ML18057A939
| ML18057A939 | |
| Person / Time | |
|---|---|
| Site: | Palisades  |
| Issue date: | 05/24/1991 |
| From: | James Gavula, Jeffrey Jacobson, James Smith NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | |
| Shared Package | |
| ML18057A938 | List: |
| References | |
| 50-255-90-25, NUDOCS 9106040174 | |
| Download: ML18057A939 (38) | |
See also: IR 05000255/1990025
Text
\\,,
. I
- . **
. U.S. NUCLEAR REGULATORY COMMISSION
REGION I I I
Report No:
50-255/90025(DRS)
Docket No:
50-255
. License No:
Licensee:
Consumers Power Company
1945 West Parnall Road
Jackson, MI 49201
Facility Name:. Palisades Nuclear Generating Plant
Inspection At:
Palisades Site, Covert, MI 49043
Inspection Cond
1uct1. S ptember 19, 1990, through April 18, 1991
r011t
~
Team LeadetJ~-* )t Ja obs on . *
.
rf~.Jcff~~
Insp~ctorL ,.J: A. 'blvul a ~ fov
Approved
.
-~~~~
J. f.i Smith.
~f
P. V.
Lougge~;d
~~?.
R. A. Langstaff
- ~:UC/312~ /q
By:
D.
~- Danielsoh, Chief
Materials and Processes Section
Inspection Summary
stt.'1~1
Date
.
. ;~~/~/
Date
5/z;/l!
Date
/
- JJl(rt(
Date
S/2y/q1
Date
f/2 yf,1
Date
5f?rh1
Date
Inspection on September 19, 1990, through April 18, 1991, (Report No:
50-255/90025(DRS)).
Areas Inspected:
Announced team inspection of activities related to
replacement of the steam generators .. Areas inspected include engine~ring,
field implementation*, and testing activities (73756, 92700, 37700,. 37701);
Results: Three apparent violations were identified; multiple examples of
inadequate design control (Paragraphs 2.a and 5); failure to follow procedures
(Paragraphs 2.a and 5); and insufficient corrective actions (Paragraphs 2.a
and 5).
No Notice of Violation is being issued at this time pending further
evaluation by the NRC .
- ~11(>6040174
F'DF,:
ADOC:r<
G!
910524
05000255
F'DR
~
The following strength was noted during this inspection:
0
The
0
0
0
0
Scheduling, planning, and coordination of outage activities were
considered excellent.
following weaknesses were noted during this inspection:
Mechanical contractor (Townsend & Bottum) welding procedures were not
sufficiently detailed to control the welding process.
Corrective actions taken to improve implementation of piping design
controls were not effective in the areas examined.
Procedural improvements in.the piping design areas did not reflect FSAR
commitments.
Although technical audits identified nonconformances in implementation
of Bechtel's design controls, effective corrective actions to prevent
recurrence did not occur .
2
TABLE OF CONTENTS
. SECTION
Executive Summary
i
1. O
Piersons Contacted ......... ~ ......... ~ .............. ' ... ~...... I
2.0 * Licen~ee Action on Previous Inspection Finding~ .............. 1
3.0
Review of Licensee Event Reports *..... ~~ ................ ~ .... 7
4.0
Steam Generator Replacement Project Implementation ........... 7
. 5.0
Design Engineering ........................................... * 13
6.0
Testing Activities ....... ~ .................................... 27
7.0
Unresolved Items ..........*............................ ;~ .... 32
a*. 0
Open Items . ................................................... 3-2
9.0
Exit Interview ............
- .................................... 32
. *
Executive Summary
This report ~ocuments the NRC inspection of the design, implementation and
test activities associated with the Palisades Steam Generator Replacement
Project (SGRP).
Palisades' planning efforts for the SGRP began several years
ago with_ subsequent design activities starting in early 1990, field-
implementation during mid to late 1990, and start-u~ tests during early 1991.
Although the steam generators were replaced with nearly identical units,
extensive modifications were required to the containment structure and
attached piping systems.
Various portions of these changes along with
modificatitins to other safety related plant equipment were reviewed during the
course of this team inspection.
_Previous NRC inspections at Palisades have identified numerous examples where
the design controls for the modification process did not perform as required. *
This weakness had been both recognized and ~cknowledged by the plant and
extensive efforts had recently been undertaken to improve this- area.
The
prob 1 ems as characterized by the previous inspect i ans were not -SO much a 1 ack
of design controls as much as poor performance of the controlling process.
While the recent improvement efforts appear to have further enhanced the
process controls, the results of this inspection indicate the performance
problems still exist.
As with the previous inspections, very few problems were noted during the
field implementation of the modifications.
In this respect, the SGRP was
considered successful with construction activities being well planned and
coordinated. Except-for the problems with the reactor coolant pipe welding
and containment liner plate welding, there were no major field implementation
setbacks.
-
Similarly, however, design control deficiencies continued to be identified in
the calculations used as a basis for the field implementation.
While the
majority of these deficiencies did not result in hardware changes and were not
of great safety significance individually, the number and nature of the design
errors was significant, was indicative that the design control process was
still not performing well, a_nd had the potential to produce significant
reductions in safety margins. Although several of the deficiencies could be
classified as documentation problems, they have an effect on the accuracy and
adequacy of existing analyses when used for modifications.
Previous NRC
inspections have noted similar deficiencies which resulted in a loss of
confidence in design basis information and subsequent extensive
reverification. This inspection again identified instances where design basis
calculations were assumed to be correct without careful verification of their
application.
For example, original seismic response spectra were utilized for
the revised piping design specification without regard to inconsistencies with
current Final Safety Analysis Report (FSAR) commitments .
- .
. '
I.
DETAILS
Persons Contacted
Consumers Power Company (CPCol
- D. P. Hoffman, Vice President, Nuclear
- .*D. W. Joos, Vice President, Energy Supply Services
- R. D. Orosz, Nuclear Engineering and Construction Manager
~ *W. Clark, Engineering and Construction Manager
- G. B. Slade, General Manager
- J. G. Lewis, Assistant Project Manager
- J. L. Kuemin, Licensing*Engineer
J. C. Nordby,- Welding Engineer
- W. L. Roberts, Licensing Engineer
T. Fouty, Senior Nuclear Operations Analyst
M. Vanek, Level II, Ultrasonics Examiner
Bechtel Construction
M. Charney, Field Services Manager
R. Steffy, Project Manager
H. Kaiser, Mechanical Engineer
E. Brueckner, Welding Engineer
G. Finnam, QC Welding Engineer
U.S. Nuclear Regulatory Commission CNRC)
- H. J. Miller, Director, DRS
- T. 0. Martin, Deputy Director, DRS
- D. H. Danielson, thief, M&PS
. *B. E. Holian, Project Manager, NRR
- J. K. Heller, Senior Resident Inspector
J. A. Hopkins, Resident Inspector
The NRC inspectors also contacted other licensee and contractor
- personnel during the inspection.
- Denotes thos~ attending the exit meeting on April 18, 1991.
2.
Licensee Action on Previous Inspection Findings (92702)
a.
(Closed) Unresolved Item (255/89007-05): *Questionable design
control practices regarding branch connection welds on Auxiliary
Feedwater bypass piping.
Full penetration welds as required by
design, were shown as fillet welds on the installation documents.
The licensee responded to this issue in a letter to the NRC dated
. December 18, 1989. According to the response, the only method to
verify full penetration for the welds in question was to employ
remote boroscopic examination. This would, according to their
response, require disassembly of adjacent valves to gain access to
the inside of the pipe and would be performed for the four branch
connectio~ welds no later than the 1990 refueling outage.
The
I
,*
licensee's above corrective actions were tracked under Event
Report E-PAL-89-030P which proposed to verify full penetration
welds for the branch connection. A reference to the licensee's
response letter was included in the Event Report's proposed
corrective action.
During this inspection, the NRC inspectrir asked to see the visual
ex~mination report documenting the v~rification of the welds in
question.
The licensee provided a one page document which was
used to close out the corrective actions for the above event
report .. The document discussed the_circumstances surrounding the
installation of the welds and concluded that full penetration
welds were installed as required. This conclusion was based on
reviews tif the available documentation as well as discussions with *
the job supervisor. Consequently, the visual ~erification as
committed to in the original response, ~as not performed.
The NRC inspector's review of this document disclosed several
inconsistencies in the licen~ee's logic. It was stated that the
connection was initially tack welded then had root passes
performed which, the licensee claimed, would not have been done
unless it was a full penetration weld.
The NRC inspector pointed
out that all of the fillet welds in this modification also showed
these same attributes in the weld sheets and therefore their logic
was fl awed ..
Once the incohsistency was presented, the licensee agreed to
radiograph the four connections in question to potentially provide
a basis for their position. The result~ of the radiographs were
documerited in a CPCo Deviation Report which identified weld No. 14
on Drawing 24804972-01-0 as having incomplete penetration and weld
No.* 1 on Drawing 24804973-01-0 as having incomplete penetration
- and/or slag inclusions;
Based qn the above discussion this Unresolved Item was considered
closed.
Instead this item was considered an example of a
violation of .10 CFR 50, Appendix B; Criteria III, in that design
control measures failed to assure that design requirements wefe
correctly translated into drawings or instructions (255/90025-
0lA).
In addition, since the licensee failed to identify the deficient
conditions and chose to close the event report item without
adquate bases, this was considered an example of an apparent
violation of 10 CFR 50, Appendix B, Criteria XVI, in that
.
corrective action measures failed to promptly identify and correct
a condition adverse to quality (255/90025~02A).
Furthermore, the licensee's corrective action program stated that
if the corrective actions taken differ from the.proposed actions
specified by the Plant Review Committee (PRC), the Event Report
shall be returned to the PRC for concurrence. This process was
2
a.
. *
not followed when, instead of verifying that the welds were full
penetration as described in their response, th~ item was closed
using reviews and discussion~. This was considered an example of
an apparent violation of 10 CFR 50, Appendix B, Criteria V, in
that the procedure for corrective actions was not followed
(255/90025-03A).
.
.
The licensee subsequently issued work requests to repair the two
welds in question prior to the end of the outage.
b.
(Closed) Open Item (255/89007-04):
Inconsistency between the
codes of construction specified in the FSAR and the implementing
procedure.
c.
The license~'s response to the NRC, dated August 10, 1989, stated
that the Palisades staff would complete a reconciliation of all
piping codes of construction to the latest edition of ANSI 831.1,
"Power Piping" and would revise the Updated Final Safety Analysts
Report (UFSAR) to identify applicable codes and standards.
Revision 11 to Palisades UFSAR, dated December 18, 1990, included
a paragraph stating that the code of record for new and existing
piping was changed to ANSI 831.1 (1973) Power Piping Code with
Summer of 1973 Addenda.
This item is considered closed*.
However, during the NRC inspectcir's review of the licensee's code*
reconciliation document, the.licensee ~tated that the
incorporation of th~ "0.75 i" factor for piping stress
jntensification was previously adopted without including any of
the Code related requirements.
It stated that the previous UFSAR
design criteria contained one factor out of the-new Paragraph
- 104. 8 of the Summer 1973 Addenda and none of the re 1 ated
requirements in Paragraph 119.6~4, Figure 119.6.4(8), Table D-1,
Table D-2 or Figure D-1.
When this problem was discussed with licensee representatives, it
was stated that the stress intensification issue was discussed in
an April 1980 meeting between the NRC and CPCo and found to be
acceptable.
Furthermore, in CPCo's submittal dated October 24,
1980; it was stated that the use of different code editions was:
found to be acceptable and reviewed in accordance with
Based on the current code reconciliation document,
the acceptability of this position was questionable and
p6tentially indicated that the 50.59 review for that UFSAR change
was inadequate.
Pending a review of the licensee's action to reconcile the impact
of the improper use of the "0.75 i" fattor in all affected
analyses, t~is was considered an Unresolved_ Item (255/90025-04). *
(Closed) Violation (255/89024-01):
Design bases were not
correctly translated into desjgn documents and the design adequacy
was not sufficiently documented .
3
- -
. **
litensee's above corrective actions were tracked urider Event
Report E-PAL-89-030P which proposed *to verify full penetration -
welds for the branch connection. A reference to the licensee's
response letter was included-in the Event Report's.proposed
corrective acti~n.
During this inspection, the NRC inspector asked to see the visual
examination report documenting the verification of the welds in
-question.
The licensee provided a one page document which was_
used to close out the corrective actions for the above event
report. The document discussed the circumstances surrounding the
installation of the welds and concluded that full penetration
welds were installed as required. This conclusion was based on
reviews of the available documentation as well as discussions with
the job supervisor. Consequently, the visual verification as
committed to in the original response, was not performed.
The NRC inspector's review of this document disclosed several
inconsistencies in the licensee's logic. It was stated that the
connection* was initially tack welded then had root passes
performed which, the.licensee claimed, would not have been done
unless it was a full penetration weld.
The NRC .inspector poin_ted
out that all of the fillet welds in this modification also showed
these same &ttributes in the weld sheets and therefore their logic
was fl awed .
. Once the inconsistency was presented, the licensee agreed to
radiograph the four connections in question to potentially provide
a basis for their position. The results of the radiographs were
documented in a CPCo Deviation Report which identified weld No. 14
on Drawing 24804972-01-0 as having incomplete _penetration and weld
No. 1 on Drawing 24804973-01-0 as having incomplete penetration
and/or slag inclusions.
Based on the above discussion this Unresolved Item was considere~
closed.
Instead this item was considered an example of a
violation 6f 10 CFR 50, Appendix B,
Crite~ia III, in that design
control measures failed to assure that design requirements were
correctly translated into drawings or instructions (255/90025-
0lA).
-
In addition, since the licensee failed to identify the -deficient
conditions and chose to close the event report item without
adquate bases, this.was considered an example of an apparent
violation of 10 CFR 50, Appendix B, Criteria XVI, in that
corrective action me.asures failed to prompt 1 y identify and correct
a condition adverse to quality (255/90025-02A).
Furthermore, the licensee's corrective action program stated that
if the corrective actions taken differ from the proposed actions
specified by the Plant Review Committee (PRC), the Event Report
shall be returned to the PRC for concurrence. This process was
2
- -
not followed when, instead of verifying that the welds were full
pen et ration as described in their response, the item was closed .*
using reviews and discussions. This was considered an example of
an apparent violation of 10 CFR 50, Appendix 8, Criteria V, in
that the procedure for corrective actions was not followed
(255/90025-03A).
.
-
The licensee subsequently issued work requests to repair the two
welds in question prior to the end of the outage.
b.
(Closed) Open Item (255/89007-04):
Inconsistency between the
codes of construction specified in the FSAR and the implementing
c.
procedure.
The licensee's response to the NRC, dated August 10, 1989, stated
that the Palisades staff would complete a reconciliation of. all
piping codes of construction to the latest edition of ANSI 831.1,
11 Power- Piping
11 a:nd would revise the Updated Final Safety Analysis
Report (UFSAR) to identify applicable codes and standards.
Revision 11 to Palisades UFSAR, dated December 18, 1990, included
a paragraph stating that the code of record for new and existing
piping was changed to ANSI 831.1 (1973) Power Piping Code with
Summer of 1973 Addenda.
This item is considered closed.
However, during the NRC inspector's review of the licensee's code
reconci 1 i at ion document; the 1 i censee stated that the *
incorporation of the
110..75 i
11 factor for piping stress
intensification was previously adtipted without including any of
the Code related requirements.
It stated that the previous UFSAR
design criteria contained one facto~ out of the new Paragraph
104.8 of the Summer 1973 Addenda and none of the related
requirements in Paragraph 119.6.4, Figure 119.6.4(8), Table D-1, *
Table D~2 or Figure D-1.
-
When this problem was discussed with 11-censee repre$entatives, it
was ~tated that the stress i~tensification is~ue w~s disc~ssed in
~n April 1980 meeting between th~ NRC and CPCo and found to be
acceptable.
Furthermore, in CPCo's submittal dated Octobet 24, *
1980, it was stated that the use of different code editions was
found to be acceptable and reviewed in accordance ~ith
- *
Based on the current code reconciliation document,
the acceptability of this position wa~ questionable and
potentially indicated that the 50.59 review for that UFSAR change
was inadequate.
Pending a review of the licensee's action to reconcile the impact
of the improper use of the
110.75 i
11 factor in all affected
analyses, this was considered an Unresolved Item (255/90025-04].-
(Closed) Violation (255/89024-01):
Design bases were not
correctly translated into design documents and the design adequacy
was not sufficiently documented .
3
..
In their letter dated April 12, 1990, the licensee responded to
the items of noncompliance and unresolved ttems documented in NRC
Inspection Report 50-255/89024 as well as the programmatic
observations made in the December 15, 1989, *NRR Audit Team Report.
For "Corrective Actions to Prevent Recurrence" for the inspection
report'~ design control violation e~ampl~ A~2 and procedure
violation example B.3, the licensee discussed their.intention to
upgrade CPCo Specification M~195 "Requirements for the Design and
Analysis of Palisades Plant Safety Related Piping and Instrumerit
Tubing."
The revision to M-195 was further discussed in the
licensee's responses to Unresolved Items No. 4 and No. 5 in the
inspection report and to Observations No. 3, No~ 5, and No. 8 in
the NRR Audit Report.
The intent of the revision was to provide
clarifi~atio~ of how seismic response spectra were to be applied,*
increase usability, highlight the use of certain seismic ~pect~a~
include integral welded attachments in the evaluation and design
of piping supports and incorporate recent lessons learned and
experiences.
Revision 1 to CPCo's Specification M-195 was issued on May 9,
1990, with extensive changes to the document. Attachment 3 of the
specification was entitled "Original Palisades Plant Response
Spectra and Building Displacements," and consisted of a set of
tables and graphs for various elevations .in the auxiliary
building~ reactor building shell and reactor building internal
' . concrete structure. This information was generated from a s.et of
Bechtel straight-edged plots which CPCo, in their August 10,. 1989,
response to NRC Inspection Report 50~255/89007, acknowledged were
very difficult to read.
The clarity and precision of these plots
may have contributed to the following problems noted during the
NRC review of this data:
(I)
Palisades UFSAR Section 5.7.1.3, "Floor Design Response
Spectra," states that peaks at building natural frequencies
were wjdened +/- 10% to account for variations in soil and
structure material properties ..
Contrary to the above, the response spectra peak for the
first natural frequency of the containment was only widened
between 6.97% and 7.77% on five of the seven floor
elevations.
For the second natural frequency of the
containment, four of the seven floor elevations were widened
less than the prescribed 10%.
For the third natural
frequency of the containment internal structure~ the peak
was not widened at all ~or elevation 649 feet.
Thi~
insufficient widening discrepancy also applied to the
respon$e spectra for all elevations in the auxiliary
building. This ~as considered an example of an apparent
violation of 10 CFR 50, Appendix B, Criterion III in that
the design basis_* information was not correctly translated
into procedures (255/90025~018) .
4
..
(2)
Palisades UFSAR Figures 5.7-9, 5.7-IO, and 5.7-I6 give
typical floor response spectra for the containment building
at elevations 590 feet and 683 feet and for the auxiliary
building at elevation 649 feet, respectively. Zero Period
Acceleration (ZPA) values given in these figures correspond
to the acceleration values given in UFSAR Figures 5.7-7 and
5.7-I5 at the appropriate elevations~
Contrary to the above, the ZPA values given in the Tables of
Attachment 3 to Specification M-I95 were lower than the ZPA
values listed in UFSAR Figures 5.7-7 and 5.7-I5 in II of the
I3 elevations given.
The worst case was a value 20% lower
than that specified for the internal structure at elevation
649 feet. This was considered an example of an apparent
violation of IO CFR 50, Appendix B, Criterion III in that
design basis information was not correctly translated into
procedures (255/90025-0IC).
- In addition to the above discrepancies, the following
inconsistencies in the response spectra data were ~oted by the NRC
- inspector:
(3)
The methodology used to widen the spectra peaks at the
building natural frequencies did not appear to be
appropriate.
The acceleratiun plots used as bas~s for the
information in Attachment 3 of M-I95 showed peak broadened
spectra with sides which were not parallel to lines forming
the original spectra peaks. This has been found by the NRC
to be an unacceptable practice in the past.
(4)
For the auxiliary building, the UFSAR gives the fundamental
natural frequency of the reinforced concrete structure as
4.1 Hz in the east-west direction and 5.4 Hz in the
north-south direction. * As stated i~ the UFSAR, the results
foi each direction were enveloped to form a*single set of
results for ,the auxiliary building.
An apparent
inconsistency was noted during a review of the auxiliary
buildin~ response spectra. If these spe~tra were enveloped
for both sets of directions, it would be expected that at
least some building amplification would be seen at both
fundamental natural frequencies.
This was not the case with
the 4.1 Hz natural frequency in the east-west direction.
Comparing values, the east-west acceleration at the
fundamental frequency was a factor of 6 to 9 less than the
acceleration values at the fundamental frequency in the
north-south direction. Although the peak acceleration value
may be bounded with the existing spectia, the frequency
i
content mai not be sufficient to envelope the seismic input
for piping analyses .
5
d.
(5)
All of the original Bechtel response spectra plots for the*
containment building and auxiliary building have**
acceleration peaks occurring at 3 Hz.
However, the lines
drawn on these spectra plots to represent the bounding input
spectra did not include these peaks.
As an example, for
containment elevation 590 feet, th~ value indicated on the .
original Bechtel plot was approximately 0.6 g ~hereas the
value given in Attachment 3 of M-195 was approximately 0.4g.
For piping systems with fundamentals frequencies close to 3
Hz, the seismic response is potentially underestimated by
(6)
50%.
-
The spectra peak value.s for auxiliary building elevations
589 feet, 601 feet *and 610 feet were given as 2.650g,
2.550g, and 2.784g respectively. Since the peak *
acceleration values increase with elevation at all other
floor locatibns~ it was inconsistent for the value at
elevation 601 feet to be less than the value at elevation
589feet. While this may indicate that the peak value at
elevation 589 feet was conservatively high, this was not
intuitively obvious and some verification would be
appropriate in light of the other inconsistencies noted
above.
Pending a review of items 3, 4, 5, and 6 above by the
licensee, with written bases for the adequacy of the data
given in Attachment 3*of Specification M-195, this .was
consid~red an Unresolved- Item (255/90025-05).
(Closed) Open Item (255/90018-01):
Although used for ~scertaining*
technical specification compliance for safety injection tank
levels, the safety injection tank level switches were not
calibrated.
Upon review, the licensee discovered that the*
- actuation level of the switches had the potential to drift as much
as two inches. Since the level switches had been set at the
technical specificatibn limits, it was possible that these limits**
may have been exceeded.
(Licensee Event Report 90-021 was
- submitted to report the discovery_.)
However, the levels would
still have been within the recently revised 1 imits, as approved by
technical specification Amendment 136.
By revision of the
technical specification limits and modification of the low level
switches (see paragraph 6.d of this feport, facility change FC-
905) such that the switches were set two inches on the
conservative side of the technical specification limits~ potential
instrument drift was actourited for.
The licensee's commitment to
develop a setpoint methodology for instrumentation was
i~ the
process of being implemented.
However, at the time of this
inspection, the methodology had not yet been applied to these
level switches.
An I&C supervisor stated that the setpoint
methodology would be applied in the future to the level switches
a$ part of a periodic review. This item is considered closed .
- 6
L.,
"°lj
-*
3.
Review of Licensee Event Reports CLER) (92700)
4.
a.
(Closed) LER (255/90021-LL) See Open* Item 255-90018~01, paragraph
2.d of this report. This item is considered closed.
b.
'(Closed) LER (255/900i8~LLl ."Inadequate Flow Through PCS Hot Leg
Injection Check Valves."
During performance of R0-65 "HPSI/RHPSI Check Valve Test" on
September 30, 1990, the hot leg injection (HLI) flow rates were
less than required by the acceptance criteria for full stroke
verification of the HLI check valves.
As a result of this test
failure, the licensee questioned the adequacy of the acceptance
criteria~
The licensee subsequently implemented corrective actions to ensure
that the acceptance criteria for full stroke veri fi cation reflects
the design basis requirements.' The licensee's review concluded
that the failure of the valves to full stroke was attributed to
improper application of the valve design.
Valve design
modifications were being planned by the licensee.
The NRC inspectors noted that the licensee could have avoided this
event if actions taken to address Generic Letter (GL) 89-04 had
been completed earlier.
GL 89-04 required licensees to review
their implementing pro~edures and programs and amend them as
necessary to ensure conformance with the positions stated in GL 89-04.
The licensee's review was ongoing at the time of the
event, and this procedure had not yet been evaluated.
Steam Generator Replacement Project (SGRP) Implementation (37701)
a;
Background-
Steam generator (SG) tube degradation appeared.early in the
Palisades Plant operating life. The first incidence of tube
corrosion failure occurred after less than one year of operation*
(1972). After approximately four years of operation, 1,775 tubes
out of a total 8,519 (over 20 percent) had been plugged due to
corrosion. A change in water chemistry in mid-1974 slowed the
corrosion rate down such that only 378 additional tubes were
plugged during t~e period 1976 to 1987.
However, d~ring the 16
month period beginning in December 1987, seven tube leaks resulted
in six forced outages.
A decision was made to replace the SGs in conjunction with the
plant refueling outage beginning September 15,
1990~ and
continuing through February 1991. lhe replacement SGs were
fabricated by the original equipment manufacturer, Combustion
Engineering.
7
l, .
To date, six U.S. plants (all Westinghouse design) have completed
SG replacement projects.
For those projects, the containment
equipment hatch openings were of sufficient size to accommodate
the SGs.
The Palisades replacement was unique in that the
significantly larger size of these SGs prevented removal from the
containment through.the equipment hatch. "This required that a
special construction opening, 26 feet QY 28 feet, be cut in the
containment wall.
Another unique feature of the Palisades effort was the first U.S.
application of the narrow groove welding technique developed in
. Germany.
This technique was utilized to reattach the primary
coolant piping to the SG nozzles.
b.
Procedure Review
In preparation for the SG replacement activities, the following
Bechtel procedures and safety evaluations were reviewed:
(1)
Construction Work Plans/Procedures:
WP/P NO.
TITLE
REVISION
1
Establishment, Control, and Implementation
0
of the Construction Work Plan/Procedure
Program .
2
Bechtel Site Orgahization
0
3
Indoctrination and Training
0
4
Document Control
0
5
Field Change Request/Notice
0
6
Field Sketch Preparation
0
7
Deviation Control
0
8
Field Material Requisition
0
9
Material Control
0
11
Housekeeping
0
12
Construction Safety Tagging and Clearance
0
Procedures
14
Control of Temporarily Remoyed Permanent Plant
0
Material
8
i.,_
' .. J
I
15
Maintenance of Materials and Equipment While
0
in.Storage
18
System Turnover to CPCo
0
19
Construction Test Program Requirements
0
20
Field Fabrication
0
21
Safeguards Information
0
22
Installation of Concrete Expansion Anchors
0
P-3
Piping System Cleanness During Construction
0
(2) * Welding Procedures (Automatic)
o
General Welding Standard (Narrow Gap Technique)
GWS-NC, Revision 1
o
Gener~l Welding Standard~ GWS-1, Revision 2
o
Narrow Groove Welding Procedures
o
Pl-G2 (Clad)-NGT-Ag (CVN), Revision 1
0
0
P3-G3 (Clad)-NGT-Ag (CVN), Revision .1
Pl(G2) NGT-Ag (CVN), Revision 1
o
P3(G3) NGT-Ag (CVN), Revision 1
o
Pl-T Clad-0, Revision 1
o
P3-T Clad-0, Revision 1
o
P3-G3 (Clad), Pl-G2 (Clad)-NGT-Ag (CVN), Revision 1
o
P3(G#) and Pl(G2) NGT-Ag (CVN), Revision 1
(3)
Welding Procedures (Manual)
o
Pl-A-Clad, Revision 0
o
Pl-T-Clad, Revision 0
o
P3-A-Clad, Revision 0
o
P3-T-Clad, Revision 1
0
Pl-G2 (Clad)-T (CVN), Revision 1
9
r---- --
-
I
I
0
- P3-G3' (Clad)-T (CVN), Revision 1
o
Pl-A-Lh (CVN), Revision 2
.(4)
Replacement bf St~am Generators (safety evaluations)
o
Containment Construction Opening
o*
Replace 2 Inch Slowdown Piping with 4 Inch
o
Remove and Reroute Auxiliary Feedwater Piping
o
Rigging* and Transportation of Heavy Loads
All procedures and safety evaluations reviewed were found to be
well documented and to demonstrate sound technical rationale.
c.
Material Documentation
The replacement SGs were manufactured by Combustion Engineering in
Chattanooga, Tennessee, and shipped to Palisades by barge in
September 1989.
Improvements incbrporated in the riew SGs include
increased blowdown ~apacity; improved flow characteristi~s, and
optimized tube support design.
The SG shells are fabricated from
ASME SA-533, Grade A, Class 1 low alloy steel .. Th~ to~ arid bottom
heads are fabricated from ASME SA~516, Grade 7n and SA-533, Grade
B, Class 1 low alloy material respectively.
Tube ~aterial .is
Inconel
~00 (SB-163) and remains unchanged from the original SG
design.
The NRC inspector reviewed the Code Data Form (N-1} for
the SGs and traced the documentation path*back to select certified
Material Test Reports (CMTRs).
The documentation for the SG
materials was found to be traceable and in conformance with Code
requirements.
d.
Containment Opening and Heavy Lifts
The containme*nt opening was prepared by utilizing a combination* of
~ore drilling, concrete saws and specialized rope saw equipment.
During the course of initial core drilling, one of th~ horizontal
tendons used to post-tension the containment structure was
damaged.
Deviation Report No~ 90-267 was issued to disposition
. the damaged tendon which was subsequently replaced.
The NRC inspector observed a portion of the sawing effort and
removal of the 170 ton concrete block.
B6th the cutting and
lifting oper~tions were found to be perform~d_in accordance with
the approved.procedures and industry standards.
Rigging International was the responsible contractor charged with
erection of the lifti~g devices, performing the lifts, and ground
transportation of the SGs .. The NRC inspector observed portions of
the erection of the outside rigging platform and lifting device,
10
. '"'->
and the in-containment semi-gantry crane. All rigging appeared to
conform to the approved design drawings.
Review of the rigging
design was included in NRC Inspection Report 50-255/90003.
The
NRC inspector also observed portions of the SG lifts and ground
transportation. All observed SG movements were performed in
accordance with approved procedures.
The SG movements and lifts
were considered to be well coordi~ated and executed ..
e.
Wel di nq
Throughout the course of this project; the NRC inspector observed
a sample of various portions of the welding process. Observations*
included welder qualifications, weld edge preparation, fit-up,
welding, and final examination.
As previously mentioned, one of the unique features of this
project was the use of the "narrow-groove" (NG} welding technique .
. The NG welding process requires very precise weld groove prepara-
tion and highly specialized automatic welding equipment.
The
application of this process to heavy wall nuclear system *piping at
Palisades was the first of its kind in the United States. The
process had been applied in Europe with a high degree 6f success
and most recently at the Ringhals facility in Sweden.
Initial NRC
review of this process was documented in NRC Inspection Report
50-255/90003.
The NG welding procedure was based on the automatic gas tungston
arc welding (GTAW} process with; as stated above, specialized*
equipment enabling the welding torch to reach deeply into a narrow
weld groove preparation.
The NG welding process was in concert
with the ALARA concept in that the welding equipment was operated
remotely and the narrow weld-groove required significantly less
weld volume which reduced welding time.
Initial application of this process at Palisades was the welding
of the hot leg piping to the SG nozzles.
Both of these weld
joints were considered successful in that only minimal near-
surface porosity repairs were required .. At this time the cause of
the porosity was hypothesized to be a disturbance of the shielding
gas flow due to the mismatch of the outside diameters of the pipe
and SG nozzle.
Further application of this process on the cold leg connections
resulted in severe porosity defects in the completed welds.
Due
to the joint geometry, the porosity appeared as l in.ear i ndi cations
on the Code radiographs.
Skewed techni~ue radiographs were then
obtained which identified a "~all" of scattered porosity along the
fusion line. Although in-process radiography was* a topic of dis-
cussion with the NRC during early reviews of the NG weld process,
it was not utilized on this project. The licensee elected not to
perform in-process radiography due to the relative difficulty of
obtaining meaningful radiographs coupled with the history of
11
-
.
- -
!
success with this Welding process. Since the .cause of the defects
were unknown, a decision was made to grind out the defects and
reweld the joints using the manual shielded metal-arc welding
(SMAW) process.
To date, the root cause of the NG welding defects
is unknown.
The NRC inspector reviewed the final radiographs of the reactor
coolant piping welds and main steam line modifications and found
them to meet Code acceptance criteria. In general, the overall
controls a*pplied to piping fit-ups and to activities associated *
with the NG welding process appeared excellent.
The NRC inspector reviewedBechtel NCR Nos. 14 and 15 which were
cissued to disposition ultrasonic examination (UT) indications on
the steam generator nozzles.
During clad buildup for weld prep,
welding difficulties necessitated an informational UT on the "B"
hot leg. During this examination sev~r~l areas of lack of bonding*
of the clad were identified.
As a result of this finding, all
other nozzles were examined and lack of bonding was also
identified on the "B" cold leg of "A" steam generator.
Thoµgh desirable, a 100% bond is not required for structural
- integrity. However, to facilitate the. NG ~elding of the nozzle
joint, the affected clad was removed and restored to a desirable
conditidn.
Several other NCR's were reviewed and found to be adequately
documented and prudently dispositioned.
Per NRR letter from J. Knight to D. Dutton, NCIG, dated June 26, .
1985, licensees wishing to commit to the Visual Weld Acceptance
Criteria for Structural Welding, must document this commitment in
the form of an amendment to the FSAR. This request was docketed on
November 5, 1990.
The NRC inspector, in accordance with the above
letter, reviewed the licensee's training program and obs~rved a
portion of the training effort and found it to be acceptabl~.
The NRC inspector also reviewed a s~mple bf the mechanical
contractor's (Townsend and Bottum) welding procedures.
During
this review, several ASME Code deficiencies were noted.
While
these deficiencies ap~ear to be documentation errors, the lack of
guidance contained in the procedures is considered to be the more
significant issue.
The procedures, as written, give little or no
direction to the welder.
The CPCo welding procedures were briefly review~d and found to be
somewhat better; however, these procedures are still considered
deficient with respect to controlling the welding process.
Deficiencies in the control. of the welding process contribute to
such current issues as the undersize socket welds and branch
connection discrepancies .
12
....
5 .
')
Design Engineering (37701)
An extensive amount of piping and pipe support modifications were
required to support the SGRP.
Changes in nozzle orientations,
locations, and sizes required reanalyses of several. existing piping
systems. *Changes in seconpary chemistry design required analyses of
se~eral completely new piping systems.
In addition to modifications
associated with SGRP, changes were also made to other plant systems
which required reanalysis of piping and pipe supports.
a.
Bechtel Engineering Efforts
Modifications to piping systems associated with the SGRP ~ere
designed and analyzed by Bechtel's Gaithersburg Office. These
efforts were controlled by design criteria given in CPCo
Specification M-195, "Requirements for the Design and Analysis of
Palisades Plant Safety Related Piping and Instrument Tubing;"
Revision 1, and.by Bechtel's Specification 20557-G-OOlP, "Design
Criteria Documents for Consumers Power Company Palisades Nuclear
Plant Steam Generator Replacement Project", Revision 3.
Paragraph, 5.10.4.2, "Seismic Anchor Movements," (SAM) of M-195 *
specifies that SAMs for the original seismic criteria shall be
taken from Attachment 3 and SAMs for Code Case N-411 seismic
criteria shall be taken from Attachment 4.
For the containment
structure the SAM displacements given in Attachment 3 are lower
than the SAM displacements given in Attachment 4 by a factor of
approximately 3. Paragraph 4.4.2.4.2, "Seismic Anchor Movements,"*
of 20557-G-OOlP specified that displacements to be used were given
in Appendices D and F.
The values given .in these appendices
correspond only to the original response spectra values.
The SAM
displacements for Code Case N-411 seismic criteria were not
included in 20557-G-OOlP. * Therefore, any analysis that used the
N-411 seismic criteria, utilized the wrong SAM values.
This was considered an example of an apparent violation of
10 CFR 50, Appendix B, Criteria Ill, in that design basis
information from M-195 was not correctly translated into
Specification 20557-G-OOlP (255/90~25-0lD)~
This deficiency was corrected with the issuance of 20557-G-OOlP,
Revis'ion 4, January 21, 1991. All analyses that utilized the
incorrect N-411 SAMs were reviewed and revised if necessary.
No
modifications were required as a result of this review.
The following facility change (FC) packages, prepared by Bechtel,
were r~viewed by the NRC inspector:
13
( 1)
FC-911 Main Steam System
As part of the SGRP, *the Main Steam Piping System .was
modified to accommodate the higher nozzle location of the
steam generator. This change r~quired the installation of
an approximately 32 inch piping sptiol piece. The followi~g
documents associated with this facility change were reviewed
by the NRC inspector for comp*l i ance with NRC requirements
and licensee commitments.
(a)
Calculation No. SGRP-PDS-033, "Pipe Stress Analysis of
Steam Generator E50A Main Steam System," Revision 1,
September 6, 1990, and Revision 2, January 21, 1991. ...
The stated purpose of this calculation was to analyze
the piping in accordance with Design Criteria
20557-G~OOlP and CPCo Specification M-195.
During this review the fol1o~ing discrepancies were
noted:
l
In Revision 1, steam generator nozzle SAMs were
taken from structural elevation 649 feet. A
footnote in the calculation stated that the*
nozzle location was at elevation 676 feet 11
. inches *and that the structural movements were
used in accordance with CPCo Specification
M-195.
Specification M-195 discussed SAM
displacements for branch piping decoupled from
run piping and stated that the total seismic
displacement will be used.
By using SAM
displacements from elevation 649 feet, the
. analysis neglected the additional displacement
caused by the 28 feet difference in elevation.
This was considered an example of an apparent
violation of 10 CFR 50, Appendix B,
Criterion III, in that design verification
activities failed to assure that appropriate
design values were used in the pipe stress
analysis (255/90025-0lE) ..
Following discussions with the NRC, the licensee
revised the analysis to account for the
increased SAM displacements.
Using an
extrapolation technique, the licensee concluded
that the SAM displaiements would increase by
approximately 30%.
However, due to the flexible
nature of the main steam piping system,
reanalysis demonstrated that stresses were still
14'
within allowable limfts. This was apparently a
generic problem for any piping system attached
to the steam genetator above elevation 649 feet.
Response spectra given in Paragraph 3.7,
"Applicable Seismic In~ut;" were taken from
structural elevation 649 feet.
However, the
main steam piping is attached to the steam
generator ntizzle at elevation 676 feet 11
inches.
Paragraph 4.4.2.4.1, "Seismic Inertia,"
of Bechtel's Design Criteria 20557-G-OOlP stated
that input response spectra would be developed
by.enveloping the applicable response spectra
for all structures and elevations supporting the
piping.
This w.as an example of an *apparent violation of
10 CFR 50, Appendix B, Criteria III, in that
design verification activities failed to assure
that appropriate design values were used in the
.stress analysis (255/90025-0lF).
A meeting b~tween the licensee and NRC was held.*
on February 8, 1991, to discuss _the above issue.
The presentation given by Bechtel, on behalf of
CPCo, focused on the licensed seismic design
basis for Palisades and then went on to discuss
the main steam line seismic design.
The
presentation concluded by stating that the
response spectrum at elevation 649 *feet was
"representative" of the seismic input at the
main steam nozzle elevation 677 feet.
Although engineering principle~ dictate that
seismic acceleration will be greater at higher
elevations for certain_structural frequencies,
the licensee's presentation did not attempt to
quantify this aspect.
On this basis the
significance of the issue could not be assessed
during the meeting.* The licensee contended that
the current analysis was more conservative than
the original analysis and therefore the analysis
was appropriate.
The fundamental issue was the acceptability of*
using the structur~l model results for
elevations which were significantly higher than
the model elevations. The accuracy of the
structural model to ~redict maximum stresses in
the containment structure was not in question.
However, conservative results for maximum
stresses at the base of the structure ao not
15
- -.
ensure conservative accelerations and
displacements at the top of the "structure", if
the structural location is not included in the
model.
Following the meeting, the NRC inspector noted
the following discrepancies in the licensee's
presentation. First, it appears that the
original main steam analysis was not performed
in accordance with the Palisades FSAR.
Paragraph 5.7.4 of the FSAR states that piping
systems spanning two or more elevations used
seismic curves closest to and higher than the
center of mass of the piping system.
Since the
highest internal structure spectra was for
elevation 649 feet and the center of mass of the
main steam piping was approximately 665 feet,
the FSAR statement was apparently not met.
Secondly, based on a review of the original
analysis, the spectrum for elevation 649 feet
wasn't even used.
Instead seismic spectra from
elevation 608 feet was used.
This second aspect
is another example of deficient I.E. Bulletin
79-14 implementation by Bechtel and will be
considered as a basis to expand the ongoing
safety related piping reverification program
(Refer to NRC Inspection Reports 50-255/89024
- and 50-255/90002).
Further discussions between the NRC and licensee
disclosed that an evaluation had been performed.
to assess the response of the main steam piping
using response spectra extrapolated up to the
677 feet elevation. This evaluation, submitted
on February 20, 1991, concluded that since the
main steam piping natural frequencies did not
correspond to the higher extrapolated spectra
peaks, the piping stresses and support loads
would not exceed allowable limits.
However, to
date the calculation has not been revised to
incorporate this aspect.
(b)
Calculation No. MSA-PD-EB1-H3, "Pipe Support Design
for Main Steam System, Steam Generator E50A, EB1-H3,"
Revision 2, January 21, 1991.
This calculation evaluated a modification to an
existing stanchion located on the lower elbow of the
. main steam riser. This integral welded attachment has
two non-standard lugs attached to the stanchion to
prevent uplift during a seismic event .
16
During this review the following discrepancies were
noted:
l
Paragraph 5.4.13.1.4, "General Requirements for
Integral Welded Attachments," of Bechtel's
Design Crite~ia 20557-G-OOlP states that "only
one-half of the lugs used shall be considered
effective," and- later states~ "when more than
half of the lugs are considered effective the
flexibility of each load path shall be evaluated
and the load distributed accordingly."
Contrary to the above, the analysis assumed that
the restraining force *would be equally
distributed between the two load paths without
considering the flexibility of each load path.
The pipe support was analyzed using a finite
element frame analysis program in which an
identital force was applied to both sides of the
supporting structure. The resulting
displacements indicated that one side of th~
structure was approximately twic~ as fle~ible-as
the other side. * Based on this, the assumption
that the force would be equally distributed was
invalid.
This was considered an example of an apparent .
. violation of 10 CFR 50, Appendix b,
Criterion III, in that design control activities
failed to appropriately verify.the adequacy of
the support design (255/90025-0lG).
Paragraph 5.7.1, "Deflection Requiremerits,
General Requirement," of CPCo's Specification
C-173(Q), Revision 1, stated that "the total
deflection of the pipe support, in the direction
of the restraint, at the point of load ... shall
not exceed 1/16 inch."
Contrary to the above, the analysis failed to
evaluate the total deflection of the pipe
support which, according to the displacements
given in the analysis, exceeded the 0.063 inch
acceptance criteria. The erroneous method used
in the analysis considered the displacement of
the lug separate from the displacement of the
structure. The total displacement of the pipe
support will be the 0.056 inch displacement from
the structure plus the 0.025 inch displacement
from the lug .
17
This was considered an example of an apparent
violation of 10 CFR 50, Appendix B,
Criterion Ill, in that
design control activities failed to
appropriately verify the adequacy of the support
design (255/90025-0lH).
~
Paragraph 5.4.17.1.1.ii of Bechtel's Design
Criteria 20557-G-OOlP, which discusses baseplate
and expansion anchor bolt design, stated that
the analysis must account for expansion anchor
bolt flexibilities.
Contrary to the above, the baseplate analysis
used ancho~ bolt stiffness values derived from
expansion anchor data, which were not applicable
to the four through-bolted one inch diameter
rods attaching the baseplate to the concrete.
This was considered an example of an apparent
violation of 10 CFR 50, Appendix B,
Criterion III, in that design control measures
failed to verify the adequacy of the design
values used in the stress analysis
(255/90025-0lI) .
(c)
Safety Review, "Main Steam System, FC-911,
Revision O," PS&L Log No. 90-0797, Re~ision O,
September 28, 1990.
This was the safety evaluation of the modification to
the Main Steam System which considered the criteria
for an unreviewed safety question as prescribed in
During- this review the following discrepancy was
noted:
Paragraph 5.9.2, of CPCo's Administrative Procedure
No. 9.03A, "Facility Change for SGRP," Revision 0,
March 8, 1990, stated that "Safety evaluati.ons shall
be performed in accordance with Pali~ades
.
. Administrative Procedure 3.07, "Safety Evaluations."
Paragraph 5.2.4 of the above procedure stated that
"When ~nswering each safety review question, the
preparer shall list FSAR ... Sections reviewed as well
as those affected by the item under review~"
Contrary to the above, the safety evaluation did not
review Section 5.7.4, "Seismic Analysis of CPCo Design
Class 1 Piping," and subsequently failed to note that
18
FSAR Section.5.7.4.1 and Figure 5.7-27 were affected
by this change to the facility.
This.was considered an example of an apparent
violation of 10 CFR 50, Appendix 8, Criterion V, in
that the safety evaluation ~~s not accomplished in.
accordance with the procedure (255/90025-038).
As a result of the above, the licensee committed to
make the appropriate changes to their FSAR.
(2)
FC-893 Steam Generator Slowdown System
As .part of the SGRP, the existing 2 inch bottom blowdown
piping was completely removed and replaced with 4 inch
piping. This modification started at the generator. nozzle,
continued thrriugh modified containment penetrations, and
ended at the first containment isolation valve in each li~e.*
The changes were part of a secondary side chemistry control
improvement program and were needed to increase the capacity
of the blowdown system.
The following documents associated with this facility change
were reviewed by the NRC inspector for compliance with NRC
requirements and licensee commitments:
(a)
Calculation No. SGRP-PDS-003, "Pipe Stress An~lysis of
Steam Generator E50A Slowdown Piping Inside
Containment," Revision 5, August 21, 1990.
Section 4.2, "SIF for Branch Connections," of the
calculation provides the bases for the stress
intensification factors (SIF) used. at three welded
branch connections in the pipe stress analysis .. The
calculation concluded that an SIF of 1.0 could be
assigned using the ANSI 831.1 Code equation for branch*
connections.
However, footnotes in the Code stated
that the equation was applicable only if certain
configurational conditions were met.
Contrary to the above, the restrictions for the inner
and outer radii of the branch connection as given in
the Code equation were not specified by any
installation or fabrication document;
This was considered an example of an *apparent
violation of 10 CFR 50, Appendix 8, Criterion III, in
that design control measures failed to assure that the
design basis were correctly translated into drawings
or instructions (255/90025-0lJ).
19
. '
This deficiency potentially applied to other branch
connections designed by Bechtel since this approach
was taken from generic communications applicable to
all projects. The internal memo which discussed the
Code equation did not indicate that the 4se of the
formula had any installation *restrictions ..
Revision 6 t6 the calculation was subsequently issued
on January 30, 1991, with the SIF based on formulas
given by the manufacturer of the branch fitting.
Using this formula, the SIF was increased to 1.367;
however, there was sufficient margin in the original
design to accomodate the 37% increase at the branch
connections without exceeding_ allowable stresses.
(b)
Drawing:No. MlOl-6010, "S.G. E50A Blowdown, Pipe
Support No. H9," Revision 3, Novembe~ 10, 1990.
This was a new support for the new blowdown system
which provided vertical restraint to the piping. It
was constructed fro~ tube steel members with kickers
in the vertical and horizontal direction.
Paragraph 5.4.19.1, "Weld Design; Codes and Symbols,"
of Bechtel's Design Criteria No. 20557-G-OOlP, stated
that welds for pipe supports shall be designed* in
accordance with the American Institute of Steel
Construction (AIS.C) Manual.
Part 4 of the AISC Manual
under "Prequalified Welded Joints" states that fillet
welds for skewed T-joints are limited to a.minimum
angle of 60° and that .for angles less than 60° .the
weld is considered a part_ia_l penetration groove weld.
Contrary to the above, Field Change Notice (FCN) No.
293, December 10, 1990, changed the attachment point
of the horizontal kicker which decreased the angle
between the kicker and baseplate to *approximately 49°
without changing a portion of the field weld from a
fillet weld to a partial penetration groove weld.
The
groove weld required different qualifications for the
.welder and had different limitations regarding base
metal thickness and effective throat measurements.
This was considered an example of an apparent
violation of 10 CFR 50, Appendix B,
C~iterion III in
that design verification activities did not assure
that field changes were subject to the same design
control measures as the original .design
(255/90025-0lK) .
20
(3)
FC-894 Steam Generator Recirculation System
- As part of the SGRP,
th~ existing 2-inch surface blowdown
piping was completely removed and replaced with 4 inch
recirculation piping. This modification started at the
generator nozzle, continued through modified containment
penetrations, and ended at the containment isolation valve
in each line. The changes were part of the secondary side
chemistry control improvement program.
The following documents associated with this facility ~hange
were reviewed by the NRC inspector for compliance with NRC
requirements and licensee commitments.
(a)
Cal~ulation No. SGRP-PDS-002~ "Pipe Stress Analysis of
Steam Generator E50B, Recirculation Piping Inside
Containment," Revision 8, January 30, .1991.
Paragraph 4.4.2.4.1, "Piping Analysis, Seismic Inertia
Loads," of Bechtel's Design Criteria No. 20557-G-OOlP
stateo that input response spectra will be developed
by enveloping the applicable response spectra for all
structures and elevations supporting the piping.
Contrary to the above, the enveloped response spectrum
did not consider accelerations at the steam generator
nozzle elevation of 661 feet but instead only used
spectral values from elevation 649 feet.
No basts or
justification was given for this discr~pancy.
This was considered an example of an apparent
violation of 10 CFR 50, Appendix B, Criterion III in
that design verification activities failed to assure
that appropriate design values were used in the stress
analysis (255/90025-0ll).
(b)
Calculation No. SGBR-PD-Hl4, "Pipe Support Design for
Steam nenerator E50B Recirculation System Support
Ml01...;6075-Hl4," Revision 2, January 30, 1991.
During as-built walkdowns, the NRC inspector noted
that this support had a significant shear cone overlap
condition between the lower bolts on the upper
baseplate and the upper bolts of the lower baseplate.
Review of the support analysis revealed that the shear
cone overlap had not been evaluated for this confirmed
calculation. This deficiency had been noted twti weeks
earlier by the CPCo technical reviewer during -the
augmented reviews of Bechtel calculations.
Paragraph 5.4.17.3.1, "Anchor Bolt Capacity Reduction
for Shear Cone Overlay," of Bechtel's Design Criteria
21
' *
J
No. 20557-G-OOlP referred to Tables 84 and 85 which
stated that if smaller spacing was used, the allowable
design capacity shall be reduced in proportion to the
ratio of the spacing provided to the required spacing. *
Contrary to the above, with an anchor bolt spacing of
5.3 inches and a required spacing of 7.5 inches for
- 3/4 inch Hilti bolts and 6.0 inches for 1/2 inch
Drillco bolts, the stress cone overlap was not
evaluat~d in Revision 2 of the calculation.
This was considered an example of an apparent
violation of 10 CFR 50, Appendix B, Criterion III in
_that design verification activities failed to assure
the adequacy of the anchor bolt design
(255/90025-0lM).
The above calculation was subsequently revised on
March li 1991, with an evaluation of the stress cone
overla~~ The methodology used in this evaluation was *
questioned by the NRC inspector because of its
unconventional approach.
The NRC inspector questioned
how concrete could be "reserved" for the Hilti anchor
bolts and not experience any.load from the. Drill co
bolts in this reserved area.
Bechtel's response to
the question stated that the analysis methodology
complied with the requirements of ACI 349, Appendix B .
The application of the refere~ced ACI 349, Appendix B
methodology was not apparent since the document did
not discuss the "reservation" approach used by
Bechte 1 . Sect i. on B. 4. 2 *of the above reference stated
that the effective design strength of the concrete
area is limited by overlapping stress cones. Although
the concrete design strength was conservatively
1
calculated for the Drill co bolt, it was not calculated
for the Hilti bolt even though the two stress cones
6verlapped significantly. The design strength area of
the Hilti bolt was not determined using the
overlapping stress cone method prescribed in the
reference nor was the capacity reduced in accordance
with the controlling design specification.
As discussed above, the evaluation of the anchor bolt
stress cone overlap was not performed in accordanc~
with the applicable design standards.
This was considered an example of an apparent
violation of 10 CFR 50, Appendix B, Criterion III in
that design verification activities failed to assure
that the analysis used an accepted design methodology
(255/90025-0lN) .
22
\\
I
b .
Consumer Power Engineering Efforts
Modifications to piping systems which were not directly associated
with the SGRP were designed and analyzed by CPCo personnel. These
efforts were controlled by design criteria given in CPCo .
Specification M-195, "Requirements for the Design ~nd Analysis *Of
Palisades Plant Safety Related Piping and Instrument Tubing."
The following modifications were reviewed by the NRC inspector:
(1)
SC-90-083 Auxiliary Feedwater Turbine Replacement
This modification upgraded the turbi~e .driver for the P-8B
auxiliary feedwater pump.
The turbine casing pressure
rating was increased from 250 psig to 675 psig.
The NRC
inspector reviewed the following calculation for compliance
with_NRC requirements and licensee commitment5:
Calculation EA-SC-90-083-01, "Change K-8 Turbine to Clais II
(675 psi/6S0°F)," Revision 2, November 27, 1990. This
latest re¥ision deleted proposed changes to pipe support
EB13-H924A and changed the type of reducer on the steam
inlet pipe.
Because of dimensional differences in the
attachment flange, the 4x6 inch reducer was changed from a
concentric to an eccentric reducer.-
Paragraph 6.4.2.b, "Detailed Technical Review," of Palisades
Administrative Procedure No. 9.11, "Engineering Analyses,"
Revision 4, December 28,
1989~ stated that detailed reviews
shall verify the accuracy, completeness and adequacy of the.
- engineering analysis.
Contrary to the above, the detailed technical review
performed for this calculation on November 27, 1990, did not
consider the effect of the additional moments caused by the
offset of the eccentric reducer nor the effect on the-SIF
for a .component which was not defined in the piping design
Code.
This was considered an example of an ~pparent violatio~ of
10 CFR 50, Appendix B, Criterion III in that the effects of
the eccentric reducer were not adequately considered d~ring
the design verification activities (255/90025-010).
Subsequent reviews by the licensee, documented in* a CPCo
Internal Memorandum from R. Jenkins to D. Bixel,
December 12, 1990, stated that the reducer change-out was
acceptable from a stress analysis perspective.
Based on
this, the significance of this deficiency is minimized .
23
. I.*
'
I
t
(2)
SC-90-032 Replacement of Valves tV-1103 and CV-1104
This modification replaced the two existing c6ntainment sump
drain isolation valves, CV-1103 and CV-1104.
The new 4 inch
valves we~e heavier than the existing valves and the piping
stress analysis was rerun to evalu~te these effects. Piping
stresses were found to be acceptable, and as expected* some
support loads increased.
The NRC inspector reviewed the pipe support calculation
No. EA-03340-HC12-Hl, Revision 3, dated May 28, 1990, for
compl~ance with NRC requirements and licensee commit~ents.
This support was a double rod hanger with a riser clamp and
had a load increase of approximately 50%.
Paragraph 5.11.5, "Rbd Hanger," of CPCo Specification C-173,
"Technical Requirements for the Analysis and Design of
Safety Related Pipe Supports," Revision 1, stated that when
double rod hangers were used on a vertical riser, the *hariger
components and supporting structure were to be designed to
take the total design load on one side.
Contrary to the above, the analysis did not evaluate the
hanger components and structure with the total design load
on one* side.
Instead the design load was divided by two and
the cbmponents were evaluated for this lesser load .
This was considered an apparent violation of 10 CFR 50,
Appendix B, Criterion III, in that design verification
activities failed to assure that an appropriate design
methodology was used in the calculation (255/90025-0lP).
Subsequent reviews performed by the licensee concluded that
as a result of the load increase, the anchor bolts on the
support exceeded al1owable loads and had to be replaced.
This additional modification was performed prior to plant
startup.
c.
. CPCo Quality Assurance Audits
Because of the extensive engineering and construction efforts
associated with the SGRP, CPCo conducted multiple QA audits of
Bechtel's work.
These audits started prior to the beginning of
the design efforts and continued through the closeout of the
construction packages. A total of four technical audits were
performed by CPCo with audit teams consisting of technical experts
from CPCo and Bechtel.
Based on the types of issue§ disclosed during these efforts, the
NRC inspector concluded that the audits were extensive in nature
and performed by competent engineers.
The details discussed in
the audit reports showed an excellent trend_toward performance
24
based inspections.
Detailed reviews done by the technical
auditors were intended to verify the accuracy and adequacy of the
engineering calculations and drawings.
The NRC inspector*
considered this a positive approach by the licensee to assess the
effectiveness of design controls and design verification
activities.
However, because of the number of discrep~ncies found during this
inspection combined with the findings and observations documented
in all four of the licensee's audits, the NRC inspector questioned
the overall effectiveness of actions taken by the licensee to
improve the design controls.
In their response to the design
control violations cited in NRC Inspection Report 50-255/89024,
the licensee acknowledged that their previous modification program
had insufficient controls and was not implemented acceptably.
While the corrective actions taken by the license.e significantly
increased the controls within the modification program, there
contin~ed to be *problems associated with effective implementation.
of these controls.
The discouraging aspect of all of this was
that even though the licensee, through their audits, had
recognized the weak program implementation at Bechtel, apparently
- sufficient meaningful actions to correct these problems were not
taken.
In the first SGRP audit of Bechtel in December 1989, Observation
No. 1 was written for a failure to identify and correct two
calculational errors during the final checking process:
CPCo
requested that Bechtel evaluate the cir~umstances surrounding the
events which resulted in the problem and provide reasonable
assurance that this was an isolated case and not a programmatic
problem.
/
The second audit conducted in February 1990, documented six
findings and 23 observations. The audit report stated that, while
none of the findings or observations by themselves rendered any
work product completely unacceptable or useable, the number of
concerns raised by the audi.t required prompt Bechtel management
attention ..
In the third SGRP audit of Bechtel in July 1990, additional .
discrepancies were documented in which calculations were found to
be unclear, contained errors or failed to provide justification
for analytical assumptions.
The deficiency further noted that
although individually the items were not considered significant;
however, when taken collectively the conditions were considered a
failure of engineering to provide adequate attention to detail
during calculation preparation and checking.
-
In an*internal memorandum dated August 29, 1990, from CPCo's SGRP
Quality Assurance Manager to the SGRP Project Director, it was
noted that Bechtel's calculation packages appeared to suffer from
the same lack of clarity and completeness that contributed to the
25
I*
I
I
..
1179-14" situation.
He continued by stating' that the noted
concerns did not appear to have affected the technical adequacy of
the design itself. Later he stated that it appeared that
Bechtel's reviews had been ineffective, 'for whatever reason, since
the audits and SGRP reviews continued to disclose concerns.
Further, it was stated that since most of the engineering design
work was nearly completed, .it did not seem beneficial to him to
request that Bechtel institute any new reviews at that time.
Instead he recommended a final audit be performed at the end of
the *project.
The fourth SGRP audit was subsequently conducted from February 1
to February 21, 1991.
As documented in the March 25, 1991, audit
report, one finding was cited by the audit team. The finding
stated that contrary to Bechtel's quality assurance commitments,
calculations did not have adequate detail to permit a technical
review without recourse to the originator. Over 100 comments,
questions or concerns were documented in the audit team finding.
In the transmittal letter, the licensee stated that the question
of completeness and clarity of design documents was a repeat issue
from earlier audits; however, the technical adequacy of the design
product did not appear to be in question.
Paragraph 4.0 of CPCo's audit report stated:
11The finding addressed a significant number of problem~ and
questions involving most*of the calculations reviewed.
Discussion
with the Techni~al Specialists and Bechtel Engineering has
revealed that individ~ally or collectively, the content of the
finding does not jeopardize the adequacy or integrity of the
design basis. Concern is nonetheless expressed over the number ofr
problems, regardless of the degree of severity.
As the problems.
cited represent a sample, the question remains regarding the
status or quality of th~ balance of the calculations *we did not
examine..
S~RP Engineering routinely performs a technical review
similar to that done by the Technical Specialists on this audit
and have, in fact, discovered many of the same types of.problems
and discrepancies as were uncovered during the audit. Through
their comments, Bechtel has provided appropriate explanations and
corrections.
To this extent, the Audit Team is ~atisfied that the
final calculations will be of sufficient accuracy and detail and
will not request that Bechtel provide corrective action to prevent
recurrence for the finding."
Based on the above.discussions, it is apparent that Bechtel's
design control program did not produce the quality of documents
expected by CPCo or committed to in Bechtel's Quality* Assurance
Manual.
This problem was detected early in the project by audits
performed by the licensee and apparently continued throughout the
entire SGRP.
The deta i 1 ed reviews performed by the 1 i censee' s
SGRP engineering group identified and corrected the majority of
26
- .
'l
J
.
6.
the deficiencies in Bechtel's calculations. These additional
reviews were needed because Bechtel's internal reviews failed to
identify discrepancies in the calculations.
The issue that these discrepancies were simply clarity and
crimpleteness problems is somewhat misleading.
If an error was
noted in a completed, reviewed Bechtel calculation and that error
did not result in a hardware problem, then this could be
considered simply as a documentation problem.
Even though this
would indicate that the design control process was not functional,
the calculation could be revised and no other corrective actions
need to be taken based on lack of significance .. If this were an
infrequent occurrence, some justification could potentially be
given; however, since this appeared to be a prevalent occurrence
in the Bechtel analyses, a programmatic breakdown in the design
corttrol process was indicated.
- Con~u~ers Power Comp~ny Quality Assurance Program CPC-2A,
Section 16 stated that conditions adverse to quality of safety
related activities shall .be promptly identified and corrected~
Contrary to the above, corrective action measures, while promptly
identifying conditions adverse to quality in Bechtel's design
control implementation, failed to correct these conditions. This
is considered an example of an apparent violation of 10 CFR 50,
Appendix B, Criterion XVI (255/90025-02B).
The lack of sufficient corrective actions on the part of CPCo, to
correct the programmatic problems with Bechtel's design control,
is considered a significant lack of corporate management
involvement in assuring quality work at Palisades.
Testing Activities
a.
- Inservice Testing Inspection (73756, 92700)
- This portion of the inspection included selected sections' of the
inseryice testing of pumps and valves and activities related to.
such testing.
The NRC inspectors observed the.applicatioh of the
Valve Operation and Test Evaluation System (VOTES) to a valve
l oc-ated inside containment.
During the calibration portion of the
test, the operators detected anomalous information in the test *
data and concluded that the system could not be properly
calibrated without reinstalling a strain gage on the valve yoke.
Although the testing was not completed, the inspectors were
convinced that the operators were familiar with the equipment and
that they exercised good judgement in ter~inating a test when the
results revealed small, but reproducible anomalies.
The NRC inspectors also reviewed the details of an incident
regarding spray pump operability which occurred in February 1990.
One of three such pumps fell into the required action range of the
Inservice Testing (IST) program vibration criteria. The
27
~J. ( **.)' .) *
'
-~.
. .
- inspectors determined that the licensee had established the root
cause of the problem and had provided suitable corrective action .
The corrective action cited in the licensee's Event Report E-PAL-
90-0034H is considered to be capable of ~reventing a repetition of
this problem.
b.
Review of Structural Integrity Test (37700)
The structural tntegrity test (SIT) was considered a post-
modi fication test-for modification FC-914.
Its purpose was to
show that containment structural integrity was not impacted by the
SGRP a~tivities.
The NRC inspectors reviewed the licensee's procedure 20557-SIT
"Primary Reactor Containment Structural Integrity Test Procedure,"
Revision 3, as prepared by Bechtel Power Corporation.
The
inspectors had no comments on the procedure.
The NRC inspectors witnessed the containment pressurization,
including licensee hold points, and inspected the full pressure
crack-mapping activities. The inspectors reviewed the
displacement gauge data, both during the pressurization and hold
periods, and during the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> partial depressurization hold
period.
One gauge exceeded the expected maximum displacement
value.
The NRC inspectors discussed the behavior of the gauge
with the licensee during the test. The licensee'_s prelim.inary
conclusion ftir the increased displacement value was that the
containment was slightly ovoid, rather than perfectly round, and
the increase~ pressure was causing the structure to "round out.
11
This conclusion was substantiated by the behavior of the other
three gauges in the area.
No other problems were identified with
these activities.
The NRC inspectors reviewed a prel imina-ry report of the fi.nal
results.
The report concluded that the gauge discussed above was
acceptable, as the acceptance criteria was based on the average
displacement values being below the calculated maximum.
This
criterion was ~et. The inspectors had no problems with this
conclusion.
c.
Tendon* Surveillance Requirements for the Replaced Tendon (37700)
During inspection docume~ted in NRC Inspection .Report
50-255/90017, and subsequent conversations with both licensee and
Bechtel SGRP personnel, the possibility of Palisades using new
replacement tendons was discussed. Specifically, it was_ noted
that Palisades technical specifications no longer contained the
requirement for demonstrating containment tendon structural
28
L
I
t
,)
- ~ ~*
'*.
d.
integrity at the end of one, three, and five year intervals
following the structural integrity test, and every five years
thereafter. This subject was left for discussion in the event
that any tendons were actually replaced during the SGRP.
During
the course of the SGRP~ one tendon was damaged during the cutting
of the containment opening and had to be *replaced with.a new
cable.
In a letter from Consumers Power dated February 12, 1991, the
licensee acknowledged NRC questions in regard to tendon
inspections~ The letter stated that one vertical and two hoop
tendons from the SGRP opening, as well as one tendon which had
experienced low lift-off readings, would be examined during the
scheduled 1992 surveillance. The NRC inspectors questioned
whether the replaced tendon was one of those to be included in the
surveillance program.
During further discussions, the licensee
confirmed that the new tendon was not included in the 1992
surveillance.
It is the NRC's position that the replaced tendon should be.
included in the 1992 scheduled surveillance. This is based on the
NRC technical staff and Bechtel experience which indicates that
the majority of a new tendon's detensioning occurs during the
first year or two.
Therefore, an inspection in 1992 would show if
the replaced tendon had detensioned.
The NRC staff i~ not .
requiring that the on-going surveillance program be adjusted for a
single tendon, but does consider inclusion of this tendon in the .
1992 surveillance to be prudent.
Inclusion of this tendon into
the 1992 planned surveillance is considered an Open Item
(255/90025-06).
The NRC, in a letter from B. Holian, Palisad~s Project Manager,
NRR to G. Slade, dated November 20, 1990, reserved the final
acceptability of the containment structure until a review of the
structural integrity test was performed.
This inspection
constituted the review specified in that letter.
No violations .or deviations were identified in this area.
Review of Facility Change FC-905, Safety Injection Tank Lower
- Level Alarm Switch Modification:
This modification altered the low level switches for the safety
injection tanks such that the switches would actuat~ at a lower
level. The change in actuation level was made to correspond to a
revised minimum l~vel incorporated into technical specification
Amendment 136. Originally, the modification set the new actuation
point at the revised technical specification minimum level.
However, due.to concerns regarding potential instrument drift of.
the level switches (see paragraph.2 of this report, Open Item 50-
255/90018-01)~ the new switch as~emblies were modified to actuate
two inches above the revised technical specification minimum level
29
\\:. ,, : "
thus accqunting for potential instrument drift .
The NRC inspector reviewed the de$ign package for this
modification, the associated licensing history, and the procedure
for calibration of the safety injection tank level transmitters.
The 10 CFR 50.59 evaluations, associated *safety analyses, and
design inputs for the modification were generally found to be well
justified and documented.
However, several discrepancies were
noted during the NRC inspector's ~~views of the* following
documents:
(1)
Test Procedure T-FC-905(a)-001, "SI Tank Low Level Float
Ass~mbly," Revision 0, July 24, 1990.
(2)
One of the purposes of the test was to verify proper
operation of the newly installed low-level float assemblies.
As part of this procedure, the SI tank levels were to be
gradually raised to elev~tion 737 feet and then clear~nce of
the low-level alarm was to be verified. However,
Engineering Design Change 03 revised the level switch
actuation point by 2 inches, but the test procedure was not.
revised to account for this change.
With the elevation 737
feet specified in the procedure, the level switches would
not have actuated and the low-level alarm would not have
- cleared as specified in the test .
10 CFR 50, Appendix B, Criterion III requires that design
control measures be applied to the delineation of test
acceptance criteria and that changes be subject to controls
commensurate with the original design.
Contrary to the above, when the ~loat design elevation was
changed, the test elevation acceptance criteria was not
.revised. This was considered an example of an apparent *
violation of 10 CFR 50, Appendix B, Criterion III
(255/90025-0lQ).
.
After the NRC inspector informed the licensee of this error,
the test procedure was promptly revised.
In this case the
significance was reduced since the error would have been
self disclosing during the test.
Calculation EA-FC-905-003, "Safety Injection Tanks (T-82)
- Level Calibration Calculations," Revision 2, November 30,
1990.
Calculation assumption No. 2 stated that the existing
atialysis was correct. This 1978 calculation determined the
initial relationship between the differential pressure and
the SI tank level. The calculation did not account for the
effects of the weight of the pressurized nitrogen inside the
30
tanks nor the containment temperature on the differential
pressure and level relationship. Based on this, the
accuracy of the calculation was in question and the validity
- of the assumption was suspect.
Pending the licensee's verification that the design basis
accuracy of the level. indication instrument loop had not
been exceeded by neglecting the effects of the nitrogen
weight and containment temperature, this was considered an
Unresolved Item (255/90025-07).
The significance of this issue was minimized since the
technical specification compliance was based on the level
switch actuation and not on the level indication from the
transmitter.
e.
Review of Facility Change FC-852, Addition of Second PCS Level.
f.
Indication:
This modification added a second means of reactor coolant. system
level indication. The modification provided 30 inches of level
indication centered on the hot legs for use during mid-loop
operation.
The modification added a differential pressure
transmitter which tapped into existing instrument lines. The
second means of level indication was added in response to Generic
Letter 88-17.
The NRC inspector reviewed the safety analysis, design inputs, and
installation. The 10 CFR 50.59 evaluations, associated safety
analysis, and design in~uts for the modification were found to be
well justified and documented.
No concerns were identified.
.
.
. Review of Procedure RT-70F, Primary Cool~nt System:
Procedure RT-70F was written as a technical specifications
surveillance procedure to verify leak tightness of the primary
coolant lines .. The NRC
in~pector reviewed the procedure for
adequacy as part of the post-modification testing for several
modifications.
The procedure satisfied the American Society of
Mechanical Engineers (ASME)Section XI Code requirements for a
hydrostatic test and an inservice system leak test.
No concerns
were identified .
31
- .t- *. \\'
....
.....
- ..
"'
,,.
I,
7.
Unresolved Items
An unresolved item is a matter about which more information is required
in order to ascertain whether it is an acceptable item, an open item, a
deviation, or a violation. The unresolved items disclosed during this
. inspection are ~iscussed in Paragraphs 2.b, 2.c, and 6.d of this report.
8.
Open I terns
Open items.are matters which have been discussed with the licensee,
which will be reviewed further by the inspector, and which involves some
action on the part of the NRC or licensee or both .. The open item
disclosed during this inspection is discussed in Paragraph 6.c of this
report.
9.
Exit Interview
The Region III inspectors met with the licensee representatives (denoted
in Paragraph 1) at the conclusion of the inspection on April 18, 1991 as
well as periodically during the course of the inspection.
The
inspectors summarized the purpose and findings of the inspection.
The
licensee representatives acknowledged this information .. The inspector
also discussed the likely informational content of the inspection ~eport
with regard to documents or processes reviewed during th~ inspection~
The licensee representatives did not identify any such
documents/processes as proprietary .
32