ML19325E446
| ML19325E446 | |
| Person / Time | |
|---|---|
| Site: | Palisades  |
| Issue date: | 10/27/1989 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML18054B069 | List: |
| References | |
| NUDOCS 8911070103 | |
| Download: ML19325E446 (50) | |
Text
o l
ENCLOSURE 1 l
[
SAFETY EVALUATION BY THE OFFICE OF NUCLEAR _ REACTOR REGULATION i
REPORT ON A5YMMETRIC LOCA L0kDS i
7AI.15ADE5 NUCLEAR POWER PLANT l
NERAhlCAL ENGTNERTNU TRANCTI i
DIVISION OF ENG!NEERING & SYSTEMS TECHNOLOGY j
1.0 INTRODUCTION
On May 7,1975, the Nuclear Regulatory Commission (NRC) was informed i
that asymmetric loading on the reactor vessel supports resulting from f
f a postulated reactor coolant pipe rupture at a specific location (e.g.,
[
the vessel nozzle) had not been consfdered in the original design of the reactor vessel support for North Anila Units 1 and 2.
It had been indenti-fied thdt in the event of a postulated, instantaneous doeble-ended offset shear pipe break at the vessel nozzle, asymmetric loading could result i
from forces induced on the reactor internals by transient differential t
pressures across the core barrel and by forces on the vessel due to trans-
+
ient differential pressures in the reactor cavity. With the advent of more sophit,ticated computer codes and the development of more detailed analytical mocels, it became apparent that such differential pressures.
l although of short duration, could place a significant load on the reactor vessel supports and other components, thereby possibly affecting their integrity. Although this potential safety concern was first identified during the review of North Anna facilities, it was determined to have generic implications for all pressurized water reactors.
i s
i 8911070103 891027 PDR ADOCK 05000255 P
t o
i it e !
i In October of 1975, the NRC staff notified each operating Pressurized Water Reactor (PWR) licensee of a potential safety problem concerning the design of their reactor pressure vessel support system.
From this
[
survey it was discovered that these asynenetric loads had not been con-I sidered in the design of any PWR primary system, in June 1976, the j
t NRC requested all operating PWR licensees to evaluate the acequacy i
of reactor system components and supports at their facilities, with respect to these newly identified loads.
Licensee and vendor responses to this request were proposals to augnent inservice inspection and/or i
probability studies that supported no analyses due to the low probability
{
t of the pipe breaks at a particular location. Although the NRC recognized some merit in these proposals, they determined that the more fundanental questions still remained unanswered.
Therefore, licensees of PWR plants were notified by letter dated January 20, 1978 that the evaluation of their primary systems for asymmetric LOCA loads would be required.
l Although the NRC Staff's original emphasis and concerns were focused primarily on the integrity of the reactor vessel support system with j
respect to postulated breaks inside the reactor cavity (i.e., at a nozzle), it became apparent that significant asynenetric forces could also be generated by postulated pipe breaks outside the cavity and that s
the scope of the problem was not limited to the vessel support system, itself.
The staff, after reviewing this problem, determined that a re-evaluation of the primary system intergrity of all PWR plants to with-1
-.~.- --,, -.,,-
L c
l
- c 3-f t
l stand these loads was necessary.
Therefore, in January of 1978, the NRC l
i Staff requested each PWR licensee to submit additional information in accordance with the expanded scope of the problem.
Those letters outlined l
the present scope of the problem specifying a minimum number of pipe break locations to be addressed and the reactor system components to be evaluated.
)
The asymmetric loading on the primary system that was determined by NRC to
(
have generic i.nplications for all PWRs, was formally identified in Task Action Plan A-2 Unresolved Safety Issue (USI), " Asymmetric Blowdown Loads l
on Reactor Primary Coolant System," as published in NUREG-0371, " Task i
i Action Plans for Generic Activities (Category A), USNRC, November 1978.
Since the identification of the asymmetric load problem in May 1975 EG&G Idaho, Inc. has performed a number of independent audit analyses to verify t
licensee submittals on this problem.
A total of six anaylses have been completeo (one linear elastic and one nonlinear-inelastic analysis of re-actor coolant loop (RCL) for each of the three major reactor vendors).
t Based on these analyses and additional NRC staff investigations, criteria and guidance for conducting an evaluation of asymmetric loss of coolant accident (LOCA) loads were developed.
USI A-2 was resolved in January 1981 with the publication of NUREG-0609 (reference 3).
This document provided
[
3 i
i t
-,...,,--,-e.
o k
j rc I
-4~
I I
l r
an acceptable basis for performing and reviewing plant analyses for asym-metric LCCA loads and affected all operating and future PWRs.
During the course of the work on USI A-2, it was demonstrated that there were only a very limited number of break locations which could give rise
{
to significant loads. Subsequently, af ter the development of substantial new technical work, it was demonstrated that the rew techniques for the i
analysis of piping failures assured adequate protection against failures in primary system piping in Pressurized Water Reactors.
This was reflected in a revision of GDC-4 published in the Federal Register in final form on April 11,1!iB6, and in a further revision to GDC-4 pub-f lished in the Federal Register on July 23, 1986.
In addition, it has also been satisfactorily demonstrated in the course of the A-2 effort I
i that there is a very low likelihood of simultaneous pipe loading with both LOCA and SSE loads.
I t
For Combustion Engineering plants of the pre-CESSAR vintage without the I
l SSE-LOCA load combination, the loads on primary system piping would not result in pipe breaks which could lead to significant loads on the core l
structure.
Accordingly, for these facilities the staff had concluded that the potential for asymmetric loading on the core structure resulting from primery system piping LOCA, need not be considered in the design of
[
the core structure.
i
n-i c.
- o l
5-f In June of 1980, combustion Engineering (CE), on behalf of the Baltimore Gas and Electric Company, a member of the CE Owners Group, submitted a final asymmetric LOCA loads evaluation report (Reference 1), applicable to the Palisades power plant.
This material, submitted in response to the January 1978 letter from NRC, was reviewed by the NRC staff and its con-sultants.
Upon review of the submittal, it was determined that additional information was required to satisfy the established guidelines and accept-f ance criteria.
On February 23, 1981, the NRC staff notified CE of the additional requests, and the response (Reference 2) was submitted in
[
August of 1981, t
The Palisades final submittal and supplement represent the limiting cases for the asymmetric LOCA loads evaluation and have been reviewed in conjunc-tion with the criteria outlined in NUREG-0609 (Reference 3).
Subsequent sections of this safety evaluation report summarize the evaluations per-formed by the licensee for subcooled blowdown loads, cavity pressurization, and structural response.
The staff's evaluation includes an assessment of the licensee's compliance with the acceptance criteria.
2.0 DISCUSSION I
l C
The licensee's analysis procedure including analytical models, computer j
i methods and analytical results is discussed in the following paragraphs.
l l
l r
I I
i
i O
The analytical methodology primarily consists of (a) development of thermal hydraulic loads for the rea: tor coolant system (RCS) structural analysis, (b) Calculation of the steam generator and reactor cavity pressures, and (c) Calculation of the loads and stresses on the various components and supports of *he f.CS which include the vessel and steam generator supports, vessel internals, fuel assemblies, control element drive mechanisms (CEDM) and emergency core cooling system (ECCS) piping.
2.1 Thermal Hydraulic loads Analysis The CEFLASH-48 computer code (Reference 4 and 5) was used to predict the transient hydraulic response of the reactor primary coolant system to the most critical postulated pipe breaks.
These were the guillo-tine pipe breaks with a break opening area of 0.10A8 (or 135 in.2) in 0.20 sec at the reactor vessel outlet nozzle and the 2.0A (or 1414 in.2) break in 0.23 see at the reactor vessel inlet nozzle.
The CEFLASH-4B analysis is based on the volume-flow path concept.
This involves simultaneously solving the conservation equations of mass, momentum, and energy, and the fluid pressures, densities and enthalpies.
All of these parameters are assumed to exist in a state l
of thermodynamic equilibrium.
The CEFLASH-4B code also assumes the a.
A break opening areawill be referred to as a inultiple of the cross-sectional flow area (A) of the pipe at the specified location.
A full guillotine off-set break is defined as 2A.
l
l c
i e
7 1
l fluid boundaries to be rigid and at rest, thereby excluding fluid-l structure interaction effects of the core barrel-reactor vessel relative motion on the downcomer pressure transients in the subcooled l
4 loads hydraulic analysis.
The results of the thermal-hydraulic analy-l sis provide the time history forcing functions applied to the reactor vessel and internals in the reactor coolant system (RCS) structural analysis.
2.2 Cavity Pressurization Analysis The subcompartment pressurization analyses of the reactor cavity and steam generator subcompartment were performed in a two-step f
procedure.
First, the blowdown mass flow rates and energy release i
i rates were calculated for the RCS. Then cavity pressures were l
computed using these release rates to determine component support l
and compartment wall loading transients.
[
l l
Mass and energy release rates were calculated using the modified i
CEFLASH-4 computer code based on four design basis postulated pipe i
ruptures; which were determined to be the most limiting breaks.
l m,.
r; 1.
2.0A (or 1414 in.2) break in 0.023 see at the reactor vessel inlet nozzle.
2.
0.10A (or 135 in. 2) break in 0.020 see at the reactor vessel outlet nozzle.
0.70A (1000 in. 2) break in 0.024 sec at the steam generator 3.
inlet nozzle.
2.0A (or 1414 in.2) break in 0.020 see at the steam Generator 4.
outlet.
(
The J. FLASH-4 code was modified to incorporate a critical flow l,
correlation subroutine that maximizes the blowdown rates.
This was l
achieved by utilizing a combi ted Henry /Fauske and Moody critical flow correlation with a flow multiplier of 0.7 throughout the blow-down transient.
The subcooled and low quality fluid conditions used the Henry /Fauske correlation, while the Moody correlation was used for i
the remainder of the saturated regime.
l l
Calculation of rtactor pressure vessel (RPV) cavity pressures was performed using the RELAP4-M006 computer code.
Pressures were determined Ith the mass and energy releases from the design breaks at ths rarctor vessel iniet and outlet nozzles.
e :
3 The reactor cavity has a net free volume of about 9000ft.
The input I
t model contains 36 volume-nodes, determined from sensitivity studies
{
to be detai' led enough to provide a convergent solution.
Resultant force and mou nt time histories on the reactor vessel are shown in Figures 1 and 2, respectively, for the hypothesized hot leg break of f
0.10A.
The 2.0A double-ended cold leg break results are provided in Figures 3 and 4.
f i
l l
Calculation of steam generator compartment pressure was performed using the DD1FF1-MOD 7 computer code.
Pressure time histories were determined with the mass and energy releases from the design breaks
[
at the steam generator inlet and outlet nozzles.
The steam genera-tor cavity is illustrated in Figures 5 and 6 using an elevation view and several plan views.
Each plan view in Figure 6 is keyed to an elevation shown in the elevation view of Figure 5.
The nodalization shown in the figures refers to the input model of the cavity.
Based on the geometry of the cavity and sensitivity studies, 36 volume-nodes were determined to be adequate.
i 2.3 Structural Analysis The Licensee's structural analysis was performed utilizing two primary finite element models and several component and support detailed finite element models.
The subsystem nodels were used to
r 9
', l i
develop input to the primary models and to calculate component and f
support loads and stresses for detailed evaluations.
The mathematical modeis to which asymmetric LOCA loads were applied are described in l
l the following subsections.
The general plant layout is shown by the i
i illustration of Figure 7.
I t
1 2.3.1 Reactor Coolant System Analysis As shown in Figures 8 and 9, the mathematical model consists of l
the reactor vessel, simplified reactor internal components,
[
reactor coolant pumps, steam generators, and the interconnecting i
piping.
The RPV support characteristics were determined from t
an individual detailed model (described in Section 2.3.3).
Responses of the modeled syetem components were calculated using the STRUDL, DAGS, and FORCE computer codes for the four design basis pipe breaks.
STRUDL developed the system stiffness matrix which was supplied to DAGs, a nonlinear structural code.
Along
-l with other pertinent structural parameters and the applied LOCA i
forcing functions, DAGs determined the RCS time dependent motions.
Applied loads consisted of cavity pressurization and pipe tension release forces, and for pipe breaks within the RPV cavity, internal blowdown loads were also included.
The nonlinear analysis con-t tained gapped support definition and included the effects of hydrodynamic mass.
The DAGs time history motions were supplied l
t
O 0
+
0 to the post-processor code FORCE, which calculated maximum pipe nozzle loads and support loads. For the breaks at the steam generator (SG) inlet and outlet nozzles, the SG was modeled in greater detail to better understand its response.
Included in the modeling were the SG internals and support non-linearities as shown in Figure 10.
The resulting RCS analysis response time historieswereutilizedinthesubsystemevaluations(Section 2.3.3) of the vessel supports, vessel internals, fuel assemblies, control element drive mechanism (CEDM) and emergency core cooling system (ECCS) piping to determine final components and support qualifications.
2.3.2 Primary Shield Wall Analysis The ability of the primary shield wall to sustain the worst case pipe rupture loads was determined from a linear elastic, three dimensional model of the wall using the Bechtel Corporation's BSAP computer code. The finite element model used hexahedron brick elements to represent the shield wall, the steam generator pedestals, and the north and south refueling pool walls. Loads on the model consisted of reactor vessel support reaction loads l
and reactor cavity pressurization loads.
The loads were applied l
as static loads with appropriate dynamic load factors, l
{
i
.C i
I 1 j 2.3.3 Subsystems Analysis i
Numerous smaller, more detailed mathematical models were used in the LOCA analysis to provide representative and meaningful responses to the applied loadings. One model was developed to detennine the component support stiffness to be used in the system analyses as well as to qualify the supports once system responses were obtained.
The MARC computer program was used i
for the analyses of the reactor vessel supports, located below l
the two cold leg nozzles of Loop 1 and below the hot leg nozzle of Loop 2.
Figure 11 illustrates the supports, and Figure 12 f
t represents the mathematical model using elastic-plastic three-dimensior.31 elements. The resulting cold leg and hot leg support i
load deflection relationships supplied to the system model of
[
Section 2.3.1 are shown in Figures J3 and 14, respectively.
Several other models were utilized for detailed qualifications of particular components. Applied loadings and/or motions to I
i these models were rJsponses from the system dynamic analysis.
[
l The detailed models are as follows.
1.
The reactor vessel internals were evaluated for guillotine pipe breaks at the reactor vessel inlet and outlet nozzles, employing l
lateral and axial mathematical models. Nonlinear analyses were
[
performed in accordance with established procedures (Reference 6) f using beam elements, gap elements, and linear and nonlinear spring 1
1 i
- o i
elements.
Hydrodyne;nic coupling effects'were included in the hori-zontal model. Botn models were subjectM to a combination of ap-plied forces and excitations. The time history forces applied to vessel internals resulted from the LOCA blowdown analysis described in Section 2.1, and the time history motions of the reactor vessel resulted from the RCS analysis described in Section 2.3.1.
The lateral and axial models are shown in Figures 15 and 16 respect-ively. Results of the analyses were time dependent member loads and nodal displacements, velocities, and accelerations.
In ad-dition to the horizontal and vertical responses of the vessel internals, vibration and stability analyses were performed on the core support barrel (CSB) to determine possible contributing barrel stresses. The shell mode response of the barrel due to LOCA pressure loads applied to the barrel from the break at the RPV inlet nozzle was analyzed. Axial type loadings on the CSB from a pipe break at the RPV outlet nozzle were investigated with the aid of the SAMMSOR/DYNASOR computer code (References 14 and l
- 15) and the buckling potential of the barrel was determined.
Figures 17 and 18 show the axismmetric vibration and stability l
models of the barrel.
l
.s
-i i
14 -
a i
2.
The CEDMs were evaluated with an elastic plastic finite element model using the MARC computer code.
Time history motions of the
{
reactor vessel head, determined from the RCS analysis, were applied to the base of the CEDMs.
The controlling section of the compo-i nent is the CEDM nozzle, near the interface'with the PPV head.
f 3.
The integrity of the ECCS pipino was evaluated for asymmetric LOCA l
' loadings by performing an elastic analysis using the STRUDL and
?
I DAGS computer codes.
Input excitation to the analysis was provided by the time history motions of the ECCS nozzle, resulting from the system LOCA response.
The motions were directly computed at the appropriate location on the reactor coolant pump (RCP) dis-charge leg.
The mathematical models are shown in Figure 19.
j 2.4 Summary of Licensee's Analytical Results The basic criteria for acceptability of the p h nt for the postulated i
faulted condition is to provide high assurance that the reactor can be brought safely to a cold shutdown condition.
The licensee con-cluded that overall acceptability of the plant for the postulated 1
i-LOCA was met.
This was demonstrated by the following component and l
structure evaluations believed by the licensee to be the worst or L-limiting cases.
A summary of load and stress results from the LOCA l
analyses is presented in Table 1.
l L-I lp l
l l
1 l
(
L l'
~ 2.4.1. Reactor Yessel Supports The primary supporting system for the reactor vessel consists of three nozzle supports: beneath two cold leg nozzles and one hot leg ~ nozzle.
Using the system model previously aescribed in Section 2.3.1, the greatest loads on the supports were horizontal forces, resulting from a postulated break at the reactor vessel inlet nozzle. The criterion for the supports based on an instab-ility analysis cs des::ribed in Section 2.3.3, and according to the ASME Code, 'Section III Appendix F, loads should not exceed 70%
of the plastic instability load. As it can be seen from Table 1, this criterion has been met. Therefore, the RPV supports are considered acceptable for LOCA loadings by the licensee.
2.4.2 Steam Generator Supports The SG supports were evaluated using the RCS model with the SG described in Section 2.3.1.
Subcompartment pressurization j
i considering pipe breaks at the SG inlet and outlet nozzles
]
providing the primary source of applied loadings in the analysis.
Support components evaluated consist of the lower pads, lower i
i stop, lower and upper keys, holddown bolts, and snubbers. Ac-ceptability of the support components was based on a comparison of the results to design loads criteria. The design loads shown j
b m
l l
' i in Table 1, are less than or equal to 90% of yield values, except for the bolts and snubbers which are compared to yield and actual 1
test values, respectively.
2.4.3 Reactor Coolant Pump Supports The RCP pump supports consist of four lugs welded to the scroll and were evaluated directly in the RCS analysis for the pipe l
rupture at the RPV inlet nozzle.
Design loads were used as the acceptance criterion, and as it is shown in Table 1 the resulting shear loads on the support lugs were well below the acceptance
. limits.
2 4.4 Reactor Coolant Piping The primary coolant piping was expected to be most highly stressed at component nozzles.
Considering the four design basis pipe breaks, resultant loads on the RPV nozzles, SG nozzles, and RCP nozzles were determined from the RCS analysis described in Section 2.3.1.
All loads result in stresses which meet the faulted limits set forth by the ASME Code,Section III, Appendix F.
Table 1 indicates the acceptability of the primary piping.
f~
c t
I.4 l
- l'/ -
j.
I 2.4.5 Reactor Internals The three major parts of the internals consist of core support barrel, the lower core support structure, and the upper guide l
structure. These components were evaluated with the. mathematical models described in Section 2.3.3, and the results are shown in Table 1, as percent margins only. The shell vibration response was combined with the axial and lateral barrel beam response, and barrel stability was investigated and found to be too far removed from stability considerations. The internals responses
~
to the asymmetric LOCA loadings are shown to be acceptable com-pared to the ASME code, Appendix F allowables:
2.4 S, for mem-brane stress intensity and 3.6 S, for membrane plus bending stress intensity.
2.4.6 Control Element Drive Mechanism l
The evaluation of the CEDM is based on an elastic-plastic instah-ility analysis (Section 2.3.3).
The maximum moment at any section.
l due to the RPV motions during the transient, occurs at the base l
of the CEDM nozzle.
From Table 1 it can be seen that this moment j.
is less than 70% of the plastic instability load; therefore, the component is acceptable.
C q
y
-,y..
.r
-.u-+,
w
-. v--
e
p -1 1
p.
q t
s 2.4.7 ECCS and Connected Piping The ECCS piping was evaluated based on the results of an elastic dynamic analysis using the substructural models described in I
Section 2.3.3.
Motions from the RCS dynamic analysis provided the input excitations to the models, and the calculated maximum piping moments were shown to meet the ASME Code, Appendix F limits.
The results are given in Table 1.
The maximum moment in an elbow occurred from the break at the RPV inlet nozzle, and the maximum moment in a straight run of pipe occurred due to the break at the RPV oulet nozzle.
il 2.4.8 Primary Shield Wall The reactor cavity wall was evaluated for cavity pressurization loadings and vessel support reaction loads resulting from pipe breaks at the vessel inlet and outlet nozzles.
A static analysis was performed with the structural model described in Section 2.3.2.
I Worst case reaction loads were applied to appropriate nodal points, and asymmetric pressures were distributed across the surface of affected elements. The results consisted of moments, axial loads, and shears evaluated at the hoop and vertical sections of the shield wall. High hoop stresses o: curred at the elevation just below the pipe penetration, and in some locations the high tensile
. ~.
~
i.l g::T' P
i
! [
P loads slightly exceeded the reinforcing steel yield limits.
This may result in some local cracking but the structural integrity of the shield wall would be maintained.
The evaluation of the vertical sections indicated that the bending and shear
'f results meet the requirements of the American Concrete Institute Code ACI-318.
It was concluded that the biological shield wall has the capacity to' resist the severe loads imposed by a design basis pipo rupture.
3.0 STAFF EVALUATION The licensee's analysis. procedure including analytical models, computer methods, and acceptance criteria have been evaluated by the staff for asymmetric LOCA loads.
The staff evalsation was accomflished by reviewing the licensee's submittel and using the independent audit calculations performed by the staff or its consultants.
In general, the staff has concluded that the licensee's assessment of the asymmetric LOCA loads prob-lem is acceptable.
The staff evaluation of each specific analysis phase 1
is addressed in subsequent paragraphs, following the guidelines set forth l
by NUREG-0609.
1
)
'O-
' 3.1 Thermal Hydraulic Blowdown Loads The thermal hydraulic blowdown calculation portion of the Palisades
(
asymmetric LOCA load submittal has been reviewed and is cansidered to be acceptable to the staff.
The basis of this acceptance is the staff's re-view and approval of the CEFLASH-4B computer code used for the internal 4
hydraulic loads calculations.
Independent audit calculations for the CE 2570 MW plant by the staff's consultant aided in approval of the CEFLASH-4B application to subcooled bicwdown.
The code does not consider fluid-i structure interaction, and the structural boundaries are assumed rigid and at rest.
Such conditions normally give rise to tonservative pressures and loaas.
A significant number and location of postulated pipe breaks were analyzed to determine worst case loadings on the primary coolant system.
Size and length of break openings consisted of reasonable and realistic values.
Nodalization and modeling were also developed in a manner that provided reasonable representation of the existing system.
l 3.2 Cavity Pressurization Analysis The licensee's reactor cavity pressurization analysis of the Palisades plant for postulated breaks at the reactor vessel inlet and outlet nozzles l
has been reviewed and is considered to be acceptable by the staff.
The i
basis of this acceptance is the staff's review and approval of the CEFLASH-L 4 and RELAP4-M006 computer codes used for calculating LOCA mass and energy re-lease rates and cavity pressure loadings, respectively.
Although the
m
- 4.
licensee used RELAP4-MOD 6 instead of REl.AP4-MOD 5, the code and its applica-tion were concluded to be acceptable.
The licensee used a flow multiplier of 0.7 instead of the recommended value of 1.0 in the CEFLASH-4 calculations.
The value 0.7 was justified by comparison of the data with the critical flow correlations.
The nodalization of the input model is acceptable based on the staff's review of input data and sensitivity studies performed by the licensee.
The SG subcompartment pressurization analysis of the Palisades plant for postulated breaks at the SG inlet and outlet nozzles has been reviewed and is considered acceptable.
Acceptance is based on the staff's review of v
the data provided by the licensee and its previous review and approval of the DDIFF1-MOD 7 code for calculating LOCA cavity pressure loadings.
The nodalization of the input model is acceptable based on review of the input 5
data and sensitivity studies performed by the licensee.
3.3 Structural' Evaluation 3.3.1 Evaluation of-Methods ard Models The structural computer codes cited in the licensee's report are l
found to be acceptable to the staff.
The c.) des (STRUDL, DAGS, NASTRAN, MARC, CESCH0CK, ASHSD, SAM 450R/DYNASOR, and ANSYS) util-ized in the LOCA analyses have been bench marked in a satisfactory I
manner to the staff.
The methods used in performing the required l
l l
j
-t
! l l
structural analyses are acceptable to the staff in as much as they l
conform to the accepted state-of-;Se-art standards, and. regulatory codes.
Based on a review of the submittals (References 1 and 2), the detail employed in the system and subsystem structural finite ele-ment models is considered acceptable by the NRC staff for predicting the mechanical response.
l t
The staff evaluation in this report has considered the need to combine LOCA and. safe shutdown earthquake (SSE) loads in the design of the RCS-piping.
The staff belic,rs that there is sufficient technical evidence (Reference 9) which demonstrates that the SSE and LOCA for the main loop piping in PWR plants may be considered as independent events in determining the appro-priate combination of the effects of accident conditions and natural phenomena as required by GDC 2.
In its load combination program, as a part of generic issue B-6, l
Lawrence Livermore National Laboratory (LLNL) conducted a program to estimate the probability of a double ended guillotine break (DEGB) in the reactor coolant loop piping of PWRs.
1 The results of the LLNL investigations indicate that the probabil-ity of a direct seismically induced DEGB is extremely small.
The i
best estimate probabilities of direct DEGB using the medians of l
l l
r
I
?.--
1 l
23 -
j b
the distribution of the modeling uncertainties range from 5 x 10' N 12 to 7 x 10 per plant year for both Westinghouse and Combustion Engineering plants.
From the uncertainty analysis, considering the whole range of modeling uncertainty, it is concluded that a direct
-10 DEGB probability of 3 x 10 per plant year can be considered as the absolute upper boand for Westinghouse and CE plants.
Indirectly induced DEGB in the reactor coolant loop piping (de-fined as a DEGB in the reactor coolant loop piping as a result of I
an impact with a large component or structure, e.g., a falling
_ polar crane) is a'more likely event compared to direct DEGB; how-ever..the probability of indirect DEGB is also very small.
For the lowest seismic capacity Westinghouse plant, the median proba-
-6 bility of DEc3 is 3.3 x 30 per plant year.
The corresponding
[
-5 t
indirect DEGB probability at the 90th percentile is 2.3 x 10 Even for this lowest capacity plant, these probability values are still very small.
For all 46 Westinghouse units east of the Rock-ies as a whole, the median probability is more than one order of I
magnitude lower. The probability values for the Combustion En-gineering plants are also very low.
The upper bound probability i
values for the Combustion Engineering plants are comparable with those of the Westinghouse plants.
Based on the results of these l'
probability studies, che analysis submitted by the licensee and independent assessments by the staff and its consultants, the staff 1
has concluded that the licensee has provided adequate justification 1
l l:
v e
ee m
- ~ -, -
v e,
kh t
}
i i
for the documented deviations from the requirements of Standard 1
Review Plan 3.9.3.
The instability approach in the analyses of the RCS supports, CEDM.
and ECCS piping is acceptable since it complies with the ASME Code,Section III, Appendix F guidelines.
Determination of the total stresses 'in the core barrel resulting from the asymmetric depressurization using decoupled beam and shell modes.is acceptable since this procedure has been shown to be mathe-matica11y exact. for linear analyses.
Analysis of the ECCS piping is acceptable based on the bounding l
analysis performed by the licensee. This analysis consisted of a dynamic analysis of the most highly stressed ECCS lines for motion of the ECCS nozzles on the cold leg piping as determined i
i from the RCS dynamic system analysis.
i.
Acceptability of the shield wall analysis is based on the conser-l l
vatism employed in the structural model and the applied loads.
l L
L 1
l L
b f
w
!L..
3.3.2 Compliance with Accept 6nce Criteria l
Except for the fuel assembly grid design, the licensee's stress and/
er load evaluation of the reactor system components is acceptable I
to the staff. The criteria used in the evaluation are, in general, in agreement with industry standards and meet the acceptance criteria
~
outline in NUREG-0609. Although some exceptions to the outlined criteria occur, functionality of each analyzed reactor system component is demonstrated.
l The licensee has indicated (Reference 15) that the Asymmetric Blowdown 9
I Loads issue as it affects fuel assembly grid design would be resolved L
by demonstrating that leak-before-break assumptions are valid for the Palisades Plant. The licensee has also provided documentation (Refer-
[
L.
ence 16) of the leak-before-break evaluation. Appendix H of that sub-mittel verifies the applicability to the Palisades Plant.
l l
The specific concern raised at Palisades was that the fuel assembly grid design could not withstand the asymmetric blowdown loads. This t
issue is resolved by the referenced analysis and leaves seismic loads as the only remaining isn e on grid design. The Palisades fuel vendor j
has developed a new grid design for inclusion in ali reloads beginning with Reload M (Cycle 9). This new grid design will provide improved thermal performance and test results show it is a significantly stronger
[E
)
I n+
)
- 1 i
grid than the presently installed fuel assemblies.
Four demonstration assemblies, having this new grid design, have been loaded in the Palisades core at the last refueling outage with Reload L (Cycle 8).
Before the Reload M fuel is fabricated, the fuel vendor will have completed seismic analyses of the new grid design using the NUREG/CR-1833 methodology.
The results of the seismic analysis of the grids will be submitted by the licensee for review as soun as they are available.
These analyset will be reviewed by the staff and if found acceptable, the issue can then be closed.
The reactor vessel supports meet the ASME Code, Appendix F criteria based on 70% of the instability load.
Therefore, the component is acceptable.
l l
The licensee's stress and/or load evaluations of the reactor vessel inter-I nals, primary piping, CEDMs, and ECCS piping is acceptable since ASME Code, Appendix F criteris are me1..
l Acceptability of the steam generator support evaluation is based on the comparisor. of the calculated support loads to design loads, yield capac-ities, and test lads.
The LOCA results for the supports are well within their allowable limits.
- s. i Acceptance of the shield stall stress evaluation is based on the demonstration r
that the structural integrity of the shield wall would be maintained and therequirementsofAmericanConcreteInstituteBuildingCode(ACI-318) would be met.
i CONCLUSION With the exception of the concerns regarding the seismic grid design as discussed in Section 3.3.2, the licensee has provided reasonable evidence that the Palisades reactor system would withstand the effects of asymmetric LOCA loads and that the reacter could ba brought to a cold shutdown condition safely, f
y f
" 'f _
28 4
120.
J i-l 1
100. -
1 80.-
- 60. -
Z wS 40.-
u O
l
_9 e
E
- 20. -
t i
ow-x
.}
O.- )
Y j
-20. -
n.
..W r
I l
-40.
i.
0.0 0.5 1.0 1.5 2.0 2.5 3.0 Time (seconds) l Figure 1.
Hot leg break force time history.
' f, Ji
g
.. _. _~
i:- t
.)
29 f#..
"j 4
'00.j o
j 150.' -
l 4
)
l 100.-
I g,
50.
i y
'o i'
Q : _ :_
i y
L O
i
- q
-50._
I.
g.
i I
-100.
H.
- 1
- f-
-150.
i1 i ',
i
-200.
i i
0.0 0.5 1.0 1.5 2.0 M
M M
Time (seconds)
Figure 2.
Hot leg break moment time history.
1' W
.j r.__.
30 400.
j 300. -
]
g
- Z 200. -
4 1
1 100. -
l l
I Ni
'o 0.
y M
O2
~.
8 -100..
g-
_ x w
-200. -
l
-300..
k 1
-400.
0.0 0.2 0.4 0.6 0.8 1.0 1.2 Time (seconds)
Figure 3.
Cold leg break force time history.
$s&
A, ;..
W 4 z 31
',' 4-'
3 400.
i 300..
t 200..
100..
'o o-
}
x E
g l'
2, l
t
\\
-100. -
l.
u e
c
_ y
-200..
I v
-300..
N 0U
-400.
c 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Time (seconds)
Figure 4.
Cold leg break moment time history.
-...a......a--
i i
0 32 1
1 1
1
'1 1
t
/
' NCOE 34 3
/
i s.-
EL $3'
+
l
~...
.n' NC088 ll:.
30 TO 35
-o NOCES
= ro n l
l' STRAf.1 l
__ a_s_u.sna_r_e_r. ' 'u,,:
. L a..-
NCCts 13 TO ls l
t L to.c-f 6
1:
- ..? r
^
4 j
ll t L 1.s*
l.
poco ya
.!2 l
IL-0.564' g
R EA CTCR
'.1.3 l NCOES VE3SEL l
E L 3.5 l
3 2570 28 t
e..
- ..i..
oct n
i 1
1 Figure 5.
Steam generator compartment, elevation view.
l l
l.
i l
f I.
~ ~ ~ ~ ~ - ~ '
i 33 ELEVATICN 90.593' TO 63'~
ILEVAT10N 35.l* TO H.!I
- ?y in
- ..,.
m.?
l
- 0, 31, = =
==s 0
21 hll
)
I ~ 's I
f e
'Je, 25
% gI 3 6.
l
( 25 fl' s *:
j
,4
,ei, s,s l* C,14
' 'cy" i
s, q..
l.,t'.
. '...l..
ELEVATION 20.0* TO 35.5' 8LIVATION 2.5' TO 20.0'
\\-
i 18 13 14 it
, 'a cP ',,
,, n e,
. REACTOR i
/
8
's l
VEMEL j
17 15
,c,
. ; /.. /
- s$.I'7:7
... '#D ::
s
/
[
6 7
12
' I.'
O* s,,,
- ',.N
',[f
- 10 I
/ s ELEVATION -0.t937' TC 2.5'
./g,q,'s a
s N'
g N
STEAM
,7G EN E R ATOR f
1
$UCING 4
5 g
SASE
/
s.'
's,
,{:/s' L
O
' +l ELEVATION -3.5' TO -0.55:7'
.s..,
e
.'...i' 2
1
..s..
i. :.
Figure 6.
Steam generator compartment, plan views.
)
i
ma-
+
e.
,,m_,a 34 9
h A
E 8.
\\_
e i:
c 4
i W
J o
I
> ))
U o
1
't>
jJP J
g
+----~n--
-~n-,.~-----,-,,--n--,~v---
,---r,v,-,----~~-w-were w
.e
+
e~
o--,+,w-
l i
2.
N 4
.l) p ea e
J x
a g.
f 5
5 5
=
=
m v
t 4
i o
m' I
6 C1c 4>
Ci W
_ L c
3 I
\\ \\,
i-1 G,
Y v
~,
7
~36 1
o l
i 4
Jk 4H i
HD t
I.
Upper Guide o
999,-
Structure O Core N
Support L.
Barrel 1
L
[
J I
4 t
p g2 O
O t'p L
Reactor 4
o Core Vessel Shroud Fuel 4
O o
o e
O
.O.
(
r-a m
ff A
26 I;
O Lower Support Structure i.
Figure 9.
RCS reduced internals model, t
I 1
J i-.-
37 l
ro:
O es b
e o
C i
4)
)
o e
c' 3
0 5
3 i
r 4)
SG Tube Shell Bundle 1
I L
a Hot Leg P
A*
r v
Figure 10. Detailed steam generator model, i
- - - - - - - - - - - ~ ~ - ~ ~ ~ ~ ~ ~ - ~ - ~ ~ ' ~ ' ~ ' - ~ ~ ~ ~ ~ ~ ~
~
ea
'C 38 9
o d
.4 s-w w
U V.%
zQ\\
e ss e /,,
s) d s
k/ /
g ldE
./
e M
\\. ts V
I if h
Nw y.
-l!
m _.: _ -,.
,i,
u.
- I,:n' i
~N
- ...a s
8
\\ \\}. e,# /*
- ms y
8 3
g
- f..
~
5 i**
.a
~
L g,
E g1 m
=
~
l ii 1
i s
E Q
=-
O O
P E
w
.s
39 s
/ /\\ /
N
//
V/'
- >\\
//
/ \\/ /
N
=
j?
h, 'l i 5
//\\/
\\
1 F
I s
a
,,A 5
E j
g g
i a
c' d *
( i.i i
E k /\\\\
\\'V\\
JV
\\A \\\\
\\A \\
/
= '! I' 'i 8
I
\\\\
- /\\ \\ \\
I i
f'l
,\\1
\\
i\\
1 9
W\\
8 t
l l
8 I-I II I
e I
N
\\
l l
I lt 2
l l
Figure 12. Reactor vessel support model.
l l
l' 1
40 7800. =
t t t 7200. -
2 6600. -
6000.-
5400. -
0.7 Pult = 5180 kips 4800. -
ll 2
h a
{
6 4200..
1 '
8 o
g u
o 3600. -J[
3352 kips RPV inlet nozzle break 3000.. o o
2400. - o 1800.
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 Displacement in stop direction (inches)
Figure 13. Cold leg support foot load-deflection relationship.
~
~
~
l n
41 l
I
.i a,
4800.
4500. -
i i'
4200..
3900. -
.I
. l.
t l
3600."
3300. -
0.7 P
= 3318 kips ult h
f i
2000..
3130 kips i
l RPV inlet nozzle break 1
C.
lI o
n 2700,-
l l
2400..
o 2100. -
o 1800.
0.0' 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0
)
Displacement in stop direction (inches) l' Figure 14. Hot leg support foot load-deflection relationship.
)
~
l
.,,,,-n-,w n
wc~--
w~-*
--r-'-*~="""''#"
' " * ~ ' ' ' ' ' ' ' ' " ' ~ ~ ~ " ~ " ' " " * * ^ ^
-. - ~
. e i
i I#
42 l
i Frietion l
Nonlinger it... Sprinc + osp j
si Hysteresis j
Rotational a
I C
Top UGS 1 [.. c.._
6 cylinder O
M
> RPV 4\\
ledge m
=
l o
o o
o CSB upper i
flange o
j o
o o
o I
o o( I o o
CEA se I
shrouds i
Outlet
[
o o
o o
nozzles
,e CSB l
o 1
o o
o o
FAP
,^t.
o l
H }--dHH6 4Hl l }--d >
0 i
l H )-<HHb Hj -
lH i o
4 o
tH }--MHH(
Hj "
{ f-o t
o--iHHHi 4Hl jf-d >
0 l
d!'
IH Fuel '
b i
.j P
o 4H }--d>-4H>
Hj' -JH I
i Core HHHHi throud <qj:
jH i o
i o
<HHHHu e
Hl>
IH l
0 tHHHHi 4Hl' j }--M o
CSP t
o f hY,'.
f LII Snubbers i
e Figure 15. Lateral internals model.
h
43 l
i Espansioi y
Compensating j
J, Ring g,
j US$ Flange t
l 3
i
/
l l Linear Spring d.
C$3 Flange 7 q
t Non-linear t
f j /
l PV
$pring o
W M9e i
I 1
Axial Gap 7
o 0
RPV Ledge l
, t qpFrictionElement if i
o i
15 i
i
- f i
CEA o
g E
i t 1 '
o o
CSB l
ll REV
/,
tSupport
' t l >
0 l
O i
A i
7 jl b
i N
o j
4
/
w e
t.,'
Reactor F
Guide j
Vessel a
o ik JP Tuces O
O j
((
i ras,i 1,
It:5 it Core P
4i Shrouc o 1;
i g" Tie o
l
=
i g
D 1 ;
lC Rod 8
{M 0
tolts 0 q
e 16
,f O
-.=
r l ;
Lower o
Support Structure o Cs3 L wer Flange m
1 Figure 16. Axial internals model.
k F~ =
44
- 4. W n
74.875" R 2"
104.25 1
)
75.125" R 5.,
318.50
.i e
I
-1.5" 140.75
[
(
i i
t i
l a
b 86 r
72.375" R we gg
~~
n
_ g.,
74.375" R 36.00
?
r Q
P o
Figure 17. CSB axisymetric vibration model.
1
F 1
_ _. _. = =
a d
45 4
o o
2 g
M1.75" 1
e i
r 74" R IJ" --e 2tSJ"
)
I e
i l
i o
a n
e t
e 121.25" 3
e y.
74"R 1.75" -e a
l g@
i j
n 5.5"
(
i i
r73.5" A i
2.25" - * *
=-
r i;
\\
_ 9.m..
t 4" 4 o
i Figure 18. CSB axisy metric stability model.
r
?
- _ - - - - - - - ' ' ~ ~ ~ ~
f a
e H
w
(' av 1
1 i l 4 9 il l1 g
$U Nx i
f c,
r
)
I, 4>
Y h
p I
i
, 3 9, Ects pip D9 *
~
.....____.._.m.,
4 o
41 1
-I 2
=..
1'
~* ] }T1 9
2R n
=
t-
- 41. E5&EE g
> 4 444. }
t' 3: 65..:
2:
5
-i ecos g q-
$! (. 3-e!fff. ~a.
l < <.,5., f X
..5 5
I
<=
- i 5 ' C*
3.:,.""2.R.
%.33.%. S. R. 2. ", R. ".
C.;.
EE..
Ell
\\
hk.22 22 2,5 4 4 i
l I. Y a s12 33333h Shth saaama smaa
..."o..k.a. - -
r.
87 g~.
--,"- -#f..=g h,
'i R
!E-it $$
5 su u
c.
t t
T
,th && 22 2
i
....a.
, me 4.
4 4,
s Si 121331 tith e
- e. 4e 4 4 4.s e...s 4 a
5 7
- c. e l
er-~.l ~.-,.
-= 5eg g; g-2,!
- >i c"
- SSR; R
- st C"
a,5 e
o
- e. ~
.s, 5
I e
5 5
.=....
v
- 5 p
I, 6c.
-y
'{#i
't j 12 :
e 1
i a. :1,'h.8t a'*
g i 2E i
8 t i', C, 2
C.::2 g'
.t.a
- j r
- e
.s r
-~-,
t gh-.,~9-,bE w
s
.y
=$
b EIa I
b 8
FFgF e.
}
g s
y a
s a
a sa
.a-a o
e E
3 6
e e
E i
2 l
6 4*
g
.T 6
6 8
t D
3 6
w o
a e
ip d
.t
.?
3 I
I. S.
2 s
e 8
m
+
w h
t
,,,-,_..--,.,,,---,--.._,,.,--s
N j
o c
I REFERENCES I
1.
" Reactor Coolant System Asymetric Loads Evaluation Program, Final Report, Calvert Cliffs 1 and 2. Fort Calhoun, Millstone 2, Palisades,"
Vol. 1, 2, and 3. Combustion Engineering, Inc., June 30, 1980.
I 2.
" Response to Questions on the Reactor Coolant System Asymmetric Loads Evaluation Program, Final Report " Vol. 1, 2, and 3. Combustion
]
1 Engineering, Inc., June 1981.
(Volume 3 Proprietary).
3.
"Asynynetric Blowdown Loads on PWR Primary Systems," NUREG-0609 U.S.
Nuclear Regulatory Commission. Office of Nuclear Reactor Regulation, January 1981.
t 4.
" Method for the Analysis of Blowdown Induced Forces in a Reactor Vessel,"
CENPD-252-P. Combustion Engineering. Inc., December 1977. (Proprietary) 1 5.
"Hethod for the Analysis of Blowdown Induced Forces in a Reactor Vessel,"
l l
CENPD-252-P, Amendment 1-P, Combustion Engineering, Inc., August 1978.
(Proprietary) l i
6.
" Topical Report on Dynamic Analysis of Reactor Vessel Internals Under Loss-of-Coolant Accident Conditions with Application of Analysis to I
CE 800 MWe Class Reactors," CENPD-42 Combustion Engineering, Inc.
I l
(Proprietary) l
r j
n i 7.
J. C. Watkins, "Subcooled Blowdown Analysis for a Cmbustion Engineering 2570 MW Pressurized Water Reactor," RE-A-78-248. EG&G Idaho, Inc.,
November 1978.
i 8.
"CEFLASH.4A: A FORTRAN-IV Digital Computer Program for Reactor Blowdown f
Analysis (Modifications)," CENPD-133P, Supplement 2. Combustion Engineer-ing Inc., November 1978.
9.
"RELAP4-M006--A Computer Code for Transient Thermal-Hydraulic Analysis of Nuclear Reactors ar,d Related Systems," User's Manasal CDAP-TR-003, l
EG&G Idaho. Inc., January 1978.
10.
" Reactor Plant Subcompartment Analysis," CENPD-141, Levision 2 Combustion i
Engineering, Inc., Marsh 1978.
11.
" ICES-STRUDLII, The structural Design Language " Engineering User's Manual, First Edition, Massachusetts Institute of Technology, November 1968.
12.
" Design Basis Pipe Breaks for the Combustion Engineering Two Loop Reactor Coolant System " CENPD-16SA, Appendix A, Combustion Engineering, Inc.,
June 1977.
j i
13.
" MARC-CDC Nonlinear Finite Eiement Analysis Program." Control Data Corp-oration, Minneapolis, Minnesota, 1976.
t 6
(.
a
'O I
i,
- 14. American Society of Mechanical Engineers, "ASME Boiler and Pressure Yessel Code "Section III. Division 1, 1977 Edition.
13.
Letter from Consumers Power to NRC dated March 16, 1988, regarding the final disposition of the asymmetric blowdown loadi as it affects fuel
{
- desiga, 16.
Letter CE0G-87-662 from J. K. Gasper, Chairman of the CE Dwners Group to l
L James A. Norberg (NRC) submitting CEN-367 " Leak-Before-Break Evaluation of Primary Coolant Loop Piping in Combustion Engineering Designed Nuclear Steam Supply Systems", dated November 20, 1987.
l r
l L
t i
i l
l l
I T
- - -