Assessment of shutdown dose rates at the ESS target cooling system using SCALE6.2

Abstract

The production of high-energy neutrons at the European Spallation Source through the spallation process may cause an erosion of the tungsten target. The eroded particles could be released into the target helium cooling system which contains four kind of filters. Among them, the auxiliary filters called “getters” are designed to capture volatile elements and remaining dust. In this work, the ORNL’s SCALE6.2 modelling and simulation suite for nuclear safety analysis is applied to assess shutdown dose rates and determine if added shielding and/or robotic arms are needed for their maintenance. SCALE6.2 is well adapted to treat this problem as it allows for isotope selection regarding source term calculation. Dose rates are determined by an ORIGEN2 source term and a MAVRIC shielding sequence calculation. As SCALE6.2 is non-standard software for ESS, the results are verified against MCNP, which is the baseline tool for neutronics analysis at ESS. Dose rate calculations show that additional shielding and/or robot arm are not needed to remove the getters after 3 months of cooling time, following 5400 h of operation at 5 MW beam power. At a distance of 1 mm from the getter, the dose rate is 0.2 mSv/h in the most conservative estimation.

Keywords

European Spallation Source getters gamma dose rate SCALE6.2 MCNP6.2

1. Introduction

The European Spallation Source (ESS), now under construction in Lund, Sweden, will be the world’s most intense neutron source in term of brightness and most advanced neutron scattering facility. In order to produce high-energy neutrons through the spallation process, a 2 GeV proton beam will interact with a rotating tungsten target. To dissipate the heat generated in the target wheel, the Target Station incorporates a helium-based cooling system capable of sustaining a beam power of 5 MW [2].

The beam-target interaction may cause an erosion of the target and the release of radionuclides into the helium cooling system which is shown in Fig. 1.

Fig. 1.

3D schematic of ESS target wheel and its helium cooling system.

The total mass of tungsten contained in the target wheel is 3.1 tonnes and during 5400 hours of system operation at 5 MW beam power it is assumed that up to 10 g of this tungsten may be eroded [5]. The cooling system contains a system of filters intended to capture larger particles from the target wheel, while auxiliary filters named “getters” are used to catch remaining gases and dust. Regularly, the getters have to be replaced. Only one of the two getters present in the helium purification system will be exchange at a time. Since a getter exchange is planned every 6–12 months, a decay time of at least 3 months could be scheduled before getter exchange. In order to determine if shielding or robotic equipment are needed for this operation, gamma shutdown dose rates are determined. The SCALE6.2 simulation suite developed by ORNL is chosen for these calculations because it allows for isotope selection [10].

As the baseline tool for neutronics analysis at ESS is MCNP [4], this study is used to validate the non-standard software SCALE6.2 in the context of this ESS application.

In Section 2 the system of filters within the helium cooling system and the anticipated radionuclide content are described. The methodology applied using SCALE6.2 is described and results presented in Section 3. Finally, SCALE6.2 results are benchmarked against MCNP6.2 in the case of a pure ¹²³I photon source.

2. ESS target cooling system

The ESS target helium cooling system contains the following four kinds of filters as shown in Fig. 2:

The main filter is intended to catch small particulates;

A mesh filter removes large particles that may come from the turbo compressors. It is installed to avoid these particles from reaching the target.

HEPA filters installed in the helium purification system protect the getters from receiving the smallest particulates. As a side effect, it also cleans the main loop from these small particulates;

Finally, the auxiliary filters called “getters” or “absorbers” are used to capture the remaining volatile elements and gases, in particular tritium.

Fig. 2.

Schematics of the ESS helium cooling system [3].

Within the system, the helium mass flow is 3 kg/s. The pressure is 1 MPa and the inlet/outlet temperatures are 230°C and 40°C, respectively [12].

The getter is a MC14K-902 F type filter provided by the SAES Pure Gas Group. Table 1 presents the properties of the getter as provided by the manufacturer.

Table 1

Characteristics of getters given by SAES Pure Gas Group [9]

Gases purified	N2, Ar, He, Kr, Ne, Xe, CH₄, C₂H₆, C₃H₈, C₄H₁₀, SF₆, fluocarbons
Impurities removed	H₂O, O₂, CO₂, H₂ to < 100 ppt 1 ; volatile acids, organics, refractory compounds to < 1 ppt; volatile bases < 5 ppt, metals < 1 ppb (volume) 2
Particle filtration	2 micron or 0.003 micron metal

parts per trillion

parts per billion by volume

Each absorber, which is considered a consumable, has a capacity corresponding to 6 months of operation at the nominal beam power of 5 MW [3].

2.1. Radionuclide content

It is assumed that:

90% of the particles released from the target into the helium loop are of size larger than 5 μm;

100% of the large particles ( $> 5 μ m$ ) are caught by the main loop filters;

The remaining 10% of smaller particles ( $< 5 μ m$ ) are caught by the HEPA filter installed in front of the getters;

There is no filter which is able to catch the noble gases, not even the getters.

The Helium filters do not have a fraction capture of 100% at each passage, but, since the total helium volume is filtered at least 5 times each minute no large particles will remain in the helium cooling system. Therefore the purification rate becomes practically 100%.

For the calculation of the shut-down dose rates at the getters, conservative hypotheses are assumed:

The noble gases are caught by the getters;

100% of the small particulates ( $size < 5 μ m$ ) are caught by the getters (not by HEPA).

The radionuclide content in the target cooling system and the corresponding activities after operation are determined by MCNP and CINDER’90 activation calculations [8].

3. Gamma dose rate assessment using SCALE6.2

SCALE6.2 (Standardized Computer Analyses for Licencing Evaluation) is the latest version of the modelling and simulation suite for nuclear safety analysis and design developed by ORNL [10]. This computational package is used for criticality-safety purposes, reactor physics, radiation shielding calculations, source term characterization and sensitivity-uncertainty studies.

SCALE6.2 can compute gamma dose rate maps by the shielding sequence MAVRIC while this sequence needs a source term as input [10]. This source term can be calculated with ORIGEN2. Fig. 3 illustrates this workflow.

Fig. 3.

Dose rate map calculation using SCALE6.2.

3.1. Gamma source term calculation using ORIGEN2

In order to perform the source term calculation with ORIGEN2, the first task is to select the isotopes expected to reach the getters. In the context of this study, the radionuclides of interest are those which are gamma emitters. The 33 radionuclides with the highest activity are presented in Table 2 [8].

Table 2
Isotopes expected to reach the getters and corresponding activity after shutdown

Isotope Activity (Bq) in getter after 5400 h of irradiation

¹¹C 7.45e+02

¹³N 8.41e+03

¹⁶N 1.54e+08

¹⁵O 5.58e+02

¹⁸F 2.93e+07

⁴¹Ar 2.34e+08

⁶⁷Ga 1.72e+08

⁷⁵Se 2.51e+08

^81mKr 2.86e+08

⁸⁵Kr 2.20e+06

^85mKr 6.76e+07

⁸⁷Kr 1.46e+07

⁸⁸Kr 1.77e+00

⁸⁹Kr 1.83e+07

¹⁰³Ru 1.28e+08

¹¹¹In 8.38e+08

¹²²Sb 1.22e+08

¹²⁴Sb 2.88e+05

¹²⁵Sb 9.18e+03

^123mTe 4.86e+07

¹²⁷Te 2.98e+06

¹²⁹Te 9.86e+04

¹³²Te 1.81e−04

¹²³I 3.28e+09

¹²⁴I 4.19e+08

¹²⁵I 5.62e+09

¹²⁹I 4.22e−07

¹³¹I 1.46e+05

^131mXe 1.30e+08

¹³³Xe 8.56e+06

^133mXe 1.56e+06

¹³⁵Xe 2.53e+02

^135mXe 1.01e+02

Isotope	Activity (Bq) in getter after 5400 h of irradiation
¹¹C	7.45e+02
¹³N	8.41e+03
¹⁶N	1.54e+08
¹⁵O	5.58e+02
¹⁸F	2.93e+07
⁴¹Ar	2.34e+08
⁶⁷Ga	1.72e+08
⁷⁵Se	2.51e+08
^81mKr	2.86e+08
⁸⁵Kr	2.20e+06
^85mKr	6.76e+07
⁸⁷Kr	1.46e+07
⁸⁸Kr	1.77e+00
⁸⁹Kr	1.83e+07
¹⁰³Ru	1.28e+08
¹¹¹In	8.38e+08
¹²²Sb	1.22e+08
¹²⁴Sb	2.88e+05
¹²⁵Sb	9.18e+03
^123mTe	4.86e+07
¹²⁷Te	2.98e+06
¹²⁹Te	9.86e+04
¹³²Te	1.81e−04
¹²³I	3.28e+09
¹²⁴I	4.19e+08
¹²⁵I	5.62e+09
¹²⁹I	4.22e−07
¹³¹I	1.46e+05
^131mXe	1.30e+08
¹³³Xe	8.56e+06
^133mXe	1.56e+06
¹³⁵Xe	2.53e+02
^135mXe	1.01e+02

Due to a lack of decay data in the origen.rev04.endf7dec library, ¹¹C, ¹⁵O, ¹⁸F are not taken into account in this ORIGEN2 calculation; still 99.8% of the total activity just after operation is included in the dose calculation.

In the calculation 27 photon energy bins ranging from $10^{- 11} MeV$ to 20 MeV are used for the source. Four cooling times are considered for the dose calculation: 1 day (minimum cooling time before technical maintenance), 1 month, 3 months, and 6 months.

The source terms are stored from the ORIGEN2 calculation in a .f71 binary file. In order to take into account the decay history and the correct normalization of the source in MAVRIC for each cooling time, the ORIGEN2 calculation is split in different cases corresponding to each considered cooling time.

3.2. Dose rate maps using the MAVRIC shielding sequence

MAVRIC (MONACO with Automated Variance Reduction using Importance Calculation) is based on the CADIS (Consistent Adjoint Driven Importance Sampling) and FW-CADIS methodologies [7,13,14]. The sequence automatically performs a three-dimensional, discrete ordinates calculation using the DENOVO 3D discrete-ordinates code to determine the adjoint flux as a function of position and energy. The adjoint flux is then used by MAVRIC to construct a space and energy-dependent importance map (target weights for weight windows) and a mesh-based biased source distribution. MAVRIC passes the importance map and biased source distribution to the fixed-source Monte-Carlo radiation transport code MONACO to complete the particle transport problem [7]. The workflow of the sequence is presented in Fig. 4.

Fig. 4.

MAVRIC shielding sequence workflow.

The getter is mainly composed of porous zirconium alloy surrounded by a thin coating of stainless steel. The composition of the zirconium alloy is not available so this material is assumed for the SCALE6.2 modelling to be a standard cladding zirconium material available in the material library, however, with a low density of 1.29 g/cm³ to account for the porosity.

As the available description of the getter is not more precise that the one described in Table 3, the modelled source from ORIGEN2 is isotropic and uniformly distributed through the getter. The filter is surrounded by a $4 m \times 4 m \times 4 m$ box filled with air. Table 3 summarizes the modelling hypotheses of the MAVRIC calculation and the getter dimensions.

Table 3

SCALE6.2 modelling hypotheses

Filter Materials	Porous Zr alloy density 1.29 g/cm³
Filter Materials	Coating SS316
Filter Geometry	Outer diameter 152.4 mm
	Vessel thickness 2.77 mm
	Length 1190 mm
Source Description	Homogeneously distributed in filter
Dose calculation	Dose rate in air
Flux-to-Dose factors	ANSI-1991 [1]

The DENOVO calculation will use the coarse-group shielding library (27n19g) for all of the importance map calculations. In the calculation, 10000 batches of 100000 photons each were simulated. Using these parameters, the total calculation time using 1 processor on a laptop is 8 hours. Cross-sections libraries based on ENDF-BVII.V0 were used.

The gamma dose rate is evaluated at two distances from the getter:

1 mm, corresponding to the case of a worker removing the filter wearing gloves;

10 cm, corresponding to the case of a worker removing the getter using handles.

After 1 day of coling time, the dose rate is $2.73 mSv/h \pm 0.02 mSv/h$ at a distance of 1 mm from the getter, decreasing to $1.08 mSv/h \pm 8 μ Sv/h$ at a distance of 10 cm from the getter as illustrated on the dose rate map presented in Fig. 5. The getter is marked as a black circle in the figure.

Fig. 5.

Gamma dose rate map in conservative case after 1 day of cooling time.

The mesh tally used to assess the dose rate only considers the air region which is surrounding the getter. Hence, due to the optimization process that focused the MONACO Monte-Carlo calculation on dose rates in the air region, values of the dose rate inside the getter are underestimated and should not be considered [10]. This calculation assumption leads to a lower and uniform dose rate inside the getter. After 6 months of cooling time, the dose rate values decrease to $0.09 mSv/h \pm 0.0008 mSv/h$ and $0.04 mSv/h \pm 0.0004 mSv/h$ at a distance of 1 mm and 10 cm from the getter, respectively, as shown in Table 4.

Table 4

SCALE6.2 gamma dose rate results for the conservative case (HEPA not catching smallest particulates)

Distance from the getter (cm)	Gamma Dose Rate (mSv/h)

	Cooling Time	1 day	1 month	3 months	6 months
0.1		2.73 ± 0.02	0.4 ± 0.003	0.2 ± 0.002	0.09 ± 0.0008
10		1.08 ± 0.008	0.1 ± 0.002	0.07 ± 0.0008	0.04 ± 0.0004

In Table 5, these gamma dose rates can be compared to the values obtained in a case considering that the getter is not catching the smallest particulates which are trapped by the HEPA filter before reaching it (normal condition case).

Table 5

SCALE6.2 gamma dose rate results for normal conditions case

Distance from the getter (cm)	Gamma Dose Rate (mSv/h)

	Cooling Time	1 day	1 month	3 months	6 months
0.1		1.63 ± 0.01	0.06 ± 0.002	0.03 ± 0.0005	0.009 ± 0.00032
10		0.65 ± 0.005	0.03 ± 0.0005	0.01 ± 0.0003	0.004 ± 0.00009

Note that the dose rate values in this case are halved after 1 day of cooling time. The remaining dose rate is due to iodine isotopes which are the major contributors to the dose rate and which are intended to be only catched by getters.

The dose limit for exposed workers is given by Swedish regulations to be 20 mSv/year [11].

With respect of the ALARA principle, ESS is committed to limit exposure even further than the legal limits. In this context, a dose constraint of 2 mSv/year in planned operations should be considered for exposed workers [6]. When planning work, this dose limit should be considered.

Considering a decay time of 3 months before getter exchange, in the conservative case, the dose rates will respectively decrease to $0.2 mSv/h \pm 0.002 mSv/h$ at 1 mm from the getter and $0.07 mSv/h \pm 0.0008 mSv/h$ at 10 cm. Changing a getter is taking only 5 minutes so these results lead to the conclusion that there is no need of additional shielding or robotic equipment for the maintenance of the getters if 3 months of cooling time after irradiation are considered to limit the exposure [6].

4. SCALE6.2-MCNP6.2 code-to-code comparison

As the Monte-Carlo transport code MCNP is the baseline tool for neutronics assessment at ESS, this code is used to benchmark the non-standard software SCALE6.2. In order to perform this verification, only one of the released isotopes is considered for the source term.

¹²⁵I is the major contributor to the total activity just after shutdown, as shown in Table 6, but ¹²³I is interesting after shut-down because of its half-life (13.2 h) and its main 158.97 keV gamma ray (abundance 83.3%) that can be modelled for the comparison. This isotope is selected as the source term.

Table 6
Contribution to the total activity of the selected isotopes in function of cooling time

Nuclide % of total gamma source term

Shut-down 1 day cooling time 6 months cooling time

¹³N 0 0 0

¹⁶N 3.4 0 0

⁴¹Ar 1.65 0 0

⁶⁷Ga 1.19 1.40 0

⁷⁵Se 2.81 4.06 15.63

^81mKr 1.16 0 0

⁸⁵Kr 0 0 0.04

^85mKr 0.37 0.01 0

⁸⁷Kr 0.17 0 0

⁸⁸Kr 0 0 0

⁸⁹Kr 0.30 0 0

¹⁰³Ru 0.64 1.02 0.45

¹¹¹In 10.96 12.45 0

¹²²Sb 0.81 0.91 0

¹²⁴Sb 0 0 0.01

¹²⁵Sb 0 0 0

^123mTe 0.33 0.84 0.55

¹²⁷Te 0 0 0

¹²⁹Te 0 0 0

¹³²Te 0 0 0

¹²³I 28.65 11.84 0

¹²⁴I 4.20 5.17 0

¹²⁵ I 43.17 62.05 82.5

¹²⁹I 0 0 0

¹³¹I 0 0 0

^131mXe 0.40 0.54 0

¹³³Xe 0.04 0.05 0

^133mXe 0.01 0.01 0

¹³⁵Xe 0 0 0

^135mXe 0 0 0

Nuclide	% of total gamma source term
¹³N	0	0	0
¹⁶N	3.4	0	0
⁴¹Ar	1.65	0	0
⁶⁷Ga	1.19	1.40	0
⁷⁵Se	2.81	4.06	15.63
^81mKr	1.16	0	0
⁸⁵Kr	0	0	0.04
^85mKr	0.37	0.01	0
⁸⁷Kr	0.17	0	0
⁸⁸Kr	0	0	0
⁸⁹Kr	0.30	0	0
¹⁰³Ru	0.64	1.02	0.45
¹¹¹In	10.96	12.45	0
¹²²Sb	0.81	0.91	0
¹²⁴Sb	0	0	0.01
¹²⁵Sb	0	0	0
^123mTe	0.33	0.84	0.55
¹²⁷Te	0	0	0
¹²⁹Te	0	0	0
¹³²Te	0	0	0
¹²³I	28.65	11.84	0
¹²⁴I	4.20	5.17	0
¹²⁵ I	43.17	62.05	82.5
¹²⁹I	0	0	0
¹³¹I	0	0	0
^131mXe	0.40	0.54	0
¹³³Xe	0.04	0.05	0
^133mXe	0.01	0.01	0
¹³⁵Xe	0	0	0
^135mXe	0	0	0

A total of $10^{9}$ photons are simulated using MCNP6.2. The mcplib84 cross-section data library is used for photons transport. A multiplier is used in both cases to get the corresponding activity value at time 0. The effect of using continuous energy distribution (CE) instead of multi-group energy distribution (MG) in the MAVRIC Monte-Carlo transport simulation is evaluated. To reproduce the same type of physics used in the continuous-energy SCALE calculation, the MCNP simulation uses the following options: photon cutoff energy of 0.01 MeV, $EMCPF = 100 MeV$ (detailed physics), $IDES = 1$ (no electrons/no bremsstrahlung), $NOCOH = 0$ (coherent scattering occurs), $ISPN = 0$ (no photonuclear collisions), and $NODOP = 1$ (no Doppler energy broadening). Flux-to-dose factors are the same in the two modellings and comes from ANSI-1991[1]. Mesh tallies are used both in SCALE6.2 and MCNP6.2. The comparison of the SCALE6.2 results with those obtained using MCNP6.2 shows an agreement between the two simulations, see Table 7. In particular, SCALE6.2 results using continuous energy transport are included in the MCNP6.2 uncertainty range.

Table 7

Dose rates from ¹²³I calculated using SCALE6.2 and MCNP6.2

Distance from the getter (cm)	SCALE6.2 MG Dose rate (μSv/h)	SCALE6.2 CE Dose Rate (μSv/h)	MCNP6.2 Dose rate (μSv/h)
0.1	204 ± 0.3	327 ± 1.4	329 ± 5
10	81.36 ± 0.1	129 ± 0.3	131 ± 6

5. Conclusion and perspectives

The getters are auxiliary filters placed in the helium purification system of the ESS target, intended to capture gases, volatile acids, organics and refractory compounds. They have to be replaced after a minimum of 24 hours of cooling time after every 6 months of operation. This operation only takes 5 minutes.

The SCALE6.2 simulation suite is well suited to assess gamma dose rates in this context as it allows for isotope selection for source term calculation. The gamma source term is determined in an ORIGEN2 decay calculation which is exported to the MAVRIC shielding sequence for dose rate calculation. Under conservative assumptions, considering the legal dose limit of 20 mSv/year and the ESS dose constraint of 2 mSv/year for exposed workers, the calculations show that there is no need for added shielding or robotic equipment to remove the used getters if maintenance is planned after 3 months of cooling time. Indeed, after 3 months of cooling time, the dose rate is $0.2 mSv/h \pm 0.002 mSv/h$ at 1 mm from the getter and $0.07 mSv/h \pm 0.0008 mSv/h$ at 10 cm.

As MCNP is the reference tool at ESS for neutronics analysis, the SCALE6.2 results are compared to MCNP in the context of a pure ¹²³I gamma source. This comparison shows the agreement between the results, especially when using continuous energy distribution in the MAVRIC transport calculation. In that case, this result was expected due to similarity in method. However, the ratio of MCNP/SCALE results is not planned to be adopted as “safety factor” in dose rates estimates at the moment.

In order to examine the sensitivity of the shutdown dose rates to the assumed inventory of radionuclides reaching the getter, a case considering that the HEPA filter is catching the smallest particulates is compared to the more conservative case, thereby reducing the shutdown dose rates next to the getters by around 40%. However, other parameters are also subject to uncertainty; in particular the amount of tungsten particles released from the target by the erosion process is rather uncertain.

Footnotes

Acknowledgements

The authors gratefully acknowledge Jens Harborn, Thomas M. Miller and Sigrid Kozielski from ESS for respectively providing technical details on the helium cooling system operation and characteristics, advice regarding SCALE6 options and practical elements on radiation protection purpose.

References

ANSI/ANSI-6.1.1-1991, Neutron and Gamma ray flux-to-dose-rate factors, American Nuclear Society, La Grange Park, Illinois, 1991.

Åberg

et al., ESS technical design report, 2013.

Emås, SDD-Sol System 1015 – Target primary cooling purification system, ESS-0051446 technical report, 2019.

J.T.

Goorley

et al., Initial MCNP6 release overview – MCNP6 version 1.0, LA-UR-13-22934, 2013.

Haines, ESS target facility design, in: Proc. of Int. Conf. IPAC, Richmond, VA, USA, May 3rd–8th, 2015.

Hansson, General safety objectives, ESS-000004 technical report, 2020.

D.E.

Peplow, Monte-Carlo shielding analysis capabilities with MAVRIC, Nuclear Technology174 (2011), 289–313. doi:10.13182/NT174-289.

Radionuclides content in Target cooling system – Dose to worker at maintenance, ESS-0061660 Technical report.

SAES Pure Gas, Ambient Line Purifier, 902 Purification media specification.

10.

SCALE code system, ORNL/TM-2005/59, Version 6.2.3, 2018.

11.

Swedish Radiation Safety Authority, SSMFS 2008:51.

12.

Target primary cooling loop circulator – Technical description, ESS-0051820, 2016.

13.

Wagner

et al., Automated variance reduction applied to nuclear well-logging problems, Nuclear Science and Engineering168 (2009), 799–809. doi:10.13182/nt09-a9309.

14.

Wagner

et al., FW-CADIS method for global and regional variance reduction of Monte-Carlo radiation transport calculations, Nuclear Science and Engineering176 (2014), 37–57. doi:10.13182/nse12-33.

Nuclide	% of total gamma source term

	Shut-down	1 day cooling time	6 months cooling time
¹³N	0	0	0
¹⁶N	3.4	0	0
⁴¹Ar	1.65	0	0
⁶⁷Ga	1.19	1.40	0
⁷⁵Se	2.81	4.06	15.63
^81mKr	1.16	0	0
⁸⁵Kr	0	0	0.04
^85mKr	0.37	0.01	0
⁸⁷Kr	0.17	0	0
⁸⁸Kr	0	0	0
⁸⁹Kr	0.30	0	0
¹⁰³Ru	0.64	1.02	0.45
¹¹¹In	10.96	12.45	0
¹²²Sb	0.81	0.91	0
¹²⁴Sb	0	0	0.01
¹²⁵Sb	0	0	0
^123mTe	0.33	0.84	0.55
¹²⁷Te	0	0	0
¹²⁹Te	0	0	0
¹³²Te	0	0	0
¹²³I	28.65	11.84	0
¹²⁴I	4.20	5.17	0
¹²⁵ I	43.17	62.05	82.5
¹²⁹I	0	0	0
¹³¹I	0	0	0
^131mXe	0.40	0.54	0
¹³³Xe	0.04	0.05	0
^133mXe	0.01	0.01	0
¹³⁵Xe	0	0	0
^135mXe	0	0	0