Technologies to Detect Materials for Nuclear/Radiological

Report
Technologies to Detect Materials
for Nuclear/Radiological Weapons
Gerald L. Epstein
Senior Fellow, Center for Strategic and International Studies
and
Adjunct Professor, Georgetown Security Studies Program
November 10, 2004
November 10, 2004
1
Outline
• Detection principles
• System considerations
• Nuclear radiation and radioactivity
• Technological approaches and limits
• Can address chemical and biological
detection in discussion
November 10, 2004
2
Detector Principles
• Detectors are physical systems measuring
noisy phenomena amidst backgrounds
• Sensitivity and selectivity must be
considered together
– It’s easy to make a detector with a 100%
detection probability (perfect sensitivity)
– It’s also easy to make one with a 0% false
alarm rate (perfect selectivity)
– The trick is doing them at the same time
November 10, 2004
3
What’s Measured vs. What’s Real
What’s Real
(+)
In fact
(-)
In fact
November 10, 2004
What’s Measured
(+)
(-)
Reported
Reported
Correct
False
detection:
negative:
p(D)
1 – p(D)
False
True negative:
positive:
1 – p(FA)
p(FA)
4
Three Useless Detectors and an
Impossible One
• One that never misses
• One that never falsely detects
• One that’s somewhere in between
• One that’s perfect
November 10, 2004
5
Useless Detector 1:
Always Reports Detection
Detector Report
Reality
(+)
Reported
(-)
Reported
(+)
In fact
1.00
p(D)
0.00
(-)
In fact
1.00
p(FA)
0.00
November 10, 2004
6
Useless Detector 2:
Never Reports False Alarms
Detector Report
Reality
(+)
Reported
(-)
Reported
(+)
In fact
0.00
p(D)
1.00
(-)
In fact
0.00
p(FA)
1.00
November 10, 2004
7
Useless Detector 3:
Randomly Reports Detection
Detector Report
Reality
(+)
Reported
(-)
Reported
(+)
In fact
X
p(D)
1-X
(-)
In fact
X
p(FA)
1-X
November 10, 2004
8
Unattainable Detector:
Perfect Sensitivity and Selectivity
Detector Report
Reality
(+)
Reported
(-)
Reported
(+)
In fact
1.00
p(D)
0.00
(-)
In fact
0.00
p(FA)
1.00
November 10, 2004
9
Actual Detectors Trade Off
Selectivity and Sensitivity
As threshold T decreases
from T1 to T2, more signal
peaks are detected (PD
increases) but more noise
peaks are detected as
well (PFA increases too).
Source: Robert J. Urick, Principles of
Underwater Sound (New York: McGraw
Hill, 1983), p. 381
November 10, 2004
10
“Receiver Operating Characteristic”
• Obtained by plotting PD vs.
PFA as detection threshold
varies
• Curves force PD and PFA to
be examined simultaneously
• The better the detector, the
more that PD exceeds PFA
• Name derives from early
days of radar / sonar
Source: same as previous
November 10, 2004
11
“Receiver Operating Characteristic” (2)
• Any one curve
represents a single
detector with different
thresholds
• Different curves
represent different
detectors
• Parameter “d” here
describes how close to
ideal a given detector is
Source: same, p.382
November 10, 2004
12
Significance of Detection Depends
on Number of Expected Positives
Case 1: Medical condition expected 5% of the time
N=10,000 patients; p(D) = 0.9; p(FA) = 0.01
(+)
In fact
(+) Reported
fraction | events
0.90 | 450
83% (+)’s correct
(-) Reported
fraction | events
0.10 | 50
0.5% (-)’s wrong
500 actual
positives
(-)
In fact
0.01 | 95
17% (+)’s wrong
0.99 | 9,405
99.5% (-)’s correct
9,500 actual
negatives
505 positive
reports
9,455 negative
reports
10,000
patients
November 10, 2004
13
Significance of Detection Depends
on Number of Expected Positives (2)
Case 2: Medical condition expected 0.1% of the time
N=10,000 patients; p(D) = 0.9; p(FA) = 0.01
(+)
In fact
(+) Reported
fraction / events
0.90 | 9
8.3% (+)’s correct
(-)
In fact
0.01 | 100
0.99 | 9,890
91.7% (+)’s wrong 99.99% (-)’s correct
109 positive
reports
November 10, 2004
(-) Reported
fraction / events
0.10 | 1
0.01% (-)’s wrong
9,891 negative
reports
10 actual
positives
9,990 actual
negatives
10,000
patients
14
Detector Systems
•
Context; expected threat; suite of potential
response options; operational protocols and
doctrine; all affect choice of detector technology.
If you can’t act on the information, do you want it?
•
Must consider how system will be used, by whom;
for what; and at what cost; answers will force
tradeoffs
•
Real world environment and operations are quite
different from laboratory conditions
•
Testing and verification are necessary
November 10, 2004
15
Nuclear Radiation
•
Alpha particles
– Energetic helium-4 nuclei emitted from certain radioactive elements
– Cannot penetrate sheet of paper or much air; cannot remotely detect
•
Beta particles
– Energetic electrons emitted from certain radioactive elements
– More penetrative but still do not extend very far through air; cannot
remotely detect directly
•
Gamma rays
– Electromagnetic radiation (like light, but much higher frequency); can
be considered to come in packets (photons)
– Highly penetrating; range depends on energy.
•
Neutrons
– Produced spontaneously by plutonium but very rarely by other
radioactive materials, natural or man-made
– Penetrative, including through materials that shield gamma rays
November 10, 2004
16
Intensity vs. Energy
• Energy (of a particle or photon)
– Determines how far it can penetrate and how much damage it
individually can do
– Measured in “electron-volts” – the amount of energy one electron
can get from a one-volt battery. Typical values for radioactive
decay are thousands to millions of electron volts (keV to MeV).
– That’s a lot for an electron but tiny for us. Dropping a paperclip
(~500 mg) a distance of 1 cm releases 3 x 1014 ev = 3 x 108 MeV
• Intensity (of a radiation source)
– Determines how dangerous the source is or how easily it can be
detected
– Depends on energy of each particle times numbers of particles per
second
• A low-intensity source can produce high-energy radiation, and
vice versa
November 10, 2004
17
Nuclear Materials of Concern
• Nuclear weapon materials
– Highly enriched uranium (U-235); emits relatively lowenergy gamma rays
– Weapons-grade plutonium (Pu-239 with some mixture
Pu-240 and others); emits gamma rays and neutrons
• Radioactive dispersal device (“dirty bomb”)
materials, with key threats including
–
–
–
–
Co-60, Cs-137(primarily gamma emitters)
Ir-192, Sr-90 (primarily beta emitters)
Pu-238, Am-241, Cf-252 (primarily alpha emitters)
However, these materials or their decay products often
also emit gamma rays
November 10, 2004
18
Radiation Spectrum
•
•
•
•
•
Each radioactive substance emits particles or gamma rays
with characteristic energies
Graph of the intensity of the radiation of a given source as
a function of the emitted energy is the source’s energy
spectrum
The energy spectrum of a source generating gamma rays
at 400 keV would show a single peak centered at 400 keV.
Detectors do not measure the energy of a radiation source
precisely; even for sources at precise energies, they show
energies over some range. The narrower the range, the
better the energy resolution
The better the resolution, the better the source
identification
November 10, 2004
19
Gamma Ray Spectrum at
Different Resolutions
HPGe
NaI
HPGe: High Purity Germanium detector (high resolution)
NaI:
Sodium Iodide detector (medium resolution)
Source: ORTEC Corp.: http://www.ortec-online.com/pdf/detective.pdf
November 10, 2004
20
Shielding
•
Gamma radiation and neutrons are attenuated by
surrounding material
– Gammas or x-rays of different energies attenuated by different
processes, some depending essentially on the mass of the shielding
and some depending on the composition (atomic number)
– Possibility of shielding strongly influences detector system design
•
Things that shield gammas well shield neutrons poorly, and
vice versa
– High-Z (atomic number) materials absorb gammas but only deflect
neutrons
– Low-Z materials slow down and absorb neutrons (possibly below
detection thresholds) but affect gammas less
•
There is very little legitimate neutron background; any
neutron sources is of high interest
November 10, 2004
21
Backgrounds
•
Naturally occurring radioactive materials
–
–
–
–
–
–
–
Potassium nitrate fertilizers (40K)
Granite or marble (Ra, U, Th)
Vegetable produce (40K or 137Cs from Ukraine)
Old camera lenses (Th coatings)
Thoriated tungsten welding rods or lantern mantles (Th)
Certain glasses or ceramic glazes (U, Th)
Porcelain bathroom fixtures (concentration of backgrounds)
• Individuals treated with medical isotopes
•
Legal shipments of radioisotopes
November 10, 2004
22
Detection Process: Ionization
Ionizing radiation produces ions along its direction of
travel that can be collected and measured by:
– Geiger-Muller counters
Each photon or ionizing particle registers as a single count or click.
Measures rough estimate of intensity of radiation but provides no
information about type or energy of radiation or source
– Proportional counters
Chamber – usually gas-filled tube – measures the amount of ionization
formed by incident particle or photon, which is proportional to incident
radiation’s energy. Collecting many such measurements produces source
spectrum
– Solid-state crystals (e.g., germanium)
Measure energy spectrum with much higher resolution. The highestresolution detectors need to be cryogenically cooled
November 10, 2004
23
Detection Process: Scintillation
• Ionizing radiation passing through certain
substances produces flashes of light whose
brightness is proportional to the energy of the
radiation
• Flashes of light amplified by photomultipliers
• Energy resolution is modest at best
• Different types of scintillator
– Sodium-iodide or other scintillating crystal
– Liquid scintillator
– Plastic scintillator
November 10, 2004
24
Scintillator Detector Examples
Radiation
“Pagers”
November 10, 2004
25
Scintillator Detector Examples
Portal radiation detectors (yellow) at Blaine, WA
Source: Physics Today 11/2004
Port of Entry
November 10, 2004
26
Detection Process: Dosimetry
Dosimeters measure total dose over some period
of time; not real-term measurements. Types
include
• Photographic film
• Thermoluminescent dosimeters
November 10, 2004
27
Detection Process: Active
Neutron Interrogation
• Neutrons can induce reactions in materials that
produce secondary neutrons and gamma rays,
which can be detected. This approach can be
used to search for explosives or other distinctive
materials
• Nuclear weapon materials are particularly
sensitive to this approach, since they react
strongly with neutrons
• Technique not effective for other radiological
materials
November 10, 2004
28
Active Neutron Interrogation
• Lawrence Livermore
National Laboratory
concept now being
prototyped
• Neutrons irradiate
cargo from below
• Liquid scintillator used
in side detector arrays:
cheap and responsive
November 10, 2004
29
Futuristic Concept: Muon
Deflection
• Cosmic ray muons (charged particles produced
in the atmosphere by incoming protons)
constantly bathe the earth and are highly
penetrating
• They are deflected when they pass through
matter – more by high-“Z” (atomic number)
materials such as uranium, plutonium, or lead
used for shielding, than by low-Z materials
• Measuring incoming and outgoing muon
directions can locate high-Z materials
November 10, 2004
30
Muon Deflection
Source:
http://www.lanl.gov/quarterly/q_spring03/muon_deflections.shtml
November 10, 2004
31
Muon Deflection
Source: Borozdin, K.N. et al. “Radiographic imaging with cosmic-ray muons,” Nature,
422, 277, (2003)
November 10, 2004
32
Conclusion
• Technologies exist to detect radioactive
materials remotely from modest distances
(several meters)
• Particularly if shielded, signals from these
materials are weaker than materials from
legitimate background sources. Therefore,
discriminating threatening materials from
backgrounds is essential
• Issues for mass deployment include background
rejection; cost; and system design
November 10, 2004
33

similar documents