ML20236A915
| ML20236A915 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 10/13/1987 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML20236A899 | List: |
| References | |
| NUDOCS 8710230232 | |
| Download: ML20236A915 (10) | |
Text
"MS, ;, G,f <,J.
(
n
+
w.
- c. -. df
~~-
O
( f $<
m, p
'l W
.Enclosuref kY$_bATION SAFET
- 1 MAIN STEAM LINE, SUPPORT MODIFICATIONS
,I
. TROJAN, NUCJdAR PLANT
'o;p 1
1 DOCKETfid. 50-344
>< u o
nw
!?
f
/ '/
(f
1.0 DESCRIPTION
Or, EVENT y
v 0(i During-the.1987 refueling odtage, a discrepancy was noted during the inspection of main steam line (MSL). support EBB-1-1-SS-31-(SS-8L).
The discrepancy consisted of separ'ation between the' baseplate and the baseplate' The grout, and between the grout-and the concrete wall in several locations.
discrepancy was evaluated by the Portland General Electric Company (PGE), the licensee. During their evaluation of the condition, design criteria for the-support vera revi9wed and'it was deterrtined that the design of the support l
anchorage was;f p'dequate for the specid ed dynamic load.
As a result of this observation, PGE coriducted a detailed inspection of MSL -
)
supports and issued a nonconformance report NCR 87-214. Calculations l
performed by Bechtel,the Trojan A-E, indicated that a number of MSL surnorts.
PGE l
were grossly underdesigned for. the originally specified dynamic loads.
indicated that modifications of seversy MSL supp,o'rts were proceeding based on f
-newly define & dynamic loads, and these modificahtons will be compinced prior to restarting,the Trojan plant from its 1987 fefueling outage.
j i,
s 2.0 LICEYSEE : CORRECTIVE ACTION PROGPAM
+<
,i Following the discovery of the HSL pipe support design deficiency,' the licensee directed $he A-E to determine the root cause of the deficiency j
and to develop a program to detgrainerthe extent of the deficiency. The A-E determined that,the destgn error was made.by their Civil Engineering l
Group and ccyniste& of a failure to incorporate the system dynamic loads l
i in the support desih,n2 The A-E proposed a program to verify all support
-designs performe6 by their Gibil Engineering Group. The scope of this f
programirclu<)ed64 safety-relatedMainSteamLinesupports(34 designed l
by a PGE subcontractor, and 30 desiened partially by the subcontractor and f
s pa:tially by the A-E), 33 Civil-designed safety-related pipe anchors, 265 Civiladesigned-tafety-related pipe snubber and restraint anchorages, and five! Civil-dcsi'gned nonsafety-related pipe support structures providing anchorage f/or 41 supports.
- hhe 1 ensee ditecteh thu A-E to proceed ':ith their proposed program and also mtp3nded the scope to include verification of all saiety-related supports witu'dynimic loads. Tbjs group included 76 supports of which 26 were l
n1 ready, included in the group of 265 Civil-designed snubber and restraint ancho'ryges. This Pipe Support Desif;n Verification Program was to~ ~
demonstrate that eac M urport an cede allowable stress under the loads and it hd combinations desc'ribed in the Trojan FSAR.
i i
i.
i i
/
A $
h l
8710P30232 871013
- /
s, l
PDR 'ADOCK 0500034
'j PDR > '
p
. [a
)
a 1
3s 1
h ification program identified ten Main Steam Line supports which l
.T e ver
/
- were not adequately designed' for the specified dynamic load. However,
~
these loads were determined to be conser a*ive based on turbine-trip testing at Trojan. New dynamic loads were calculated which were found to be generally lower.than.the original specified loads. Using the new l
loads, three of the ten, supports were determined to be acceptable without' modi:ication.- The remaining seven supports were modified. All ten of the Main Steam Line supports discovered to be inadequately designed were part
'f of a group of 14. supports designed separately in late 1975 for dynamic
' load concerns.
Sixteen of the 33 supports in the Civil-designed pipe anchor category were 4
inadequately designed for the. original loads in the A-E files. However, the l
finding'was preliminary pending confirmation of the actual loads for those supports by the NSSS vendor. Upon receipt of those loads during the current I
verification' effort it was demonstrated the supports were adequate for the actual loads.- Four of the 33 supports in the original verification program were on piping that was disconnected during an unrelated design modification
-complete.d during the 1987 outage. The verification of these anchors was not required..
l 3.0 STAFF EVALUATION i
3.1 Audits on Pipe Support Design on June 22-25, 1987, the NRC staff performed an audit on MSL support design in the Trojan A-E's office and also inspected MSL supports at the plant site.
Activities associated with the MSL support modifications were reviewed.
During the audit, a number of discrepancies that were deemed safety-related were found (Reference 1). The staff concerns fell into the following three 1
areas:
(1) Pipe supports were inadequate to meet the originally specified design loads.
(2) The original design loads were derived inappropriately.
(3) There were no ceiginal design calculations.
On July 9,1987, pursuant to 10CFR50.54(f) (Reference 2), the NRC issued a letter to PGE requesting information to assess the impact of these concerns to plant safe operation and their potential generic implications.
PGE responded to the NRC by a letter on July 10, 1987 (Reference 3).
In this response, PGE described an action plan for implementing a pipe support design
. verification program (see section 2.0 above) which was developed by FGE and A
the Trojan A-E to verify the design of possibly affected pipe supports.
total of over 400 pipe supports were evaluated. They identified that the pipe supports in question were all on the main steam system and were all designed v
--m.....
.m
-7 y
U i
' L e,
and 'insta11ed ' late in construction in 1975.
PGE concluded.that the design problem was limited to the A-E' Civil Engineering Group and specifically to the
~
main steam pipe support designs performed in 1975.
- 0n July 21-23,1987, the staff conducted another audit on pipe support design
'in.the Trojan A-E office.- The audit was intended to ensure adequacy and
- acceptability of the. licensee's pipe support design verification program.
Sample calculations of selected piping systems in the verification program and documents related to implementation of IE Bu11etins.79-14 and 79-02 were
.j reviewed.- The staff found that the verification program was near completion and the results'were generally acceptable. However, the audit resulted in a number of additional. concerns (Reference 4).
Some issues of concern'were considered by the staff as essential to plant operation safety and needed to be resolved prior to restartAng the plant from its 1987 refueling outage (the restart issues)...The remaining issues of concern may be resolved in the near.
future after-restarting the plant.
In Sections 3.2 and 3.3, a description of these concerns,1PGE's responses, and staff evaluations are.provided.
3.2 Resolution of Restart Issues As indicated in Section 3.1 above, the staff audits resulted in a number of j
concerns, some of which were considered safety essential and requiring resolution prior to restarting the plant. The staff reviewed PGE responses Staff l
to-these issues (Reference 5, 6 and 7) and they were all resolved.
l
! evaluation and resolution of these issues are summarized below:
A.
Design Verification Program Scope Concerns In the original design of the Trojan Plant, pipe supports were generally designed by the pipe support vendor. The A-E Civil Engineering Group became involved on an exception basis, participating in the design of less then. ten percent of the supports. Since the design deficiency was I
traced back to the A-E, the verification program was intended to verify I
all A-E pipe support designs. During the second staff audit, the method which was used to identify the A-E designed supports was questioned.
The licensee was requested to provide additional assurance that the scope of l
]
the verification program indeed included all A-E designed supports.
3 Licensee Actions
~
In response to the concern, PGE conducted an additional documentation search to identify all A-E designed pipe supports. Documents reviewed included A-E Civil Discipline Calculations, Civil design drawings and l
Civil field change request drawings. This search encompassed the original j
)
R i
y l-l l
n
{ =
civil design activity on pipe supports at Trojan. This' review dis'covere'd an additional 40 safety and non-safety related supports designed by the A-E.
The verification program was expanded to include these additional supports. As a result, ten of these supports were modified to meet FSAR design criteria.
Staff Evaluation Based on the additional A-E efforts in 1dentifying all A-E designed
~
supports, the staff concludes that the licensee has provided reasonable assurance that all potentially deficient supports were included in the verification program and all deficient supports were modified prior to the restart. The staff will further verify this by audits during its review of the long term program.
B.
Safety Margin on Williams Rock Anchor Bolts Concerns The staff questioned the adequacy of safety margins in_the allowable design load for Williams Rock Anchor Bolts. The embedment depths were less than the recommended values given in the manufacturer's current catalog. The licensee provided a calculation to justify the design allowables but a review of the calculation raised additional questions as follows:
The embedment length used in the concrete shear cone calculation was o
not in accordance with the manufacturer's recommendation I
The calculation did not consider the bond strength between the bolt and o
l, the grout The strength reduction factor did not meet current ACI-349 code i
o requirements In addition, the staff audit of sample calculations identified concerns regarding the proper consideration of cone overlap area, shear allowable and edge distance criteria.
Licensee Actions In response to the various NRC concerns, the licensee recalculated rock j
anchor bolt capacities, performed a test program, ultrasonically measured i
embedment depths, and re-evaluated capacity versus demand for $11 affected
)
anchorages.
l l
The anchor bolt capacity recalculation used more conservative criteria i
than had been used previously. The criteria were based on ACI 349, l
l
..m bi..
A -
- .- 3 I
Appendix B guidelines'and the method currently recommended by Williams l
for calculating shear cone effective' area.
For a one-inch diameter. rock' Jbolt'with eight inch embedment in 5000 psi design' strength concrete, a.
i 32.7. kip ultimate capacity was.cpiculated. This is less than the bolt yield value of 37 kips.
l Ultrasonic ~ measurements of rock bolt embedmenc depths indicated that some bolts had less ' than the nominal eight inch embedment. In order to provide assurance that. capacities corresponding to embedment. less than eight l
inches are' adequate, all accessible rock bolts with eight inch design j
i embedments were ultrasonically measured.
Pull-out capacities were then l
calculated based on actual embedment depth using the criteria discussed above. For inaccessible bolts, a high confidence' level-minimum expected embedment depth.was determined based on a statistical evaluation of the accessible bolt data.
d i
The test program: involved pullout tests on nine rock bolts installed'in the anchorages.of safety-related pipe supports at the Trojan Plant.
All nine bolts withstood a test load of 34.5 kips without failure and with no' evidence of initiation of a bond failure mechanism' developing at the grout interface.- One bolt which had a measured embedment depth of 6 3/4" was retested to failure. This test resulted in a tensile failure of.the rock bolt steel at a test load of 51.5 kips with no pullout of the anchor bolt
' head.
The capacity versus demand re-evaluation for rock bolt anchorages considered calculated capacities based on measured or minimum embedment depths with appropriate reduction taken for closely spaced bolts. Based on this re-evaluation, four pipe supports were modified to increase the anchorage capacity. The remaining anchorages were shown to have a minimum safety factor of two under SSE loadings against pullout failure.
Staff Evaluation Based on a review of the licensee's re-evaluation and testing results of Williams rock anchor bolts, the staff concludes that the licensee has shown that safety margins are adequate for at least short term operation.
The test program has provided convincing evidence that demonstrates the conversatism of the calculated ultimate capacities based on the revised criteria used by the licensee.
The stoff feels that a minimum safety factor of two is adequate _and j
consistant with regulatory requirements for short term operation. The l
safety margin issue will be further reviewed by the staff in the long term program to insure that the final safety margins are consistent with requirements stated in IE Bulletin 79-02.
l l
1
at
-O
. =>
C.
Verification'of Weld Design Concern During the July 21-23 staf f audit, a number of pipe support design calculations and drawings were reviewed. The review identified a number l
of instances where calculations did not address weld adequacy.
Furthermore, a number of civil drawings did not identify weld type, size i
or configuration.
1 Licensee Actions The licensee responded to these concerns in References 5, 6 and 7.
The licensee stated that the welds were verified for acceptability to applied loads. The verification process involved either engineering judgement or Engineering judgement was used in cases where the weld a nalysis.
capacity was clearly greater than that of the structural member (such as full penetration' welds) or in cases where the load was much smaller than the capacity furnished by the weld. Calculations were performed for the The licensee remaining support welds to demonstrate their adequacy.
i provided a table in Ref erence 7 which summarized the weld design verification method for each support in the verification program.
Weld configurations were either called out on the drawing details, described by notes on the drawings, or shown by reference to another_
In a few cases, field measurements were necessary to confirm the drawing.
as-built weld configuration. During the PGE/NRC meeting of August 20, 1987, pipe support construction drawings were reviewed which satisfactorily identified weld sizes, type and configurations and resolved questions raised at the audit.
l l
Staff Evaluation l
l l
Based on the review of the additional information provided by the licensee, the staff concludes that welds have been adequately addressed in the design verification program. The weld concern, however, was a result I
of an apparent weakness in the QA documentation in regard to traceability of input data and documented bases for engineering judgements. Adequacy of QA documentation is discussed in item E and will be further reviewed l
l during the long term program.
D.
Verification of Support Loads Concern j
During the staff audit, a number of questions related to use of proper support loads were raised and it became clear that load verification was Loads generated by not part of the pipe support verification program.
other vendors were never confirmed. In some calculations, superseded l
l i
4 loads vere used. A number of calculations used OBE loads instead of SSE loads.
Licensee Actions 1
The licensee revised the calculations to include the latest loads including'those generated by the A-E's consultants. Support loads' j
i generated by the NSSS vendor were not on file with the A-E so they had to be requested. All supports were re-evaluated to the latest loads.
The pipe snubber and restraint anchorages were found to be still acceptable for the revised loads. Sixteen of the 33 safety-related pipe anchors which were found inadequate f or the original loads were shown to be adequate for the actual loads. The support calculations which considered only OBE loads were revised to considet SSE loads. All supports in that group'were found to be acceptable.
Staff Evaluation j
Based on the review of the information provided by the licensee, the staff concludes that concerns related to load verification have been properly addressed by re-evaluation to current loads.
E.-
A-E's Quality Assurance Program j
Concern i
During the staff audits, various deficiencies related to the A-E's f
verification effort were noted which raised questions about the adequacy of the A-E's Quality Assurance Program. Examples of these types of deficiencies were lack of procedures to control design interface, poorly documented calculations, insuf ficient data on drawings and calculation errors.
Licensee Actions In response to NRC concerns, the licensee contracted a consultant, Impell Corporation, to perform an independent audit of the A-E's Quality Assurance Program during the Pipe Support Design Verification Program.
The audit was performed from August 1 through August 10, 1987. The consultant and licensee developed the audit plan, QA checklist, and technical quality review criteria checklist. The audit scope included a thorough review of the verification process for a sample of supports and addressed specific concerns /wcaknesses during the performance of the verification program.
n-____
e
.. m The audit report and licensee assessment of the recommendations were submitted to the staff (Reference 7).
The audit identified a number of documentation deficiencies related to technical procedures, insufficient drawing data and poor documentation. The licensee concluded that the documentation deficiencies indicated a partial breakdown of the A-E's QA Although documentation deficiencies exist, the licensee and program.
consultant concluded that they have not affected the conclusion that the supports, except those previously identified as deficient, are acceptably designed.
The licensee has committed to correct all documentation deficiencies during the long term program and will track the implementation of the corrective actions and verify the results by audit.
Staf f Evaluation The results of the independent audit of the A-E's QA program are consistent with the staff's audit findings. The staff concludes that these documentation deficiencies should not affect safe short term operation of the plant. The staff will perform additional audits during the long term program to assure that the QA documentation deficiencies will be corrected.
3.3 Issues Remaining To Be Resolved As indicated in section 3.1, a number of concerns identified during the staff audits will need to be resolved in the near future afte. restarting the plant. Specific concerns raised during the NRC audits were documented in NRC trip reports (Ref 1, 4). The licensee has committed to respond to these concerns by September 30, 1987.
In addition, the licensee will provide a 1
description of the scope and schedule of the long term program to resolve all J
remaining open issues by September 30, 1987. The remaining concerns are summarized below.
A. Documentation Original vendor-designed pipe support calculations were not on file with the licensee or the A-E.
In addition, a number of A-E calculations generated during their efforts related to IE bulletin 79-14 were incomplete.
These and the other QA documentation deficiencies identified in Section 3.2 must be addressed in the long term program.
I
- r i
i
_9 en B.
Implementation of IE Bulletin 79-14/79-02 Programs During the staff audits, a number of questions were raised in the review of the 79-14/79-02 program calculations. Although these programs were previously reviewed and accepted by the staff, a detailed review of l
calculations raised questions regarding proper implementation.
Questions
(
included the use of center-of-mass response spectrum for piping analysis, the use of proper anchor bolt allowables, and proper consideration of baseplate flexibility. These concerns must be addressed during the long term j
program.
l C.
Design Methodology l
The version of the piping analysis computer program used by the A-E (ME101) i between March 1984 contained some errors. The licensee must assess the impact of these errors on the calculation of the Trojan pipe support loads. Othur design methodology items which should be addressed in the long term program include consideration of snubber / strut angularity and use of an elliptical t
interaction formula for anchor bolts.
D.
Remaining Audit Questions A number of questions were raised during the staff audits regarding specific j
calculations reviewed. Each question should be resolved and generic 1 implications, if any, should be evaluated and addressed in the long term j
program (for a list of questions, see reference 4).
4.0 CONCLUSION
S l
Based on the results of the staff audits and evaluation of the licensee's responses to restart issues of concern, the staff has determined the following:
1.
The licensee has corrected the design deficiencies in the Hain Steam Line Pipe Supports. The corrective actions are adequate and acceptable.
1 2.
The licensee has provided a reasonable level of assurance to conclude that the design deficiencies were limited to pipe supports designed by the A-E Civil Engineering Group.
3.
The pipe support design verification program has provided a sufficient level of confidence to conclude that all safety-related pipe --
supports not originally designed to meet the specified loads have been identified and corrected.
4.
The licensee has resolved all audit issues of concern considered essential to plant operation safety requiring resolution prior to plant restart.
j l
h
- 1 a
t =
Based on these findings and the licensee's commitment to conduct a long
-term program to resolve all other issues discussed in Section 3.3 of this report, the staff concludes that it is safe for the Trojan Plant to commence l
heat-up and return to power.
l The long term program, when implemented, should ensure that all safety-related i
pipe supports are in' full compliance with FSAR requirements.
The licensee has committed to complete the program by July 1, 1988. The staff will evaluate this program and its implementation by conducting additional audits l
to verify that all concerns are resolved and that the design of all safety-related pipe supports is adequate and properly documented.
1 i
REFERENCES 1.
Trip report, S. Hou (NRC) to L. B. Marsh (NRC), " Audit and Site Inspection for Trojan Main Steam Line Supports", July 13, 1987 2.
Letter, Per 10CFRSO.54(f), D. M. Crutchfield (NRC) to D. W. Cockfield (PGE), " Adequacy of Pipe Support Design and Analysis for Safety Related Systems", July 9,1987 3.
Letter, D.W. Cockfield (PGE) to NRC, Response to NRC letter per 10CFR50.54(f), July 10,1987 l
4.
Trip report, S. Hou (NRC) to L. B. Marsh (NRC), "NRC Audit of Trojan Pipe Support Design Verification Program", September 25, 1987 1
5.
Letter, D. W. Cockfield (PGE) to NRC, Response to NRC Audit Concerns, July 27, 1987 6.
Letter, D. W. Cockfield (PGE) to NRC, Response to NRC Audit Concerns, July 31, 1987 7.
Letter, D. W. Cockfield (PGE) to NRC, Response to NRC Audit Concerns, August 18, 1987
~ _ - - - _ _ _ _.. _. _.