A Mobile System for Detecting A Mobile System for Detecting Gamma-Radiation SourcesGamma-Radiation Sources

Part I.Part I.

Physical and Statistical BackgroundPhysical and Statistical Background

Ludmil Tsankov Dept of Nuclear Engineering, University of Sofia, Dept of Nuclear Engineering, University of Sofia,

BulgariaBulgaria ludmil@phys.uni-sofia.bg

Mityo MitevDept of Electronic Technique, Technical Dept of Electronic Technique, Technical

University of Sofia, BulgariaUniversity of Sofia, Bulgaria mitev@ecad.tu-sofia.bg

Tchavdar Lenev

Institute Institute forfor Nuclear Research and Nuclear Nuclear Research and Nuclear Energy, BAS, Sofia, BulgariaEnergy, BAS, Sofia, Bulgaria

lenevlenev@@inrneinrne.bas..bas.bgbg

IntroductionIntroduction

Two general methods are used by the national nuclear safety services to recover out-of-control ('orphan') gamma-radiation sources:

o passive ('trapping‘) methods - to set up large high sensitive stationary detectors at the borders, at the entrances of the scrap recycling facilities and at other potentially suspicious sites;

o active ('hunting') methods - to use mobile equipment able to discover signals from the sources during a survey made either by car (CGS) or by aircraft (AGS).

PurposePurpose

To perform a theoretical analysis of a CGS system in order to express the minimum detectable activity as a function of its basic parameters

To develop algorithms for real time data processing in order to come near to the theoretical limits

Statement of the problem Detector D has effective sensitive area S and moves across

a point source s with a constant activity A.

Assumptions:

• the effective area of the detector S does not depend on the source-detector disposition;

• the detector moves uniformly with respect to the source at a velocity V;

• the absorption of the gamma-rays emitted from the source in the air is negligible.

Integral countingIntegral counting

- radiation flux density

- flux

Signal-Noise Ratio

The background is estimated from preceding measurements and extrapolated: 2

00 ,B

)(4)(

222 tVd

SAt

d

Vtarctg

d

Vtarctgt

dV

AStt 12

021 4),(

212

2012021

21

)()(),(

),(

ttttBtt

ttR

Optimal Optimal systemsystem

R() has always a maximum

V=10m/s

A=3.7.107Bq (1mCi), S=0.0058m (3"x3"), d=10m,

B0=800cps,

{B0}=1.63cps

(tBG=300s)Vd /4.1max

Nearly-optimal Nearly-optimal systemsystem

Optimal system

Nearly-optimal system

for all 0.10

1.00

10.00

100.00

1000.00

0 5 10 15 20 25

V [m/s]

tau

[s]

3)( R

max)( R

R() depends on V:

s1smV /20

Adaptive algorithms for data analysis

Starts e e d B S

e s tim a te

R ece ive d a ta

(T-Tes t)

20,, BB Bn

),( 21 ttN

0

?

0 BS

u p d a te B S 2

0021 ,,),( BnttN B

N o

Yes

Yes

N o

A la rmu p d a te S S

2021 ,,),( SS SnttN

re s e t S S

5?

Sn

N ew B ackg rou nds ave B S

1?

Sn

BSSS

Yes

N o

1. Two samples are created:• Background (BS)• Signal (SS)2. New data are first regarded as a signal3. While the process is stationary, BS is extended

1?

00 ||t

BSt

Spectral registration of gamma-quanta

NaI(Tl) detectors have a good energy resolution which can be used to improve sensitivity.

Statistical hypothesis is changed:

now we have to compare not two numbers but

two probabilities distributions:

)()( 2 testtestT 2?

1

2)(1

r

k kk

kBkS

SB BS

SnBn

nn

Possible errors in decisions based on statistical tests

Type I (false alarm) Type II (to bypass a source)

Both types of errors have to be minimised simultaneously

Relationship between Type I and Type II errors depends on the choice of the significancy level (SNR threshold value)

Repeated test at yields a higher sensitivity

than a single test at at the cost of 1s reaction

delay .

)3(10 3 R610

Conclusion

Spectrometric registration mode is more complex, but yields a higher sensitivity

The problem for the optimal setting of the alarm threshold level is consistently deduced from the theory of statistical hypotheses testing

Acknowledgements

This work is supported by the Bulgarian Nuclear This work is supported by the Bulgarian Nuclear Safety Authority under contract NoSafety Authority under contract No  286-02.286-02.

To be continued …

A Mobile System for Detecting Gamma-Radiation Sources:

PART II. Design and First Experiments

(At the poster session)