IR 05000324/1993043
| ML20059G504 | |
| Person / Time | |
|---|---|
| Site: | Brunswick  |
| Issue date: | 10/13/1993 |
| From: | Blake J, Coley J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| To: | |
| Shared Package | |
| ML20059G483 | List: |
| References | |
| 50-324-93-43-01, 50-324-93-43-1, 50-325-93-43, NUDOCS 9311080173 | |
| Download: ML20059G504 (9) | |
Text
-
.
UNITED STATES
[. p meo%,
NUCLEAR REGULATORY COMValSSION y ".-
-t REGION 11
E 101 MARIETTA STREET, N.W.. $UITE 2900 g
p ATLANTA, GEORGIA 303234199
\\...../
Report Nos.: 50-325/93-43 and 50-324/93-43 Licensee:
Carolina Power and Light Company P. O. Box 1551 Raleigh, NC 27602 Docket Nos.:
50-325 and 50-324 License Nos.: DPR-71 and DPR-62 facility Name: Brunswick I and 2 Inspection Conducted:
September 11-13 and 20-26, 1993 Inspector: \\[
J. LMoley Jr.
g'\\ _
/O - f -73 cr a
Date Signed Accompanying Personnel:
S. R. Doctor, Ph.D., NRC Consultant Battelle Pacific Northwest Laboratories Approved by:
-
u-
/
3
. _.
J. J./Blake, CMef Date Signed t#ials and Processes Section-ngineering Branch
Division of Reactor Safety l
l SUMMARY Scope:
'f
!
This special announced inspection was conducted in the area of ultrasonic
.
sizing for depth, crack indications which had been visually identified in the Unit I reactor vessel shroud.
The September 11-13, 1993, portion of this inspection was conducted at the General Electric (GE) Nuclear Energy facility in San Jose, California. The inspection conducted in San Jose, CA. was performed to determine whether GE's Smart 2000 automated data acquisition
-
system and underwater scanning fixture could accurately size notches in a
.
mockup section of the Brunswick reactor vessel shroud. The demonstration also
!
-would determine GE's ability to repeatedly position the scanner through vessel
'
limitations and-attach it at a precise location on the shroud, which was-located 50 feet underwater.
\\
Results:
'
In the areas inspected, violations or deviations were not identified, f
l
!
93110B0173 931020
$e
]
<
PDR ADOCK 05000324 G
PDR i
-
-
.
'
.
.
Since the ultrasonic equipment, inspection techniques, and inspection proce-
'
dures were in the process of development to specifically deal with plant configurations and limitations, some problems were encountered during the demonstration at San Jose and at the Brunswick facility. However, management and technical representatives from Carolina Power & Light (CP&L), and GE i
response to these problems was considered a strength by the inspector.
J Another strength observed by the inspector was the ability of the GE level III, data analyst, to consistently and correctly resolve computer screen presentations involving abnormalities as well as crack tip signals. Several
'
weaknesses however, were identified by the inspector in the inspection procedure, and a data acquisition examiner was found to be unfamiliar with the
~
requirements of the inspection procedure during calibration of the system.
l The licensee and GE took immediate corrective action which included revising the inspection procedure and conducting training for all personnel involved in the sizing examinations. This training was audited by the inspector to insure
,
that the requirements and techniques delineated in inspection procedure were adequately covered and comprehended by the examiners.
'
'
l
i
.i l
l
(
-
.
.
,
.
REPORT DETAILS 1.
Persons Contacted
Licensee Employees j
i
- R. Anderson, Vice President
- E. Black, Nondestructive Examinat'on (NDE) Level III r
- B. Grazo, Manager, On-site Nuclear Engineering Department
- C. Hinnant, Director of Site Operations
- L. Langdon, NDE Supervisor
- W. Levis, Manager, Regulatory Affairs
- J. McGowan, Specialist, Regulatory Affairs
!
- G. Miller, Manager, Technical Support
- C. Osman, NDE, Corporate Manager
- M. Turkal, Manager, Licensing Other licensee employees contacted during this inspection' included engineers, technicians, and administrative personnel.
Other Organizations General Electric Nuclear Energy (GENE)
- T. Brinkman, Project Manager, NDE
- T. Brinkman, Level III Examiner
'
- R. Cowan, Manager, BWR Technology
- E. Dykes, Manager, NDE Development l
- G. Gordon, Design Engineer
~
- S. Ranganath, Design Engineer
- J. Self, Manager Inspection Services
- J. Tung, Engineer, NDE Development
Electrical Power Research Institute (EPRI)
'
- G. Selby, NDE Engineer
NRC Resident Inspectors
'
- P. Bryon, Resident Inspector
- R. Prevatte, Senior Resident Inspector t
- Attendeo exit interview on September 24, 1993
_
-
- Attended exit interview on September 24 & September 13, 1993 i
- Attended exit interview on September 13, 1993
,
2.
Inservice Inspection - Ultrasonic Sizing of Cracks in the Unit I Reactor Vessel Shroud - (73753)
During the 1993 Brunswick Unit I refueling outage.a circumferential
crack that appeared to extend 360* around the ID surface of the shroud was revealed by the in-vessel visual examination. The circumferential crack is located in the heat affected zone on the inside surface of the top guide support ring above the weld which connects the shroud to the
!
i
._
-
.
.
.
.
'
support ring. This weld is designated H-3 at this plant. The visual examinations also detected a 1 inch long axial crack on the outside of the shroud adjacent to circumferential weld H-4.
H-4 is located in the central core region of the shroud. Ultrasonic examination was performed on the shroud OD and ID surface with a transducer mounted on a pole and monitored with an underwater camera. Ultrasonic crack depth measurements taken on weld H-3 ranged from.180".to.400" and measurements taken on H-4 were.250".
For further background
,
infonnation and previous inspector surveillance of this problem, refer
>
to Region II Inspection Report No. 50-325,324/93-34.
,
!
A boat sample was taken at weld H-3 and the sample revealed a crack
depth greater than.800" with the crack still running in the area where the sample was taken. A boat sample was also taken at weld H-4 and this sample revealed a crack depth of.600".
The cracking mechanism was also identified from the boat samples to be intergranular stress corrosion cracking (IGSCC).
With the large difference between the boat sample depths and the ultra-sonic depths a decision was make by CP&L to re-examine the shroud welds
with an automated ultrasonic system similar to one that had been used by a foreign utility.
CP&L's goal was to accurately examine the H-3 crack to the maximum extent possible using the backscatter tip diffraction technique as the primary examination method. A forward scatter time of flight technique
-
(T0FI) was also planned to be used to provide confirmatory information.
.
The examination was to be performed utilizing a remote scanner developed specifically for this application. The scanner used position feedback devises which facilitate the use of GE's Smart 2000 multi-channel imaging system. The scanner was designed to examine a vertical length
of 4 inches and a horizonal leagth of 12 inches. Three scanners were manufactured in order that 2 scanners could perform examinations simultaneously and a third could be used as a spare. The scanner
,
allowed for the examination to be performed from above and below the weld using 45* and 60' refracted longitudinal (RL) wave transducers.
!
The flaw would be sized by analyzing the crack tip diffracted signals.
A late modification to the scanner also attached the tandem pair of time-of-flight transducers (T0FT) transducers to the scanning fixture.
.t The T0FT utilizes two angle beam search units operating in the pitch-catch mode. The transducers are positioned astride the crack and'
pointing at each other. The depth of the crack is calculated from the transit time of the signal diffracted at the crack tip. Another feature
,
of this technique is that the shape of the crack face can be seen and sized by its shadow in the acoustic noise of the weld. This feature was
!
of particular interest to the inspector because of its potent *al to shadow small facets of secondary cracking near the crack tip.
j
i i
_
_.
_
-
.
.
.
GE was responsible for the development of the scanner, the data acquisi-tion system and the inspection procedures.
In addition, the GE Smart 2000 operators and analysts that conducted the demonstration on the mock-up specimen of the Brunswick reactor vessel shroud specimen would
!
be the same individuals that would be used at the Brunswick facility.
j On September 11, 1993, the inspector and NRR consultant arrived at GE's
!
San Jose facility, to observed GE demonstrate their complete ultrasonic
.
system on a mockup of the Brunswick shroud H-3 weld and support ring t
configuration. During the entrance meeting with senior CP&L and GE management the inspector was informed that the demonstration would not be conducted until September 12, 1993, because the examiners needed
additional time to practice with the system and to analyze the
.
inspection data.
The additional time was also helpful for the inspector
!
and consultant since more time could be spent observing the examiners
practice acquiring data with the equipment and evaluating the computer presentations. The time also allowed the inspector to address the
specific concerns delineated below:
i
'
Determine how GE proposed to position the scanning mechanism
in order to obtain accurate X/Y measurements that could be reproduced during a subsequent inspection.
,
Determine how stable of the scanning mechanism was and the
extent of inspection overlap coverage each transducer would
!
'
obtain in the horizonal position.
I Determine if the speed of scanning used on the mockup was
the same as that described in the inspection procedure
,
,
.
Determine if plant limitations and configurations would be
properly dealt with in the demonstration.
.
'
Determine whether the calibration block configuration was
representative of the part to be examined.
Determine whether the system and personnel could properly i
size the notch reflectors.
Determine whether the scanner could be adapted for use on
,
other shroud weld configurations.
'
.
On September 12, 1993, the demonstration was conducted, data obtained, j
and a preliminary review of the data by GE indicated that everything had i
gone as planned.
However, questions raised by the. inspector concerning j
limitations at Brunswick that had not been demonstrated resulted in GE repositioning the scanner and taking additional data. This data revealed that the lower 60* transducer was not acquiring data. This i
transducer would be the most important transducer for sizing deep cracks if the stroke of the transducer fixture was limited because of plant limitations and these limitations were expected to occur.
In addition,
_
._
,
_ _ _.
.,
.
.
'
.
.
when GE re-verified the data taken during the demonstration it revealed that the 60* transducer "looking up" had not seen the notch reflectors during the demonstration.
In addition to the above finding, the inspector addressed the concern
,
that the demonstration was not conducted in an environment which simulated real plant conditions for positioning the scanner (such as manipulating the scanner 50 feet underwater through the top guide and applying it properly on shroud support ring). The inspector was also
,
concerned that GE had not demonstrated how they intended to mark the scanner's position on the shroud in order that the examination could be reproduced during subsequent examinations. The fourth concern addressed by the inspector was that the inspection procedure was still in the process of development and GE's response to several questions raised by the inspector was that the inspection procedure would address the concern.
During the exit meeting with senior CP&L management on September 12, 1993, each of the above concerns were discussed. As a result of these
'
discussions, the inspector concluded that if the inspection was extended by one day GE could correct the equipment problems and perform a demonstration that would satisfy the addressed concerns. Therefore, the
>
inspection was extended and with the exception of not having completed the inspection procedure, GE was successful in effectively dealing with each of the inspector's concerns.
On September 20, 1993, the inspector and the NRR consultant arrived at the Brunswick site to observe GE perform the sizing examinations on the Unit I shroud. The examination equipment had arrived from San Jose and was in the process of being setup.
During the delay the inspector and the NRR consultant reviewed a draft cqy of the automated ultrasonic
,
inspection procedure (GE-UT-222) to determine if concerns addressed during the demonstration at San Jose had been incorporated into the procedure. As a result of this review the following comments were identified and corrections incorporated into the procedure by the licensee:
'
Paragraph 1.1 under procedure scope: The procedure should
make this procedure applicable to irradiation assisted
.
'
stress corrosion cracking (IASCC) as well as intergranular -
stress corrosion cracking (IGSCC) since both cracking mechanisms are apparently involved.
Paragraph 3.5 should be added to delineate the
qualifications of the licensee's consultants who were to i
assist in the analyses of the time-of-flight data.
Paragraph 4.3.2 which described the refracted longitudinal
i too general in that it allowed any size or MHZ transducer to
'
search units to be used for the time-of-flight technique was be used. This examination technique had been successfully
,
demonstrated at San Jose with t inch diameter X 10 MHZ 60'
_
.
.
-
.
,
refracted longitudinal wave transducers. Therefore, these transducers should be used; or, if for any reason SE intended to deviate, a nominal value should be delineated in the inspection procedure.
Paragraph 6.1 only described the scan axis for the pulse
echo transducers. The scan axis for the time-of-flight technique was not described but would be at right angles to
the pulse echo.
Paragraph 6.3 define the scanning rate as 3 inches per
second. However, during the demonstration GE only used a scanning speed of % inch per second.
Paragraph 6.4.1 entitled, " Scanning Sensitivity," should
clarify when and how the lowering of gain to an acceptable level will occur.
,
P On September 22, 1993, GE started calibrations for the shroud at cell 46-11. The calibration block used was the mockup of weld H-3 that had been used in San Jose. This block had 5 notch reflectors ranging from %
inch to 1% inches in depth. The 1% inch reflector represented a flaw
greater than the critical flaw size allowed for the reactor shroud by GE's design engineers. The crack was expected to grow.6 inch during the next 18 month fuel cycle and the total material available was 2.2 inches. GE had previonly stated that, an average of.2 inch of material was needed 360* to maintain an acceptable safety limit.
As a result of a late decision to add the time-of-flight transducers to the scanning fixture instead of performing a separate examinatinn to
,
confirm the pulse echo tip diffractio cxaminations, the scanning fixture now had 6 transducers and lateral movement of the transducers had been reduced to 6.8 inches. This meant that a distance of only 2.1 inches would see all 6 transducers. After the system had been calibrated the inspector requested that the examiner verify the
'
calibration. As a result of this reverification, the inspector discovered that the examiner could not have calibrated on the reflector
'
that he thought he had calibrated on because the surface distance for this reflector was incorrect. CP&L's NDE technical director was notified and the calibrations were terminated until I hcurs of training was conducted for all examiners and other personnel associated with the examinations. The inspector attended the training session to ensure i
that the applicable material was covered and that, examiners from the
'
backshifts also received the training. The training and resulting
-
discussions were found to be very beneficial.
The subsequent examination performed on cell 46-11 produced preliminary data that indicated a crack depth ranging from 1.45 inches to 1.6 inches. This data was called preliminary because limitations only allowed the transducers to make a 1 inch vertical stroke. With this limited stroke only the 60* pulse echo transducer looking up could size the indication. The examination also established that the time-of-t
,
i
_ __
-
.
'
.
.
'
t
flight transducers would not provide useful information because the %
inch 5 MHZ transducers that had been substituted for the % inch 10 MHZ transducers used at San Jose, transmitted through the crack and no shadow effect was revealed by the crack face.
In addition, the tip signal could not be discern by the time-of-flight technique because, the depth of the crack was the same dimens;cn as the thickness of the shroud
whose signal obscured the crack tip signal.
l i
Although the above data was considered to be preliminary, the system was calibrated and therefore, the crack was not going to be recorded any
,
smaller. However, it could be recorded larger when the transducer stroke was increased and the other transducers could see facets of the
'
crack.
l Having a deep crack in only one area of one quadrant was in-sufficient data to implement a repair, however by Sunday, September 26, 1993, the inspector had observed preliminary data that had been taken in each quadrant which indicated the following:
Cell No.
Crack Deqth limitation
!
!
46-11 1.45" to 1.6" 1" Scan length, Only One Transducer Seeing Crack Tip 42-47 Non-conclusive A Scan Length of Only.2" But Deep Could be Obtained, No'
f Transducer Could Obtain Crack Tip Signals
,
10-07
.9" to 1.35" Scan 'imited, But Since
.
Indic Sion Is Not As Deep As
!
Recorc'd In Other Cells It Can
!
Et renfirmed from Two Sides (A
'
Crack Confirmation Asset)
l.35" to 1.6" Scan Limited to.8"
'
.
t As stated above this data was preliminary because of scan limitations.
GE intended to adjust the stand-off on the scanner and re-examine each cell.
However, as stated previously the system was calibrated when the
'
above data was taken, therefore, if there is a change in the above information it should indicate greater crack depth not less.
,
The above data indicated that a permanent repair would be the most
!
prudent option. CP&L however, had not made this decision at this point
,
because all depth measurements were preliminary.
,
,
Within the areas examined, no violation or deviation was identified.
.
!
.
,
a
.
...
_-
..
.
.
.
!
-
-
-
,
,
l
,
3.
Exit Interview l
The inspection scope and results were summarized on September 24, 1993, with those persons indicated in paragraph 1.
The inspector described the areas inspected and discussed in detail the inspection results.
,
Proprietary information is not contained in this report. Dissenting-
>
comments were net received from the licensee.
l l
[
.
!
,
!
i t
.
+
i F
,
!
!
l
!
!
L
.;
.
.
r m
.-
-