IAEA

Report
Implementation Concepts for
Unattended Measurement Systems at
Enrichment Plants
L. Eric Smith, Alain Lebrun
IAEA
January 2012
IAEA
International Atomic Energy Agency
IAEA’s “Model Approach for GCEPs”
High-capacity plants pose implementation challenges for current approaches.
Safeguards objectives: Timely detection of…
• Diversion from declared input and output
• Undeclared (excess) production of normal enrichment levels
• Higher-than-declared enrichment (e.g. HEU)
Implementation objectives
• Reduce need for routine measurements, sampling during inspections*
• Ease and expedite cylinder release process for facility operators
How might unattended measurement systems contribute?
IAEA
*Related work by Boyer, et al. (IAEA Symposium 2010)
Potential Roles: Unattended Measurement Systems
Process MBA
Pressure
Cylinder
Storage MBA
Pressure
(Temp)
UF6
Load Cell
NaI(Tl)
m Computer
E(t)
Collimator
Load-cell monitoring
Online Enrichment Monitor (OLEM)
Unattended Cyl. Verification Station (UCVS)
M(t) for each cylinder
High-accuracy E(t) for each cylinder
Continuous gas monitoring
High-accuracy net mass
“NDA Seal” for CoK on cylinder contents
Assay of blended cylinders
Ecyl = E(t)*M(t)
MU
IAEA
M235 = Ecyl * MU
Concept: Load-Cell Monitoring
tstart
IAEA
M(t)
tend
Concept: On-Line Enrichment Monitor
E(t) ∝ Rgas_186keV (t) * rgas (P, T, t)
CEMO
OLEM
Cascade 1
NaI(Tl)
Cascade 2
Cascade 3
Pressure
Temperature
P ~ 40 Torr
Load Cell
Cascade 4
Header Pump
Gas Sampling
P ~ 4 Torr
UF6
Header Pipe
IAEA
Cylinder
1)
2)
M(t)
Mass Spec Analysis
High-accuracy E(t) for product and tails
Continuous monitoring of gas
OLEM Viability Studies: Examples
Statistical uncertainty only--systematic uncertainties are not addressed.**
10.0
Tails
9.0
Total Uncertainty (%)
.
8.0
Low P: 10 Torr
High P: 50 Torr
Low D: 100 mg/cm2
High D: 1000 mg/cm2
Low P
High D
E = 0.2%
High P
Low D
7.0
Feed
6.0
Product
E = 0.711%
Low P
High P
High D
Low D
5.0
4.0
3.0
2.0
E = 5.0%
Low P
High P
High D
Low D
E = 2.0%
Low P
High P
High D
Low D
Performance
Targets
Tails: sT < 3%
Feed: sF < 2%
Product: sP < 1%
1.0
0.0
IAEA
**Plot from Smith and Lebrun (IEEE Nuclear Science Symposium, 2011)
Related work by Ianakiev (ESARDA 2010) and March-Leuba (personal communication, 2012)
Concept: Unattended Cylinder Verification Station
1)
2)
3)
4)
Apply and verify “NDA Seal” at MBA boundaries (CoK)
Unattended NDA of M235 for blended cylinders
Recovery of CoK on cylinders
Platform for weight, NDA verification during inspections
Mass: Shared-use or IAEA scale
NDA**: Hybrid (PNNL), PNEM (LANL), other?
Cylinder ID: L2IS, Global Bar Code, other?
Surveillance: NGSS
**from Smith (INMM 2010)
IAEA
**Related Work
Smith (IEEE TNS 2010, INMM 2010), McDonald (INMM 2011)
Miller (ESARDA 2011)
UCVS Viability Studies: Example
“Hybrid NDA” for 235U Assay (30B cylinders)
Intl. Target Value: sP ~ 5%
Hybrid NDA (preliminary)
sP ~ 2.5%
sF ~ ??
sT ~ ??
Other NDA methods?
sP = 2.5%
IAEA
8
NDA Seal?
**Plot from Smith et al. (INMM 2010)
UMS Implementation Concepts
“Special” treatment of feed
• Challenges
•
• Largest 235U flow rate
• Poor assay accuracy (OLEM wall-deposit issues, UCVS > 6%)
Advantages (assuming natural feed)
• Isotopics are precisely known
• Cylinders should be homogeneous
Baseline Concept
•
•
•
•
No quantitative assay of feed  assume Ecyl = 0.711%  sF ~ 0.0%...if
UCVS verifies that Ecyl_UCVS is consistent with feed-cylinder profile
OLEM only on product and tails header pipes
UCVS quantitative NDA on blended product cylinders
IAEA
Implementation Concepts: Viability Analysis
Overview
Scenario: Diversion into MUF or D
• 235U bias defect in product and tail cylinders
• SQ = 75 kg 235U (LEU, NU, DU)
Viability Metric: Fidelity of 235U mass balance (“IMUF”)
• Assume no waste, scrap, etc.
• IMUF = F – (P + T)
• sMUF2 = sF2 + sP2 + sT2
• Threshold = 3*sMUF
• PD for 1SQ diversion?
PD
IAEA
**from C. Norman, IAEA
Implementation Concepts: Viability Analysis
Reference Facility: 4,000,000 SWU/year, 0.711%, 3.0%, 0.25%
Feed
6 000 000
kg U/year
Product
1 000 000
Tails
5 000 000
Feed
710
cylinders/year
Product
667
Tails
592
Analysis variables: OLEM s , UCVS sP , blend fraction, balance period
Balance Period = 1 month
OLEM s
Concept
ID
1
2
3
4
5
F
0.0
0.0
0.0
0.0
2.00
IAEA
P
1.0
0.5
1.0
1.0
1.0
UCVS s
T
3.0
1.5
3.0
3.0
3.0
P
3.0
3.0
3.0
6.0
0.0
= Baseline Concept
Blend
Fraction
F
0.05
0.05
0.30
0.30
0.00
0.00
0.00
0.00
0.00
0.95
s (SQ)
P
OLEM UCVS
0.32
0.05
0.16
0.05
0.23
0.30
0.23
0.60
0.33
0.00
T
s (SQ)
Total
PD (% )
Total
0.42
0.21
0.42
0.42
0.42
0.53
0.27
0.56
0.77
1.09
13.6
77.4
11.0
4.5
1.9
Implementation Concepts: Viability Analysis
Balance Period = 1 week
OLEM s
Concept
ID
1
2
3
4
5
F
0.0
0.0
0.0
0.0
2.00
IAEA
P
1.0
0.5
1.0
1.0
1.0
UCVS s
T
3.0
1.5
3.0
3.0
3.0
P
3.0
3.0
3.0
6.0
0.0
Blend
Fraction
F
0.05
0.05
0.30
0.30
0.00
0.00
0.00
0.00
0.00
0.22
s (SQ)
P
OLEM UCVS
0.07
0.01
0.04
0.01
0.05
0.07
0.05
0.14
0.08
0.00
T
s (SQ)
Total
PD (% )
Total
0.10
0.05
0.10
0.10
0.10
0.12
0.06
0.13
0.18
0.25
100.0
100.0
100.0
99.6
83.7
Conclusions
• High-capacity plants require new instruments and approaches
• Integrated UMS: “Independent” 235U and U balances on 100% flow
• NDA Seal for cylinder CoK
• Special treatment of feed
• PD values (scoping) for protracted diversion are encouraging
• UMS Role: Rule out protracted diversion between inspections
• Machines do routine measurements
• Inspectors do what humans do best (investigate)
•
Many questions and issues ahead…for example
•
•
•
•
Relevance for diversion and excess production scenarios
Realistic OLEM and UCVS uncertainties
Data security for shared-use instruments
Operator impacts, acceptability
IAEA
Additional Information
IAEA
Potential Impacts to Operators
Potential Impact
Eased and expedited cylinder release process
Reduced physical presence of inspectors
Reduced sampling requirements on cylinders
Cylinder tracking infrastructure
OLEM for process control and criticality control
Load-cell (and accountancy scale?) data sharing
OLEM nodes installed on header pipes (2 per unit); additional P gauges
UCVS installation(s)
UCVS scans on cylinders moving in/out of MBAs
IAEA
Material Flow and Data Streams
Unblended Product and Tails Cylinders
Process MBA
UCVS
Storage MBA
Load Cell
OLEM
Load Cell: M(t)
OLEM: E(t)
Ecyl_OLEM = E(t)*M(t)
Ecyl_OLEM : sP < 1%, sT < 3%
NDA Seal
Scale: Mempty , Mfull , sM < 0.1%
M235_OLEM = Ecyl_OLEM * MU
M235_OLEM : sP < 1%, sT < 3%
Facility-Level Data: MU , M235_OLEM , NDA Seal
IAEA
Material Flow and Data Streams
Feed Cylinders
Process MBA
UCVS
Storage MBA
Load Cell
Load Cell: M(t)
Ecyl = known = 0.711%
Ecyl : sF ~ 0.0%
NDA Seal: “nominal” feed?
Scale: Mempty , Mfull , sM < 0.1%
M235 = Ecyl * MU
M235 : sF ~ 0.1%
Facility-Level Data: MU , M235 , NDA Seal
IAEA
Material Flow and Data Streams
Blended Product Cylinders
Process MBA
UCVS
Storage MBA
Blending
Station
Quantitative NDA of Ecyl_UCVS : sP ~ 3 - 6%
NDA Seal
Scale: Mempty , Mfull , sM < 0.1%
M235_UCVS = Ecyl_UCVS * MU
M235_UCVS : sP ~ 3 - 6%
IAEA
Facility-Level Data: MU , M235_UCVS , NDA Seal
Implementation Concepts: Viability Analysis
Balance Period = 2 weeks
OLEM s
Concept
ID
1
2
3
4
5
F
0.0
0.0
0.0
0.0
2.00
IAEA
P
1.0
0.5
1.0
1.0
1.0
UCVS s
T
3.0
1.5
3.0
3.0
3.0
P
3.0
3.0
3.0
6.0
0.0
Blend
Fraction
F
0.05
0.05
0.30
0.30
0.00
0.00
0.00
0.00
0.00
0.44
s (SQ)
P
OLEM UCVS
0.15
0.02
0.07
0.02
0.11
0.14
0.11
0.28
0.15
0.00
T
s (SQ)
Total
PD (% )
Total
0.19
0.10
0.19
0.19
0.19
0.24
0.12
0.26
0.35
0.50
86.9
100.0
80.0
43.1
15.7
UCVS Technical Objectives
• Quantitative assay of cylinder enrichment  M235 in each cylinder
•
•
•
•
•
Measurement scenario: Single measurement of many different cylinders
Key metric: Absolute accuracy for quantification of M235
Preliminary accuracy targets: sP < 3%, sF < 6%, sT < 9% for M235
Full-volume interrogation (i.e. sensitive partial defect detection)
Unattended operation
• NDA Seal  Continuity of knowledge on cylinder contents
•
•
•
•
•
•
Measurement scenario: Repeated measurements on a single cylinder
Key metric: Reproducibility of key signatures and attributes
Candidate attributes: E, MU, 234/235, 232/235, 235 spatial distribution
Preliminary uncertainty targets: TBD, but likely < 0.5%
Full-volume interrogation (i.e. sensitive partial defect detection)
Unattended operation
The NDA Seal is a recent addition to the potential roles of the UCVS. The
concept requires a viability assessment based on measurements and modeling.
IAEA
“NDA Seal”
Collection of distinguishing signatures and
attributes that can be used to provide and recover
CoK of the cylinder contents.
Reproducibility of these attributes is the key metric.
IAEA
UCVS: Signatures and Attributes
For 235U NDA and NDA Seal
Traditional 186-keV g  U-235 concentration in outer UF6
Direct measure of U-235, but weakly penetrating
Array of spectrometers  axial distribution of U-235
Induced-fission neutrons  U-235
Direct measure of U-235
For thermal interrogating neutrons, only outer layer of UF6
Neutrons from F-19 (a, n)  U-234
U-234 is primary a emitter
Neutron escape: ~0.80  full-volume
Indirect measure of U-235
Indicator of feed type
Neutron-induced g  U-234
Iron as n  g converter
Fe-56 + n  Fe-57 + g (7.63,7.65 MeV)
Indirect neutron detection
IAEA
2614-keV g  U-232 “flag”
Presence of U-232  reactor recycle feed
Performance Metrics for Quantitative Assay
Assay Enrichment (%)
~ ssys_cal
~ ssys_ran
sNDA2 = sstat2 + ssys_cal2 + ssys_ran2
Declared Enrichment (%)
23
sstat
IAEA
Prediction: s
sys_cal
> sstat and ssys_ran
Performance Metrics for NDA Seal
Attribute
~ ssys_ran
sseal2 = sstat2 + ssys_ran2
Number of Measurements on Same Cylinder
24
IAEA
Prediction:
ssys_ran can be small, so must minimize sstat
OLEM Uncertainty Budget
Product Material
3.0
Uncertainty Contribution (%)
.
2.5
Pressure
Temperature
186-keV: Deposit
186-keV: Gas
Total
E = 2.0%
P = 10 Torr
D = 1000
E = wt% 235U for gas, deposit
P = gas pressure (Torr)
D = deposit thickness (mg/cm2)
2.0
1.5
E = 5.0%
P = 10 Torr
D = 1000
OLEM target for sE
E = 5.0%
P = 10 Torr
D = 100
1.0
E = 5.0%
P = 50 Torr
D = 1000
0.5
0.0
E = 2.0%
P = 10 Torr
D = 100
IAEA
E = 5.0%
P = 50 Torr
D = 100
E = 2.0%
P = 50 Torr
D = 1000
E = 2.0%
P = 50 Torr
D = 100
*From Smith and Lebrun, IEEE Nuclear Science Symposium, 2011

similar documents