Activation of Air and Concrete in Medical
Isotope Production Cyclotron Facilities
CRPA 2016, Toronto
Adam Dodd
Senior Project Officer
Accelerators and Class II Prescribed Equipment Division
(613) 993-7930 or [email protected]
nuclearsafety.gc.ca
Outline
•
•
•
•
•
Motivation
Methods
Air activation
Concrete activation
Conclusions
Canadian Nuclear Safety Commission
2
Motivation
• Neutrons produced in cyclotron target via (p,Xn)
– can activate materials in vault
– air activation can pose a radiological hazard to workers
– concrete activation can become a disposal problem
• Why the renewed CNSC interest?
– cyclotrons have become much more powerful
– old days ~ 60 µA and 2 hour runs for F-18 FDG production
– now could have 500 µA and 6 hour runs for Tc-99m production
– 25 times more activation!
Canadian Nuclear Safety Commission
3
Methods (1/2)
Monte Carlo simulations
– MCNP5 (1% statistics)
– proton beams 18 & 24 MeV
– simplified vault geometry
– F-18 and Tc-99m neutron source spectra – assumed
isotropic
– explore sensitivity of results to
• changes in vault design
• source energy
• source location
• polyethylene (with/without boron) shielding around
targets
Canadian Nuclear Safety Commission
4
Methods (2/2)
Collaboration with UOIT co-op students
– Rob Shackelton (air activation)
– Devon Carr (concrete activation)
– Audrie Ismail (concrete activation)
Canadian Nuclear Safety Commission
5
Air activation
• Nitrogen 78%of air
– N-14(n,p)C-14 (78% of air)
– C-14: 5,730 year half-life, soft β – little radiological consequence
• Oxygen 21% of air
– O-16(n,p)N-16 (21% of air, 10 MeV threshold)
– N-16: 7s half life, 6 MeV γ – little radiological consequence
• Argon ~ 1% of air
– Ar-40(n,γ)Ar-41 (0.93% of air)
– Ar-41: 1.8h half life, hard β, 1.3 MeV γ
Dominant hazard is Ar-41 activity
Canadian Nuclear Safety Commission
6
Geometry and Materials
Go from complex to extremely simple
Reveals essential features and save computing time
Concrete
Air
6m
Iron
Void
Neutron
source
3m
2m
x
12 m
y
Canadian Nuclear Safety Commission
7
Neutron Source Spectra (1/2)
• Neutron point source emulates target during irradiation (isotropic)
• F-18 thick-target spectrum from Mendez et. al. [1]
– approximated by Maxwell fission spectrum
– 150 logarithmically spaced energy bins
• Tc-99m spectrum from nested neutron spectrometer data [2]
– Histogram representation
• All spectra automatically normalized by MCNP
Canadian Nuclear Safety Commission
8
Neutron Source Spectra (2/2)
Canadian Nuclear Safety Commission
9
Neutron Source Placement
Effect of Source Placement on Ar-41 Production
• 3m x 5m x 5m vault
• Starting at origin, moving
diagonally into corner
• Considering statistical error,
effect of source location is
negligible
Percent difference from mean Ar-41
production
• Eight source positions
1.00%
0.50%
0.00%
-0.50%
-1.00%
-1.50%
0
0.5
1
1.5
2
2.5
3
3.5
Distance from Origin (m )
Canadian Nuclear Safety Commission
10
Vault Size
Deviation from
linearity for irregular
shapes and very
small vaults
Ar-40 captures vs. cube bunker edge length
Ar-40 Capture density
Linearity result of
large neutron mean
free path in air
~62 m
0
Canadian Nuclear Safety Commission
1
2
3
4
5
6
Cube edge length (m)
7
8
11
Cyclotron and Source Energy
• Iron cyclotron added to centre of the room
– Slight decrease in production
– Cyclotron density has no effect
• Source energy tests: (3m x 6m x 12m vault with cyclotron)
Neutron energy
F-18 spectrum
Tc-99 spectrum
Isotropic 1 keV
Isotropic 0.025 eV
Canadian Nuclear Safety Commission
Ar-41 production/incident
neutron/cm3
3.0 E-5
3.6 E-5
5.2 E-5
12.0 E-5
12
Target Clamshell
Concrete
• Polyethylene target shielding
– 5, 10, 50cm thickness
– No boron
Air
6m
• Critical thickness
– Thin shield thermalizes neutrons
– >10cm necessary to capture
Iron
Void
3m
2m
x
Source
Polyethylene
clamshell
12 m
y
• γ shielding requirements
Canadian Nuclear Safety Commission
13
Partition Walls
• Wall placement below resulted in 15% decrease in Ar-41 production
Concrete
Air
• Objects in vault
reduce air activation
• Justifies simplified
geometry
7m
Iron
Void
Source
3m
2m
x
20 m
y
Canadian Nuclear Safety Commission
14
Ar-41 Activity
• Results are in captures/neutron - How many neutrons per μA of beam?
– IAEA TRS-468x gives saturation activity of F-18 at different beam
energies; at saturation
• Neutron production rate = F-18 decays/s (Bq)
– extrapolated to 24 MeV → 1.6 E10 n/s/μA
• For Mo-99 (NNS @19 MeV) → 3.2 E10 n/s/μA
• At saturation
• F-18 at 150 μA - (3 x 5 x 5 m vault) ~2 mCi
• Tc-99 at 750 μA – (3 x 6 x 12 m vault) ~ 12 mCi
Canadian Nuclear Safety Commission
15
Dosimetry – Dose rates
All dose rates in
μSv/h
F-18 @ 150 μA
3m x 5m x 5m
Tc-99 @ 750 μA
3m x 6m x 12m
External gamma3
5.1
17.3
Inhalation4
0.1
0.3
Skin5
0.9
2.2
• Assumes saturation production
• Skin dose – upper limit – assumes no clothing
Canadian Nuclear Safety Commission
16
Air Activation - Summary
1. Objects in vault reduces air activation
• Justifies simplified geometry
2. Air activation is not a problem for F-18
• Results are for saturation of Ar-41 (half-life 1.8 h)
• Runs are ~ 3 hours and typical F-18 runs are shorter
3. For Tc-99 may be a problem
• Runs are ~ 6 hours and beam current ~ 3 times higher
4. Ventilation reduces problem dramatically
• Less time to build up Ar-41 in vault and less exposure time –
1 hour air exchange time reduces dose by a factor of ~10
Canadian Nuclear Safety Commission
17
Concrete Activation
This produces radioactive waste, affecting decommissioning costs
1. How deep does it go?
2. What do polyethylene layers (with/without
boron) do?
3. Is it on all inner vault surfaces? Or is it
localized?
Canadian Nuclear Safety Commission
18
Literature Review (1/2)
Decommissioning a cyclotron [6]
20-year-old 17 MeV Scanditronix cyclotron (~40 µA)
• 40 tons of low-level radioactive waste including the concrete vault wall
• Activities with τ½ > 1 year
Isotope
Measured activity (Bq/g) UCL (Bq/g)
Co-60
0.068
0.1
Cs-134
0.005
0.1
Eu-152
0.083
0.1
Eu-154
0.010
0.1
Mn-54 *
0.016
0.1
Total
0.18
0.1
Canadian Nuclear Safety Commission
UCL: unconditional clearance level
at which material can be thrown
out as non-radioactive,
* Made by fast neutrons via
(n,p) reaction rather than (n,γ)
19
Literature Review (2/2)
Reactor study [7]
Ordinary concrete sample in TRIGA reactor in Slovenia
30 minute exposure @ neutron flux of 6.8 1012 n/s/cm2
Principal activities found Eu-152 and Co-60 at 6 Bq/g
Conclusion – concrete activation could be a
problem with the new cyclotrons
Canadian Nuclear Safety Commission
20
Absorption depth
Neutron capture density
Neutron capture density in concrete (F-18 production)
No Polyethylene
10 cm thick Polyethylene
10 cm thick Borated
Polyethylene
1
3
Canadian Nuclear Safety Commission
5
15
25
35
45
Concrete depth (cm)
55
65
80
100
21
•
•
•
•
Polyethylene around target
Poly layer thickness (cm)
5
10
15
20
Percentage of neutrons
captured in poly layer
17.1
52.0
72.5
82.6
Percentage of neutrons
captured in borated poly
layer
32.0
65.8
82.9
91.1
Results show how deep a sacrificial layer should be (if used)
Regular poly is almost as good as borated poly
Either option much cheaper than sacrificial layers in the vault
And can be put in after vault construction
Canadian Nuclear Safety Commission
22
Spatial Distribution in Vault
We now know activation goes down to depth ~20cm
But this was averaged over whole inner vault surface
Is it on all inner vault surfaces ? Or is it localized?
Investigated
1. Tc-99 vs F-18
2. Moving the source position inside the vault
3. Lateral distribution of neutron capture density within
1 side of the wall
Canadian Nuclear Safety Commission
23
1. Neutron Escape Percentage for Regular Poly
Around Target - Tc-99 versus F-18 source
1. F-18 neutrons escape the polyethylene layer more
easily than Tc-99’s higher energy (more penetration)
Canadian Nuclear Safety Commission
24
2. Moving
the Source Position on Y-Axis
With 20-cm Thick Polyethylene Layer
75
Canadian Nuclear Safety Commission
250
425
600cm
25
2. Relative Capture Density of the Left
Wall With Respect to Tc-99 Source
Position
Capture Density relative to when the source is in
middle position
3.5
3.0
y = 3.4e-0.002x
R² = 0.99
2.5
2.0
1.5
1.0
0.5
0.0
0
100
Canadian Nuclear Safety Commission
200
300
400
Distance between source and left wall (cm)
500
600
700
26
3. Lateral Distribution of Neutron Capture Density in
Capture density in the front wall
Near Wall
-50
250
200
Inverse Square Law
150
50cm away WO poly
50cm away WITH poly
100
middle WO poly
50
0
0
50
100
150
200
250
300
At 50-cm distance
radius of activation
~ 150 cm (10%)
Lateral distance on wall from target (cm)
Canadian Nuclear Safety Commission
27
Regulatory Impact
1) Compare cyclotron neutron flux with TRIGA reactor flux
• After 1 year at full operation → 0.35 Bq/g of Eu-152 and
0.33 Bq/g for Co-60 (measurable)
• After 25 years operation → 7 Bq/g for Eu-152 and
3 Bq/g of Co-60
2)100 x regulatory limit for disposal as non-radioactive waste
3) If no steps are taken, it will impact decommissioning cost and
possibly financial guarantee
Canadian Nuclear Safety Commission
28
Conclusions
• Air activation – Not a big deal and easily controlled through
– ventilation
– restricted access for a few hours (normal to allow cyclotron to cool off)
– detection of Ar-41 by area monitor in vault
• Concrete activation
– could be a challenge for decommissioning
– localized to concrete near target – including floor
For both problems, suggest borated poly around target
• Experiment – activate sample of vault concrete & analyze by γ spectroscopy
• Reactor neutron spectrum not quite the same as cyclotron neutron spectrum
• Your concrete may have different impurities
Canadian Nuclear Safety Commission
29
Questions?
Canadian Nuclear Safety Commission
30
References
[1] Mendez R. et. al. “Study of the neutron field around a PET cyclotron”. IRPA11 conference presentation,
May 2004.
[2] Debeau J. et. al. “The Measurement of Neutron Energy Spectra in the High
Neutron Flux Environments of Medical Accelerators Using Nested Neutron
Spectrometer” (2013 CRPA conference poster).
[3] Ar-41 source modeled as uniform equivalent sphere. Specific Gamma constant from
Delacroix D. et al., Rad. Prot. Dosimetry v. 98, no. 1 pp. 9-18 (2002). [4] Thériault B. “Inhalation dose
coefficient for Ar-41”, private com. (2012).
[5] β skin dose coefficient linearly extrapolated from data of - Fell TP, Rad. Prot. Dosimetry V. 36 No. 1, pp.
31-35 (1991).
[6] Sunderland J. et al. “Considerations, measurements and logistics associated with low energy cyclotron
decommissioning (2011)”, AIP Conf. Proc. 1509, 16 (2012): http://dx.doi.org/10.1063/1.4773931.
[7] Zagar T., Ravnik M., “Measurement of Neutron Activation In Concrete Samples”, Proc. Int. Conf.
“Nuclear Energy In Central Europe 2000”.
Canadian Nuclear Safety Commission
31
Initial Results
• 3m x 5m x 5m empty vault with point source
O16(n,p) Spectrum in Vault Air
Ar40 Capture Spectrum in Vault air
0.00003
0.000015
0.00001
0.000005
0
1.E-09
1.E-07
1.E-05
1.E-03
Neutron Energy (MeV)
1.E-01
6E-14
O16(n,p) (per source neutron)
0.00002
Ar41 captures (per source neutron)
0.000025
7E-14
5E-14
4E-14
3E-14
2E-14
1E-14
1.E+01
0
9.0E+00
1.0E+01
1.1E+01
1.2E+01
1.3E+01
1.4E+01
1.5E+01
1.6E+01
Neutron Energy (MeV)
Canadian Nuclear Safety Commission
32