ML19344D483
| ML19344D483 | |
| Person / Time | |
|---|---|
| Site: | Maine Yankee |
| Issue date: | 03/06/1980 |
| From: | Roudenko A, White R Maine Yankee |
| To: | |
| Shared Package | |
| ML19344D481 | List: |
| References | |
| NUDOCS 8003120466 | |
| Download: ML19344D483 (31) | |
Text
. - -. - -
..i.;
i.
gp n i
I EVALUATION OF LOCA-RELATED LOADINGS ON THE REA' TOR COOLANT SYSTEM COMPONENTS, C
SUPPORTS, AND PIPING AT MAINE YANKEE by 4
A. V. Roudenko R. E. White
+.
4 Y
1
+
i I
2 I,.
j..
?
l l
8008120 %;;g i
I ':
- +..,,
__..mm,
-,.,,m_m,4,._.m._
y_,_..m...m...,
..._.-..,.y-,-
..3.,,-v~..--r
,w--c,
DISCLAIMER OF RESPONSIBILITY _
This document.was prepared by Yankee Atomic Electric Company on This document is believed behalf of Maine Yankee Atomic Power Company.
ifically by Yankee Atomic to be completely true and. accurate for use spec priate Electric Company, Maine Yankee Atomic Power Company and/or the appro l
subdivisions within the Nuclear Regulatory Commission on y.
With regard to any unauthorized use whatsoever, Yankee Atomic i
- fficers, Electric Company, Maine Yankee Atomic Power Company and the r o r nty directors, agents and employees assume no liability nor make any war a document or to its or representation with respect to the contents of this accuracy or completeness. 1
ABSTRACT This report,is prepared on behalf of Maine Yankee Atomic Power Company, in response to an NRC requirement for licensee evaluation of the impact of asymmetric LOCA loads on plant safety. The purpose of this report is to demonstrate that asymmetric LOCA loads do not result in any violation of Maine Yankee FSAR design bases. Pipe rupture restraints were installed at Maine Yankee to greatly diminish asymmetric LOCA load severity. This report describes the evaluation program in detail. Original design analysis results and safety margins, results of analyses performed for similar plants which were applicable to Maine Yankee; and as necessary bounding, conservative thermal hydraulic and structural analyses were employed in identifying existing design margin for accommodation of asymmetric LOCA loads. Results of conservative evaluations show that as a result of asymmetric LOCA loads, the structural integrity of essential primary components, component supports, and structures is not compromised, nor is the accident conseauence mitigation ability of plant safety systems changed from that described in the Maine Yankee FSAR.
-iii-i t
J
'~
v TABLE OF CONTENTS _
fage_
11 DISCLAIMER OF RESPONSIBILITY.............................
iii ABSTRACT.................................................
iv TABLE OF C0NTENTS........................................
v LIST OF TABLES...........................................
vi LIST OF FIGURES..........................................
1 INTRODUCTION.............................................
1.0 3
REACTOR PRESSURE VESSEL LOAD DEFINITION..................
2.0 METHOD OF EVALUATING THE ADDITIONAL ASYMMETRIC 18 LOCA LOADS ON COMPONENT SUPPORTS....................
3.0 26 METHOD OF EVALUATING SUB-COMPARTMENT BREAKS........
4.0 EVALUATION OF ECCS AND BRANCH PIPING
(*).................
5.0 CONTROL ROD DRIVES EVALUATION
(*)........................
6.0 EV ALUATION OF INTERNALS AND NUCLEAR FUEL..............
7.0 PIPE RUPTURE RESTRAINTS
(*)...............................
8.0 f
RESULTS OF ANALYSES(*)..................................
f 9.0 I
SUMMARY
OF CONCLUSIONS(*)...............................
l
10.0 REFERENCES
11.0
(*) In course cf preparation.
-iv-l l
4 i
I LIST OF TABLES i
P, alge, Number Title 8
Comparison of Operating Conditions and Dimensions.....
2.1-Shield Tank 3.1
. Original Design Margin for Neutron 22 Including Original L0CA...............................
Original Design Margin for Neutron Shield Tank 3.2 23 Excluding Original L0CA...............................
4.1 Originator Design Margin for Steam Generators 28 Including Original L0CA...............................
D 8
s 1
4 f
e a
i '
t O
a 4
4.
e
-V-n
LIST OF FIGURES P_ag Number Title Maine Yankee Reactor Cavity Elevation.................
2.1 Details of Lower Cavity Maine Yankee Reactor 2.2
.Cavity................................................
Maine Yankee Reactor Cavity Details of Upper 2.3 Cavity - P1an.........................................
Maine Yankee Reactor Cavity Subcompartment Model 2.4 Schematic Plan (Upper Cavity).........................
Maine Yankee Reactor Cavity Subcompartment Model 2.5 Schematic Section.....................................
Schematic Representation of the NASTRAN Model of the Reactor Shell, Internals and Reactor Supports.....
2.6 Structural Model of Cold Leg Piping and Reactor 2.7 Coolant Pump Support..................................
Structural Model of Hot Leg Piping, Steam Gcnerator 2.8
& S.G. Support........................................
Model of Neutron Shield Tank Structure and 2.9 Rodo-Foam.............................................
Neutron Shield Tank...................................
3.1 Pressure Vessel Supports......................
3.2 Reactot Steam Generator Lower Support Plate...................
4.1 Lower Support Plate - Bottom View.....................
4.2 Support Plate on Concrete Base........................
4.3 i
l
-vi-
4
1.0 INTRODUCTION
l In 1975 the NRC'was informed that design loads on reactor pressure d reactor vessel (RPV) supports for North Anna Units I and 2 from postulate had been coolant system (RCS) pipe ruptures inside the reactor cavity Since that time, evaluation of damage potential due to underestimated.
hypothetical pipe ruptures to RPV supports, other RCS component supports ECCS piping, reactor internals and fuel, and control rod drive mechanisms was required by the NRC on a generic basis.[1]
Evaluations were performed for Maine Yankee (MY), consisting of i
f MY a pipe rupture event probability study and structural evaluat on o Based upon their reactor pressure vessel supports subjected to LOCA loads.
d not review, the NRC concluded that the results of probability studies di ient establish an acceptable basis for long term operation without trans 4
Positive results of probability studies did, however, load risk assessment.
i h tima establish a basis for continued short-term operation during wh c The RPV support additional evaluation of asymmetric LOCA loads was required.
ii l design evaluation for Maine Yankee showed that loads in excess of or g na Additional investigation was therefore considered loads could result.
necessary to demonstrate acceptability.
In mid-1979 a program of action toward resolution of the LOCA load Two simultaneous program activities issue was presented to the NRC staff.
evaluation of structural integrity were identified and have been pursued:
d design, of the reactor coolant system subjected to asymmetric LOCA loads; an f abrication and installation of pipe rupture restraints.
3 Pipe rupture restraints limit the area available for fluid blowdown This breaks inside the primary shield wall.
- into the reactor cavity due to Break limiting resu.lts in a substantial reduction of asymmetric LOCA loads.
restraints are installed at each RPV nozzle location in each loop, k
appropriately sized to withstand loads resulting from postulated pipe brea s The pipe rupture without compomising the integirty of existing structures.
Items addressed restraint design is discussed in a subsequent section.
include design basis, limiting constraints and method of thermal and Pipe rupture restraints have no effect on the RCS structural analysis.
during normal operation and do not adversely affect the ability to perform required inservice inspection.
The structural evaluation portion of the program makes extensive use of original design calculations and identification of existing safety Where possible, margins to accommodate additional asymmetric LOCA loads.
load definition and determination of system integrity is based upon f
comparhtive analyses with results of evaluations performed for plants o similar vintage and design, subjected to identical loading conditions.
As necessary, limited additional analytical work was performed to provide a basis for evaluation of structural integrity and verification of This simplified evaluation program, incorporating conservative assumptions.
the break area limiting effects of plant modifications, pipe rupture f
restraints, was met with general NRC staff agreement.[2]
l i
W h
REACTOR PRESSURE VESSEL LOAD DEFINITION 2.0 Asymmetric LOCA loadings on the reactor pressure vessel consist These are (a) of two simultaneously occurring transient load components.
i tr external asymmetric load (EAL) resulting from time-varying asymnetr c reac o cavity pressurization, and (b) internal asymmetric load (IAL) which is caused The by depressurization wave propagation internal to the reactor shell.
latter total IAL includes the LOCA thrust load component originally considered in the Maine Yankee design analysis, while the former EAL occurs Significant strain energy release only for breaks inside the reactor cavity.
load components are incorporated in the IAL forcing function applied to This is accomplished by considering the initial pressure unbalance the RPV.
This at the time of the rupture (with yet undeveloped blowdown flow).
pressure unbalance as defined in the hydraulic analysis is applied to the Local " compatibility of displacement" shear RPV and is discussed later.
and moment release load components at the ruptured nozzle, whose magnitudes are inconsequential, are neglected. For postulated pipe rupture outside ih the primary shield wall, IAL of diminished magnitude exists together w t The compartment pressurization effects, as discussed in a later section.
governing loading condition considered in this analysis is that due to a cold leg break inside the reactor cavity, with limited vent area due to pipe rupture restraints.
Critical parameters.which determine RPV support reaction load magnitudes due to limited and full guillotine pipe rupture and resulting Break opening time is defined IAL are blowdown area and break opening time.
at the time required for piping to undergo that axial and lateral relative displacement which results in the maximum blowdown area for the pipe rupture ~
v
and restraint condition under consideration.
Studies have shown that for limited vent area pipe rupture inside the reactor cavity (typically requiring approximately 10 milliseconds (msec) to reach 100-200 in2), resulting RPV support loads due to IAL are nearly equal to those due to pipe rupture outside the reactor cavity reaching full guillotine break areas in approximately 30-36 msec.
By including the (limited) EAL with the former, inside reactor cavity break IAL, the resulting RPV support loads, due to the latter, full guillotine IAL, are conservatively enveloped for subsequent use in evaluation of system structural integrity.
The pipe rupture blowdown area considered in this evaluation was initially selected to assure that conservative component support reaction loads are developed. Therefore, the break area used was greater than that used for currently licensed plants (typically 1.0 ft2 ).
Break opening time was determined mechanistically by subjecting the ruptured piping to a load which was assumed to be fully developed within one millisecond; i.e., instantaneous rupture.
Several significant conservatisms must be noted.
(1) The computed break opening area is less than previously assumed and used in development of the applied loadings.
(2) The load applied to the ruptured piping for determination of the break area versus time relationship was equal to 1.0 PA, instantaneously developed and maintained throughout the transient.
In fact, there is a substantial diminution of the load and the resulting pipe motion is over predicted while break opening time is underestimated.
D The break time used in the hydraulics analyses correspond to the (3) mechanistic break opening which resulted in a smaller break area.
Therefore, for the break area used in developing the RPV support loads, a break opening time in excess of that predicted mechanistically, could be used with ample conservatism.
In summary, a conservatively large break area was used with a shorter than anticipated break opening time.
The EAL force time history was developed from mass and energy release data for limited break area LOCA using the Maine Yankee RELAP4/ MOD 3 Based upon the blowdown mass and energy release data, a reactor model.
6 compartment pressurization history was developed using the RELAP-4/ MO The Maine Yankee reactor cavity was simulated by a lumped computer code.
ih parameter model of subcompartment volumes linked by flow junctions w t This is a classical modeling technique used in flow properties.
Figure 2.1 shows the RPV shell, subcompartment pressurization analyses.
Figures 2.2 and RPV supports, Neutron Shield Tank and the reactor cavity.
The Maine 2.3 detail the lower and upper reactor cavity, respectively.
Yankee reactor cavity subcompartment model plan and elevation views Upper cavity subdivision respectively are shown in Figures 2.4 and 2.5.
li RPV modeling was subjected to sensitivity studies to assure that resu t ng Computed support reaction loads are computed in a conservative manner.d to obtain pressure differentials acting on RPV ' Projected Areas" were use Resulting EAL an integrated force time history acting on the RPV shell.
h l
ts, peaks computed for Maine Yankee have been compared to those of ot e i
have and accounting for break area and volumetric differences, compar sons r.
s shown close agreement.
The Mcine Yankee IAL was characterized by hydrodynamic forcing functions acting on the RPV shell and reactor internals. A thermal hydraulic analysis was performed simulating the RPV and internals by an internals model originally prepared for analysis of a Combustion Engineering 800Mw plant, (CE800Mw), whose internals dimensions and operating parameters are nearly identical to those of Maine Yankee. Comparative data is presented in Table 2.1.
Using the CE800Mw plant internals hydraulic model with Maine Yankee operating parameters:
approximate break opening times and limited break area, the internal forcing functions on the RPV
-shell and CSB were computed using a similar methodology to that used in development of the EAL; computing force time histories from transient pressure differentials acting on components' projected areas. The strain energy release effect due to pipe rupture is accounted for in the IAL.
This " classical" thrust load'(for which Maine Yankee was originally designed using an equivalent static load, amplified by a dynamic load factor) is considered in the IAL by including the projected area of the nozzle and subjecting that area to the differential pressure history developed in the hydraulics analysis. This results in an initial instantaneous value of 1.0P A o.
The CE800Mw plant is a two loop design with two cold legs and one hot leg per loop, whereas Maine Yankee is a three loop plant with one hot leg and one cold leg per loop. Otherwise, geometric similarities between Maine Yankee and the CE800Mw reactor pressure vessel and core support barrel i
I are evident from an examination of their respective drawings. Radii, heights, component thicknesses and relative nozzle locations are very i
6 As Six nozzles are located sixty degrees apart in both plants.
similar.
the two plant a result of RPV inlet.and outlet location differences between d at desi,gns, some normal pressure distribution differences are anticipate However the pressure transient within the vessel the nozzle elevation.
It is the after the LOCA is classified as a wave propagation phenomenon.
ing factor propagation of the depressurization impulse that is the govern Therefore, considering similarities in determination of resulting loadings.
determines in reactor coolant system temperatures, since this effectively i tics, the relative density, fluid sonic velocity and choking character s d to be very two plants' LOCA pressure transient behavior is anticipate similar.
hird In the Maine Yankee three loop configuration, the " additional" t The hot leg is located directly opposite from a cold leg break location.
l remaining two hot legs, which may tend to obstruct the interna l
d depressurization wave are circumferentially equidistant from the pos For CE800Mw plant cold leg breaks, two hot leg break, sixty degrees away.
h depressurization obstructions also exist, however one is at sixty and The amount of decompression wave is at 120 degrees from the break.
be approximately propagation retardation in the two plant configurations cand a series compared using simplified equations based on acoustic theory an It should be noted that in the of simple one dimensional wave models.
is made to predict absolute wave discussion which follows, no attempt Instead, a semi-quantitative characteristics or hydrodynamic behavior.
i sion in one assessment of relative obstruction to rarefaction wave transm s dimension, based upon area and source / obstruction proximity dif ference is sought.
1 Given an arbitrary rarefaction wave intensity, P, and assuming o
that rarefaction wave propagation between the break location and the downcomer annulus region at the nozzle elevation has already occurred, d i subsequent circumferential depressurization is observed from the stan po nt Based of rarefaction transmission ability in the presence of obstructions.
d upon the CSB internal volume height dimensions relative to the rupture i
nozzle, CSB outlet nozzle diameter and downcomer annulus width, reflect on l
coefficients for the interface region areas were computed for each p ant Using the reflection coefficients, transmission factors configuration.
h were computed to facilitate determination of pressure wave intensity c anges.
Locally, a 20% difference in transmission factors per obstruction was observed, with the Maine Yankee system being less restrictive.
Two wave models were considered, with and without wave intensity In both cases a series of diminution as a function of distance traveled.
h d by partially closed, then partially open interface regions was approac e a single rarefaction wave of initial ~ intensity P, with pressure intensity o
factors.
changes occurring as determined by the aforementioned transmission Close examination of the variations which occur show that the effects are. highly localized and extremely short-lived.. Repeated wave when intensity changes due to obstructions " advancing" and " passing ~
observed without regard for intensity changes over distance, reveal Pressure intensity.
negligible overall intensity variations, Po = constant.
Those differences between the two plant designs are generally less than 1%.
4 differences which do occur do so on a time scale which results in RPV sup reaction load differences which are well outside the sensitivity range Therefore, the internal loadings considered significant in this analysis. -.
&~
~
i l transmitted to the RPV supports due to limited break IAL are conservat ve y characterized by the hydraulic forcing functions obtained using the CE800Mw transient parameters.
internals model with Maine Yankee Using this fluid forcing function IAL was calculated assuming flows A finite element and pressures were independent of structural oscillations.
program is used to account for the coupling that exists between structures This fluid-and adjacent fluids undergoing relative, unsteady motion.
structure interaction is included in the structural model as an "added m matrix". [3]
Hydraulic forcing functions applied to the RPV shell and internals result in linear (direct asymmetric pressure) and non-linear (reaction loads through gapped interfaces) RPV support horizontal reaction force components.
Parametric studies have, demonstrated that the non-linear gap effects do ih not substantially affect the largely linear response of the internals w t Vertical response is entirely regard to resulting RPV support loading.
As a result, forcing functions linear as no non-linear interfaces occur.
on the RPV (EAL) and; RPV and internals (IAL) could be computed separately osed in the structural model to develop total RPV and subsequently superimg support reaction loads. [4]
Reaction loads at six RPV supports were computed by dynamic, non-linear finite element analysis, using a structural model developed to conservatively predict maximum loadings on the Maine Yankee RPV supports.
The structural model simulates the mass and stiffness characteristics d
the reactor pressure vessel,~RPV supports attached to the Neutron Shiel Tank via gapped interfaces, Core Support Barrel (CSB) and RPV internals;
_9
4 The RPV and CSB were modelled as and RCS piping and component supports.
d )
concentric simple elastic beams (neglecting shell deformation mo es i
This connected through a gapped interface at the RPV snubber elevat on.
CE800Mw RPV support modelling technique is similar to that used for the h RPV reaction analysis except for the arrangement and properties of t e Loop piping and component support effects are included.
support system.
800Mw plant RPV An approximate representation, similar to those used in CEl and support reaction load analyses, is used to simulate reactor interna s A schematic representation of the fuel stiffness and mass properties.
structural model is shown in Figure 2.6.
Linear-elastic finite element models were developed to simulate d Reactor Coolant the stiffness characteristics of (1) Cold Leg Piping an and (3) Neutrpn (2) Hot Leg Piping and Steam Generator Support, Pump Support, h Primary Shield Tank and interface Rodo Foam between the N.S. Tank a Structural models are schematically shown in Figures 2.7, Shield Wall.
Static unit displacements were applied at points of attachment 2.8 and 2.9.
between the above loop stiffness and N.S.T. models and the non-linear Each degree of freedom at dynamics model of the RPV and its internals.
ith all other such points was individually subjected to a unit displacement w J
Resulting degrees of freedom at that node held at zero magnitude.
ised equilibrium loads at the displaced node were tabulated and compr h RPV and stiffness matrices which were subsequently assembled into t e internals model.
2 0
f :
~
e
==,t The forcing functions due to EAL' and IAL, computed as previously described and combined by direct super position were applied to the A non-linear dynamic analysis, incorporating fluid structural model.
f r ed structure interaction in the form of an added mass matrix, was per o m This methodology is similar to that using the NASTRAN computer program.
Results of the analysis include RPV support used in CE800Mw plant analysis.
l tion reaction force time _ history and corresponding displacement and acce era time histories at each support.
e l
4 4
4 TABLE 2.1 Comparison of Operating Conditions and Dimensions CE 2-Loop Maine Yankee 800Mwt plant Data Operation Conditions:
2250 psia 2250 psia Normal operation pressure 6020F 592oF Normal operation temperature (H.L.)
5500F 543oF Normal operation temperature (D.L.)
6 lbs/hr 122 x 106 lbs/hr 129.6 x 10 Total flow rate 1200F 1200F Ambient temperature 2500 psia 2500 psia Design Pressure Dimensions:
33 1/2 in 42 in H.L.
I.D.
33 1/2 in 30 in D.L.
I.D.
172 in 172 in R.V.
I.D.
8 5/8 in 8 5/8 in RV thickness 141 3/4, 148, 2 1/2 141 3/4, 148, 2 1/2 CSB upper section Ht, ID & th 121 1/4, 148, 1 3/4 121 1/4, 148, 1 3/4 CSB center 65 1/2, 148, 2 1/4 62 1/2, 147, 2 1/4 t
CSB lower RV nozzle support interface 1/8" 1/8" tangential gap --
~.
4
- +
METHOD OF EVALUATING ADDITIONAL ASYKHETRIC LOCA LOADS
' 3.0 Original Neutron Shield Tank, Reactor Coolant Pump Support and Steam Generator Support design analyses are used as the basis for developing new total loads at the Reactor Pressure Vessel Supports and other Reactor Coolant These analyses were based on design forces, System component supports.
thermal expansion, seismic, and LOCAs at several points including deadweight, in the RCS, originally developed by the NSSS vendor, Combustion Engineering.
Total loads were obtained by direct combination of the above Stress calculations were contributing forces at each support point.
performed generally by conservative hand calculations on a linear elastic Stresses resulting from all load combinations including those basis.
containing LOCA and seismic were limited to values less than the material Original LOCA reaction loads were based on yield strength, per the FSAR.
applied loads equal in magnitude to twice the normal pressure times area (2PA) piping thrust load, applied at the pipe rupture location, accounting for the assumed instantaneous nature of the jet thrust force which was reacted at the component supports.
The high degree of conservatism in the original calculations is Methods, assumptions, conservatisms, utilized in the current evaluations.
and design loads are kept intact with the exception of the LOCA loads.
As described in Section 2.0, a bounding thermal hydraulic analysis was performed which provided not only_ reactor cavity asymmetric pressure loads, This analysis and associated but also more realistic loop hydraulic forces.
structural analyses considering the stiffnesses of loops, supports and components, are based on current mass and _ energy release data for limited --
i Loads resulting from this analysis are used in the current break LOCA.
evaluation.
Thus, the current evaluation of RCS primary component supports thermal, and seismic loads with basically combines the original deadweight, Calculational methods utilized new LOCA loads obtained by detailed analysis.
to evaluate stresses are identical to those in the original calculations as long as the conservatisms in those calculations do not unnecessarily These conservatisms are discussed in Section 3.1.
1 effect support adequacy.
Each component in the RCS is evaluated somewhat differently according For instance, the steam to the magnitude of the additional asymmetric load.
generator compartments have much more free volume available for blowdown Since the magnitude of the external asymmetric pressure than the RPV cavity.
load is very dependent on free volumes, the forces across the Steam Generator are much less than across the Reactor Vessel. Thus, less detailed analysis of LOCA related loads in the sub-compartments was required to demonstrate The following sections describe the means of structural integrity.
evaluation of each RCS component.
3.1 Neutron Shield Tank Evaluation Data from the original calculations, including high stress points, safety factors, allowable stress limits, and contributing loads has been
(
tabulated so that critical points could be identified and location of worst As discussed New LOCA loads are included in these tables.
break determined.
in Section 2.0, a cold leg break within the RV cavity has been determined j
to produce maximum IAL & EAL and is considered here.
i With all contributing loads tabulated, the evaluation of support adequacy is performed as follows:
Combine new LOCA loads with Deadweight, Thermal and Seismic (SSE) a.
Calculate stresses and compare to existing allowables.
loads.
These calculations are performed according to the analytical methods used in the original analysis.
If adequacy cannot be shown at all locations, remove some of the b.
conservatisms in the original analyses and re-evaluate.
Repeat this procesr using good engineering judgment until adequacy c.
can be shown at all points retaining sufficient conservatism.
Some of the conservatisms present in the original analysis which,
can be removed, if necessary, without extensive analytical work are:
Per NUREG-0484 LOCA
- 1. Direct addition of LOCA and Seismic loads.
and SSE Loads can be combined by SRSS.
ii. The 30% additional deflection allowed for the NST-Rodo-Foam a the deflection which results from calculated forces, All allowable stresses, even for accident and faulted conditions, iii.
are limited to values less than S. Per ASME Code rules, faulted y
condition stress limits could be used, resulting in stresses computed on an elastic basis, exceeding the material yield i
strength.
,- s Following is a summary of the original safety factors at the NST thermal expansion, supports for the direct combination of Deadweight,
- Also, se,ismic, and a cold leg LOCA (original, based on 2PA jet thrust).
This a corresponding table of safety factors without LOCA is compiled.
comparison illustrates the significance of the LOCA load at each location.
The tabulated Factors of Safety are based upon allowable stress y and 0.54Sy for normal and shear stress components, limits of 0.9s Safety Factors in Table 3.2 reflect stresses resulting from respectively.
The maximum stress thermal expansion, and maximum seismic.
Deadweight, locations are shown in Figures 3.1 and 3.2.
i
)
- L
.o TABLE 3.1 ORIGINAL DESIGN MARGIN FOR NEUTRON SHIELD TANK INCLUDING ORIGINAL LOCA Safety Description of Location Where Factor Stress Analysis Was Performed 10.70 SHELL-(BASE) - Shear (4b) 1.84 MAX (4a) 2.20
~
ANCHOR STUD. Tensile 4.35
- Stud Thread Shear 42.00 4
- Nut Thread Shear 1.83 SHEAR KEY - Bearing 7.34 REACTOR SUPPORT STUDS - Tensile 1.69 REACTOR SUPPORT SHEAR KEY - Shear 1.51 RETAINING BLOCK - Bearing 1.74
- Capscrews Shear 1.93 3
- Capscrews Tensile 5.42
- Screw Thread Shear 1.62
- Thread Shear o
4.,
- x TABLE 3.2 ORIGINAL DESIGN MARGIN FOR NEUTRON SHIELD TANK EXCLUDING ORIGINAL LOCA Safety Description of Location Where Factor Stress Analysis Was Performed 40.00 SHELL (BASE) - Shear (4b) 2.30 MAX (4a) 59 ANCHOR STUD - Tensile
>59
- Stud Thread Shear
>59
- Nut Thread Shear 1.93 SHEAR KEY - Bearing 17.20 REACTOR SUPPORT STUDS - Tensile 7.62 REACTOR SUPPORT SHEAR KEY - Shear 6.84 RETAINING BLOCK - Bearing 2.41
- Capscrews Shear 2.20
- Capscrews Tensile 8.30
- Screw Thread Shear 2.54
- Thread Shear i.
3 EVALUATION OF SUB-COMPARTMENT BREAKS _
4.0 4.1 Steam Generator Supports When there is a break in either hot leg or cold leg outside the reactor cavity, or crossover leg, a differential pressure occurs across The magnitude of the P is the Steam Generator and Reactor Coolant Pump.
dependent on the proximity of the components to the shield walls, free vol of the sub-compartments, vent area from the compartments and mass and energy Maine Yankee primary components should release for the specific break.
not see significant asymmetric loads in sub-compartments since, The Steam Generators and Pumps are located about midway in their a.
compartments (typically 39 feet from the RV centerline) 1 There is more than adequate vent area from the compartments at an b.
elevation of 46 feet.
The Steam Generator Compartments are relativ sly large with significant c.
free volume.
)
Since small asymmetric loads are expected, a detailed thermal-hydraulic fluid analysis was not performed, but rather the results of a f
detailed analyses performed for sub-compartment breaks in a plant of similar 1
configuration are scaled considering free volumes, vent areas and peak flow rates from the mass and energy release.
1 Support loads resulting from this scaling are conservatively combine to develop the total with the original LOCA loads based on 2PA jet thrust, LOCA load on the Particular support.
4 Stresses at the support points are evaluated in the same manner It is expected that adequacy can be shown in this
,as the RPV supports.
co'nsistent manner without removing any of the conservatism in the existing calculations since'a 25% increase in load is required to influence original Original safety factors for the Steam Generator supports stress margins.
and the locations of maximum stress. points are presented in Table 4.1 and Figures 4.1, 4.2, and 4.3.
4.2 Reactor Coolant Pump Supports The RC pumps are located approximately in the center of the Steam Generator compartment, well away from shield walls, with considerable free The volume below the pump casing and to the sides of the pump support.
pump supports were originally designed for a 2PA jet thrust in combination with other design loads including the uplift pressure force resulting from Maximum stress at any point on the supports results cavity pressurization.
in a factor of safety of 1.13 produced by an applied load of about 7.2 A simple analysis of an asymmetric pressure distribution million pounds.
P = 1000 psi would have to occur in order across the pump indicates that a to overstress the pump supports.
Since Ps of perhaps 0 to 5 psi are expected across the pump based on other analyses of similar plants, adequacy of the coolant pump supports for the additional asymmetric loads is evident.
4.3 Subcompartment Walls Differential pressure loadings on subcompartment walls occur due The magnitudes of to postulated pipe rupture outside the reactor cavity.
i the pressure differentials have been conservatively bounded by comparisons -
4 O
of geometry and anticipated mass and energy release for Maine Yankee and s
The subcompartment walls are capable of withstanding similar plants.
differential pressure loading substantially greater than those anticipated.
Therefore, subcompartment walls' structural adequacy is not adversely affected by asymmetric LOCA loads at Maine Yankee.
9 4
l 2
l I f
~
t e
TABLE 4.1_
ORIGINAL DESIGN MARGIN FOR STEAM CENERATOR SUPPORTS INCLUDING ORIGINAL LOCA Safety
-- Description of Location Where Factor Stress Analysis Was Performed 1.91 (1) 4" DIAMETER ANCHOR BOLTS 3.47 4" DIAMETER NUT - Shear 2.69 (2)
ANCHOR BOLT WASHER - Compressive 9.90 (3)
(4)
CONCRETE BASE - Bearing 2.03 (5) 3" DIAMETER STUDS - Tensile 2.61 (6) 3" DIAMETER HEAVY NUT - Shear 2.45 (7) 30" SHEAR KEY - Shear 3.44 (8) 30" SHEAR KEY - Bending 5.70 (9)
SUPPORT PLATE - Compressive 1.26 (10) 18" SHEAR KEY - Shear 1.30 (11) 18" SHEAR KEY - Bending 2.90 (12) SUPPORT PLATE - Compressive 7.30 (13) CONCRETE - Compressive Fx 1.98 (14) CONCRETE - Compressive FZ 6.70 SHEAR KEY BOTTOM WELDS - Shear 4.30 (15)
(16) LOWER SUPPORT - Tensile i
l
> t C
e f
EVALUATION OF INTERNALS AND NUCLEAR FUEL 7.0 As discussed previously, utilization of results of original and existing plant analyses in establishing a basis for concluding structural
] This adequacy of components was met with general NRC staff approval. [2 approach is used for evaluation and determination of acceptability of asymmetric LOCA-related loading on reactor internals and nuclear fuel.
A blowdown analysis simulating a full double-ended guillotine rupture at the inlet nozzle was performed for Maine Yankee prior to development The Maine Yankee and utilization of the models discussed in Section 2.0.
reactor coolant system, including loop piping, vessel, and internals was The methodology used in simulated using the WATERRAMMER computer code.
the blowdown analysis as well as selection of upper bound pipe rupture condit.lon with regard to internals loadings were as described in Reference 15}.
Results of the blowdown analysis above, transient differential pressures acting across the CSB supported by a " rigid" RPV, were applie l
to a detailed non-linear structural model of the Maine Yankee interna s.
Dynamic response analysis revealed that maximum spacer grid impact load resulting from the CSB loadings were less than the minimum grid impact These results indicate that under the influence of the crushing strength.
most severe internals LOCA loading condition in that analysis, (1) no loss ble of structural adequacy occurs, and (2) fuel rods are maintained in a coola array.
Using the hydraulic model discussed in Section 2.0, maximum
. ',2\\ -
~
o differential pressures acting across the Core Support Barrel in the region In subsequent structural analysis, of the vessel nozzles were computed.
These maximum loads due to IAL and EAL on the CSB were developed.
represented two types of CSB loadings, (1) differential pressure without regard for vessel /CSB interaction, and (2) total load due to EAL and IAL A direct comparison of maximum considering component interaction.
differential pressure loadings across the CSB can be made from the results from differential of each analysis. Further, total loads on the CSB resulting 3AL may be compared with maximum pressure and reactor vessel response t loads due to the differential pressure obtained in the WATERHAMMER analys For CSB loadings bounded by those originally obtained in the Maine Yankee internals analysis considering maximum LOCA loadings, the following The severity of internals response and magnitude of grid can be concluded.
impact loading are not anticipated to exceed those originally considere Therefore, the structural integrity of the internals, concluded based upon i
nchanged the original internals analysis considering LOCA loading, rema ns u since the original loading severity is not exceeded.
. ',Wm O
11.0 REFERENCES
1.
NRC letter to All'PWR Licensees, dated January 25, 1978.
NRC Memorandum from S.B. Hosford to L.C. Shao, " Summary of Meeting 2.
with Maine Yankee Atomic Power Company on October 18, 1979 to discuss their Asymmetric LOCA Loads Evaluation, in Bethesda, Maryland."
"ADMASS, A Computer Code for Fluid-Structure Interactions Using the 3.
Finite Element Technique", by S.H. Shaaban.
" Structural Sensitivity Studies Using the Indian Point Unit 3 Finite 4.
Element Model", Report No. RE-A-78-030 by EG and G Idaho, Inc.
" Topical Report on Dynamic Analysis of Reactor Vessel Internals Under 5.
LOCA Conditions With Application of Analysis to CE800Mwe Class Reactors", CENPD-42, by Combustion engineering Inc. --