is made of concrete and the walls of galvanised steel TM plates in layers: steel-foam-steel. and Styrofoam

The ceiling is made of aluminium tiles with foam lining. Above the floor level there is an elevated platform (irradiation platform) made of aluminium at a height 2 m. The LSL dimensions are much larger than the minimum room lengths recommended by the International Organization for Standardization (ISO) 8529-2(1). However, even under these conditions, scattering cannot be avoided. The neutron spectra at the detector/sample position are always affected by walls, floor, roof, air and any other structures around the experimental set-up(2). At LN, for a source –detector distance of 1.00 m, the scattering contribution is included in the uncertainty, but no correction is made. Besides, the LSL has passed through a complete reform, which modified the set-up, changing the scattering. The aim of this work is to study the neutron scattering contribution for the quantities fluence and H*(10) and Hp(10) rates at the new set-up of the LSL in order to diminish LN uncertainties in calibration and irradiation. METHODOLOGY The study was carried out with a 241Am–Be radionuclide neutron source (cylinder of 3.5 cm diameter by LNMRI. It has an ` 3.5 cm height) standardised emission rate of 1.06 `  107 n s21, with a relative expanded uncertainty of 1.3 %. Three different methods were used for the evaluation of the neutron scattering contribution: shadow cone and two distance variation methods. The Bonner sphere spectrometer (BSS) employed in this work consists of six spheres of polyethylene (0.95 g cm23) with a thermal neutron detector at its centre.

# The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

Downloaded from http://rpd.oxfordjournals.org/ at Northeastern University Libraries on October 22, 2014

*Corresponding author: [email protected]

NEUTRON SCATTERING CONTRIBUTION

Spheres of 2, 3, 5, 8, 10 and 1200 diameters were used. Counts without any sphere have also been made. The thermal neutron detector employed was an 6LiI(Eu) cylindrical scintillator with 96 % 6Li enrichment (4 mm diameter ` 4 mm height) connected to a photomultiplier via light pipe. The scintillator pulses are registered by a Canberra MCA and analysed by the Genie-2000w spectroscopic software. It allows discrimination of photons and electronic noise. Shadow cone method

MD ðlÞ ¼ ½MT ðlÞ  MS ðlÞ  FA ðlÞ;

ð1Þ

where FA (l ) is the coefficient of correction for air attenuation (air outscattering) which can be calculated as described in Annex C of ISO 8529-2(1) standard. Its values are 1.01 and 1.02 for l values of 1.0 and 2.25 m, respectively. The shadow cone used in this work follows the recommendations of ISO 8529-2(1). It consists of two parts: the first one is 20 cm long, made entirely of iron and the second one, 30 cm long, made of polyethylene with 5 % boron. The shadow cone used has a front face diameter of 4.1 cm and a rear face diameter of 41 cm (a cone angle of 40.58). Its shadow cone angle is very large, given overshadow in the BSS. The recommendation is to use at least two different sizes of cones(2), but only one was available for this work. Distance variation methods Another technique used to determine the neutron scattering contribution employs distance variation measurements. It is implemented in an attempt to obtain the fraction of the scattered neutrons for various positions. The start point for these methods is the assumptions that the neutron scattering is composed of a constant plus a component that varies with the inverse of the distance(3).

MT ðlÞ ¼

k þ SðlÞ; l2

ð2Þ

where k is a source –detector characteristic constant (corrected reading multiplied by l2) and S(l) is the scatter contribution (scattering reading divided by direct reading). In this method, it is assumed that the air scatter is negligible compared with the total reading of the detector and that no geometrical correction is necessary(3). MT (l ) is plotted as a function of distance l and the parameters k and S (l ) are determined by weighted least-square fitting of Equation (2). S (l ) is described as a linear function of l. The generalised fit method has also been applied to the measurement data. This method assumes that the instrument reading, MT (l ), as a function of distance l, due to the total radiation field can be represented by a second-order polynomial:   MT ðlÞ F1 ðlÞ ¼ RF þ xl þ yl 2 ; F FA ðlÞ

ð3Þ

where F is the free-field fluence (direct fluence), RF is the detector free-field fluence response, F1 (l ) is the geometric correction for size of the source and the detector, and x and y are the adjustment parameters. Experimental set-up For comparison between the several scattering methods used in this work, only the 800 Bonner sphere around the 6LiI(Eu) scintillator detector was used to measure the fractional scatter contribution. This detector, named here 800 detector, was positioned along the axis of the centre of the source and its centre. It has been exposed to neutrons from a standardised 241 Am –Be neutron source. The source was positioned approximately in the centre of the laboratory and the detector moved horizontally at the same height. For application of the distance variation methods, the measurements were performed at 11 points from

177

Downloaded from http://rpd.oxfordjournals.org/ at Northeastern University Libraries on October 22, 2014

The shadow cone method evaluates experimentally the contribution of the scattered neutrons. When the shadow cone is interposed between the source and the detector, it captures any direct neutron emitted by the source, so that only the scattered contribution reaches the detector. Direct neutrons are those that arrive at the detector without undergoing any interaction. Without the shadow cone, the detector sees the direct plus the scattered neutron. Therefore, for a distance l between the centre of the source and the centre of the detector, the amount of direct neutrons measured, MD (l ), is given by the difference between the counts measured without the interposed shadow cone (total contribution), MT (l ), and the counts measured with the interposed shadow cone (scattering contribution), MS (l ), multiplied by a correction factor, FA (l ):

It is based on the assumption that the radiation field at a position at distance l from the source consists of three components: the source neutrons (decreasing with l 22), air-scattered neutrons (decreasing with l 21) and wall/floor-scattered neutrons (independent of l ). The measured readings as a function of distance are analysed by mathematical model fitting(4). The evaluations were performed by two methods: reduced-fitting method and generalised fit method, both recommended by ISO 8529-2(1). For the reduced fitting method, the instrument reading MT (l ) due to the total radiation field (source neutrons plus scattered neutrons) can be represented by Equation (2), as a function of distance l:

B. M. FREITAS ET AL.

Neutron spectra evaluation The spectra were obtained using the software Bonner sphere Unfolding Made Simple (BUMS)(5) with the input of the neutron counts rate measured with BSS. It has an interface in HTML and is built on the framework of the BUNKI code developed by Johnson and Lowry at the Naval Research Laboratory. BUMS software provides access to a wide array of unfolding algorithms, response matrices, starting spectra, dose–response functions and detector response functions. In this work, a 241Am –Be ISO starting spectrum, SPUNIT unfolding method and UTA4 matrix response were employed. The UTA4 matrix response was chosen because it has been calculated for the same type of neutron detector and of polyethylene spheres densities used in this work. The unfolded spectra were calibrated, measuring the global efficiency of the 6LiI(Eu) scintillation detector using spectra measurements for reference distances. This procedure for experimental assessment of the neutron spectra was validated through comparison of the unfolded 241Am –Be direct neutron spectra with the ISO reference spectra for a 241Am–Be source. The direct spectrum was adjusted using the difference of the counts obtained without and with the shadow cone.

methods and with the shadow cone method (1.00 and 2.25 m distances) using the 800 detector. The uncertainties reported in those graphs are only from statistical sources, and are mainly associated with the fitting procedure in the variation distance methods and with the statistical counts of the 800 detector in the shadow cone method. The results for reduced-fitting and generalised fit methods agree between them. Using the shadow cone method, the scattering fraction measured at a source –detector distance of 1.00 m is similar to the one obtained by the variation distance methods, if all the sources of uncertainty were taken into account. However, at a distance of 2.25 m, the value of the measured scattering fraction is much lower than the one evaluated by the variation distance methods. This is due to the overshadowing caused by the high dimensions of the shadow cone used, as the overshadowing increases with the increasing of the source– detector distance. Then, this method cannot be used for source –detector distances .1.00 m at this geometry of source –detector. Figures 2 and 3 show, respectively, for 1.00 and 2.25 m, the unfolded neutron spectra obtained with (scattering spectrum) and without the shadow cone (total spectrum) and the one adjusted by the difference of their counts (direct spectrum). The direct 241 Am –Be neutron spectra have also been plotted as the difference between the neutron spectra obtained without and with the shadow cone, and they are statistically equal. Observing these spectra, one can see that the neutron spectrum of the source without the shadow cone (direct plus scattered neutrons) and the spectrum of the direct neutrons are more similar at 1.00 m. At 2.25 m, as the scattering increases, the difference is more noticeable. The obtained values of fluence-averaged neutron energy EF , fluence to ambient dose equivalent conversion coefficient h*(10) and fluence to personal dose equivalent conversion coefficient hp(10) for the direct 241Am– Be spectrum at both tested distances

RESULTS Figure 1 shows the results of the ratio of inscattered neutrons to direct neutrons at various source– detector distances obtained from the distance variation

Figure 1. Results of scattering fractions obtained by different methods with the 800 detector in the LSL of LN.

178

Downloaded from http://rpd.oxfordjournals.org/ at Northeastern University Libraries on October 22, 2014

source-to-detector distances of 0.52– 3.00 m, with a statistical uncertainty of 0.6 % for smaller distances and 2 % for the greater ones. The shadow cone method has been used for the source–detector distances of 1.00 and 2.25 m. These distances were chosen because they are the most often used in LN procedures. The statistical uncertainty of these counts was ,1 %. A distance of 4 cm from the source to the front face of the shadow cone was used as optimum separation distance. As all neutron measurements depend on the detector in order to quantify absolutely the scattering contribution on the quantities fluence and ambient and personal dose equivalent rate, the BSS was employed. This is necessary for the dissemination of these quantities in irradiation procedures of LN. All spheres were positioned like the 800 detector and exposed separately to the same 241Am–Be source at source– detector distances of 1.00 and 2.25 m. These exposures were performed with and without the shadow cone, in the same way described above. The counting time was always set in order to give ,1 % statistical uncertainty.

NEUTRON SCATTERING CONTRIBUTION Table 1. Evaluated and reference values for 241Am– Be.

ISO-8529(6) 1.00 m 2.25 m

EF (MeV)

h*(10) (pSv cm2)

hp(10) (pSv cm2)

4.16 3.54 3.47

391 390+23 388+29

411 408+24 406+30

Table 2. Scattering factor for 241Am– Be at LSR/LN. Scattering factor Fluence Figure 2. Total, scattering and direct spectra at a source– detector distance of 1.00 m in the LSL of LN.

1.00 2.25

Hp(10)

H*(10)

1.09+0.09 1.07+0.01 1.08+0.01 1.27+0.15 1.19+0.01 1.20+0.01

irradiation at LN. At 2.25 m, the values measured in this work underestimated the necessary scattering correction and has to be recalculated. CONCLUSION

Figure 3. Total, scattering and direct spectra at a source– detector distance of 2.25 m in the LSL of LN.

are compared in Table 1 with ISO reference values(6). Discrepancies in fluence-averaged neutron energy are in the order of 15 %, which is acceptable. The calculated EF value is lower due to the use of only one large shadow cone. This spectrum shifted to lower energies, when compared with the reference spectrum, as had already been reported by Mirzajani et al.(2) for neutron spectrum obtained applying only one shadow cone, with overshadow. They show that it is necessary at least to have two shadow cones for the measurement with BSS. Nevertheless, the values of the h*(10) and hp(10) agree with the reference values, within their uncertainties. Using the BSS and the shadow cone method, the scattering contribution in the quantities fluence, Hp(10) and H*(10) rates for 241Am– Be at the LSL of LN has been estimated for both distances. The values of the scattering correction factor calculated are presented in Table 2. The value obtained at 1.00 m can be used for correction of the theoretical values of these operational quantity rates for all device

Comparison of the results of scattering contribution obtained in this work with distance variation methods for the 800 detector show good agreement from 0.52 to 3.00 m. Then, for calibration of any detector at the LSL of LN, the distance variation method has to be used for each of them. Using the shadow cone method, the agreement of the results is poor at the distance of 2.25 m because of the overshadowing. The too large shadow cone reduces significantly the neutron scattering measured by the BSS behind it. Therefore, for the studied configuration and distances, the correct values for fluence, Hp(10) and H*(10) quantity rates using this shadow cone may be employed only for a source – detector distance of 1.00 m. The use of this measured scattering correction factor may contribute to the reduction of the uncertainty of the operational quantity rate values at 1.00 m, stated in the CMC of LN. For evaluation of the scattering contribution in the neutron spectra for higher distances, it is necessary to use other sizes of shadow cones. FUNDING B.M.F. received financial support from Capes Brazilian Agency. REFERENCES 1. International Organization for Standardization. Reference neutron radiations. Calibration fundamentals of radiation

179

Downloaded from http://rpd.oxfordjournals.org/ at Northeastern University Libraries on October 22, 2014

Source –detector distance (m)

B. M. FREITAS ET AL. protection devices related to the basic quantities characterizing the radiation field. ISO 8529, Part 2 (2000). 2. Mirzajani, N., Ciolini, R. and Curzio, G. Analysis of the application of the shadow cone technique for the determination of neutron spectrum with Bonner sphere spectrometer. Nucl. Instrum. Methods A 722, 24– 28 (2013). 3. Gressier, V. and Taylor, G. C. Calibration of neutronsensitive devices. Metrologia 48, S313– S327 (2011).

4. Schuhmacher, H. Neutron calibration facilities. Radiat. Prot. Dosim. 110, 33–42 (2004). 5. Sweezy, J., Hertel, N. and Veinot, K. BUMS—Bonner sphere unfolding made simple: an HTML based multisphere neutron spectrometer unfolding package. Nucl. Instrum. Methods A 476, 263–269 (2002). 6. International Organization for Standardization. Reference neutron radiations. Characteristics and methods of production. ISO 8529, Part 1 (2000).

Downloaded from http://rpd.oxfordjournals.org/ at Northeastern University Libraries on October 22, 2014

180

Study of neutron scattering contribution on Hp(10) and H*(10) calibration in the Brazilian National Low Scattering Laboratory.

The neutron scattering at the Low Scattering Laboratory of the Brazilian National Neutron Laboratory has been studied using three different methods. T...
152KB Sizes 4 Downloads 3 Views