continuous analytical performances monitoring at the on ... · overview -analytical techniques at...
TRANSCRIPT
IAEAInternational Atomic Energy Agency
Continuous Analytical Performances Monitoring at the On Site Laboratory Through Proficiency, Inter-Laboratory testing and Inter-comparison Analytical
Methods
J.-G. Decaillon*, G. Duhamel*, S. Dashdondog, A. Toervenyi, C.-K. Kim (IAEA)
S. Hara, S. Kato, T. Kawaguchi (NMCC)K. Matsuzawa (JSGO)
Symposium on International Safeguards 20-24 October 2014
IAEA
Overview - Analytical Techniques at OSL
Overview - Analytical Techniques at OSL
Introducing the term performance
Introducing the term performance
ITV CriteriaITV Criteria
Proficiency test
&
Inter-laboratory results
Proficiency test
&
Inter-laboratory results
ConclusionsConclusions
Outlines
Symposium on International Safeguards 20-24 October 2014
IAEA
Overview: Analytical Techniques at OSL
• OSL is the third Safeguards laboratory of its kind
• IAEA operates the On-Site Laboratory jointly with NMCC (Nuclear Material Control Center)
• Main mission is to support IAEA inspection activities related to nuclear material inventory flow verification and confirmation of the operational status.
• Analysis request, sample type and concentration level drive the selection of the analytical method
• Whenever possible, the samples are measured by two independent methods (i.e. KED Vs IDMS, Pu(VI) Vs XRF-Pu)
• Least significant difference is introduced to validate the measure
Symposium on International Safeguards 20-24 October 2014
IAEA
HRGS• Pu Isotopics
• U/Pu ratio
Spectrophotometry• Pu ~ g/L level
Density meter• Density
Alpha Spectrometry• Correction Pu-238
• Pu low level
AnalyticalTechniquesAnalytical
Techniques
Neutron Counter• Cm ~µg/L
K-Edge• Pu ~ 40-270 g/L
• U ~ 40-410 g/L
Thermal Ionization Mass Spectrometry (TIMS)• U ~ 1-410 g/L
• Pu~ 1-270 g/L
X-Ray Fluorescence• Pu ~ 0.5-6 g/L
• U ~ 1-10 g/L
AnalyticalTechniquesAnalytical
Techniques
Overview: Analytical Techniques at OSL
Symposium on International Safeguards 20-24 October 2014
IAEA
HKED
Pros
• Fast
• Accurate
• U & Pu simultaneously
• No sample discarding issue
Cons
• Calibration effort
• No Isotopic info
• Stability of the XRF beam
Symposium on International Safeguards 20-24 October 2014
IAEA
TIMS
Symposium on International Safeguards 20-24 October 2014
Pros
• Robust
• Very Accurate
• U & Pu Isotopics
• Standard Equipment
Cons
• Time consuming and labor intensive (Sample preparation and separation)
• Pure Solution
IAEA
Introducing the Term Performance
• To meet the provision of safeguards objectives (Objective C) analytical methods must be accurate and precise
1
• Assessing/Judging the reliability of analytical techniques is of paramount importance2
• To meet verification regime, analytical techniques should work according to certain standards3
• Performance = serie of criteria used for making decision about the correctness of our measurement
4
Symposium on International Safeguards 20-24 October 2014
IAEA
Verification is achieved by cross-checking results by two independent methods: i.e. KED vs IDMS, XRF vs Pu(VI)
Result is validated when difference
(method1 –method2) < Least Significant Difference D
� � 2 � ������ ��� � ������ ��
�
Introducing the term Performance
Symposium on International Safeguards 20-24 October 2014
IAEA
ITV Criteria
“ITV values represent estimates of the ‘state of the practice’ which should be achievable under routine measurement conditions”*.
ITVsITVsITVsITVs
• ITVs are uncertainties
• Used as a reference of the quality of the measurement achievable
• Ur & Us are defined for each technique & element
Symposium on International Safeguards 20-24 October 2014
*Excerpt of International Target Values 2010for Measurement Uncertaintiesin Safeguarding Nuclear Materials
IAEA
ITV Criteria
� � 2 � 0.18 � � 0.28 �
� � 0.67%
e.g. U concentrated solution measured by KED-U & cross-checked with IDMS. Calculation of the least significant difference
Symposium on International Safeguards 20-24 October 2014
IAEA
Interlaboratory results
Off Site Shipment filament results OSL-NML
• Goals: 1. Inter Laboratory Comparison (OSL – NML): Ur, Usdetermination for both laboratory
1
• 73 Pu and 32 U loaded filaments were shipped to NML Seibersdorf2
• Data set: 2012 and 2013 inspection samples3
• Instrumentation: TRITON TIMS4
Symposium on International Safeguards 20-24 October 2014
IAEA
Interlaboratory results
• Samples were sent in duplicate (2 filaments loaded with same sample) to NML in order to be able to evaluate NML random error
• Results are decay corrected for better comparison
• New Agency (IFC) software package used (OPTANOVA) for evaluation of random and systematic errors
Off Site Shipment filament results OSL-NML
Symposium on International Safeguards 20-24 October 2014
IAEA Symposium on International Safeguards 20-24 October 2014
Interlaboratory results
Year 2013 235U/238U 240Pu/239Pu 242Pu/239Pu
Random error (OSL) 0.026 0.007 0.036
Systematic error (OSL) 0.009 0.003 0.009
Random error(NML) 0.026 0.014 0.043
Systematic error (NML) 0.009 0.005 0.011
Systematic error (OSL-NML) 0.012 0.005 0.015
Combined error 0.039 0.017 0.0580.017
Combined errors of 240Pu/239Pu, 242Pu/239Pu and 235U/238U ratios are well-below the limits given in ITV-2010 (0.18% at 1 σ)
0.0580.039
Combined Error
• Combined Errors for 240Pu/239Pu, 242Pu/239Pu and 235U/238U ratios< 0.18% (ITV-2010)
IAEA
Interlaboratory results
OSL: Comparison KED-Pu vs IDMS
• Check the accuracy & performance of KED-Pu.
• No Reference material for concentrated Pu solution
• Optanova based on 53 pairs comparison (data 2012-2014)
Symposium on International Safeguards 20-24 October 2014
IAEA
Interlaboratory results
OSL: Comparison KED-Pu vs IDMS
-1.000
-0.500
0.000
0.500
1.000
1.500
-5 5 15 25 35 45 55IDM
S-K
ED d
iffe
ren
ce (
%)
Sample ID
2012-2014 (August)
difference IDMS-KED
New KED-Pu Cal
Symposium on International Safeguards 22 October 2014
Method Material
Uncertainty component (%rel)
ITV (%rel)
Ur Us
IDMS Pu
Glove box condition
0.15 0.1 0.18
KED-PuGlove box condition
0.3 0.3 0.42
� � 2 � 0.18 � � 0.42 �
� � 0.91%Max least significant difference
Min least significant difference
IAEA
Interlaboratory results
OSL: Comparison KED-Pu vs IDMS (Optanova analysis)
Estimation of the Long-Term Systematic Error
Estimator is d% = 0.30 %
Sigma ( d% ) = 0.06 %
Absolute t-value = 5.155
Optimal Estimators for Difference IDMS KED-Pu
(a) Random Error Standard Deviation σr 0.36 % 0.25 % 0.25 %
Standard Deviation of σr sd (σr) 0.04 % 0.13 % 0.13 %
(b) Systematic Error Standard Deviation σs 0.07 % 0.05 % 0.05 %
Standard Deviation of σs sd (σs) 0.02 % 0.01 % 0.02 %
(c) sqr ( σr2 + σs
2 ) 0.36 %
sd [sqr ( σr2 + σs
2)] 0.04 %
Optimal Estimators for Difference IDMS KED-Pu
(a) Random Error Standard Deviation σr 0.36 % 0.25 % 0.25 %Standard Deviation of σr sd (σr) 0.04 % 0.13 % 0.13 %
(b) Systematic Error Standard Deviation σs 0.07 % 0.05 % 0.05 %
Standard Deviation of σs sd (σs) 0.02 % 0.01 % 0.02 %
(c) sqr ( σr2 + σs
2 ) 0.36 %
sd [sqr ( σr2 + σs
2)] 0.04 %
Random Error
IDMS Random Error > ITV Ur (0.15%) but in the uncertainty. KED-Pu: OK
Optimal Estimators for Difference IDMS KED-Pu
(a) Random Error Standard Deviation σr 0.36 % 0.25 % 0.25 %
Standard Deviation of σr sd (σr) 0.04 % 0.13 % 0.13 %
(b) Systematic Error Standard Deviation σs 0.07 % 0.05 % 0.05 %
Standard Deviation of σs sd (σs) 0.02 % 0.01 % 0.02 %
(c) sqr ( σr2 + σs
2 ) 0.36 %
sd [sqr ( σr2 + σs
2)] 0.04 %
Symposium on International Safeguards 20-24 October 2014
IAEA
Proficiency Test
REIMEP-17uranium and plutonium amount contents typical for undiluted spent nuclear fuel input solution
uranium and plutonium amount contents typical for undiluted spent nuclear fuel input solution
U & Pu concentrationU & Pu concentration
Pu Isotopic determinationPu Isotopic determination
U Isotopic determinationU Isotopic determination
9 Laboratories participated9 Laboratories participated
Symposium on International Safeguards 20-24 October 2014
IAEA
Proficiency Test
REIMEP-17: OSL detailed results vs Ref values (Pu)Reported: Certified low Certified: Certified high Relative diff %
Pu [mmol/g] 9.1751 9.1511 9.1561 9.1611 0.208
Uncertainty value 0.0205 0.005
n(238Pu)/n(239Pu) 0.04264 0.0426 0.0426 0.0426 0.119
Uncertainty value 0.00062 0.00004
n(240Pu)/n(239Pu) 0.4788 0.4786 0.4787 0.4788 0.020
Uncertainty value 0.0004 0.00006
n(241Pu)/n(239Pu) 0.1261 0.1255 0.1257 0.12560 0.296
Uncertainty value 0.0003 0.0002
n(242Pu)/n(239Pu) 0.1375 0.1374 0.1375 0.1375 0.025
Uncertainty value 0.0004 0.000038
Coverage factor (k) for all
measurand 1
Reported: Certified low Certified: Certified high Relative diff %
Pu [mmol/g] 9.1751 9.1511 9.1561 9.1611 0.208Uncertainty value 0.0205 0.005
n(238Pu)/n(239Pu) 0.04264 0.0426 0.0426 0.0426 0.119Uncertainty value 0.0006 0.00004
n(240Pu)/n(239Pu) 0.4788 0.4786 0.4787 0.4788 0.020Uncertainty value 0.0004 0.00006
n(241Pu)/n(239Pu) 0.1261 0.1255 0.1257 0.1260 0.296Uncertainty value 0.0003 0.0002
n(242Pu)/n(239Pu) 0.1375 0.1374 0.1375 0.1375 0.025Uncertainty value 0.0004 0.00003
Coverage factor (k) for all
measurand 1
All differences are smaller than the calculated least difference. Good agreement for Pu conc & Pu isotopics with Ref values
Symposium on International Safeguards 20-24 October 2014
IAEA
Proficiency Test
REIMEP-17: OSL detailed results vs Ref values (U)
Reported: Certified low Certified: Certified high Relative diff %
U [mmol/g] 0.8446 0.8429 0.8434 0.8439 0.138
Uncertainty value 0.0019 0.0005
n(234U)/n(238U) 0.000068 0.000064 0.000066 0.00006 4.038
Uncertainty value 0.0000004 0.0000015
n(235U)/n(238U) 0.0068 0.0068 0.0068 0.0068 -0.020
Uncertainty value 0.000017 0.0000057
n(236U)/n(238U) 0.0000034 0.0000014 0.0000029 0.000004 16.482
Uncertainty value 0.0000000 0.0000015
Coverage factor (k) for all
measurand 1
U234/U238 & U236/U238 difference large. However difference is smaller than the least significant difference (due to large uncertainty (e.g. ~51% Unc for U236/U238 )
Unc rel ~ 2.28%
Unc rel ~ 51.7%
- Good agreement for U conc
- For U235/U238
very smalldifference
OSL Vs
Ref value
Symposium on International Safeguards 20-24 October 2014
IAEA
Proficiency Test
REIMEP-17: OSL results vs others laboratories
OSL
OSL uncertainty reported as the corresponding ITV method uncertainty (IDMS-Pu LSD: 0.18%)
Symposium on International Safeguards 20-24 October 2014
IAEA
REIMEP-17: OSL results vs others laboratories
OSL
Positive Bias ?
All Laboratories but one reported U concentration > Ref value
Symposium on International Safeguards 20-24 October 2014
Proficiency Test
IAEA
REIMEP-17: OSL results vs others laboratories
OSL
240Pu/239Pu ratio used for IDMS calculation
Symposium on International Safeguards 20-24 October 2014
Proficiency Test
IAEA
REIMEP-17: OSL results vs others laboratories
OSL
Symposium on International Safeguards 20-24 October 2014
Proficiency Test
IAEA
Eqrain n°°°°12OSL: XRF & Pu(VI) results
XRF Random component determination
OSL vs Overall results
Proficiency Test
IAEA
Eqrain n°°°°12Test
sample
Plutonium content
AverageTotal
uncertaintyg.l
-1g.l
-1g.l
-1(at k=2)
1 2.321
2.341 0.1312 2.3353 2.3354 2.3585 2.359
Reported result based on 5 replicates
XRF
Ref Value: 2.358 g/L (2.129 g.Kg-1)
Reported Result based on 4 replicates
Pu(VI)
Test samplePu content
g.l-1 ** g.kg-1
12 2.3795 2.14813 2.3627 2.13294 2.3679 2.13765 2.3804 2.1489
Average = 2.3727 2.1419
Symposium on International Safeguards 20-24 October 2014
Proficiency Test
IAEA
Eqrain n°°°°12 (OSL code: 12 for XRF, 15 for spectrophotometry)
OSL
ζ ���� ! �"#
$�� � � $"#
�Zeta score
Satisfactory result when ≤ 2
Symposium on International Safeguards 20-24 October 2014
Proficiency Test
IAEA
Eqrain n°°°°12: Random component determination
Ampoule n°1Repeatability test (n=19)
XRFRandom
component
UUUU(r)(r)(r)(r)====0.76 0.76 0.76 0.76 %%%%
U(r) = Standard deviation of the samples distribution
UUUU(r)(r)(r)(r)OSL vs ITVOSL vs ITVOSL vs ITVOSL vs ITV
U(r)OSL < ITV XRF
Average(n=19) = 2.3474 g/L
Symposium on International Safeguards 20-24 October 2014
Proficiency Test
IAEA
Eqrain n°°°°12 (OSL code: 12 for XRF, 15 for spectrophotometry)
OSL zeta score are skewed low due to use of ITV uncertainty.
i.e. with “real uncertainty” Zeta score for XRF = -0.51 (-0.25 with ITV)
Symposium on International Safeguards 20-24 October 2014
Proficiency Test
IAEA
Conclusions
• Random & systematic error for isotopic ratio are very low
• U & Pu concentrations are well in the ITV criteria
• Overall good agreement with REIMEP 17IDMS
• Random and systematic error are in the ITV criteria (<0.30%)
• Long term systematic is constant over the years and calibrationsKED
Symposium on International Safeguards 20-24 October 2014
IAEA
Conclusions
• Good agreement with Eqrain ref value• Random component Ur calculated from
repeatability test better than XRF ITVXRF
• Good agreement with Eqrain ref value• Possible small positive bias (all
replicates > ref value)Spectrophotometry
Symposium on International Safeguards 20-24 October 2014