U.S.

U.S. Nuclear Regulatory Commission Meeting
\jWREcKj
with Nuclear Energy Institute, Material Reliability Program, and
Operating Pressurized Water Reactor Licensees'
0
(nlal
C
Thursday, November 8, 2001
1:00 p.m. - 5:00 p.m.
Commissioners'HearingRoom
Purpose:
To discuss NRC staff's technical assessment for vessel head penetration nozzle cracking
associated with NRC Bulletin 2001-01, "Circumferential Cracking of Reactor Pressure Vessel
Head Penetration Nozzles."
Success:
NEI, MRP, and external stakeholders have a clear understanding of the staff's technical
assessment and its' basis.
Introduction:
Jake Zimmerman
1:00 p.m. - 1:10 p.m.
Bulletin 2001-01 Overview:
Allen Hiser
1:10 p.m. - 1:30 p.m.
Discussion of Crack Growth Rate:
Dr. William Shack
1:30 p.m. - 2:00 p.m.
Discussion of Crack Initiation:
Dr. William Shack
2:00 p.m. - 2:30 p.m.
-
BREAK
2:30 p.m. - 2:45 p.m.
-
Discussion of Stress Analysis
and Crack-Driving Force:
Dr. Gery Wilkowski
2:45 p.m. - 3:15 p.m.
Discussion of Critical Crack Size:
Dr. Gery Wilkowski
3:15 p.m.
Discussion of Deterministic Assessment:
Allen Hiser
3:45 p.m.
-
4:15 p.m.
Discussion of Probabilistic Assessment:
Allen Hiser
4:15 p.m.
-
4:30 p.m.
Discussion of Inspection Timing:
Allen Hiser
4:30 p.m.
-
4:45 p.m.
4:45 p.m.
-
5:00 p.m.
Comments/Questions from
External Stakeholders:
3:45 p.m.
-
The NRC staff will be available immediately following the meeting to speak with members of the public.
Attachment 1
OVERVIEW OF BULLETIN 2001-01:
"CIRCUMFERENTIAL CRACKING OF REACTOR
PRESSURE VESSEL HEAD PENETRATION NOZZLES"
Allen Hiser
Nuclear Regulatory Commission
Office of Nuclear Reactor Regulation
Division of Engineering
November 8, 2001
Attachment 2
Typical Reactor Vessel Head - Oconee Unit 1 (Babcock & Wilcox)
CRDM
Insulation
.Thermocouple
Typical Location
of PWSCC
/
•.•Vessel
Head
z
-1-
Schematic View of B&W Design
CRDM Nozzle Area
SA-182 F304
ERNiCr-3
(Alloy 82)
SB-167 UNS N06600
(Alloy 600)
-
>
Outer Surface of RPV Head
RPV Head
(SA-533 Gr. B CI. 1)
Shrink Fit
Cladding)
-J-Groove Weld
EniCrFe-3
(Alloy 182)
-2-
OVERVIEW OF BULLETIN 2001-01
Bulletin was issued on August 3, 2001
Bulletin requested information on:
* All plants:
Plant-specific susceptibility ranking
• VHP nozzles (number, type, ID and OD, materials of construction)
• RPV head insulation type and configuration
, Recent VHP nozzle and RPV head inspections
• Above the head structures, missile shield, cabling, etc.
"*
Plants that have found cracking or leakage:
N Extent of cracking and leakage
Inspections, repairs and other corrective actions
• Plans and schedule for future inspections
How plans will meet regulatory requirements
"*
Other plants:
N Plans and schedule for future inspections
• How plans will meet regulatory requirements
-3-
QUALIFICATION OF EXAMINATION METHODS
"*
Verify compliance with regulatory requirements through QUALIFIED examinations
o. Graded approach depending on PWSCC likelihood
,0 Examinations of 100% of all VHP nozzles
4 Based on statistics and no identified preferential cracking tendencies
4 All VHPs - similar materials, etc., only failure consequences vary
"*
Effective Visual Examination
,. Capable of detecting small amounts of boric acid deposits and discriminating deposits from
VHP nozzle and other sources
"*
Plant-Specific Visual Examination Qualification
Plant-specific demonstration that VHP nozzle cracks will lead to deposits on the RPV head
(interference fit measurements, etc.)
, Must be capable of reliable detection and source identification of leakage (insulation, pre
existing deposits, other impediments)
"*
Volumetric Examination Qualification
, Demonstrated capability to reliably detect cracking on the OD of VHP nozzles
Appropriate if Visual Examination cannot be Qualified
-4-
REVIEW OF BULLETIN 2001-01 RESPONSES
Bulletin places PWR plants into 4 groups based on relative susceptibility ranking:
"*
Plants that have found Cracking or Leakage - 5 plants
Suggests qualified volumetric examination by end of 2001
,. Staff accepted qualified visual examination at last outage
"*
Plants with High Susceptibility (within 5 EFPY of Oconee 3) - 7 plants
Suggests qualified visual examination by end of 2001
- Staff accepted qualified visual examination at last outage
"*
Plants with Moderate Susceptibility (between 5 and 30 EFPY of Oconee 3) - 32 plants
Suggests effective visual examination at next RFO
• Staff accepted effective visual examination at next RFE
"*
Plants with Low Susceptibility (more than 30 EFPY of Oconee 3) - 25 plants
Suggests no additional actions required
No requirement to provide plans or schedule
Staff has addressed clarifications to Bulletin responses, and numerous licensees have
provided revised or supplemented Bulletin responses
-5-
PLANTS THAT HAVE PERFORMED "BARE METAL" VISUAL INSPECTIONS
Most Recent Inspection
Plants
Summary of Cracked or Leaking CRDM Nozzles
I
N
Circumferential
Number
T
NNozzle
Cracks
Repaired
Date
Method & Scope
Oconee 1
11/2000
Qualified Visual - 100%
1*
0
1
Oconee 3
02/2001
Qualified Visual - 100%
9
3**
3
ANO-1
03/2001
Qualified Visual - 100%
1
0
1
Oconee 2
04/2001
Qualified Visual - 100%
5
1
5
Robinson
04/2001
Qualified Visual - 100%***
0
0
0
North Anna 1
09/2001
Qualified Visual - 100%***
8
0
0
Crystal River 3
10/2001
Effective Visual - 100%****
1
1
1
TMI-1
10/2001
Qualified Visual - 100%
8*
0
6
Surry 1
(in progress)
10/2001
Qualified Visual - 100%***
10
TBD
5
North Anna 2
10/2001
Qualified Visual - 100%***
(3)
TBD
TBD
(inprogress)
*k
**
Thermocouple nozzles also cracked/leaking: Oconee 1 (5 out of 8), TMI 1 (8 out of 8)
The size of 2 out of 3 circumferential flaws were identified from destructive examination.
Pending acceptability of licensee's supplemental response
****
The highest ranked MODERATE susceptibility plant.
Moderate susceptibility plants that have completed effective visual examinations in Fall 2001 with no evidence of boric acid deposits: Beaver
Valley 1, Farley 1, Kewaunee, and Turkey Point 3
-6-
OVERVIEW OF STAFF PRELIMINARY TECHNICAL ASSESSMENT
Summarizes available data and evaluations related to:
"*
"*
"*
"*
*
Environment in CRDM annulus region
Crack initiation
Crack growth rate
Stress analyses and crack-driving force
Critical crack size
Deterministic assessment
Probabilistic assessment
Inspection timing
-7-
CRDM SUPPORT STUDIES
W. J. Shack
Argonne National Laboratory
November 8, 2001
* Technical Issues addressed by ANL
- Distribution of crack growth rates in Alloy 600 nozzle materials
- Impact of potential crevice environments on expected crack growth rates
- Probabilistic models for initiation of cracks in CRDM nozzles
- Conditional probability of failure for nozzles
- Integrated models to estimate probability of failure including initiation and
growth
Argonne National Laboratory
Crack Growth Rates in Alloy 600
"* Scott's
results on SG tubes suggest strong heat-to-heat variations in CGR
are likely. Measurements on nozzle materials support this expectation.
"* Critical
issue is to estimate range of CGRs that will be encountered in the
population of materials in service, not the range of CGRs in the limited
number of heats being tested
- "Weighting" the data by heat gives a better picture of CGRs of population
than counting all data points as equal (or even weighting data by quality)
-
Better to determine dependence on basic parameters like K, T, by
examining individual tests and test series with better controlled variables
than statistical analysis of too large a data bin
Argonne National Laboratory
1-9
-I
Heat 69
0
0
Crack growth rate as a function of
stress intensity factor K for Alloy 600
Heat 69. For this heat Scott A =
5.54 x 10-12. A is chosen as the
parameter to characterize a heat.
Scott A=5.54 x 1012
10-10
E
(.)
10-11
..0I
10
0
10
I..
20
I .
. I
K (MPa.m
1-
L
a
0
E.
I
I
I
I
11 2
I,
70
60
50
40
30
. I . ..
...
I
da- A(K- 9)1.16
dt
,
80
)
/•IIlI
///
I
I
0.8
/13
0.6
Cumulative distributions of A at 325 0C
for Alloys 600 and 182
=1
0
0.4
3
/0
0
E
0).2
o0
1
o/
S-
--
00
I
I
I
0-1
10-12
I
I
Alloy 600
Alloy 182
I0
10-11
Scott A
Argonne National Laboratory
4d
1
U- 0.8
LL
.0
.
•
"
0.6
0.6
a
a)
0.4
I
1
E,
E
0
0.2
Q
0.8
0 1I
I
I
I
I
I
III
0.2
10.11
0
1011
0-12
10
1o12
0.4
Scott A
Scott A
"U. 0.8
Monte Carlo analysis performed to
assess effect of uncertainties.
Changes in distribution of A are
.0
•
0.6
0.4
E
0.2
E
0.2-
0
Ayrelatively
600 MVC
/Alloy
Alloy
r
I
I
I
I
i
182 MC
small especially in region of
greatest interest.
iil
10.11
10.12
Scott A
Scott A parameter for Alloy 600 nozzle materials at 325 0C
,JAraonne
SJ
National Laboratory
Confidence
Level
50
67
90
95
1.8
2.3
4.2
5.6
Population
90
1.2 x 10-11
1.5 x 10-11
2.4 x 10-11
3.1 x 10-11
95
x 10-11
x 10-11
x 10-11
x 10-11
Percentage
67
5.1 x 10-12
5.8 x 10-12
7.5 x 10-12
8.4 x 10-12
3.1
3.5
4.4
4.8
50
x 10-12
x 10-12
x 10-12
x 10-12
10.8
109
E
v10-11
10-11
10-12
L
0
20
40
60
K (MPa .m11 2)
Argonne National Laboratory
80
100
lu
0
10
1 60
40 K (MPa.m
50
20
/2 )
Intensity
Stress 30
70
80
Crevice chemistries
and head form a tight crevice. FEA analysis suggests 0-4 mil gaps
at pressure. Deposits suggest leakage even for a nozzle with a 1650 crack
was less than 1 gallon over the cycle. Depending on the tightness of the
crack, the tightness of the interference fit, and blockage by deposits or
corrosion products environment can be steam, close to primary water, or
concentrated solutions
"* Nozzle
-
pH changes are limited by precipitation of insoluble species. Industry
calculations with MULTEQ show that depending on location of boiling pH
can become alkaline (~8.6) or acid (~4.6).
-
Because MULTEQ models don't deal with reactions with iron and nickel
components these calculations probably overestimate pH shifts
-
Samples from actual crevices are needed to substantiate these
preliminary conclusions.
shifts can accelerate CGRs by a factor of ~2. This can affect
initiation and throughwall growth of cracks. Once a significant throughwall
crack has formed, crevice has good communication with bulk and water
chemistry is even more likely to be close to primary water.
"* Predicted
Argonne National Laboratory
7
ý`-z
Probabilistic initiation models
o Mechanistic initiation models require more knowledge of local stresses and
material microstructure. Probabilistic models use inspection data
I
1
a
!
0.8
~I
.0
I
Il I,
I IlI~
I
I I
I,
I
I , , II
0
E~
b
, •
F median ranks
Weibull fit
- --....
0.4
{
Lognormal fit
0.2
0
0
b(x
p(t)= - x) expQ
- -QJb
1-exp
F(t) =
0.6
U-
b
Il
L,
0
II
I
1
i
. I
2
. . . I
3
i
b-1
Argonne National Laboratory
I
4
ii
,
ii
i
5
. .
ii
6
Weibull probability density and
cumulative probability functions.
Distribution of the Weibull slope b for
cracking of steam generator tubes.
* Estimates of Weibull parameters for plants that have been inspected and an
associated distribution of values
Plant
95th
Median
5th
Leaks
0.4
2.0
9.3
EFPY
at 6000F
20
20
20
EFPY at 1st initiation
b=3
23
13
8
b=1 .5
28
12
10
0
608.5
209.8
72.3
* Choice of b has impact on when 1st initiation occurs and the credit that can
be expected for shorter operating times
Plant
Argonne National Laboratory
Number of leaks expected
5th
Median
95th
b=1.5
5
10
0.05
0.15
5
10
0.01
0.06
0.25
0.71
b=3
0.03
0.27
1.24
3.46
0.17
1.33
NRC-FUNDED
CRDM STRESS ANALYSIS, CRACK-DRIVING
FORCE, AND LEAK-RATE ANALYSES
by
Engineering Mechanics Corporation of Columbus*
G. Wilkowski, D. Rudland, and Z. Feng
and
ORNL
R. Bass and P. Williams
Presented by
Gery Wilkowski
11/8/01 - NRC/Industry meeting
Involvement to Date
(1) Expert panel assistance on CRDMs started in late June.
Work involved reviewing industry documents and
assisting NRC staff with technical information.
Emc 2 technical efforts involved:
"* Stress analysis aspects - weld residual stress
"* Crack-driving force and crack-opening displacement
"* Leak-rate analyses
"* Critical crack length calculations
ORNL technical efforts involved:
"* Stress analysis aspects
"* Crack-driving force and crack-opening displacement
(2)
Technical assistance for plant specific assessments
(3)
Future CRDM efforts
4TU4•2-
1
Initial Stress Analysis Efforts
Residual Stresses
Some aspects of the review of work to date:
"* The residual stress analysis for this problem is complicated
by highly 3D aspect of the geometry. Industry efforts to
date are good considering the time frame of efforts.
" Some suggested improvements are;
• Weld simulation created a whole ring of elements (one pass)
instantaneously. (Traveling arc has heat sink in all directions not
just normal to weld path.)
* Using elastic-perfectly plastic stress-strain curve will give lower
residual stresses than one with strain-hardening.
Distribution of stresses also affected as well as peak values.
&_ MC'2
3
Initial Stress Analysis Efforts
Residual Stresses
Some suggested improvements, continued;
Effect of weld sequencing is not explored.
Sequencing could be;
the tube to the head in the radial direction
(forces stresses in the weld to be higher either
closer to the tube or the head), which could also
affect OD axial cracking.
•.Welding around the. circumference
or from uphill to
either continuously
downhill side in two half
circumferential steps on opposite sidc
of the tube. (Higher stresses at stop
Single stop-start
Separate stop-start positions
&
,2
4
2
Initial Stress Analysis Efforts
Residual Stresses
- Some suggested improvements, continued;
*:. Mesh refinement in the weld bead could perhaps be finer
SInformal survey of international weld stress analysts showed they
typically used a minimum of 12 to 20 element in a single weld bead
2
cross-section. (Consistent with Emc experience.)
Industry
analyses
Example of
ultra-fine
T"
mesh - Emc2
fAV
,
N
Total of 3,000 elements in weld
TUC,2
t~g5
Recent Stress Analysis Efforts
Thermal Expansion Stresses
Thermal expansion and RPV hole expansion from
pressure loads increase the annular clearance
•
*
Good from leakage viewpoint, but
Contributes to crack-driving force at the root of weld
Coounterbore
6
3
Recent Emc 2 Stress Analysis Efforts
Thermal Expansion Stresses
Stresses from thermal expansion and
pressure without residual stresses - simple
axisymmetric model.
S22
K1 = 18 ksi-in°-5
K2 = 12 ksi-in 0 5
&
411',2C
Recent Emc 2 Stress Analysis Efforts
Thermal Expansion Stresses
Stresses from thermal expansion and
pressure without residual stresses
Principal stress direction at
angle through the thickness?
5
K45 . = 12 ksi-in°
8
411.,C,2
8
4
Initial Stress Analysis Efforts
Cyclic Thermal Stresses
From analysis at Ringhals in early 1990's, there was concern of
cyclic temperatures from water going up and down the nozzle
region.
It was expected that the thermal stresses may not be large
enough to cause fatigue by themselves, but "may be a
contributor to cracking in cold heads".
0
Past gas pipeline work on SCC showed that small cyclic stresses
(R=0.95) can increase the crack growth rate.
&
M41 2
9
Initial Crack-Driving Force Analyses
Efforts conduced both at Emc 2 and ORNL to examine K and
COD. COD used for leak-rate analyses.
"*
Emc 2 analysis was elastic conditions with pressure only, and examined
the effect of restraining the pressure-induced bending from the presence
of a circumferential through-wall crack. No pressure on crack faces.
"* K and COD much lower in restrained condition that simulates CRDM
behavior.
0.o45
-
0.040
i0"035 -Unrestraineddi
/
.-
•U-usd
-U-Outside
0.030
6
_ 0.025
U-Middle
...--'"U-Inside
0.020 -
----
R-Outside
e
R-Middle
R-Inside
estraned
0015
"0.010
r0.005
o 0.000
0
50
100
150
200
Crack Length , degrees
250
300
M
10
(9010
5
Initial Crack-Driving Force Analyses
Efforts conduced both at Emc 2 and ORNL to examine K and
COD, continued
" ORNL effort used gap elements to restrain bending and was elastic-plastic
with pressure loading only, including full pressure on crack faces.
" Similarly showed lower COD and K values with restrained bending
particularly for longer cracks.
Refinr.d Men. B
8015 elemnents
,31,.629
80
70
.-d.
ORNL
E=2E (Unrestrainedcase and
no pressure on crack face)
60
7
ORNL (Restrained case and
p=2.250 psig on crack face)
50
5•
-Ec (Restrained case and
30
no pressureoncrackface)
20
0
0
50
100
150
200
250
2
300
Total Crack Angle, degrees
1l
Initial Crack-Driving Force Analyses
Efforts conduced both at Emc 2 and ORNL to examine K and
COD, continued
-
ORNL also examined effect of applying a residual stress equal to
yield in a simplistic manner, i.e., displacement-controlled axial
tension stress on tube
;residual+ pressure
'r
':pressure
J....
only~
(9 V,
12
6
Initial Leak-Rate Analyses
Efforts at Emc 2 using COD from Emc 2 and ORNL, continued
oFirst calculated a leak-rate for a circumferential through-wall crack
"Used statistical mean crack morphology parameters (roughness, number of
turns) for an IGSCC
"- Assumes no back pressure at exit plane
" Determined leak-rate (0.22 to 0.44 gpm) as well as pressure and temperature
of water exiting the crack plane for 180-degree crack. (127 psig and 347 F)
1.4
1
ORNL analysis
--
-$---2,250 psi crck-lace pressure
-1 - no crack-face pressure
1.2
El
Emc2 analysis
0.6
-j
0.4
0.2
0
4_.•1,Z
360
270
180
90
Circumferential Crack Angle, degrees
0
2
13
Initial Leak-Rate Analyses, continued
Calculated the leak-rate through the annular region (0.1 to 0.3 gpm)
Assumed 180-degree crack exit plane water is entrance water in annular
plane
SAssuming a radial gap of 1.2 mils on diameter (close to industry stated value)
SUsed
roughness for either drilled or reamed holes in annular area.
-- - 1em.
34SF
S3rrrc,127psi.34SF
Fps..
- .-
CflokodF-*
1.a
cŽ.
L/rcmkefcmked--
//C
0.co
0D.6c
0
0.OcS
s.c1
0ccl2
c.cc14
0=cc6
Radial Cleaar-e Beteeeta Rod and Head.Inch
9
IYUC,2
14
7
Initial Leak-Rate Analyses, continued
* Determined that for this crack size the annular leakage was limiting
* The calculated leakage rate, however, was about 24,000 time greater than
the Oconee 165-degree crack
* Could be explained if the 165-degree crack all the way through the
thickness or only over a short length?
Leakage only at a few small locations
along crack, rather than whole 165
degree length?
• Residual stresses causing crack faces to rota e, and hence pinch off flow?
• Plugging occurring at low leak rates?
15
Initial Review of Industry Work
"•Industry efforts underway are impressive and
involve a significant undertaking considering the
time frame involved.
"*Some suggested improvements if more time was
available
Redistribution of residual stresses may not be properly
handled by putting load-controlled stresses on the crack
faces from the uncracked model.
• Need to map 3D stress and strain field from weld model onto the
fracture model to properly determine the redistribution of stresses
with crack growth.
MAC, •
16
8
Review of Industry Work
Some suggested improvements if more time was
available, continued
It appears that longitudinal (axial stresses) in tube
were applied to crack face.
.
•
Crack was in helical direction, so stresses normal
to that direction should be used.
Examination of weld model may show that crack
growth may not be in a plane normal to the
thickness, i.e., the principal stresses may be
different than the axial direction or normal to the
helical plane of the crack.
•4'Y-C 2
17
Review of Industry Work
Some suggested improvements if more time was
available, continued
" Weld sequencing effects on crack-driving force not examined
* Could have high stress spots at 0 and 180 degree locations
"Needto
examine K from toe of weld as well as K from through
wall crack.
"* Will there be multiple initiation sites?
"* What is the crack growth rate in the radial versus circumferential
directions?
"* Could a complex crack form (long surface crack with through
wall crack of shorter length)?
"•t~c" 2
18
9
NRC-FUNDED
CRDM CRITICAL CRACK SIZE ANALYSES
by
Engineering Mechanics Corporation of Columbus*
G. Wilkowski, D. Rudland, and Z. Feng
Presented by
Gery Wilkowski
11/8/01 - NRC/Industry meeting
Involvement to Date
(1) Expert panel assistance on CRDMs started in
late June. Work involved reviewing industry
documents and assisting NRC staff with
technical information.
Emc 2 technical efforts involved:
*
*
*
*
Critical crack length calculations
Stress analysis aspects - weld residual stress
Crack-driving force and crack-opening displacement
Leak-rate analyses
••~2
I
Critical Crack Analysis
Limit-load analyses examined to;
"* Determine proper limit-load boundary conditions,
"* Determine flow-stress definition?
" Determine if toughness of Inconel 600 is sufficient for
limit-load to be used?
" Conduct analyses for CRDMs with ideal
'.
through-wall crack, surface crack, and
complex crack.
,,
N
rnm-ArrOd CroackFbmd m Service
&
-1,2 3
Determine Proper Limit-Load Conditions
Solutions exist for axial tension on a cylinder with a
circumferential crack
• End-capped solution most common, but allows for free rotation
of cylinder ends due to pressure-induced bending from presence
of crack.
CRDM tube restrained from bending by RPV head
.......
.......
.......
.............
....
.. .
& 4ThC
4
2
Determine Flow-Stress Definition
Axial tension tests with restrained-bending on circumferential
through-wall-cracked 4" diameter stainless steel pipe
conducted in past.
• Flow stress
=
(yield + ultimate)/2.4, which is less than average
Flow strml empirically determined to be (yleld+ ultimaNey2.4
ýeY,ga -FR
~
det.
*0
0.7
0.6
~o
0.1
02
0
45
130
90
180
225
360
315
270
Total crack length, degrees----
2
5
Determine if Inconel 600 Has Enough
Toughness to Use Limit-Load Solution
" Dimensionless Plastic-Zone Screening criterion developed to
determine toughness requirement for using limit-load analyses.
" From PIFRAC database Inconel 600 JIC = 9,310 in-lb/in 2 (1.63
MJ/m 2), which give plastic-zone parameter of 3.5 and limit-load
should work.
- ..
S--
.-
D--:,. N9
4TUC,26
3
Limit-Load Analysis with Crack-Face Pressure
Longer cracks are more affected by crack-face pressure
idealized through-wall crack example.
40.000
Using Inconel 600 properties
35.000
Using typical
at 0F
properties
actual strength
30,.000-
crack length at
2•.000Critical
3 times operating pressure
0.
crack lesign
Crtia.............
31~~~~~
eraesUdeinpan.es.s0P5
sg<
atsue(,O
25,000
t15.000
3.000
3Design
pressure
(2,5a0 psg)
length, degrees
0
Total
40
0-
180
135
through-wall-crack
225
30 0
315
270
CRDM Critical Crack Length Calculations
20,000
Using Inconel 600 Code properties at 600F
18,000
16,000
.
att=0.25 ••/
10.000
.
Idealized TWC
Cornplexcracks.
14,000
12,000
"Q.
()
.-
a/t-0.50
10,000
. ..........
:.........
"+............
. ...........
3 times designprsre(,0psg
............................
... .... .... ... ......................
2 ....
sas
-"
.... ......... .............
6.000
400 aAt=0.75
L/"'
2..
M.++
----------
4,000
200
0
45
+o+
+o&
90
135
180
225
270
315
360
Total through-wall-crack length, degrees
4
CRDM Critical Crack Length Calculations
0.7
4.
0.6
")
0.5
0.4
Q
t
0.3
0.00
45.00
g0.00
135.00
180.00
225.00
270.00
Total through-wall-crack length, degrees
315.00
;;; 'Y1"
360.00
2
Critical Crack Length Analyses
Summary
*
*
*
*
*
*
Restrained-bending limit-load solution is more appropriate for CRDMs.
Flow stress = (Y+U)/2.4 from pipe tests with similar loading (lower than
typical average of yield and ultimate definition).
Pressure on crack-face important for longer cracks.
Toughness of Inconel 600 high enough to use limit-load solution.
Ideal TWC critical length at 3*design pressure = 262 to 269-degrees
(slightly less than industry calculated value), but very fracture resistant
material.
Critical surface crack/complex crack/ideal through-wall crack lengths at
design pressure given.
Critical 360-degree surface crack of constant depth would be
90-percent of thickness. Difficult to get such a crack geometry
without getting some through-wall component.
&
41~L-to
5
DETERMINISTIC AND PROBABILISTIC ASSESSMENTS
Allen Hiser
Nuclear Regulatory Commission
Office of Nuclear Reactor Regulation
Division of Engineering
November 8, 2001
STAFF CONCLUSIONS
Annulus Environment
"* Not expected to be highly aggressive - normal PWR reactor coolant
"* Annulus deposits from leaking nozzles should be obtained and analyzed by industry to
provide confirmation of the assumed annulus environment
Crack Initiation
"* The operating experience of leaking nozzles appears to be well modeled by the
Weibull analysis with b = 1.5
"* New findings data will continue to be assessed
Crack Growth Rate
* Crack nrowth rate data for PWSCC is a reasonable annroximation for OD VHP nozzle
cracking
0 Analysis of data provided in Table 3 is appropriate for use at 3250C (617'F)
0 The Arrhenius relation can be used for crack growth at other temperatures
-2-
STAFF CONCLUSIONS (cont.)
Stress Analysis and Crack-Driving Force
* A single estimate for K as a function of circumferential crack length was provided (with
a value of 66 MPa /--m (60 ksiv/ in.) due to residual stresses for a crack angle of 900)
Critical Crack Size
"* Critical size with a safety margin of three on pressure is 2700
is 3240
"* Critical size for nozzle failure and possible ejection
-3-
DETERMINISTIC ASSESSMENT
Base Case - Assumptions
"*
Critical Flaw Size
,. 2700 with a safety margin of three on pressure
,. 3240 for nozzle failure and possible ejection
"*
Crack Growth Rate
,. 95/50 statistical bound
3180C (6050F)
A for Scott model is 1.303 x 10-11
•
"*
Initial Flaw Size
0. Unknown - basis for issuance of the Bulletin
No. Used as a parameter
Uncertainties and Sensitivity Studies
"*
"*
"*
Different statistical bounds to crack growth rate
Effects of temperature on crack growth rate
Initial flaw size as a parameter
-4-
60
50
",
0
*• 30
20
.,--
200
10
0
o
-,-"
.......
K Estimated
KSIA
K ORNL pressure
0
0
50
100
150
200
Crack Half Angle
Figure 15
Estimated stress intensity factor K for a CRDM nozzle based on SIA and
ORNL results.
-5-
Table 4
Summary of OD Circumferential Flaws Identified in
Spring and Fall 2001 Outages
Plant
*
Nozzle
Circumferential
Through-Wall
ID
Crack Length
Extent
Oconee Unit 3
50
1650
100%
Oconee Unit 3
56
1650
100%
Oconee Unit 3
23
660 *
35% *
450*
18
Oconee Unit 2
900 *
32
Crystal River Unit 3
Crack dimensions estimated from UT data.
10% *
50% *
-6-
CRDM NOZZLE CRACK GROWTH RATE
(Base Case, 3180C, 95/50)
3.0
cz
2.5
2E-9
-
2.0
-----
\,-------
.. . . . -- - - - - - .-. --. .-. -. .-. -. .-. -. . .
--------..
-----
----
E
-------
. . . . . -. . -. . -.',.. . . . .
--------------
(I
O1
M
0
1.5
--------------
0.
1E-9
-----/---------- -------
-6
U
1.0
0.5
U.U
1
0
60
120
180
240
300
360
U
Circumferential Crack Length (Degrees)
Figure 18
Variation of crack growth rate with circumferential crack length for the
base case of 318'C (605 'F) 95/50 curve.
-7-
FAILURE TIME EVALUATION
(3180C, 95/50)
(I)
360
0)
ci)
ci)
300
0)
240
0
-J
0
180
0
6
0 120
60
0
0
12
24
36
48
Operating Time (months)
Figure 19
Variation of time to failure as a function of initial crack length, for the
base case of 318 'C (605 TF), 95/50, crack growth rate.
-8-
60
CRACK GROWTH EVALUATION
Cl)
0.le
(Base Case, 3180C, 95/50)
360
300
Q
240
0(D
E)
180
0
12
24
36
48
60
Operating Time (months)
Figure 20
Evaluation of operating time to reach critical flaw sizes at three times
design pressure and at nozzle failure/ejection after development of a
1650 long circumferential through-wall flaw.
-9-
EFFECT OF OPERATING TEMPERATURE ON 'A'
Operating Temperature (OF)
550
625
600
575
1.0
1.0
)
0
LL
0
Cq
coJ
0.8
0.8
0
©
E, 0.6
0.6 <
E
I-
N
(0
0.4
0.4
0.2
0.2
0.0
C.
E
(
0.0
280
290
300
310
320
330
Operating Temperature (°C)
Figure 21
Lower operating temperature results in lower crack growth rates for VHP
nozzle materials, within the operating temperature range of the nozzles.
-10-
CRDM NOZZLE CRACK GROWTH RATE
3.0
co
2E-9
0OC
318 0 C (B)
-~
0
o
(5
.%
1.0
"E
0
/
1•
315 0C (M) I
0.5
0).0
0.0
'
0
I
60
a
I
120
,
I
I
,
180
240
,
I
300
,
I
0
360
Circumferential Crack Length (Degrees)
Figure 22
Variation of crack growth rates at several pertinent temperatures and using 95/50 ('B'
on the curves) and mean values ('M' on the curves).
-11
-
CRACK GROWTH EVALUATION
U,
360
a,
300
a,
I-
240
180
0
12
24
36
48
60
72
84
96
Operating Time (months)
Figure 23
Crack growth analysis using various crack growth rate assumptions, from
an initial flaw size of 165(. Although decreasing the temperature has
some effect, the most dramatic increases in failure times occur with the
mean crack growth curve instead of the 95/50 curve.
-12-
TIME TO 3 X DESIGN PRESSURE
C')
360
a)
0)
0
-0
-j
C
300
240
180
120
C:
60
0
0
24
48
72
96
120
Operating Time (months)
Figure 24
Comparison of time to reach the flaw size representing three times the design pressure,
for a variety of crack growth rates and as a function of initial flaw size.
-13-
TIME TO NOZZLE FAILURE/EJECTION
360
(D
0)
300
240
180
0
CZ
120
60
0
0
24
48
72
96
120
Operating Time (months)
Figure 25
Comparison of time to reach the flaw size representing three times the design
pressure, for a variety of crack growth rates and as a function of initial flaw size.
-14-
CONCLUSIONS FROM DETERMINISTIC CALCULATIONS
Results are very sensitive to:
"*
"*
"*
Initial flaw size
Statistical bound on crack growth rate
Temperature
Traditional safety margins may not be sufficient to account for large variability in crack growth
rates for Alloy 600 in PWSCC conditions
-15-
PROBABILISTIC ASSESSMENT
"*
A Complete Phenomenological Model
0. Requires a better understanding of the complete cracking process
,. Requires data to characterize critical parameters (means & bounds)
"*
Empirical Model
, Based on reliable data on number and size of cracks found in service
Qualification of NDE sizing an issue
Cost of destructive confirmation large
* Need to determine Frequency of Failure to estimate Core Damage Frequency
-16-
INSPECTION TIMING
Likelihood of Circumferential Cracking
High susceptibility plants - 8 out of 9 have identified cracking
, Moderate susceptibility - effective visual examinations will provide additional data
High Susceptibility Plants That Have Performed Effective Inspections
, Can use Figures 23 to 25
New circumferential cracking can initiate
High Susceptibility Plants That Have NOT Performed Effective Inspections
,. Need baseline inspection to provide basis for evaluation
Inspection Method
Qualified visual examination is appropriate
, Surface or volumetric examinations
Inspection Scope
100 percent of nozzles
, Entire surface or metal volume of interest
", "Wetted surface" - J-groove weld, nozzle OD (below the weld), and nozzle ID to a
location above the weld
, Volumetric - OD of nozzle above the J-groove weld
, Visual qualification analysis can occur ex-post facto after the inspection
-17-
FUTURE STAFF PLANS
"*
Continue development of probabilistic modeling
"*
Complete review of Bulletin supplemental responses
"*
Assemble findings from inservice inspections
"*
Issue NUREG report
"*
Long-term inspection plans
-18-
INDUSTRY INTERACTIONS
"*
Interactions on deterministic and probabilistic analyses
"*
Inspection methods and findings
"*
Destructive confirmations
,.
Flaw sizes
,.
Annular conditions
-19-.
-"P4- NRC Meeting with Nuclear Energy Institute, Materials Reliability Project
Operating Pressurized Water Reactor Licensees
.•and
CO
0.
rThursday,
November 8, 2001
1:00P.M. - 5:00 P.M.
0
I'IUU IIh
UlIIIIII I.IJIIIO
Orcaanization/Title
Name
[llu0aa III1I rI"UII
m
Phone Number/Email
Jake Zimmerman
NRC/NRR/DLPM - Lead Project Manager
(301) 415-2426, [email protected]
Allen Hiser
NRC/NRRIDE/EMCB - Lead Technical
Reviewer
(301) 415-1034,
alhl @nrc.gov
Jack Strosnider
NRC/NRR/DE
(301) 415-3298
Bill Bateman
NRC/DE/NRR/EMCB
(301) 415-2795
Keith Wichman
NRC/NRR/DE/EMCB
(301) 415-2757
Andrea D. Lee
NRC/NRR/DE/EMCB
(301) 415-2735,
adwl @nrc.gov
Jay Collins
NRC/NRR/DE/EMCB
(301) 415-1038
Nilesh Chokshi
NRC/RES/DET/MEB
(301) 415-0190
Ed Hackett
NRC/RES/DET/MEB
(301) 415-5650
Wallace Norris
NRC/RES/DET/MEB
(301) 415-6796
Shah Malik
NRC/RES/DET
(301) 415-6007
Jin Chung
NRC/NRR/DSSA/SPSB
(301) 415-1071
Ian Jung
NRC/NRR/DSSA/SPSB
(301) 415-1837
Giovanna Longo
NRC/OGC
(301) 415-3568
Darl Hood
NRC/NRR/DLPM/PDIII-1
(301) 415-3049
Tim Colburn
NRC/NRR/DLPM/PDI-1
(301) 415-1402
K.N. Jabbour
NRC/NRR/DLPM/PDII-2
(301) 415-1496
John Goshen
NRC/NRR/DLPM/PDII-2
(301) 415-1437
Dan Collins
NRC/NRR/DLPM/PDI-1
(301) 415-1427
Brendan Moroney
NRC/NRR/DLPM/PDII-2
(301) 415-3974
Ujagar Bhachu
NRC/NRR/DLPM/PDII-2
(301) 415-3271
R.L. Clark
NRC/NRR/DLPM/PDI-1
(301) 415-2297
Ray Wharton
NRC/NRR/DLPM/PDIV-2
(301) 415-1396
Attachment 3
(614) 459-3200
Gery Wilkowski
Engineering Mechanics Corp. of Columbus
W.J. Shack
Argonne National Lab
Scot Greenlee
American Electric Power (AEP)
(616) 697-5728
Dan Garner
AEP
(616) 466-3419
S.P. Moffitt
FENOC
(419) 321-8222
Guy Campbell
FENOC
(419) 321-8588
David Lockwood
FENOC
(419) 321-8450
David Geisen
FENOC
(419) 321-8109
Ken Byrd
FENOC
(419) 321-7924
Robert Enzinna
Framatome ANP
(434) 832-2418
Stanley Levinson
Framatome ANP
(434) 832-2768
Peter Scott
Framatome ANP
(33) 147963577
Stephen Fyfitch
Framatome ANP
(412) 264-1610
Ken Youn
Framatome ANP
(434) 832-3280
Alex Marion
NEI
(202) 739-8080
Gretchen Testaye
Calvert Cliffs
(410) 495-3736
Dan Salter
HGP, Inc.
(864) 370-0213
Dick Labott
PSEG - Salem
(856) 339-1094
R. Hermann
SIA
(540) 710-6717
Bob Hardies
CCNPPI
(410) 495-6577
Jim Meister
Exelon
(630) 657-3800
Altheia Wyche
SERCH Licensing/ Bechtel
(301) 228-6401
Tom Harrison
McGraw-Hill
(202) 383-2165
Harold Chernoff
CP&L
(843) 857-1437
Shataro Mori
The Konsai Electric Power
(202) 659-1138
Paul Gunter
NIRS
(202) 328-0002
Deann Raleigh
LIS, Scientech
(301) 258-2557
Roger Huston
Licensing Support Services
(703) 671-9738
H. Fontecilla
Dominion
(703) 838-2314
Roy Lessy
(202) 887-4500
Attachment 3
PHONE PARTICIPANTS
Organization/Title
Name
Phone Number/Email
Dick Mattson
Structural Integrity Associates
(408) 978-8200
Daniel Stenger
Ballard, Spahr, Andrews &
Ingersoll, LLP
(202) 661-7617
Robert Lemberger
Florida Power, Crystal River 3
(352) 795-6486, x3862
Stephen Collard
FPL
Michael Moran
FPL
Terry Pickens
Nuclear Management
Company, LLC
Donald Bemis
CMS Energy
Richard Gerling
CMS Energy
Ed Siegel
Westinghouse
George Lavigne
NAESCO
Jeffrey Sbotka
NAESCO
Kevin Whitney
NAESCO
Scot Sulley
NAESCO
James Connolly
NAESCO
Yogen Garud
APTECH Engineering Services,
Inc
(408) 745-7000, x3060
Mark Fleming
Dominion Engineering, Inc.(DEI)
(703) 790-5618, x239
Glenn White
DEI
John Crane
Westinghouse
Greg Gerzen
Exelon
(630) 657-3845
Christine King
EPRI
(650) 855-2605
(715) 377-3390
(603) 773-7126
Frank Ammirato
fammirat@ epri.com
Ron Baker
[email protected]
Warren Bamford
bamforwh @westinqhouse.com
Jim Bennetch
jim [email protected]
Dave Berko
Daveberkode @inpo.orcq
Prasanta Chowdhury
pchowdh @enterqy.com
Attachment 3
Kurt Cozens
koc @ nei.org
John Hall
iohn.f.hall@ us.westinghouse.com
John Hamilton
[email protected]
Craig Harrington
charrinl @txu.com
Larry Mathews
[email protected]
Gary Moffatt
[email protected]
Ben Montgomery
[email protected]
Donald Naylor
[email protected]
Raj Pathania
RPATHANI@ epri.com
Jeffrey Portney
ilp4 @pge.com
Mike Pugh
[email protected]
Eric Schoonover
schoonei @ soncqs.sce.com
Michael Shields
mike shields@ rae.com
William Sims
[email protected]
Ronald Swain
rswainl @enterqy.com
Chuck Tomes
[email protected]
Vaughn Wagoner
vaughn.wagoner@ pgnmail.com
Joseph Weicks
iweicks@ enteroy.com
Attachment 3