ML20246H750
| ML20246H750 | |
| Person / Time | |
|---|---|
| Site: | Grand Gulf  |
| Issue date: | 04/26/1989 |
| From: | Blake J, Chou R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| To: | |
| Shared Package | |
| ML20246H740 | List: |
| References | |
| 50-416-89-10, NUDOCS 8905160184 | |
| Download: ML20246H750 (10) | |
See also: IR 05000416/1989010
Text
'
l3
. >
i:
8V3 EEcu *o
UNITED STATES -
,
' NUCLEAR REGULATORY COMMISSION -
)
t.
4
'
' '
$v
. REGION ll
,
.:
..
g
l j;
101 MARIETTA STREET.N.W.
8
'*
ATLANTA, GEORGI A 30323
4 ..... /
U
Report No.:
50-416/89-10
Licensee: System Energy Resources, Inc.
1
1
Jackson, MS -39205
Docket.No.:
50-416
License No.:
.
Facility Name: Grand Gulf.
Inspectio
on
te :
arch 27-31,-1989
Inspect r:
/
Y(
f:
'
,4L R.
D a~t e . igned
.
Appro ed by
( _s
5/ 16
9
J.
Jake, Chief
.Date Signed
.
Mt ials and Processes Section
n neering Branch
ivision of Reactor Safety
SUMMARY-
Scope:
This routine, unannounced inspection was conducted in the area of
review of piping system procedures and calculations.
Results:
In the area inspected, no violations or deviations were identified.
One Unresolved Item (UNR) was identified concerning the piping system
calculations,
involving five generic problems were identified.
The
licensee was very cooperative.
They are aggressive and agree to
-resolve the problems- identified and improve their calculation-
quality.
"
8905160184 890428
ADOCK 05000416
Q
- - _ - _ _ - _ -
_ - - _ _ . . _ _ _
.
__
.
r
-
7
.
REPORT DETAILS
~1.
Persons Contacted
Licensee Employees
- J. D. Bailey, Compliance Coordinator
- L. F. Daughtery, Compliance Supervisor
- N. Deshpande, Pipe Support Design Supervisor - Nuclear Plant Engineering
- W.
C. Eiff, Principal Quality Engineer
- C. R. Hutchinson, Station General Manager
D. S. Pace, Nuclear Design Manager
- J. Summers, Compliance Coordinator
- S. Tanner, Quality Services Manager
- F. W. Titus, Nuclear Plant Engineering Director
I
'
- M. J. Wright, Plant Support Manager
- J. W. Yelvesta, Plant Operation Manager
Other licensee employees contacted during this inspection included
craftsmen,
engineers,
mechanics,
technicians,
and
administrative
personnel.
Other Organization
l
- A. W. Davis, INPO Trainee
NRC Resident Inspector
- C. Christensen, Senior Resident Inspector
- Attended exit interview
2.
Piping System Calculations Review
a.
Description
During an audit of pipe supports design calculations for the
alternate decay heat removal system (ADHRS) on January 4 and 5, 1989,
by NRR, the staff found that one of the calculations presented had
not been revised to bring it up to date. When questioned about this,
l
the licensee stated that their procedure does not require a revision
!
be made if there is evidence that the change would result in larger
factors of safety. The staff views this procedure to be a potential
safety problem because the basis for the determination that the
change is in the conservative direction is not documented, reviewed,
or approved to the same QA criteria as the original calculations.
- - _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.
_
- _ . - _ _ _ _-
.
_ _ _ _ _
- - _ -
_ _ _ _ _ _ _ _ _ - _ _
-
-_ - -_
"'
- .,
.
,
.
,
2
b.
Stress Calculations Reviews
The support loadings summary sheets of Stress' Calculation Problems
Nos. 63, -Issue No. 09 for Isometric .No. H-1349A, Rev. M and 141A,
Issue No. 10, for Isometric No. H-1351H, Rev. F were partially
reviewed to check the actual load combinations based on Supplement
No.1, Rev. '0, to . Specification No. 9645-M-300.2,,- Rev.18 against the
Maximum Design ~ Loads listed in those1 summary sheets. All calculated
' actual loads are lower or equal to' the Maximum Design Loads l_isted.
The inspector questioned why the lower loads were- different. . .The
licensee's engineers replied that three cases applied, but none of
the calculated actual. loads is higher than the Maximum Design Loads
q
generally used for the pipe support design calculations.
Fi rst , the
.,
calculated actual' loads were multiplied by 110% for conservatism and.
rounded up to the closet figures.
Second, the calculated actual
loads (without adding 10's) were-close to the previous Maximum Design
Loads and the previous Max Design Loads were used.
Third,. the
calculated -. actual loads af_ter adding 10% were still lower than the
previous Maximum Design Loads and the previous Maximum Design Loads
were kept since the support calculations were designed using the
previous loads.
Since . the calculated actual loads were lower and
were not reflected in the Max Design Loads the support design
engineers were not supposed to review or evaluate the' load changes.
This practice of not revising the Max Design Loads to- reflect the
calculated actual loads will blind the review and documentation of
the pipe support design calculations due to stress load. changes and
is considered a generic problem for the stress calculations.
The individual load' figures shown on the support loadings summary
sheets for Support Nos. S-16, S-9, and'S-10 on the stress calculation
problem No. 63 were checked against the computer printout to verify
the accuracy.
All the figures were picked up correctly from the
computer printout,
c.
Support Calculation Review
The stress calculations on four isometrics were randomly selected to
compare their Maximum Design Loads, listed in the support loadings
summary sheets, to the design loads used in the support design
calculations.
The four isometrics with the corresponding stress
calculation are listed below:
1-
_ - _ - - _ .- _ -__ -___-
- - _ _ _ - - - - _ _ _ _ _ - _ _ _ _
_ - _ - _ - _ _ _ _ _ _ _ -
- _ _ _ _ - - . -
_
-
-
--
- _ _
_ _ . - - - - - - - _ - _ - - _ - - . - - - _ - _ - - _ - _ _ _ - _ - _ - . - - - - - - - - - _ - - - _ _ - - - -
.
.
'
.
3
Table 1
1
_l
l
l
Stress
l
l
Isometric
lRev l Calculation lRev.
Line Description
l
No.
lNo. I
No.
lNo.
I
I
I
I
HPCS PP Discharge to Containment l HL-1349A
1161
63
l
9
l
-1336C
l
l
l
1
-1336L
l
l
l
RHR, LPCI "A" & "B" & CTMT Spray I HL-1348F
l 15 l
141-A.
I
1
I
I
l 13
l
l
l
1
IFSK-P-1013M.001-Cl
4l
l
1
I
I
I
Standby Surface Water Loops
l H-1358K
l
7l
170
1
5
l H-1358L
l
8l
l
A total of about 80 support design calculations from the above four
stress calculations were randomly selected and the design loads used
checked against the Maximum Design
Loads listed in Stress
Calculations. The support calculations for stress calculation No. 63
were also reviewed and evaluated in detail.
'
(1) Stress Calculation No. 63
This stress calculation contains 18 support calculations.
I
Seventeen support calculations were mainly reviewed for
comparison of loads between the latest stress calculations and
the latest support design calculations. One support calculation
was not available for comparison. All support calculations were
completely or partially reviewed and evaluated for member sizes,
weld sizes, anchor bolts, base plates, standard components,.
thoroughness, clarity, consistency, and accuracy.
In general,
the design calculations were of good quality. The calculation
package contained cover sheet, computer input, computer output,
and verification of critical members, deflection, member
flexibility, welds,
anchor bolts, and base plates.
The
discrepancies or comments are listed in Table 2.
More than 90%
of the standard components, such as struts and clamps were not
verified in the support calculations for their actual loads
against the manufacturer's load capacity sheets or catalogs.
l
-- - _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ -
,
_
_
_
__.
_ ____ _ . _ _ _
4
.
p
. . -
-
-
4
l
l
i
i.
'
Table 2
Support
l
1
l
Calculation
l
Rev.
l
Stress l
No.
I
No.
I
No.
I
Discrepancy / comment
!
I
l
!
Q1E22G001C08 l
B
l
S-1
l Note 1
l
l
1
01E22G001C01 l
G
l
S-2
l Note 1
l
l
l
Unknown
l
l
S-3
l Calculation was not received.
-l
I
l
Q1E22G001C03 l
C
l
S-4
I Clamp was not verified in
i
i
I
l calculation.
I
I
i
!
'
Q1E22G001R03 I
C
l
S-5
l Note 1
I
I
I
Q1E22G001C04 l
B
l
S-6
l The resolutions of forces on
l
l
l p.5 were wrong.
Therefore
,
l
l
l the computer input and inter-
I
l
l actions of bolts were wrong.
I
I
I
l
l
l Note 1
I
I
I
Q1E22G001R02 l
D
l
S-7
l DCN #2 was not incorporated
.i
l
l in calculation.
'
l
l
1
l
l
l Note 1
I
I
I
Q1E22G001C02 l
D
l
S-8
l Note 1
,
d
I
I
I
l
l
l Reaction on Joint 1 on p. 10
,
l
l
l of computer output should be
l
l
l bearing or compression
,
l
l
l only. The wrong computer
l
l
l model produced the wrong
l
l
l output.
<
l
1
I
Q1E22G001C05 l
8
i
S-9
l Note 1
I
l
I
Q1E22G001C06 l
C
l
S-10
l Note 1
I
I
I
Q1E22G001C07 l
D
l
S-11
l None
j
l
l
l
Q1E22G001R04 l
C
l
S-12
l Note 1
l
l
1
)
__ _
_ _ _ _ _ _ - .
_ __-_______________ _____ _ _ ___ - -
_ _ - _ _ _ _ _
'
.
.
-
.
5
Table 2
Support
l
l
l
Calculation
l
Rev.
I
Stress l
No.
I
No.
I
No.
I
Discrepancy / comment
(cont'd)
Q1E22G001R05 I
B
l
S-12
i Note 1
I
I
I
<
Q1E22G001R06 l
C
l
S-13
l Note 1
l
l
l
Q1E22G001R07 l
A
l
S-14
l Note 1
l
l
l
Q1E22G001H01 l
B
l
S-14
l None
1
l
1
Q1E22G001H04 l
A
l
S-15
l None
i
I
l
Q1E22G001H05 I
B
l
S-16
l Note 1
l
l
l
l
l
l The critical weld between
l
l
l bracket of Items 3 and 2
l
l
l (w4x13)was not reviewed.
Note
1.
Sway strut and clamp were not verified in support
calculation.
(2) Stress Calculation No. 141-A
This stress calculation containing about 42 support calculations
15 support calculations were randomly selected and reviewed for
comparison of loads between the latest stress calculation and
the latest support design calculations.
Table 3 lists the
discrepancies or comments.
The standard components such as
spring cans, sway struts, clamps, etc. were also not verified in
support calculations for their actual loads against the
allowable loads stated in manufacturer's load capacity sheets or
catalog.
Seven out of 15 support design calculations had the
latest lower stress loads than the latest support design loads.
The support design calculations were not reviewed for the latest
lower stress loads and documented the evaluation.
.
l
.
.
.
6
Table 3
Support
l
l
l
Calculation
l
Rev.
l
Stress l
No.
l
No.
I
No.
l
Discrepancy / comments
l
I
l
Q1E22G015C04
l
D
l
S-12
l None
i
I
I
Q1E12G015R31
l
D
l
S-37
l None
1
l
1
Q1E12G015C06
l
C
l
S-13
l None
l
I
l
Q1E12G015C08
i
B
l
S-5
l Note 1
l
l
1
Q1E12G015H07
l
B
l
S-29
l Note 1
I
I
I
Q1E12G015H08
i
B
l
S-30
i Note 1
l
l
1
Q1E12G015H09
l
B
l
S-32
l None
l
l
l
Q1E12G015H06 & H10 l
B
l
S-28
l Note 1
i
l
I
Q1E12G015H11
l
B
l
S-4
l Note 1
l
I
l
Q1E12G015H14
l
B
i
S-6
l Note 1
I
I
I
Q1E12G015H15
l
0
l
S-33
l None
I
l
l
Q1E12G015H17
l
C
l
S-37
l None
I
l
l
Q1E12G015H19
l
G
l
S-7
l None
l
l
l
Q1E12G015H2O
I
D
l
S-2
l None
l
I
l
Q1E12G015R10
l
B
l
S-6
l Note 1
l
l
1
Q1E12G015R13
l
D
l
S-14
l None
Note:
1.
The latest stress loads were lower than the support
design loads and the effects of the lower loads on the
support design calculations were not reviewed and
documented.
(3) Main Steam Line
Both the Main Steam Line and the Recirculation, Line do not have
support design calculations.
All supports an both lines are
standard components such as snubbers or springs. Therefore, the
i
__
_
-_
__
_ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ - _ _ _ _ _ _ - _ _ _ _ - _ _ _ - _ _ _ _
- i
.
.
i
7
l
licensee generally did not create support design calculations to
'
verify the actual loads against the component allowable loads
set by the manufacturer's load capacity sheets or catalog.
Twenty-four
loads
shown
on
Drawing
i
i
No. FSK-P-1013M.001-C, Rev. 4 were compared to MPL No. B21-G006,
Installation and Instruction Manual for Main Steam System,
Rev. A, Grand Gulf I and 2 Montek Division of E-Systems, dated
November 1980.
The Manual illustrated parts listing, snubber,
and accessories.
The snubber loads shown on drawing were
correct. The snubber loads were also compared to the wmputer
printout of stress analysis performed by General Electric
Company and all loads were correct.
Ten variable springs (or supports) shown on the above drawing
were also compared to MPL B21-G002, ITT Grinnell Corporation
I
Installation Manual for Grand Gulf I & II and the attachment,
MSS
Pipe
Hangers
and
Supports
Fabrication
and
Installation
Practices,
by Manufacturers
Standardization
Society, Inc., 1978 Edition. The spring loads showed on drawing
were correct.
(4) Stress Calculation No. 170
j
Fourteen out of 35 support calculations were randomly selected
and reviewed for comparison of loads between the latest stress
calculation and the latest support design calculations.
Calculation Nos. Q1P41G010R01, Rev. A for Stress No. S-3 and
Q1P41G010C01, Rev. A for Stress No. S-4 were found to have
latest stress loads lower than the support design loads and the
support design calculations had
at been reviewed and the
evaluation documented.
d.
Procedures and Specifications Review
The following Nuclear Plant Engineering Administrative Procedures and
Technical Specifications were reviewed during this inspection:
(1) Procedure No. 305, Engineering calculations, Rev. 11
(2)
Procedure No. 307, Engineering Drawing
(3)
Procedure No. 315, Updating /As-Building Grand Gulf Nuclear
p
Station Design Documents, Rev. 7
(4) Procedure No. 323, Design Inputs, Rev. 1
l
(5) Specification No. 9645-C-103.1, Design and Installation of
Concrete Expansion Anchors, Rev. 9
__-_-_-_ _ --____----_______ -
_
_
_
_
- . _
_
___ _ _ - - _ _ _ _
_ _ _ _ _ _ _ _
_ _ _ _ _ _
,
I
.
"
w.
<
t
. ['
.
i
'(6) Specification ' No. 9645-M-300.2, Design Specification for Pipe
Hangers, Supports, Restraints,-and Anchors, Rev. 18
.
1
'(7) Specification No._9645-M-220.0, Design Specification for Nuclear
Piping Systems, Rev. 15
None of above procedures or specifications required. action, when. the
. latest stress loads were lowe- than the previous stress. loads.or the
design -loads.
The licensee's engineers replied that common sense
required them to review and revise. the support design calculation
g
when the latest stress loads were higher than the previous stress
loads or design loads, but for the stress loads. lower than the.
previous stress . loads or design loads, the design engineers thought
'that the safety factor would be higher, and while they did' review the'.
.
design calculations they did not document the evaluation since. no
4
'
procedures required them to do so, _ There were no evidence.in the
i
design calculations to prove that the design calculations were
reviewed to compare the latest lower stress loads against the higher
1'
design.. capacity.
e.
Findings and Results
After reviewing the . stress and support calculations, five potential-
)
R
generic problems were identified:
(1) The latest stress loads were lower than the previous stress
loads, the stress calculations were not revised. The. problem is
,
that the support design engineers do not know the new l stress
!
loads are lower- and do not review the support . design.
j
calculation; and document the review evaluation.
4
(2) The latest stress loads on the Maximum Design Loads were . lower
than the previous stress loads or support design loads.
The
j
support design calculations were not reviewed or their '
i
evaluation documented to determine the safety significance.
l
(3) Base plate flexibility was not considered in anchor bolt
qualifications per requirements of IE Bulletin 79-02.
The
licensee stated that the premise that base plate flexibility
.
'
could be neglected in support design calculations was approved
by NRR, but no ' approval letter was available.
The NRR
l
inspectors also found out this problem during their second
inspection on March 20-24, 1989. The NRR may take care of this
problem by inspecting the licensee A/E firm-Bechtel power
i
Corporation.
(4) Standard Components such as snubbers, sway struts, clamps,
!
springs, etc. were not verified in nearly all of the support
,
design calculations.
A few of the support design calculations
l
did verify the standard componer,ts.
- _
_ _ - _ - _ _ - - _ _ - - _ _ _ - - _ _ _ - _ _ _ _ _ - _ _ - _ - - _ - - _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ - _
_ _ _
_ -
_ _ - _
_ _ _
_
_ - - _ .
_ _ _ _ _ - _ -
..
.
.
9
(5) Main Steam Line and Recirculation Line do not have the support
design calculations.
The licensee acknowledged the above problems and agreed to solve them
by reviewing and regenerating calculations.
The procedures or
specifications will be revised to add the procedures for reviewing
and documenting the evaluation in support design calculations when
the stress calculations are revised, even when the new loads are
lower than the previous design loads.
The licensee did take quick
action and issued " Pipe Support Load Reconciliation Guidelines" on
March 29, 1989, during the inspection, to provide instructions for
the review of pipe support calculation design loads - and the
corresponding piping stress load sheets.
This review will include
all
"Q" designated pipe supports.
Pending the licensee's action to
resolve the above five potential generic problems and discrepancies
found on Table 2, this item is identified as UNR 50-416/89-10-01,
Piping System Calculation Concerns.
3.
Exit Interview
The inspection scope and results were summarized on March 31, 1989, with
those persons indicated in paragraph 1.
The inspector described the areas
inspected and discussed in detail the inspection results listed below.
Proprietary information is not contained in this report.
Dissenting
comments were not received from the licensee.
The inspector expressed
concerns about the five generic problems identified during this
inspection. The management said they would solve the problems in a proper
way by reviewing, revising and regenerating the calculations and
procedures.
UNR 50-416/89-10-01, Piping System Calculation Concerns.
!
-
- - - - - - - - - _ - - - - - - _ - _ -