ML20126B457
| ML20126B457 | |
| Person / Time | |
|---|---|
| Site: | Limerick  |
| Issue date: | 06/05/1985 |
| From: | Collins S, Gallo R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20126B428 | List: |
| References | |
| 50-352-84-48, NUDOCS 8506140108 | |
| Download: ML20126B457 (24) | |
See also: IR 05000352/1984048
Text
(
,
4 -
.
U.S. NUCLEAR REGULATORY COMMISSION
REGION I
Report No.
50-352/84-48
Docket No.
50-352
License No.
CPPR-106
Category
B
Licensee:
Philadelphia Electric Company
2301 Market Street
Philadelphia, PA
19101
Facility Name:
Limerick Generating Station Unit 1
Inspection At: Bechtel Corporation Offices
San Francisco, California
Inspection Conducted: August 27-30, 1984
NRC Personnel:
S. K. Chaudhary, Senior Resident Inspector, Limerick
S. Kucharski, Reactor Engineer
P. D. Milano, Vendor Inspector, IE
C. P. Tan, Structural Engineer, NRR
J. Rajan, Mechanical Engineer, NRR
Reviewed By:
Mk
$
BE
R. M. Gallo, Chief, Reactor Projects
dat'e
Section 2A, Team Leader
Approved by:
(MULkh///tzd
6/5I85
~
S./J. Collins ( Chibf, Projects Branch
date
No. 2
Inspection Summary:
A special announced inspection by two Region-Based Inspectors, two NRR Techni-
cal Reviewers, one IE Vendor Program Branch Inspector and one Region-Based
Supervisor of allegations related to structural and piping design activities
performed by Bechtel Engineering for the Limerick Generating Station.
The
inspection involved 180 hours0.00208 days <br />0.05 hours <br />2.97619e-4 weeks <br />6.849e-5 months <br /> at the Pechtel offices in San Francisco and 20
hours onsite by the NRC inspectors and reviewers.
Results:
Of the two major areas inspected, one issue regarding the Limerick
Feedwater Check Valve Slam Analysis was identified which was not relevant to
the specific allegations. .(Para. 4.5).
Part of one allegation was substan-
tiated, but no safety concern was identified. (Para. 4.4)
$
DR
2
0
pg
n
-
-
,
.
.
<
DETAILS
1.
Persons Contacted
J. S. Kemper, Vice President, Engineering & Research
J. Corcoran, Head, Field QA Branch
H. R. Walters, Resident Project Manager, San Francisco
G. Szonntagh, Engineer
Bechtel Power Corporation
H.'Hollinghause, Manager of Engineering
R. E. Jagels, Chief, Mechanical Engineer
G. Ashley< Mechanical Analysis Group Engineer
LC. Soppet, Project Manager
R. Schlueter, Assistant Project Engineer
~ A. Wong, Group Leader, Civil Engineering
G. Duncan, Nuclear Group Leader
H. Safwat, Mechanical Analysis Group Supervisor
E. R. Nelson, Manager, QA Division
In addition to the above, the inspectors interviewed and held discussions
with many more members of engineering and management staff of PECO and
Bechtel during the course of inspection.
2.
Background
On June 28, 1984, NRC Region V office in Walnut Creek, California, re-
ceived an allegation that structural design problems,- involving blast
~
loads of the Reactor Building south stack and strength calculations for
the 201 and 217 Reactor Building elevations, exist at -Limerick.
Respon-
sibility for follow-up on this allegation was transferred to Region I on
June 29, 1984.
On July 19, 1984, another allegation was received by Region. V that the
pipe break forcing functions per the RELAP Code used in analyses for
4
Limerick, Susquehanna, and Hope Creek were inadequate and did not conform
to . FSAR commitments.
Responsibility for follow-up on this allegation was
transferred to Region I on July 20, 1984.
Both allegers were contacted by Region I staff for additional information.
On July 18, 1984, NRC representatives met with the first alleger at' the
Region V office. On July 23, 1984, Region I representatives spoke, via
telephone, with the second alleger.
Based on the preliminary information
obtained,
Region I conducted a one-week inspection at the Bechtel
San Francisco Office with assistance from NRR and IE Vendor Inspection
Branch.
(Bechtel is the Architect / Engineer (A/E) for Limerick, Susquehanna
and Hope Creek).
- - - - - . . - . . , ._
. . - - . - . . - - - - - -
r
.
.
3
The inspection included a review of (1) the structural design applicable
to the Limerick Reactor Building and (2) the pipe break analyses used at
Limerick with sampling inspection done on Susquehanna and Hope Creek in
order to address all the concerns of the second alleger.
3.0 Structural Design Allegations
3.1 Allegation - South vent stack was not designed to resist blast loads.
The first alleger's primary concern, as understood by Region
I, was:
"The Reactor Building south stack was not designed to resist blast load
due to sudden air compression resulting from a nearby railroad accident.
Also, the design calculations were in error."
3.1.1
Scope of Inspection
The inspection was directed to ascertain the technical validity of
the allegation, to determine if designers had engaged in any improper
or inadequate design process, and to assess, if the allegation were
valid, the impact of this error on the safety of the plant operations
and the safety margin in the plant as a whole. The effort to pursue
the inspection objective consisted of the following:
visual inspection of the south stack for its relationship to
--
other plant structures, and its "as-built" geometry.
review of engineering
calculations
to determine technical
--
adequacy and validity of design approach and basic assumptions.
review of technical specifications, design and construction
--
drawings, and referenced codes and standards.
review of construction and quality control
procedures for
--
installation.
discussion with cognizant engineering and management personnel.
--
3.1.2
References
LGS South Stack Calculations:
-
t
..
.
-
4
No.
Title
Date
Revision
21.7
South Stack Truss
11/2/78
Original
8/30/84
8-
21.7.2
NRC Inquiry on Reactor
8/25/84
0
Building South Stack Due
to Blast Load (68 sheets)
21.5
Precast Panel Design -
8/22/80
0
Reactor. Building (175 sheets)
Field Changes and Supporting Calculations:
--
No.
Title
Date
C-9668F
Reactor Building #2 South Stack
7/28/82
Truss; Calc. #21.7, Revision 6,
Sheet 34-1
C-9505F
Platforms Elev. 332'- 2 3/8" through
5/13/82
378'- 9", S'outh Stack; Calc. #21.7,
Revision 6, Sheet 34-7
C-1059F
Reactor Building #2 South Stack
8/31/83
Truss; Calc. #21.7, Revision 6,
Sheets 34-8 to 34-10
Bechtel Specification 8031-A-1, " Specification for Furnishing,
--
Fabrication and Delivery of Precast Concrete."
Bechtel Specification 8031-G-41, " Specification for Furnishing,
--
Detailing, Fabrication and Delivery of Structural Steel for the
Reactor Building and Control Complex Super Structure and Rad
Waste Building."
-Bechtel Meeting Notes; Document Control No.
182634, dated
--
5/16/84.
Limerick Generating Station " Fire Protection Evaluation Report"
--
(FPER).
Bechtel Drawings:
--
C-484; C-660, Rev. 16
C-667, Rev. 10; C-791, Rev. 4
C-795, Rev. 15; C-845, Rev. 14
C-847, Rev. 12;
-QAD-108, Rev. 16; QAD-109, Rev. 10
QAD-110, Rev. 9, QAD-111, Rev. 9
r;
.
...
5
3.1.3
Conduct of Inspection
a.
Visual Inspection of the South Stack
The Limerick Senior Resident Inspector visually examined the
south stack for any obvious defects in material, construction,
or workmanship. The "as-built" configuration was also compared
with design .and construction drawings.
The orientation and
geometry of the stack was reviewed to ascertain its exposure to
the postulated blast load from a nearby railroad car accident.
This inspection was carried out at the Limerick site.
b.
Review of Engineering Calculations
The review of calculations and other pertinent documents was
carried out at Bechtel's
Home Office
in
San
Francisco,
The inspectors reviewed the calculations and held
discussions with cognizant engineering personnel to assess the
validity and technical adequacy of calculations, design assump-
tions, and approach. The original design calculations for the
structural steel framing for the south stack were performed in
early 1978. The calculations were further upgraded and expanded
to include precast concrete panels in 1980. The complete design
of the south stack was also subjected to an in depth interdis-
,
ciplinary group review in May 1984.
This review is documented
in Bechtel meeting notes (Document Control Nc. 182634) of May
16, 1984.
The inspectors also reviewed additional calculations for the
south stack that were performed by Bechtel, in response to this
NRC inquiry, to verify the validity of the original calculations.
This effort was completed by the design engineers at the time of
inspection, but all documents were not formally approved.
The
approval of all additional calculations was accomplished'during
the inspection period.
c.
Review of Technical Specifications, Design and Construction
Drawings, and Referenced Codes and Standards
The inspectors reviewed the applicable design and construction
specifications, and other documented requirements pertinent to
the design and construction of the south stack.
The specifica-
tions were reviewed to determine if they contained adequate
technical requirements, evoked pertinent codes and standards
directly or by reference, and were sufficiently detailed and
free of ambiguities to convey the technical requirements.
The
drawings were examined to determine if they contained sufficient
information and adequate details to permit acceptable construc-
tion / installation and inspection.
The referenced codes and
industry standards were reviewed to assess pertinence and ap-
plicability to the functional objectives of the construction /
installation of the south stack.
_
.
.
6
3.1.4
Findings
Based on the above reviews of documentation, and discussions with the
licensee and A/E, the inspectors determined the following:
The south stack is not a safety-related structure as defined in
--
the Limerick-Project Q-List.
The south stack is analyzed and designed as a seismic category
--
IIA structure as defined in Section 3.2.1 of the LGS-FSAR.
Because the south stack is not Q-listed and is designated
--
seismic category IIA, it is not required to withstand blast
overpressure load.
The scuth stack is not required for post-accident monitoring
--
because HVAC exhaust to this stack containing accident effluents
is automatically isolated (LGS-FSAR, Section 11.5.2.2.2).
On the basis of the above findings, the inspectors examined the
technical validity and adequacy of the south stack design-basis and
assumptions.
The south stack is located on the south side of the reactor enclosure
building between column lines 21.5 and 24.5 (east-west), and C and D
(north-south).
The stack is designed to meet the requirements of
seismic category IIA (FSAR Sec. 3.2.1).
The stack is a tall tubular
passage created by structural steel framing enclosed by precast
concrete panels attached to the structural members by high strength
bolted connections. The structural framing itself is attached to the
reactor enclosure south wall through welded connections to steel
embedments in the wall. The principal code used in the design of the
structure is AISC for framing and connections and ACI for precast
concrete panels.
The precast panels are doubly reinforced on both
faces, and are six and one-half inches (6h") thick.
Because the stack is not safety related, and its failure in a seismic
event, in itself, will not jeopardize the safe shut-down of the
plant, it is not required to withstand other than normal structural
loads. However, due to the adjacent location of the diesel generator
building that is safety related, the stack has been analyzed and
designed as a seismic category IIA structure.
3.1.5
Conclusions
The stack is not required to withstand the blast overpressure load
because its failure due to blast overpressure will not affect the
safe shut-down of the plant. The allegation was not substantiated.
No violations were identified.
.
___
.
. . -
7
,
.
3.2 Allegation - Deficiencies in the Design of Concrete Floor Slab of
Reactor Enclosure
This allegation as understood by the NRC consisted of several subparts as
follows:
a.
The amount of reinforcing steel used in the floor slabs around the
containment structure is not in conformance with ACI 318-71 Code,
Section 10.5.1, specifically the formula (P min =200f fy),
b.
The distribution of the reinforcing steel is not in accordance with
the requirements in Section 10.6 of ACI 318-71 Code.
c.
The shear reinforcement requirements as contained in Section 11.1.1
of ACI 318-71 Code are not complied with,
d.
The bundling and lap splicing of rebar is not in conformance with ACI
318-71 Code.
Due to the close interrelationship of the above subparts, they were re-
viewed as one concern.
l
3.2.1
Scope of Inspection
The inspection effort was directed to ascertain that proper design
techniques and assumptions were made, and that requirements of the
code were correctly interpreted and/or applied, and that the slab, as
designed and constructed, would fulfill the safety function it was
designed to perform.
The above inspection objectives were pursued by review and examina-
tion of documentation, discussions with cognizant engineering per-
sonnel, and an evaluation of the existing design of the slab. Addi-
tionally, the adequacy and validity of design bases and associated
assumptions used in the design and analysis were also evaluated.
3.2.2
References
LGS - Preliminary Safety Analysis Report (PSAR)
LGS - Final Safety Analysis Report (FSAR)
Quality
Control
Inspection
Reports
(QCIR)
for
preplacement
inspections of Reactor Building floor-slabs at Elv. 201 and 217.
!
-
-- -
-
. - - . .
.
.
-
-
.
.
.
.
-
r
.
.
-
8
Bechtel Design Drawings:
C-122, Rev. 22, Reactor Building Unit 1 Floor Plan,
Elevation 217' 0", Area 11
C-123, Rev. 17, Reactor Bldg., U-1, Elv. 217, Area 12
C-126, Rev. 25, Reactor Bldg., U-1, Elv. 217, Area 15
C-127, Rev. 23, Reactor Bldg., U-1, Elv. 217, Area 16
C-113, Rev. 21, Reactor Bldg., U-1, Elv. 201, Area 11
C-114, Rev. 17, Reactor Bldg., U-1, Elv. 201, Area 12
C-117, Rev. 26, Reactor Bldg., U-1, Elv. 201, Area 15
C-118, Rev. 22, Reactor Bldg., U-1, Elv. 201, Area,16
Bethlehem Steel Drawings:
8031-C-39-140-2, 134-2, 115-2, 116-2, 191-2, 191A-2, 173-3, 173A-3,
174-3, 174A-3, 169-2, 166-2, 163-2, 163A-2, 158-2, 158A-2
Bechtel Drawings:
C-601, Rev. 31
C-602, Rev. 25
Project Civil Standards
C-606, Rev. 9
Bechtel Calculations
.
VOL.
FILE NO.
SHEET NO.
TITLE
(Calc. No.)
24
23.3
1 thru 12, 12-1,
Reinforced Concrete
13 thru 19, 19-1,
Slab design at
20 thru 25, 25-1
El . 201 - 0",
thru 25-4, 26,
Reactor Building
26-1 thru 26-3,
27 thru 50, 50-1
thru 50-26, 51
thru 53, 53-1,
54 thru 63, 63-1
thru 63-6, 64
thru 68, 68-1, 69
thru 73, 73-1 thru
73-2, 74, 75, 75-1
thru 75-6, 76, 77
(Total = 127 sheets)
24
23.5
1 thru 10
Recap of reinforce-
plus attachment
ment requirement of
(Total = 11 sheets)
slab edge at reactor
building
.
-
.
r
.
..
9
3.2.3
Conduct of Inspection
a.
Visual Inspection
The inspector visually examined floor slabs for general confor-
mance with design and for any obvious defects in workmanship.
The "as-built" configuration of the slabs was compared with de-
sign and construction drawings. This inspection was carried out
at the Limerick site.
b.
Review of Engineering Calculations
The review of engineering analysis and design calculations was
performed at Bechtel's Home Office in San Francisco, California.
The inspector reviewed calculations and held discussions with
cognizant engineeririg personnel
to determine the technical
validity and adequacy of design approach, design assumptions,
applicability of codes and standards, and the governing con-
struction drawings and specifications.
The referenced codes and standards were also reviewed and
evaluated to assess their applicabili_ty to the functional
objectives of the slabs in the Reactor Building.
3.2.4
Findings
a.
Description of Reactor Building
The Reactor Building (RB) is a reinforced concrete structure,
designed as a shear wall building. It is a rectangular building
about 324' long, 138' wide and 238' in height.
Most outside
walls are three feet in thickness, and are continuous around the
building. The interior walls are intermittent. There are about
nine floors in the building; some of which are continuous around
the containment structure (drywell/wetwell), and others extend
to limited areas. A one-inch gap separates the RB floors from
the containment wall.
The RB floors consist of a concrete floor-slab resting on a
horizontal steel framework consisting of girders, beams and
joists.
In some areas, the concrete floor slab is interrupted
by steel grating.
The slab thicknesses range from 12" to 36".
The concrete floor slab and the supporting steel members are
built as a composite structural element by the use of shear
connectors.
Metal decks are used to support the slab during
construction. The concrete slab is rigidly connected to the RB
concrete wall where the wall and slab meet.
_
,
.
.
10
b.
Design Approach and Assumptions
In the design of shear wall buildings it is conservatively
assumed that the horizontal earthquake load is resisted by the
walls parallel to the direction of load.
The distribution of
load in the walls is proportional to the rigidities of these
walls at a horizontal cross-section of the building. The floor
system acts as a diaphragm to enhance the rigidity of the shear
walls, and is not considered to resist any horizontal earthquake
load by itself.
Because horizontal earthquake loads imposed on
the floor system will be transmitted to the shear walls, the
purpose of designing a concrete slab for horizontal load is to
ensure that the slab will act as a diaphragm when transferring
load.
In view of the above, the following are the major simplifying
assumptions used by Bechtel in the design:
1.
The
horizontal
loads are
resisted by the
concrete walls in the direction of the earthquake forces.
2.
The
floor
system consisting
of
the
floor
slab
and
horizontal steel framework is designed mainly for vertical
loads.
3.
In designing the floor system as a diaphragm, only the
concrete floor slab is taken into consideration, and the
horizontal steel framework and metal decking is neglected.
4.
The floor slab is idealized as a beam with fixed or simply
supported end conditions depending on the actual support
condition of the slab, with the horizontal earthquake
forces acting parallel to the plane of the slab.
c.
Design and Analysis
In view of the foregoing design assumptions, the inspectors
reviewed the analysis and the associated design calculations to
assess technical, procedural, and analytical details, and to
evaluate the design output specified in drawings and specifica-
tions for construction.
To simplify the analysis, it was assumed by Bechtel that in
transferring the earthquake load to the shear walls, the load
will be resisted only by the concrete floor acting as a dia-
phragm even though the floor-slab is a part of a composite
structural system with the supporting steel members.
the horizontal earthquake load will be resisted only by the
--
concrete floor even thought the floor-slab is a part of a
composite structural
system with the supporting steel
members.
_
m
-.
.
11
the floor slab is a beam with the thickness of the slab as
--
the width of the beam'.
depending on the boundary conditions, the beam may be
--
fixed, or fixed at one end and simply supported at the
other.
the horizontal load acting on this idealized beam is the
--
product of the sum of the unit weight of vertical wall sec-
tion plus the unit weight of floor system, and the hori-
zontal acceleration due to an earthquake at the floor under
consideration.
Based on the above, bending moments at different sections of the
beam are computed, and the required area of reinforcing steel is
determined. Because there are openings of various sizes in the
floor, additional reinforcement around these openings is also
determined.
The above observations are based on the review of documented
analysis,
a random verification of computations during the
review process,
and extensive discussions with
structural
engineers at Bechtel involved in the design and analysis,
d.
Review and Evaluation of Code Requirements
In order to ascertain the applicability of any requirement of
the code, the intent of the code provisions must be established.
The minimum reinforcement requirement in ACI 318-71 Code Sectinn
10.5 is to prevent sudden failure due to too low a steel ratio
in a flexural member. The provision in Section 10.6 of the ACI
Code is to prevent the formation of large concentrated cracks in
flexural members. The provision in Section 11.1 is to increase-
.
the ductility, that is, to prevent sudden failure of flexural
members. As indicated in preceding sections, the floor slab is
designed for vertical loads and does not carry any load in the
horizontal direction, not even its own weight.
Its function, to
act as a diaphragm to distribute the horizontal earthquake load,
is incidental.
Moreover, the reinforcing steel designed for
vertical loads and placed orthogonally at the top and bottom of
the floor slab will enhance its ductility and resistance to
cracking when the floor slab is subjected to horizontal earth-
quake
load.
Disregarding
the
contribution
from composite
action of the supporting steel framework and the metal deck
acting as permanent framework for the floor slab is conservative.
e.
Bundling of Reinforcing Steel Bars and Lap Splice Length
The inspectors reviewed the QC documentation of inspections, and
rebar placement / detail drawings (Bethlehem Steel), and compared
e-a.
r
.
.
12
them to corresponding Bechtel design drawings to verify the con-
formance of rebar detailing to the design and code requirements.
The major part of this inspection and review was performed in
conjunction with the review of floor-slab design due to its
close interrelationship to each other. The inspector observed
that there was adequate instruction provided on Drawing C-601,
detail
8,
regarding lap splices and other details to design,
fabricate and install, and veri fy the splices in installed
rebars. The installation conformance wr.s verified and documented
by QC on Quality Control Inspection Reports (QCIRs). The in-
spector also verified that there were very few bundled rebars
(not exceeding 3 bars to a bundle) in the floors in question,
and they met code requirement for bundling.
3.2.5
Conclusions
On the basis of the above examination, review, discussions with en-
gineers and independent evaluation of available data, it is concluded
that:
a.
the ACI Code 318-71, Sections 10.5, 10.6, and 11.1.1 provisions
are not applicable for the design of the RB floor-slab against
horizontal earthquake loads.
b.
the bundling and lap splicing of rebars in floor-slabs are in
compliance with the project specifications and Code. requirements.
This allegation was not substantiated.
No violations were identified.
4.0 Pipe Break Analysis Allegations
4.1 Allegation - Analyses were performed with incorrect revisions of
isometric drawings
Through discussions and interviews, the second alleger's first concern as
understood by Region I was:
" Analyses were performed using original revisions of isometric drawings
when there were many later revisions available".
4.1.1
Scope of Inspection
The inspection effort was directed to ascertain the technical
validity of the allegation, to determine if the design staff has
performed the proper or improper analysis, and to assess, if the
'
allegation was valid, the impact of this error on the safety of the
plant.
The effort to pursue the inspection objectives consisted of
the following:
,
Review of the methodology of the plant design staff.
--
__
k
f
.
.
-
13
,
Review of the calculations performed to determine the location
--
and design of the different types of restraints.
Discussions with cognizant engineering and management personnel.
--
4.1.2
References
Limerick
Pipe break valve operability
Main
Steam Line
--
-
-
Inside Containment. Calculation No. S/8031/P-013.
Limerick - Pipe break dynamic analysis for isolation valve-
--
Operability of Main Steam Outside Containment.
Calculation No.
S/8031/PB-001, Revision 2.
Limerick
Pipe Break Dynamic Analysis for Energy Absorbing
--
-
Catacomb (EAC) design for Main Steam Lines A& B outside
containment.
Calculation No. S/8031/PB-031, Revision 0.
Limerick
Pipe Break forcing function calculation for Main
--
-
Steam Lines.
Calculation No. 25-76.
4.1.3
Conduct of Inspection
The inspector reviewed Bechtel's methods of analyses for the location
and design of pipe whip restraints and valve operability restraints.
There are three groups involved in the process which include plant
design staff, plant design project, and the civil group. The plant
design staff is responsible for the analysis of Class 1 piping. The
plant design project is responsible for the analysis of Class 2 and 3
piping, and the civil group, based on the results from the other two
groups, is responsible for locating and designing the pipe whip and
valve operability restraints.
4.1.4
Findings
Bechtel's civil group varied the process of waiting for the results
'of the analysis that was to be performed by the other groups (design
staff, and plant design project) and performed its own hand calcula-
tion (which after considerable investigation, was found to be con-
servative) and designed and located the restraints as needed in the
original and later revisions of the piping isometrics.
Once the
other groups completed their analysis, the civil group would verify
and make appropriate changes where needed.
4.1.5
Conclusions
Although the procedure for the calculational method of designing and
locating the pipe whip restraint and valve operability restraints was
somewhat unorthodox, because of the order in which the analysis was
c-
.
.
14
I
performed, the final check performed by the various groups involved
was assurance that no problems existed.
The allegation was not
l
substantiated.
No violations were identified.
'4.2
Allegation - Analysis was performed with incorrect Computer Code
The alleger's concern in this area as understood by Region I is as follows:
"For the Hope Creek project, the forcing function calculations were
performed by the plant design project staff using the ' Jet' Code which is
not the proper tool to use."
4.2.1
Scope of Inspection
The
inspection effort was directed to ascertain the technical
validity of the allegation, to determine if the plant design project
staff has performed the proper or improper analysis, and to assess,
if the allegation were valfd, :he impact of this error on the safety
of
the plant.
The effort to pursue the inspection objective
consisted of the following:
Review of the methodology of the plant design project staff.
--
Review the purpose of the " JET" computer code
--
Discussion with cognizant engineering and management personnel.
--
4.2.2
References
Hope Creek - HPCI System Inside Containment Pipe Whip Design.
--
Calculation No. 625-10Q
F. J.
Moody, " Prediction of Blowdown and Jet Thrust Forces",
--
ASME Paper 69 HT-31, August 6, 1969.
Standard Review Plan 3.6.2, " Determination of Rupture locations
--
and Dynamic Effects Associated with the Postulated Rupture of
Piping".
Branch Technical Position MEB 3-1, " Postulated Rupture Locations
--
in Fluid System Piping Inside and Outside Containment".
Hope Creek - JET impingement effects.
Calculation No.12-128,
--
Revision 0.
Hope Creek Pipe Whip Restraint and Isolation Valve Operability
--
!
File No. 1085.
i,
I
-.
F
.
.-
15
,
!
4.2.3
Conduct of Inspection
The inspector reviewed the plant design project staff role in
the process for analyzing the forcing function.
The plant
design project staff is responsible for the forcing function
calculations for Class 2 and 3 piping. In their analysis of jet
impingement effects, two methods are used.
First, Bechtel
performs a hand calculation to determine the total impingement
force acting on any cross-sectional area of the jet.
This
magnitude is equivalent to the jet thrust force as defined in
SRP 3.6.2, Subsection III.2.c(4),
i.e.,
T = KpA
where
P = System pressure prior to pipe break
A = Pipe break area
K = thrust coefficient
For the thrust coefficient, Bechtel uses a " Moody multiplier" of
less than 1.2 for steam and water-steam mixtures (See SRP
no
3.6.2, Subsection III.3.f.), and a factor not greater than 2.0
for subcooled, nonflashing water (See SRP
3.6.2,
Subsection
III.2.c(4).).
This method is in accordance with MEB.3-1 and SRP
3.6.2, and is considered conservative. Bechtel used this design
method for the majority of pipe whip and valve operability
restraints.
When an interference problem occurs, due to cumbersome design
based on the conservative calculations, the staff will use a
second, more realistic method that involves computer code
calculations.
The codes used in this method were RELAP4 and
REPIPE, which give a smaller forcing function that is still
within the design safety limit.
4.2.4
Finding
The JET code was never used by the plant design project staff for a
forcing function calculation for the design of the restraints.
4.2.5
Conclusion
The allegation was not substantiated.
No violations were identified.
_ _ .
.
.
. -
..
-.
..
.
-
.-.
-
.
. . ..
.
. ..
.
.
a
. _ _ -
.
.
16
4.3 Allegation - Inconsistent Pipe Break Analysis for Similar Plants
The alleger's concern as understood by Region I was:
"There has been many more complicated pipe break analyses performed for
Limerick
than Susquehanna,
for one similar piping system 14 break
locations were analyzed for Limerick where as only 6 locations were
performed for Susquehanna".
The alleger does not understand how this
could be valid for systems so similar.
4.3.1
Scope of Inspection
The . inspection effort was directed to ascertain the technical
validity of the allegation, to determine if the design staff has
performed the proper or improper analysis, and to assess, if the
allegation were valid, the impact of this error on the safety of the
plant. The effort to pursue the inspection objectives consisted of
the following:
Review of the methodology of the plant design staff
--
Review of the design and construction drawings
--
Review of the Stress report to determine usage factors.
--
Discussions with cognizant engineering and management personnel.
--
4.3.2
References
Limerick - Stress Report Calculation No. S/8031-1803
--
Susquehanna Unit I and 2 Calculation No. SR8856-1800
--
Appendix E.
4.3.3
Conduct of Inspection
The inspector reviewed the pipe location drawings for the main steam
system, main feedwater system,
high pressure coolant injection
system, and the standby liquid control system (SBLC) for Limerick and
Susquehanna.
All of the systems were primarily the same for both
plants except for the SBLC system.
The SBLC system for Limerick is
connected to the core spray line whereas for Susquehanna it is
connected to the reactor vessel.
4.3.4
Findings
For each piping system a stress analysis is performed. Based on the
stress analysis a number of break locations have to be analyzed. A
break location is designated by . usage factor. This factor is based
on the location of piping and stress due to loads.
For values
greater than 0.1, a break location is designated for that region of
pipe.
_
_ - _ _ - _ _ _
.
-
.
17
4.3.5
Conclusion
Due to the similarity of Limerick and Susquehanna, the same number of
breaks were analyzed for the majority of systems.
Based on the
differences in the location of the SBLC system (even though the SBLC
systems are similar), more locations had to be analyzed for Limerick
because of the greater number of break locations that were identified
based on the stress analysis report.
The allegation was not substantiated.
No violations were identified.
4.4 A_11egation - RELAP Analyses
The alleger's concerns as understood by Region I were:
(a) "Susquehanna FSAR Section 3.6.1,
entitled " Postulated Piping
Failures in Fluid Systems", and the accompanying figures specify
the
pipe break locations that should have been analyzed;
however, the calculations done by Bechtel do not correspond with
the locations committed to in the FSAR."
A similar allegation was stated relative to the Hope Creek
project.
The allegation as understood by the NRC was:
"For Hope Creek, of the systems listed in FSAR Table 3.6-1, pipe
break analyses were done on only five systems.
The other
analyses had not been done as of April 1984."
(b) "The main steam and feedwater lines for Susquehanna were
originally done with hand calculations but, subsequently, RELAP
analyses were carried out for these systems.
For the HPCI
system no blowdown forces were calculated." The Standby Liquid
Control System was also mentioned as being unsatisfactory with
respect to pipe break analyses.
(c) "At Susquehanna, Hope Creek and Limerick, hand calculations
should not have been used for determining valve operability."
The alleger believes that, instead of hand calculations, RELAP
Code analysis should have been used because this analysis
identifies the frequency of vibration in the pipe so that cor-
rective valve supports can be determined.
The alleger believes
that the hand calculations cannot trace a transient wave in the
pipe and, therefore, may cause a problem determining correct
valve supports.
4.4.1
Scope of Inspection
The inspection was directed to ascertain the technical validity of
the allegation to determine if designers had employed an appropriate
.
-
F'
.
.
18
design process, and to assess, if the allegation were valid, the
impact of this error on the safety of the plant operations and/or the
safety margin in the plant as a whole.
The effort to pursue the
inspection objective consisted of the following:
Review of engineering calculations
to determine technical
--
adequacy and validity of design approach and basic assumptions.
Review of technical
specifications,
design
drawings,
and
--
referenced codes and standards.
Discussion with cognizant engineering and management
--
personnel.
4.4.2
References
Report No.
NSC-1-78-027,
Pipe Break Dynamic Analyses for
--
Mainsteam and Feedwater Lines at Susquehanna.
Stress Report No. S/8856/PB-049. Rev. O, Forcing Function Study
--
i
forRCIC(outsidecontainment)SystematSusquehanna,
Stress Report No. S/8856/PB-046, Rev. O, Pipe Whip Analysis with
--
EAC Design for Mainsteam Lines at Susquehanna.
Susquehanna FSAR, Fig. 3.6.1A, Rev. 17
--
Hope Creek FSAR, Table 3.6.1 and FSAR Section 3.6.2.
--
Hope Creek FSAR Question 210.19.
--
Calculations 101 File No. SR-114, Comparison of GAFT Code with
--
RELAP 5 Calculations for the HPCI System outside Containment at
Susquehanna.
Calculation No. 4001, HPCI Line's Susquehanna.
--
Stress Report SR 8856-1800 Appendix E.
--
4.4.3
Conduct of Inspection
The review of calculations and other pertinent documents was carried
out in Bechtel's office in San Francisco, California.
The group
reviewed the calculations
and held discussions with cognizant
engineering personnel to assess the validity and technical adequacy
of calculations design assumptions and approach. The group compared
the pipe break locations identified in the FSAR Section 3.6.1 with
those actually analyzed in the pertinent stress report (First and
third references from Section 4.4.2).
.
_ , _ _ _ _ _ _ _ _ . . _ _ _ _ . _ - - -
#
,.
-
.
.
19
The FSAR was reviewed to determine if it contained requirements
relative to the method of pipe break analysis and whether those
requirements evoked pertinent codes and standards directly or by
reference and were sufficiently detailed and free of ambiguities. The
dynamic analysis procedures were reviewed for transient pipe break
loading to verify that postulated pipe breaks at intermediate
locations and terminal ends were in accordance with MEB-3-1 and valve
operability criteria.
4.4.4
Findings
Based on the above reviews of documentation and discussions with the
licensee and A/E engineers, the group determined the following:
With respect to allegation (a), the pipe break locations postulated
and analyzed for jet impingement effects and valve operability for
the main steam and feedwater systems at Susquehanna do not correspond
with the locations shown in Susquehanna FSAR Section 3.6.1 entitled
" Postulated Piping Failures in Fluid Systems" and the accompanying
figures.
However, discussions with the licensee's design engineers
indicated that the breaks were conservatively postulated at each
fitting and terminal ends.
Rationale was also provided to justify
r
'
that the breaks analyzed in the stress reports envelope the loadings
for restraint design and jet impingement effects resulting from the
postulated breaks committed to in the FSAR.
The group concluded that analysis of breaks at each fitting and at
terminal ends is an acceptable alternative procedure which conserva-
tively envelopes the FSAR commitment.
The allegation concerning the
discrepancy between the pipe break locations shown in the Susquehanna
FSAR to those actually analyzed, is valid but alternative analyses,
discussed above, make the allegation moot.
The listing of systems that are considered in the Hope Creek
Generating Station FSAR as critical in terms of High Energy Line
Breaks are listed in FSAR Table 3.6-1.
This tabulation provides the
listing of nineteen systems that are within this category.
Within
FSAR section 3.6.2, each of these systems is described with relation
to this condition, and the actions that have been taken to resolve
'
the issue.
USNRC Mechanical Engineering Branch requested from Bechtel in its
FSAR Question number 210.19
identification of any unrestrained
whipping pipes located inside containment.
The Bechtel response to
this question, which is described in the Bechtel Meeting, held on May
9,
1984, with NRC relative to FSAR questions, indicates that five
systems are in this category.
On the final page of these meeting
notes,
the
signature
showing
acceptance
for this question as
" acceptable-as-is" was verified.
.
I
-
-
-
.
_ _ _ . _ _ _ _ - _ _
. . . . .
..
. . . . . .
_
.._
.
. . _ . .
- . . . . _
. . , . . - . -
.
.
-
-
20
N
l
The pipe break analyses for the remaining systems were verified to be
!
available by selection of a sample of four systems listed in FSAR
4
Table 3.6-1.
These systems were:
(1) main steam; (2) reactor water
clean-up (RWCU); (3) core spray injection, and (4) emergency diesel
generator starting air. With the exceotion of the emergency diesel
,
generator starting air line, calculations for pipe whip and/or valve
operability restraints and line break analyses were available.
For
the emergency diesel generator air line, FSAR Section 3.6.2 provided
the detailed justification for the lack of need for these analyses.
With respect to allegation (b), the loadings from postulated pipe
breaks for the main steam and feedwater lines were initially hand
calculated, using the largest break area and the operating pressure
in the line resulting in a conservative design of restraints.
Sub-
sequentiv, the RELAP-5 Code was used for the same calculations, which
e
resulted in elimination of some restraints. The group did not find
anything unacceptable in this procedure.
In regard to the HPCI
System, the group reviewed the blowdown calculations (fifth and sixth
reference in Section 4.4.2) and found that the blowdown forces had
been properly calculated.
I-
The review of the calculations for the pipe break analyses for the
standby liquid control system revealed no inconsistencies or non-
compliance with applicable guidelines and standards.
The group,
therefore, concluded that allegation (b) stated in paragraph 1 above
1
is invalid and without technical merit.
Relative to allegation (c), the group determined that the A/E's
approach on Limerick, Susquehanna and Hope Creek with regard to pipe
whip and valve operability is to satisfy the requirements stated in
NRC Standard Review Plan 3.6.2, BTP MEB 3-1.
During the pipe whip
event, the stress in the pipe, between the containment and isolation
'
valve due to dynamic loads, dead weight and pressure remains less
than 2.25 Sm (design stress intensity) for Class 1 or less than 1.8
Sh (allowable stress at maximum hot temperature) for Class 2 and
o
Class 3 piping. Between the isolation valve and the moment limiting
pipe whip restraints, the stresses are maintained low enough to
prevent the formation of a plastic hinge.
The pipe stress at both
interfaces with the isolation valve is limited to 1.1 times the yield
stress, thus satisfying MEB 3-1 requirements.
The A/E performed a special set of calculations to evaluate various
hand calculated and RELAP forcing functions generated for valve oper-
ability analyses.
Five reduced linear transient analyses were per-
formed using two detailed RELAP forcing functions and three simpli-
'
fied hand calculated forcing functions.
Pipe stresses and restraint
loads obtained from these analyses were compared. Differences in the
input forcing functions and differences in the results obtained were
evaluated.
The Limerick RCIC piping system was used in the forcing
f
r
,
_ . _
_ _ _ _ _
___
. _ _
_ _ _ _ _ _ .
.
.
21
function study. The results of these calculations demonstrated that
the hand calculated (1.26 Pressure x Area) step forcing function com-
pletely envelopes the results of all other forcing functions, and the
,
pipe stresses resulting from hand calculated forcing functions are
more conservative as compared to the stresses due to RELAP forcing
functions.
In addition, the licensee made a qualitative assessment of the ac-
celerations and frequencies resulting from valve operability calcu-
lations.
The bending moment in the valve supports is inversely
proportional to the square of the excitation frequency in pipe and
directly proportional to the pipe acceleration. As the frequencies
associated with pipe whip are usually much higher than the frequency
content of a seismic event (for which valves are qualified), it was
concluded that for a given allowable bending moment at the valve
supports, higher accelerations expected during a pipe whip event are
acceptable. The group concurs with the assessment and concludes that
allegation (c) is without merit.
4.4.5
Conclusions
Allegation (a)
The inspectors concluded that although the pipe break location
analyzed for jet impingement and valve operability for the mainsteam
and feedwater systems at Susquehanna did not correspond with the FSAR
break locations, the actual analysis of breaks at each fitting and at
piping system ends was an acceptable procedure that conservatively
envelopes the FSAR commitment.
Therefore, although the allegation
was substantiated no safety concern exists with regard to the
operability of the main steam and feedwater systems.
The inspectors did not identify any discrepancies relative to the
pipe break analyses listed in Hope Creek FSAR Table 3.6-1.
There-
fore, it was concluded that this portion of the allegation was not
substantiated.
!
Allegation (b)
i
The inspectors did not identify any inconsistencies and/or noncom-
al %nce with applicable guidelines and standards in the analyses and/
i
f
or calculations for HPIC or Standby Liquid Control System.
Therefore,
i
it was concluded that the allegation was not substantiated.
Allegation (c)
The inspectors concluded that the Bechtel stress analyses and calcu-
lations were acceptable because the Bechtel hand calculations for
forcing functions are r;: e conservative than calculations by the
RELAP computer code. Momver, the frequencies associated with pipe
__
-
_
.
.
22
whip were much higher than the frequencies associated with a seismic
event. The inspectors, therefore, concluded that for a given allow-
able building moment at valve supports, higher acceleration expected
during a pipe whip event was acceptable.
The allegation was not
substantiated.
No violations were identified.
4.5 Allegation - Feedwater Check Valve Slam
The alleger's concern as understood by the NRC is as follows:
"The feedwater system piping analysis for Susquehanna Steam Electric
Station ignored the effects of feedwater check valve slam.
This problem
had been worked on by Sargeant and Lundy; but the problem (check valve
slam) was still unresolved."
4.5.1
Scope of Inspection
The inspection effort was directed to determine the validity of the
allegation, to determine if feedwater check valve slam had been
considered in the feedwater pipe break analysis, and to identify any
continuing unresolved problems. The inspection effort consisted of:
--
Review
of
appropriate
documentation
for
the
Susquehanna
Feedwater Check Valve Slam Analysis.
Discussion with cognizant engineering and management personnel.
--
Review of the Feedwater Check Valve Slam Analysis for Limerick
--
Generating Station.
(This review was conducted for comparison
of Bechtel
activities
for Susquehanna
and
Limerick.
The
allegation did not mention any concerns relative to the limerick
Feedwater Check Valve Slam Analysis.)
4.5.2
References
Sargeant and Lundy Report - Susquehanna Steam Electric Station
--
Feedwater Check Valve Analysis - dated March 11, 1983.
NRC letter to Pennsylvania Power and Light Company, Feedwater
--
Check Valve Analysis, dated April 24, 1984.
Bechtel Report - Evaluation of Feedwater Containment Isolation
--
Check Valves for
a. Hypothetical Pipe Rupture Condition for
Limerick Generating Station, dated July 1983.
4.5.3
Conduct of Inspection
The inspector reviewed the 3echtel contractor analysis for the effects
of a postulated break of the feedwater system in a line outside con-
tainment.
The analysis was conducted to verify (1) the integrity of
__
(
.. _.
O
.
23
the intact portion of the feedwater system from the containment iso-
lation valves to the reactor and (2) the integrity of the check valve
for performance of its containment isolation function. The analysis
for Susquehanna was conducted by Sergeant and Lundy for Pennsylvania
Power and Light Company (PP&L). The contractor report and responses
. to NRC requests for additional information were provided to the NRC
as discussed in the referenced NRC letter to PP&L dated April 24,
1984.
For comparison, the analysis of this corresponding postulated break
in the feedwater system for Limerick was reviewed.
The Limerick
analysis was performed by Bechtel and provided in the referenced re-
port dated July 1983. The Limerick analysis was reviewed to deter-
"
mine if integrity of both the feedwater check valve and the intact
portion of the feedwater piping were maintained.
4.5.4
Findings
'ae Feedwater Check Valve Slam Analysis for Susquehanna had been
conducted by Sargeant and Lundy and subsequently submitted to the NRC
for review.
No discrepancies were identified relative to
the
Susquehanna analysis.
Based on the NRC review of the conduct of the Limerick Feedwater
Check Valve Slam Analysis four discrepancies were identified.
The
discrepancies
identified were
related to
the analysis of the
integrity of the check valve. The discrepancies were:
a.
The material properties usumed in the analysis were for SA-216;
however, the manufacturer's (Atwood and Morrill Co.) drawings
indicated the valve body and disc materials to be SA-352, Grade
LCB.
This inconsistency in the analysis was not properly
reviewed and evaluated for acceptability.
b.
The material properties used in the analysis were based on room
temperature; whereas, the system was designed for 425 F service
temperature. The higher service temperature resulting in lower
tensile properties of the material was not properly evaluated
and accounted for in the analysis,
c.
No calculation or other documented evidence was available to
establish the integrity of hinge / hinge pin during valve closure
and disc / seat deformation at the time of seating.
d.
The evaluation did not address the use of welded-in seat seal in
place of direct disc contact with the weir / orifice of the valve
body.
._.
m
.
.
24
Licensee review and revision of the analysis considering the noted
discrepancies is an unresolved item. (352/84-48-01). Based on NRC
review of the analysis and the margins with respect to ult: mate
strain limits demonstrated by the analysis it was determined that the
structural integrity of the valve would be retained.
The Limerick analysis was reviewed to determine if the feedwater
piping would withstand the pressure surge resulting from the check
valve slam. No discrepancies were identified for this portion of the
report.
4.5.5
Conclusions
On the basis of the above review, it is concluded that:
a.
The Susquehanna Feedwatar Check Valve Slam Analysis identified
no unresolved. problems and no discrepancies. The allegation was
not substantiated.
b.
The corresponding Limerick Feedwater Check Valve Slam Analysis
included discrepancies identified by the NRC. The discrepancies
identified were not associated with any allegation.
5.0 Exit Interview
An exit interview was held with representatives of the Bechtel Corporation
on August 30, 1984 at the Bechtel Engineering Office in San Francisco,
California. .The results of the inspection were discussed at that meeting.
One issue pertaining to Limerick was identified during the inspection as
described in this report (p.23).
The issue did not result from any
allegation but was discovered during comparison of Bechtel activities for
the three projects reviewed.
Also, part of one allegation was substan-
tiated, but it was determined that no safety concerns existed (p.19).
This allegation pertained to Susquehanna.
)