Part 21 Report - 1996-900
ACCESSION #: 9607260276
Westinghouse
Energy NUCLEAR SAFETY ADVISORY LETTER
Systems
Business
ESBU Unit
THIS IS A NOTIFICATION OF A RECENTLY IDENTIFIED POTENTIAL SAFETY ISSUE
PERTAINING TO BASIC COMPONENTS SUPPLIED BY WESTINGHOUSE. THIS
INFORMATION IS BEING PROVIDED TO YOU SO THAT A REVIEW OF THIS ISSUE
CAN
BE CONDUCTED BY YOU TO DETERMINE IF ANY ACTION IS REQUIRED.
P.O. Box 355, Pittsburgh, PA 15230-0355
Subject: Reactor Coolant Pump (RCP) Support Column Tilt Issue
Number: NSAL-94-025
Basic Component: RCP Supports Date: Nov. 10, 1994
Plants: See Page 2, Table 1
Substantial Safety Hazard or Failure to Comply Pursuant to 10 CFR
21.21(a) Yes [] No []
Transfer of Information Pursuant to 10 CFR 21.21(b) Yes []
Advisory Information Pursuant to 10 CFR 21.21(c)(2) Yes []
Reference:
SUMMARY
The issue concerns a situation whereby certain reactor coolant pump (RCP)
support columns may be titled beyond design parameters. Specifically,
some three and four loop plants have a loop piping layout in which the
crossover leg piping interferes with the front inside support column of
the RCP. The other two columns on the RCP are oriented to be vertical in
the normal full power condition. This condition has been recognized by
the support designer and changes have been made to the column layout to
accommodate the interference between the column and the crossover leg.
The change consists of moving the base of the one RCP column toward the
reactor pressure vessel approximately 6-12 inches so that titled column
no longer interferes with the crossover leg piping. The required tilt is
from approximately 2 to 5 degrees depending on the column length and
amount of movement. This condition was apparently reconciled during the
design phase and was not considered a significant issue. Upon further
review it has been determined that while the actual change in the value
of the vertical stiffness is small, the potential impact on the thermal
expansion of the system, and thus the loop loadings, may be significant.
Based upon the analyses performed, it has been determined that while some
of the affected plants will have completely acceptable results from their
specific RCP column tilt, other plants may have difficulty in meeting
certain margin values for such specific evaluations as Leak-Before-Break.
While such a situation may technically place a given plant outside its
licensing basis, it has been determined that adequate margins remain such
that this situation does not pose a challenge to the reactor coolant
system (RCS) pressure boundary and therefore, does not represent a
substantial safety hazard or failure to comply per the definitions
provided in 10 CPR Part 21.21 (a).
Additional information, if required, may be obtained from the originator.
Telephone 412-374-5036.
Originator(s):
J. S. Galembush H. A. Sepp, Manager
Strategic Licensing Issues
NSAL/94-025 Sheet 1 of 6
TABLE 1 PLANT APPLICABILITY LIST
Based upon design information held by Westinghouse, the following plants
are affected by the RCP Column Tilt issue:
J. M. Farley 1 & 2
V. C. Summer
Shearon Harris
Vogtle 1 & 2
Seabrook
Wolf Creek
Callaway
Comanche Peak 1 & 2
Sequoyah 1 & 2
Watts Bar 1 & 2
Sizewell B
Kori 3 & 4
Yonggwong 1 & 2
Ohi 1 & 2
Takahama 1
Maansham 1 & 2
NSAL/94-025 Sheet 2 of 6
TECHNICAL DESCRIPTION
ISSUE DESCRIPTION
As described in the "Summary" section, the issue concerns a situation
whereby certain reactor coolant pump (RCP) support columns may be tilted
beyond design parameters. Specifically, some three and four loop plants
have a loop piping layout in which the crossover leg piping interferes
with the front inside support column of the RCP. The other two columns
on be RCP are oriented to be vertical in the normal full power condition.
This condition has been recognized by the support designer and changes
have been made to the column layout to accommodate the interference
between the column and the crossover leg. The change consists of moving
the base of the one RCP column toward the reactor pressure vessel
approximately 6-12 inches so that the tilted column no longer interferes
with the crossover leg piping. The required tilt is from approximately 2
to 5 degrees depending on the column length and the amount of movement.
This condition was apparently reconciled in an informal manner during the
design phase and was not considered a significant issue. Upon further
review it has been determined that while the actual change in the value
of the vertical stiffness is small, the potential impact on the thermal
expansion of the system, and thus the loop loading&, may be significant.
Based upon the analyses performed for this evaluation, it has been
determined that while some of the affected plants will have completely
acceptable results with their specific RCP column tilt, other plants may
have difficulty in meeting certain margin values for such specific
evaluations as Leak-Before-Break (LBB). While such a situation may
technically place a given plant outside its licensing basis, it has been
determined that adequate margins remain, such that this situation does
not pose a challenge to the reactor coolant system (RCS) pressure
boundary.
TECHNICAL EVALUATION
Analyses have been performed to evaluate the impact of the RCP column
tilt on the plant with the most severe tilt. A plant with a column tilt
associated with a 12 inch movement of the column base from the front
inside pump column was chosen for this evaluation.
The additional loadings associated with column tilt originate from the
use of a support stiffness that has been rotated from vertical by a few
degrees. This routed stiffness at one of the three column locations acts
to restrain and rotate the RCP. For a loop piping system thermal
expansion of approximately 1.9 inches, the vertical displacement
associated with the rotation of an RCP column will vary from about 10
mils for a nearly vertical column to about 130 mils for an upper bound
tilted column. This difference in displacements at the three RCP columns
is what causes a rotation in the pump not previously accounted for. This
source of additional loading varies from plant to plant based on the
primary equipment support design. There are plants without pin-ended
columns that will not have this type of additional loading. That are a
number of plants that have support details not of Westinghouse design.
Many of the Westinghouse designed supports have a tilt that is
approximately half of the that analyzed as a part of this evaluation.
When this enhancement was made to the thermal analysis, differed system
loadings resulted. The bending moment at the RCP outlet nozzle for the
test case increased by more than 100%. The RCP column loads also
changed. The tilted column went from a compression load to a tension
load. The change in column loading met applicable Code allowable limits.
The loop LBB evaluation was of particular concern for system load
changes. The LBB evaluation is preformed with the goal of achieving
certain margins required by the
NSAL/94-025 Sheet 3 of 6
NRC. The new loadings were reviewed for all 12 weld locations in the
primary loop for the test plant, and acceptable margins were maintained.
The test case loop evaluation for column tilt loads was successful. The
column tilt for the test plant is approximately twice (12 inches versus 6
or 6.5 inches) that of most other plants. Unfortunately, it has been
determined that due to the overwhelming plant specific nature of both RCS
loops and the RCP column design, the test case configuration is not
generic nor necessarily enveloping, but it is representative of a typical
Westinghouse configuration. Every one of the plants affected by the RCP
column tilt issue has a plant-specific loop analysis that has a unique
set of margins. Because loadings local to the RCP can be quite different
due to the inclusion of column tilt, those plants that have small margins
in this area may be unable to meet NRC mandated LBB margins. Those
plants with small LBB margins in the cold leg may require additional
analysis. As mentioned, there are conservative NRC mandated margins that
are part of the licensing associated with LBB. Even if the licensing
margins we not satisfied, one can argue that the LBB conclusions are
still valid with smaller margins and that the LBB design basis continues
to apply.
A second test plant was reviewed, which has an LBB critical point
(location, of lowest margin) at the RCP outlet nozzle weld. This test
plant also has a significant column tilt (approximately 10 inches out-of-
plumb). Using an estimated increase in bending moment, based on the
first test case results, the critical location for the second case does
not meet NRC required LBB margins. This evaluation is preliminary
because the increase in loading was estimated. This second cue is
unusual because the critical location is at the RCP outlet nozzle. It is
possible that more refined modeling and analysis could be performed that
would change the loading and the location of the critical weld for the
second test case plant.
The inclusion of RCP column tilt has been shown to change loop loadings
by large percentages in the area of the RCP. A review of the increased
loadings on our first test plant yielded acceptable results in all areas
reviewed. For the first test case evaluation, there were significant
margins available in the areas where loadings increased. Good margins in
the RCP area are fairly typical for Westinghouse plants, but there may be
plants where margins are slim. If sped& modeling techniques are the
cause of low margins, additional analysis may reduce some of the,
potential conservatism that led to the low margins. For those plants
with low margin around the RCP, structural analysis may need to be
revisited to assure compliance with the licensing basis, but as has
already been mentioned, there is not a pressure boundary integrity
concern due to the RCP column tilt issue.
ASSESSMENT OF SAFETY SIGNIFICANCE
From information available to date, plants that incorporate an RCP column
tilt in their design may be operating in an unanalyzed condition in that
the actual tilt at hot conditions may be greater than the tilt called out
in the design drawings thereby potentially affecting the structural
reliability of the RCP supports. More importantly, regardless of my tilt
condition described in the design drawings, there may be no analysis
available to substantiate the acceptability of such an "as built" tilt in
any affected plant design. At hot conditions the plant thermal analysis
may be impacted in the following areas:
- Primary equipment support evaluation
- Primary piping fatigue evaluation
- Primary piping leak-before-break (LBB) evaluation
- Auxiliary line thermal anchor movement analysis
The above list is tied to any change to the plant thermal analysis. The
seismic and LOCA analyses are expected to be less significantly impacted
by a tilted column stiffness, but must also be included in the above
NSAL/94-025 Sheet 4 of 6
to determine exactly what magnitude of column tilt is indeed significant.
It is noteworthy that the RCP column design is determined by the seismic
and LOCA loads, therefore, even though the thermal portion of the load is
changing, the total load the column may be experiencing represents a much
smaller percent change.
The various loadings performed in a system evaluation have been reviewed
and qualitatively assessed for the RCP column tilt situation. Due to the
small angle of tilt associated with the RCP column tilt issue, the
loadings for deadweight, OBE seismic, SSE seismic, and LOCA will change
very little from the design basis analysis numbers that assumed no column
tilt. This judgement is based on the fact that a stiffness matrix
rotated by a few degrees is expected to offer the same kind of restraint
to a deadweight or a dynamic loading as the original stiffness. The
deadweight loading will load the column with an additional small shear
load because there is no lateral movement. The seismic and LOCA loading
conditions will involve some small lateral movement but the additional
lateral restraint is likely to generate a somewhat smaller seismic
response and potentially, slightly higher LOCA loadings.
The thermal condition represents the loading condition most likely to be
impacted by the column tilt. Because a normal plant heat-up and cooldown
moves an RCP by more than 1.5 inches in a lateral direction, there is a
more significant restraining load generated by the rotated stiffness as
well as a heretofore unaccounted for rotation of the RCP centerline due
to the 1.5 inch deflection. Both of these effects will likely increase
the column loadings and the adjacent piping and pump nozzle loadings.
The thermal loadings factor into the ASME Section III Code fatigue
analysis and into the leak-before-break (LBB) evaluation. The fatigue
analysis has a stress limit for which there is usually good margin and a
usage factor which, for the crossover and cold leg, there is good margin.
The LBB evaluation is performed to satisfy certain NRC mandated margins.
Even though the mandated margins may not be satisfied, it is the
Westinghouse position that the LBB conclusions will still be valid with
smaller margin such that the LBB design basis continues to apply.
However, such a situation may technically place a given plant outside its
licensing basis.
Primary equipment nozzle leads for the various loading conditions are
compared to a set of allowable nozzle loads. This comparison is made for
both a load case, such as thermal, as well as for a load combination like
normal or upset. There is typically sufficient margin in the nozzles to
accommodate a large increase in the thermal case loading. There is also
margin between the Code limit and pressure boundary failure. There are
not expected to be load magnitudes large enough to challenge pressure
boundary or shutdown requirements.
In summary, the effects of RCP support column tilt have the potential to
change the loadings on the loop piping, the primary equipment nozzles,
and the primary equipment supports and embedments. Even with the loading
changes it is judged that pressure boundary integrity will be maintained
thus allowing continued safe plant operation.
NRC AWARENESS/REPORTING CONSIDERATIONS
Westinghouse, submitted to the NRC an Interim Report of an Evaluation of
a Deviation or Failure to Comply Pursuant to 10CFR21.21(a)(2) on
September 19, 1994. This action was required since the Westinghouse
evaluation of this issue would exceed the initial 60 day evaluation
period allowed in 10 CFR part 21. Based upon our analyses and
evaluation, this issue does not represent a substantial safety hazard or
failure to comply pursuant to 10 CFR 21.21(a).
NSAL/94-025 Sheet 5 of 6
RECOMMENDED ACTIONS
While it ha been determined that a substantial safety hazard does not
exist, it is prudent that utilities review their plant specific drawings
to verify the status of the RCP columns.
NSAL/94-025 Sheet 6 of 6
*** END OF DOCUMENT ***
Page Last Reviewed/Updated Wednesday, March 24, 2021