ML20032A073
| ML20032A073 | |
| Person / Time | |
|---|---|
| Site: | Oconee  |
| Issue date: | 10/20/1981 |
| From: | Taylor J BABCOCK & WILCOX CO. |
| To: | Stello V NRC OFFICE OF INSPECTION & ENFORCEMENT (IE) |
| References | |
| NUDOCS 8110280251 | |
| Download: ML20032A073 (2) | |
Text
'
s Babcock & Wilcox woei..r Power Generation Division a McDermott company 3315 Old Forest Road P.O. Box 1260 Lynchburg, Virginia 24505 (804) 384 5111 October 20, 1981
[#
A h g lbi Mr. Victor Stello, Director I/
t Office of Inspection and 3
OCT2 7193Ib1g Enforcement E
~
United States Nuclear Regulatory u.,, dud Commission Washington, D. C.
20555 E
4
~ k/TrTTb
SUBJECT:
Broken Thermal Shield Bolts at Oconee Unit 1
Reference:
- 1) Lettar from J. H. Taylor (B&W) to Victor Stello of Augu:it 11, 1981,
Subject:
Broken Thermal Shield Bolts at Oconee Unit 1.
- 2) Letter from W. O. Parker (Duke Power Co.) to James P. O'Reilly (NRC-01E) of October 5,1981.
Copy attached
Dear Mr. Stello:
This letter is an update to Reference 1 of the same subject.
In Reference 1 I indicated B&W did not consider this matter to be a safety concern. Additional information has now been developed which has been summarized in Reference 2.
The investigation reported in Reference 2 further confirms the earlier conclusions that there are no significant safety implications associated with the thermal shield bolt failures.
B&W has provided input for and reviewed Reference 2 and we concur with the information presented in that report.
8110280251 811020 CDR ADOCK 05000269 D
PDR 1
$\\
t
o Babcock &Wilcox Mr. Victor Stello Page 2 - October 20, 1981 B&W has continued to assist Duke Power Co. in keeping our other operating plant utilities up to date on the thermal shield bolt investigations.
B&W also made detailed presentations to the Staff on October 9,1981. At this time no further B&W communications with the staff are planned but should you have questions or desire additional information, 'please contact me.
Very truly yours, THE BABC0CK & WILC0X COMPANY
/
J. H. Taylor, Manager Licensing JHT:efc cc: B&W Regulatory Response Group W. O. Parker K. S. Canady T. Novak
J.1;. 1 u.p'.
E s,,
D URI? POWl?H COM PANY
[g,I O f Jgg/
Powr.n Uciton<u 400 SouTu Cucucn STuner. CHARLOTTE, N. C, anan wiuima o. ran a ca. s a.
/<F PegocrNT
- (i.C8mO 4(. Ae ( 4 7 0 4 ser-P osa+.c=
October 5, 1981 Mr. James P. O'Reilly, Director U. S. Nuclear 21gulatory Commission Region Il 101 Marietta Street, Suite 3100 Atlanta, Georgia 30303 Re: Oconee Nuclear Station Docket No. 50-269 RO-269/81-ll, Supplement 2
Dear Mr. O'Reilly:
My letter of July 24, 1981 as supplemented on August 5,1981 provided Reportable Occurrence Report R0-269/81-11 concerning broken core barrel assembly thcrmal shield bolts. This letter supplements these initial submittals and provides information concerning the present status of examination and repair ef forts.
As previously discussed, three bolt shanks and two bolt heads were ex ained at the Babcock and Wilcox Lynchburg Research Center for failure acdc determination.
It was concluded that the failure mecht aism was either intergranular stress corrosion cracking or intergranular cerrosion assisted fatigue or some combina-tion of the two.
These laboratory results confirmed the need to perform more extensive investigations relative to the thermal shield bolt manufacturing history, the stress state of the bolts, the thermal shield manufacturing and assembly history, and most im,.artantly, the potential for this mechanism to affect other bolcs of the same material (A286) in the reactor internals. Results of these investigations are described in the attached report.
Because A286 bolts are used in four other joints in the internals, a thorough inspection and sample examination was conducted relative to these other joints.
These joints are the upper thermal shield restraint blocks, the core barrel to core support shield, the core barrel to lower grid assembly and the flow dis-tributor to lower grid assembly. These investigations consisted of ultrasonic-testing of a aclected sample of bolts in all of these bolted joints and removal of a smill sample of bolts from three of the four joints for laboratory examina-tions. All results indicated no problcus at the other joints.
As a result of the work performed since the August 5, 1981 submittal it has been confirmed that the bolt problem is limited to the thermal shield.
Bolts from all other major joints which utilize the A286 naterial have been examined and no indications of deterioration were found. Our earlier conclusion that there is no significant safety implication associated with the thermal shield bolt failures has, therefore, been confirmed.
- a GOlW h
'~
^
" ' ~ '
--e g o.
e t
Mr. James P. O'Reilly, Director October 5, 1981 Page 2 1
This report also presents a description of the design modifications for the thermal shield.
We are currently planning to attend an October 9, 1981 meeting with the staff to make a presentation and to respond to questions which the staff may have in It is any of the areas related to the Oconce thermal shield bolt problem.
hoped that the information contained herein will serve as a good foundation for that meeting.
Very truly yours,'
/
/
4 l
. w.a N. % t!LD l William O. Parker, J.
NAR/php Attachment ec: B&W Regulatory Response Group:
Director, Office of Management and Program Analyitis J. J. Mattimoe, SMUD, Chairman Mr. T. M. Novak, U. S. Nuclear cJ.cll: TaylorpB&W-W. C. Rowles, TECO Regulatory Commission i
D. C. Trimble, AP6L Mr. Bill Lavallee, Nuclear C. Beatty, FPC R. J. Wilson, GPU Safety Analysis Center e
^
r
-rr-
.n-g mg
--e y-4 pgw~,--
wy-eee -ty g,ws-v s
t
n Cctober 5, 1981 Thermal Shield Bolt Failure Supplementary Information Introduction and Summary As a result of extensiv' investigations, it has been concluded that the only distressed areas of the internals are the thermal shield attachments and restraints. All other joints using A286 bolts indicate satisfactory performance.
Therefore, the preliminary conclusion that a significant safety issue does not exist has been confirmed.
The sections below present more detailed information in each of the following areas.
Laboratory examinations Inspections onsite Bolt fabrication history Thermal shield manufacturing and assembly review Thermal shield structural analysis and testing Potential failure causes of thermal shield bolts Repair status Laboratory Examinations Laboratory examinations were performed at the Babcock and Wilcox Lynchburg Research Center on both broken and unbroken bolts from Oconee and on archive bolts which had never been installed. The objective of these examinations was to determine bolt f ailure mode and status of other joints using A286 bolting material.
The broken samples consisted of three bolt shanks and two bolt heads. Of the five fracture surfaces, two were damaged from impact, and consequently clean fracture surfaces were not available for examination. The remaining three fracture surfaces (two bolt shanks and one bolt head) were examined and found to contain similar fracture features.
All of the broken bolts examined had fractured in or near the fillet between the head and shank. An island of transgranular fracture (as shown in Figure 1) with evidence of f atigue striations and/or crack arrest marks was surrounded by intergranular fracture, with evidence of grain boundary corrosion attack.
No shear lips or other macroscopic indications of ductility were present on the fracture surfaces. The initiation and propagation of intergranular cracks with branch cracking in this material is consistent with a form of an environmentally assisted cracking process - stress corrosion and/or corrosion fatigue.
A tensile test-was performed on a specimen machined from the shank of a fractured bolt (irradiated condition) and a specimen from an archive bolt of thc aame
material (base condition). The measured tensile prcperties were within the range of expected values and met the requirements-of the applicable specifica-tion.
Neutron irradiation had virtually no effect on the tensile properties.
The other A286 bolted joints in Oconee 1 were indicated to be sound from video insp ec tions. To confirm the integrity of these other A286 bolted joints, lab-oratory examinations were performed on three thermal shield upper restraint bolts, two core barrel to core support shield bolts, and one flow distributor bolt. These bolts were examined visually, with liquid penetrant and with a scanning electron microscope. A surface contamination examination was also performed. The two intact lower thermal shield bolts were also removed, and one of these bolts was included in the above examination. In all cases, the examinations included both the head-to-shank transition and the thread regions.
The archive bolt examinations included one bolt from each of the A286 joints.
The thermal shield bolt examination included metallography, tensile, micro-hardness, spectrographic chemical analysis and reheat tests. The other three bolts received metallography tests. The main finding from these examinations was a radical transition in grain structure (size) at the fillet of the lower thermal shield bolt.
In the lower thermal shield bolts from the archive, the transition zone was marked by a band of significantly larger grain size than found on either side of the transition. This band of larger grains also extended into the fillet area. This large change in grain size in the transi-tion zone was not evident in the archive bolts from other bolted joints including the upper thermal shield bolts.
This larger grain size in the lower thermal shield bolts from the archive was seen in the one unbroken bolt examined from the Oconec Unit 1 lower thermal not shield. It is noted that this bolt was one of only two in the lower thermal shield which did not fracture. An examination of a fractured bolt from the Oconee Unit i lower thermal shield is planned. The anomaly in the unfractured-l Oconce Unit 1 lower thermal shield bolt should be confirmed upon the examina-tion of a fractured lover thermal shield bolt.
Additional confirmatory laboratory examinations are currently underway or l
planned. This additional work involves microscopic analysis of grain structure l
and scannning electron microscope examinations. Table 1 summarizes the lab-oratory examinations.
l Onsite Inspections l
l The onsite inspections resulted in the recovery of all pieces of the missing l
thermal shield bolts.
To supplement the previously reported (Reference 2) visual inspections which l
have been performed using remote video equipment, two additional types of inspections were made onsite to check for distressed arans in the Oconee 1 reactor internals. This involved in-place ultrasonic testing (UT) for cracked bolts and a combination of UT, feeler gage, and acceleroueter testing which would indicate motion and/or wear at tne thermal shield interfaces.
In addition, ultrasonic tests were attempted to check preloads on selected bolts.
I = - - -
bltrasonic testing was conducted on the following bolts for. crack indications:
- Thermal Shield Upper Restraint Belts - Approximately 40% Sample
- Core Barrel'to Lower Grid Bolts - Approximately 20% Sample
- Core Barrel to Core Support Shield Bolts - Approximately 20%. Sample
- Flow Distributor to Lower Grid Bolts - Approximately 20% Sample No crack indications were found in any of the bolts examined.
In the early visual inspections of the lower thermal shield bolts, only 13 of the installed 96 bolts appeared to be intact. Of those 13, only two did not have crack indications. The UT examination of the other 11 indicated the bolts were broken.
As input to the overall determination of failure cause and the corrective action, inspections were made to check for gaps which would indicate motian and/or wear of the thermal shield. Caps were checked in the radial direction between the outer thermal shield restraint block and the upper end of the thermal shield and axial gaps were checked between the lower end of the thermal shield and lower grid flange.
The outer therral shield restraint block gaps (thermal shield to outer restraint block) were found in the range of 0 to 160 mils. Gaps over 90 mil were confirmed with a weld wire as feeler gage. The as-built gap range was 5 to 11 mils. These gaps indicate some wear at the thermal shield upper restraint. Axial gaps vera found between the lower edge of the thermal shield and the grid flange which varied around the circumference from 0 to 40 mils. Caps over 20 mil were con-firmed by a feeler gage. The circumferential variation indicated a slight tilt in the thermal shield. These axial gaps are scheduled to be closed up prior to replacing the thermal shield bolts.
UT bolt preload testing was performed for three thermal shield upper restraint bolts, two thermal shield lower restraint bolts, one flow distributor bolt, and two core barrel to core support shield bolts. The UT was conducted before and after removal of the bolts so that a correlation could be made between bolt extension and preload. Measurements for lower thermal shield bolt preloads were i
unsuccessful, probably due to the displaced condition of the thermal shield.
l Results for flow distributor bolts indicate a preload higher than the design l
value but considerably less than yield. Results for core barrel und upper thermal shield restraint bolts indicate preloads are in the design range.
l Bolt Manufacturing History and Materials Investigation The failure of the lower thermal shield bolts as contrasted with the completely satisfactory performance of the other bolts made from the same aterial has led This to a comprehensive investigation of the A286 bolt manufacturing history.
examination has included a review of vendor records, material te-
- reports, material processing and bolt fabrication processes, as well as metallurgical examinations of archive bolts and mechanical testing of archive bolts and samples l
of material extracted from the failed thermal shield bolts. The following discussion summarizes the important conclusions from these examinations.
L
.j L -
i i
The thermal shield upper and lower restraint bolts were manufactured by Valley i
Todeco. All other A286 bolts were manufactured by Standard Press Steel. The heat of bar stock material used for the lower thermal shield bolts is different from the heat of bar stock material used in any other A286 bolts including the upper restraint bolts which were made by the same manuf acturer, Valley Todeco.
i All of the bolts were manufactured by a " hot heading" process which produces a heat affected, transition zone located at the base of the bolt head at or near j
the fillet region as discussed earlier.
In a review of the processing of the lower thermal shield bolts, it was found
]
that bar stock which was 40 to 50% cold reduced was used in the bolt fabrication.
i This cold reduced bar stock is apparently unique to the lower thermal shield bolts (none of the other bolts were manufactured from bar stock with such l
extensive cold reduction). This cold reduced bar stock is considered to be the j'
likely source of the significant change in grain structure in the head-to-shank transition zone from the " hot heading" process.
i The relationship of this duplex grain structure transition in the lower thermal shield bolts to the fracture mechanism is undergoing further investigation.
Thermal Shield Manufacturing and Assembly Review A review of the thermal shield manufacturing and assembly history was conducted l
in an attempt to discover any step in these processes which could have contributed to the f ailure of the thermal shield bolts.
It was noted that the Oconce 1 and 2 RV internals were reassembled at the reactor site following the initial hot functional testing at Oconce 1.
Oconee 3 internals were assembled in the shop.
The manufacturing review focused on the five main thermal shield interface areas.
These ere:
l 1.
The radial interference fit at the thermal shield lower end I.D.
2.
The radial interference fit at the thermal shield upper end.
3.
.The restraint block on the core barrel upper end 0.D.
4.
The preload on the upper restraint block assembly holddown bolts.
i 5.
The preload on the lower thermal shield holddown bolts.
1 This above review has been completed and did not provide any additional insight as to the cause of the bolt failures.
a Thermal Shield Structural Analysis and Testing i
A review was made of the structural analysic and structural testing which were used in the design and confirmation of the thermal shicid. Bolt preload stresses and thermal stresses were both found to be within the design liuits. Flow-induced vibration stresses were measured on two thermal shield bolts during hot functional testing and these stresses were found to be very low.
I
E Potential Failure Causes of Thermal Shield Bolts The factors which seem most likely at this time to have contributed to the lower thermal shield bolt failure are the rather pronounced microstructure transition at the head-to-shank fillet in the lower thermal. shield bolts and the potential for this " metallurgical discontinuity" to increase the sensitivity to stress corrosion cracking and corrosion fatigue.
Repair Status At the present time, redesign of the lower thernal shield has been completed.
All 96 bolts and locking clips will be replaced with stud and nut assenblics as shown in Figure 2 attached. The assembly consists of two studs with a baseplate so that 48 such assemblies will be used. Each stud, nut, locking clip and baseplate is first assembled and the assembly so then installed into the existing threaded holes in the thermal shield. A stud tensioner will then tension the studs and set the nuts. After the tensioning is complete, the locking clip which is welded to the baseplate will be crimped onto the nut at groves machined into the nut.
The stud and nut material in Inconci X750 which has been selected.
because of its availability and stress corrosion resistance and satisfactory experience in PRR environment. The locking clip and baseplate are 304SS.
The need to modif y. the upper thermal shield restraint has not been determined; i.e., final stress analyses may show that restraint at the upper end is not required, and the restraint can be left as is.
Supplementary Information Regarding Lower Internals Guide Blocks As reported previously, one guiue block was missing from the right side of the "Y" axis as viewed from the outside of the thermal shield. The missing guide block is believed not to have been in place since the time when the internals were last installed in 1976. This is because none of the attachment parts has been located despite a very thorough video inspection.
UT examinations of the bolts in the other guide blocks have been completed and j
UT indications have been reported in two bolts. The bolts with the indications are in the block which mates with the missing block and the bolt in one of the l
guide blocks adjacent to the missing one.
Video inspections of the vertical l
surfaces of the guide blocks containing the bolts with the UT indications show I
no evidence of contact between the core support lug and the guide blocks.
I I
In order to determine whether these indications are truly flaws, the bolts will be removed in two stages for supplementary examinations. The first to be removed i
will be the bolt from the other half of the missing block. If a flaw is truly l
present, the second bolt with an indication will be removed and examined.
As a part of the overall repair plan, a review has been performed concerning the i
{
guide block function. As a result of this review, it is considered unnecessary-to replace the removed guide blocks, i.e., the one believed to have been missing when the internals were installed in 1976 and either one or two additional blocks i
which will be removed for further bolt examinations.
I
I i
The as-designed function of the guide blocks is to provide a vertical guide for the core guide lugs in the remote event of a circumferential seve.cnce of the core support shicid. Removal of three of the twenty-four guide blocks will not significantly affect the function of the guide blocks.
Ilot Functional Testing flow-induced vibrational testing showed a 1 to 1.5 mil beam mode displacement of the CSA lower end which is within the 20 mil design gap between a guide lug and guide block. Additional analysis has been performed which shows that removal of three guide blocks will not significantly effect
-CSA vibrational behavior in the event contact between guide lugs and guide blocks occurs. Also, further justification is based on the acceptabic results of the analysis of the effects of asymmetric LOCA loadings (BAW-1621) 5:htun was performed taking no credit for the guide blocks.
As a result of further investigations related to the missing guide block, it has been concluded that the above plan for inspection and removal of either one or two additional half blacks is prudent and that the removal of three of i
the 24 blocks does not represent any significant reduction in safety.
W 4
i I
r i
m~
a
-4 r..e w e. M
l TABLE 1 LABORATORY BOLT flAMINATIONE Specimen-Bolt Type Source Examinationn 3 bolts MK 375 Oconee J V, L.P., S.C.
(Upper Thermal Shield Restraint) 2 bolts MK 256 (Core Barrel)
Oconee 1 Cyl. Surf.*
V.,
L.P., S.C.
1 bolt FDC 390 Oconee 1 Cyl. Surf.*
(Flow Distributor)
V., L.P., S.C.
2 bolts
}DC 380 Oconee 1 Cyl. Surf.*
(Lower Thermal Shield)
V., L.P., Met.*
~
1 bolt (A)
FDC 380 Oconee 1 V.,
Frac.
head i bolt (B)
FDC 380 Oconee 1 v.
head 1 bolt FD; 380 Oconee 1 V.,
Frac., Met.
shank (#1)
_5 1 bolt 13" 3C0 Oconee 1 V., T.T.
shank (#2) 1 bolt MK 380 Oconce 1 V.,
Frac., Met.
shank (#3) 1 bult MK 380 Archive V., T.T., Met.
Chem. Rch 1 bolt MK 390 Archive V., Met.
1 bolt FDC 375 -
Archive V., Met.
1 balt MK 256 Archiva V.,
Met.
V.
- Visual examination L.P.
- Fluorescent liquid penetrant examination for surface cracking T.T.
- Tensile test
- Surface contamination examination using replica technique S.C.
Cyl. Surf.
- SEM examination of cylindrical surface SEM examination of fracture surface FRAC Chem
- Spectrographic chemical analysis Mar
- microhardness Reh
- Reheat treatment Met.
- Metallography
3 i
l l
l l
g l 2 ne j'
e-Intergranular Surface Transgranular Island.
Final Fracture Area I
>d u;.
'1/10"'
l Figure 1.
Macrograph of Fracture Surface on Bolt Shank 1 l
l t
i r
I l
i r = - = - -
ev erev..
w
...w w+
-4 FIGUR' 2 LOWER THERMAL SHIELD STUD DESIGI' g
q.___.
1 THERMAL SHIELD N
A
'/,/' '/,(
LOWER GRID
//,/
{
' /,,
',/
SHELL FORGING
/ /.N
'!/ //(
'l l
-7 i
'l \\
I
~ '
'/
BASE PLATE %
3
{
s f
. t
\\i I
~1 LOCKING CUP i
i x
i i N n
i p
f,
! kE LCCKING CUP CRIMP
( TYPICAL TWO NUT' l
l i
l',
PLACES PER CUP )
i l
STUD '#
3
,I I
i
.