ML20147E337
| ML20147E337 | |
| Person / Time | |
|---|---|
| Site: | Black Fox |
| Issue date: | 10/03/1978 |
| From: | Gallo J ISHAM, LINCOLN & BEALE |
| To: | Purdom P, Shon F, Wolfe S DREXEL UNIV., PHILADELPHIA, PA, Atomic Safety and Licensing Board Panel |
| References | |
| TRAN-781003, NUDOCS 7810160202 | |
| Download: ML20147E337 (33) | |
Text
,
4 Nr.0 ?T E IC pccint m ROO k
,,] :.
~
ISHAM, LINCOLN & BEALE COUNSELORS AT LAW -
b toso if!..* STRect, N. w. s VcNTH FLOOR g
J wasmuotoN,0. C co36
//
d V
TcLeewout aca as e7so f /[ c*0 9/p, r
emenoo orrice s
owc riast wanon4L Plaza
' 7 jj -Q j.
roavv-seco=o rooom r
emcAco,etuhoss somoa y
\\ -.
g vetzewowc ais ree tsoo t
f E
,,h. g/
October.3, 2978
' ' ' ' * * ~ " "
',?
QM Sheldon~J. Wolfe, Esq.
Mr. Frederick'J. Shon Atomic Safety and Licensing Atomic. Safety and Licensing-Board
( Board U.S. Nuclear Regulatory U.S. Nuclear Regulatory Commission Commission Washington, D.C.
20555 Washington, D.C.
20555 Dr. Paul W. Purdom Director, Environmental Studies Group.
Drexel University 32nd and Chestnut Street Philadelphia, Pennsylvania 19104 In the Matter of-Public Service Company of Oklahoma Associated Electric Cooperative, Inc. and Western Farmers Electric Cooperative, Inc.
(Black Fox Station, Units 1 and 2)-
Dockets Nos. STN 50-556 & STN 50-557 Gentlemen:
F Due to a collating error, a number of pages were inadvertently omitted from the testimony of Lambert J.
Sobon concerning Contention 16.
A corrected copy of his testimony j
is enclosed for the Licensing Board and all parties.
1 l
I apologize-for any inconvenience.
sincerely,
(
oseph Gallo One of the Attorneys for the Applicants cc:
Service List (with enclosure) 7 &/ o/ 6 o'2
- t
[
t.'
_...,.. - - -, - - - -, - _. ~.,
20 pgBLIC DOCUMggt ROOM
/
s s-UNITED STATES OF AM 'RICA
'O g(1 h
NUCLEAR REGULATORY COi'iISSION
' l-.
BEFORE THE ATOMIC SAFETY AND L: CENSING BOARDS.3 e
~
)
In the Matter of
)
)
PUBLIC SERVICE COMPANY OF OKLAHOMA,
)
Docket Nos. STN 50-556 ASSOCIATED ELECTRIC COOPERATIVE, INC.,
)
STN 50-557 AND WESTERN FARMERS ELECTRIC
)
COOPERATIVE, INC.
)
)
(Black Fox Station, Units 1 and 2)
)
Testimony of Lambert J. Sobon Concerning Contention 16 I
l September 25, 1978 t
I l
l 1
i i
TESTIMONY OF LAMBERT J.
SOBON CONCERNING CONTENTION 16 My name is Lambert J. Sobon.
I reside at 992 Redmond Avenue, San Jose, California.
I am the Manager of BWR Containment
~
Licensing,. Containment Improvement' Programs in the Nuclear Energy Business Group.
A Statement of my background and qualifications is attached as Attachment.I to my testimony.
q t
My testimony deals with Contention 16.regarding the resolution of the following phenomena and-associated loads which relate to the design of a Mark III Pressure Suppression Containment:
~
1)
Vent Clearing 2)
Vent / Coolant Interaction 3)
Pool Swell 4)
Pool Stratification
- 5)
Pressure Loads and Flow Bypass 1.
Introduction My testimony presents technical information regarding the Mark III Containment System that may be used by a utility applicant in a Preliminary Safety Analysis Report (PSAR) design review.
This information is set forth in s
we
--Y y
4r
-Mn?
--w y
r we v.
y'M-d
=-
te e
mir W-
i Appendix 3B of the GESSAR-238 Nuclear Island Standard Design for Mark III Containment Systems for BWR-6 nuclear power reactors.*
My testimony will address Contention 16 to the extent of establishing the generic load definitions for Mark III Containment Systems.
The application of the generic load definitions to the Black Fox Station, in the context of Contention 16, is discussed by Mr. Guyot in his testimony.
My testimony will show that sufficient technical information has been. developed and documented by GE to permit the applicant in this proceeding to adequately address all of the hydrodynamic phenomena in the design of the Black Fox Station.
2.
Summary Description of the Mark III Containment The Mark III containment is a barrier to contain the energy of the reactor system and to prevent significant fission product release in the event of a postulated 238 Nuclear Island General Electric Company Standard Safety Analysis Report (GESSAR), Docket STN-50-447, Appendix 3B, "Information Report, Mark III Containment Dynamic Loading Conditions."
Appendix 3B consists of (i) Part I - load definitions for loss-of coolant-accident and safety relief valve related phenomena, and (ii) Part II - the application of the loads defined in Part I to the design of af fected structures and compo nents of the GE Mark III Containment design.
i
-3_
0 i
l loss-of-coolant 1 accident;(LOCA).*- The containment system-i employs'the. pressure-suppression concept, in which a
-l
-large ' pool of1 water ' (the suppression pool). is used to I
condense reactor steam.which issues'from a postulated)
~
N
.t i
reactor system pipe-rupture.
The suppression-pooloalso acts'as a reservoir.for reactor energy under certain~'
normal or antisipated operational conditions,Lsuch as' safety / relief. valve operation'(as would occur'during cer-r tain transients) and shutdown.'.
i 5
h The important pressure _ suppression features of the Mark
.III' containment design are theidrywell, suppression pool'.
i and' containment upper pool.
A schematic drawing ofLthe l
Mark III reactor-building which shows the-location:and orientation of the drywell, containment, suppression pool-I l
LOCA is.the sudden break of a high. energy pipe.in the j
l reactor coolant pressure boundary of the nuclear steam supply system.
The largest postulated break could be-i either the break of a main steam or a recirculation line.
.This LOCA'is the design basis accident (DBA).
Other small:
]
line breaks result in LOCAs, and although their energy;
'I release do:not result in large dynamic loadings, their thermal effects may control the design of structures'.
i The intermediate break accident (IBA). and small break
-l l
accident (SBA) fall into'this category. LThe size of the SBA is defined as that which will'not cause automatic depressurization of the reactor.
The SBA is of concern-l I
because it imposes the most severe temperature condition inside the drywell..The IBA is of concern because'it
~ i
~ ~ " ~~
is postulated to include the automatic _ actuation'of Lthe safety relief valves associated with the automatic depressurization syste'm.
.i
._;,_-._.__.-....a.._..--_._......-_...,..
1 4-
?
p 1
and upper pool: as well as of the ' horizontal vent openings
]
in the drywell wall is shown in Figure 1.
c
- =
The drywell functions to contain the transient pressure resulting from a postulated LOCA and to channel the air-I t
steam mixture to the suppression pool.
The drywell is
(
designed to withstand the pressure and temperature.,tran-
]
t sients associated'with the design basis LOCA inside'the drywell.
It is also designed.to withstand the'high tem-perature associated with the break of a small' steam line in the drywell which does not result in rapid depres-
-[
surization of the reactor pressure. vessel.
Large diameter horizontal vent openings penetrate through
[
t the lower section of.the drywell.
These-vents conduct the reactor steam to the suppression pool.
Three identical rows of vents are uniformly spaced.circumferentially around
^ the drywell.
i The suppression pool is a annular volume of water 11ocated between the drywell and the outer. containment boundary.-
h
-l This pool covers the horizontal. vent openings _ inathe dry-l L
well to maintain a water seal'between the drywell interior l
and the remainder'of the containment volume.
As'shown in f
Figure 1, a' portion of the suppression pool is located i
5
.---.-.x.~.-..
.,,,---,.-.~--...u---
..a..-..--..-..-,..-..
._5-I insidefthe drywell between.an. annular weir wall and the i
drywell wall.
Following ai postulated LOCA in the -
'5rywell,:the resultant:drywell pressure' increase forces' the water in'the weir wall annulus down, allowing the-l
~
steam / air mixture to enter the suppression pool.
The suppression _ pool condenses the= steam released in"the drywell.
Steam discharged through the safety / relief; valves during reactor transients is also-piped to the suppression l
j pool and is condensed.
The upper containment pool is a volume of water located-above the drywell.
This pool is used for fuel transfer.
during refueling operations.
Part of this water may be drained.to suppression pool to augment the long-term energy storage capability of the containment.
.l l
t 3.
Mark III Test Program L
- This section describes'the General Electric Mark III Test I
Program, which supports the conclusions in Appendix 3B.
[
h The General Electric Mark III containment pressure suppres-f i
sion testing program was initiated in 1971 with a series V
l l
of small-scale tests.
The= test apparatus consisted of small-scale simulations of the reactor pressure vessel, i
drywell, suppression pool and horizontal vents.
A total
,(
l of sixty-seven blowdown runs were made.
The purpose of
[
o I
a
_.. - _ _ ~
6'-
g
?
i theseitests was to: determine'the' behavior of.the horizontal i
vents and to obtain data for determining the acceleration
'f of the water inlthe' test section vents during initial clearing.. This information was used to establish'an analyt-
~
ical model~for predicting. vent system performance?in Mark' III and the'resulting drywell pressure response.
In November.1973,ftesting in the Mark III~ Pressure Suppres
- sion Test Facility (PSTF) began.
The PSTF consists of an 9
electrically heated steam; generator connected'to:a simu-lated drywell which can.be' heated to prevent steam condensa-tion.within its volume'during'the simulated blowdowns..The drywell is.modeled as a cylindrical vessel having a 10-foot diameter and 26-foot height.
A 6-foot' diameter vent duct-passes from the drywell into the suppression pool and con-nects-to the simulated vent system.
Pool baffles are used to simulate a scaled or full scale sector of a Mark III suppression pool.
The pool ~ arrangement is such that both vent submergence and pool areas can be varied parametrically.
The purpose of the Mark III Confirmatory: Test Program was to confirm the analytical methods used'to predict.the dry-
.well'and containment pressure. response following.the postu-
' lated LOCA.
In addition, this Test Program also was used to obtain information on the hydrodynamic loads that are_ _...._,,
generated in the vicinity of_the suppression pool during a LOCA.. It is this latter L aspect of theT Test Program that is pertinentito Contention 16.
-... - a.
.2.
4 7.--
The ful'l-scale PSTF testing performed'between November
- 1973 and February 1974:obtained: data for the confirmation
- f of the analytical model.
In March 1974 pool swell tests m.
were performed in the'PSTF.
These full-scale. tests u.
involved air. blowdown'into the drywell.and' suppression
~
pool to identify bounding pool swell impact loads and.
breakthrough' elevation, i.e.,
that elevation at which.
the water ligament!begins to break up and: impact loads are'significantly reduced.
Impact load data was obtained on selected targets. located above the. pool.
In June'of 1974, after.the PSTF. vent and pool system was converted'to 1/3-scale, four series of tests.were performed
~
.to provide. transient data.on the inter' action of poo1 swell with flow restrictions above the suppression pool. surface.
Other areas where data was obtained-included vent clearing, drywell pressurization, and jet forces onLpool walls.
The next seriesLof 1/3-scale testing began in' January 1975 and was directed at obtaining local impact pressures and total loads for' typical small structures located-over the pressure suppression pool including I-beams, pipes, and grating.
Data.from this test' series expanded the data base from the full-scale air tests.
A further series of 1/3-scale tests was.added in June 1975 to.obtain comparable
.. ~.......,
data on pool swell velocity and breakthrough elevation to
'the full-scale air tests.
i It should.be noted that although_most of the emphcsis in the testing; described above was directed at the evaluation I
of the pool swell phenomena, each test run consisted of a.
simulation of the postulated blowdown 1 transient for various postulated break' sizes up to-two times the_ Design Basis Accident for the containment.
Data was recorded at selected-locations around the test facility suppression pool through-out the blowdown so that the' hydrodynamic conditions asso-e ciated with each phase of,the blowdown is available'for selecting appropriate design-loading conditions.
General-A.
Electric has utilized this data to develop Appendix 3B.
It provides numerical information for thermal and hydrodynamic loading conditions in the GE Mark-III reference plant _ pres-sure suppression containment system during the postulated LOCA. - Appendix 3B also-presents information on thermal:
(
8 and hydrodynamic loading conditions during the anticipated j
safety relief valve (SRV). discharge and related dynamic events.
This information 'is appropriate for PSAR evalu-ations.
Separate test data has been utilized to estab-lish the SRV air clearing load-prediction model_ presented' i
in Appendix 3B to.GESSAR as well as the SRV thermal per-formance.
The GE reference plant report contains infor-l mation1and-guidance to assist the containment designer in_
[
E
_ evaluating the design conditions'for the various soructures_
l r
which form the containment system.
E 1
l-h a
t
..,,m-o w--
,,.,.+
,-,..,,,4..,4
-y-+
..,v,.,
,,7. ~, ~
..,y e
_9_
4.
Phenomena Addressed by Intervenors This section provides a description of each pressure suppression-phenomenon identified in Contention 16, and discusses how these phenomena have been evaluated for purposes of the design of Mark III pressure suppression containments.
4.1 Pool Swell Almost immediately following a postulated LOCA, the.drywell is pressurized by reactor steam, and a mixture of steam and air is directed to the suppression pool through the main vents.
The steam is rapidly condensed; but air forms large bubbles at the vents.
These bubbles cause an upward dis-l placement of the pool water above the vents.
The bubbles rise relative to the pool water, reducing the thickness of the water " slug" above the bubbles.
When the bubbles break through the water surface, an air-water froth is formed which I
L
' rises further before falling back into the suppression pool.
The initial motion of the water " slug" and the subsequent j
motion of the froth create impact and drag loads on sup-pression pool structures and components in their path, namely l
catwalks, gratings, pipes, and certain equipment.
The entire process is referred to as " pool swell."
The pool swell loads on suppression pool structures and components have been evaluated in 1/3-s'cale and full-scale experiments as part of the Mark III test program conducted
,.., v.
10 _
c.
.s by GE._
From.this information, loads are selected and specified-for'GE's standard plant in a form directly_appli-E
. cable to'the plant-design, as set forth_in Appendix 3B.
t A
The Mark III test program with respect to the pool. swell phenomenon is complete, and the. program provides adequate.
data'to, assure'that the Mark III containment pool swell.
t loads.are properly _ defined.
t The following Sections discuss the pool' swell-loadings-l identified in Appendix 3B.-
4.1.1.
Loads on Drywell During bubble formation, the outside of the drywell in the pool will be subject to. varying pressures.. A bounding range of 0.to 21.8 psid is specified on.those sections of the drywell wall below the suppression pool surface.
'The basis for this specification is-the knowledge that the minimum pressure increase is 0-psi and the maximum bubble pressure can never exceed the peak drywell pressure of 21.8 psig.
t Any structures'in the containment annulus that are-within approximately 20 ft. of the initial suppression pool surface will experience upward loads during, pool........,
swell.
If these structures are attached'to the drywell
[
1
,-1-,..
-.:., e
--...-.....-....a~
_ 11 _
s wall, thenlthe upward loads will be-transmitted into drywell structure.
In addition, the region of the 2
drywell below'the Hydraulic Control Uni" (HCU) floors will experience the wetwell pressurization transient during pool swell froth flow at the HCU floor.
4.1.2 Loads on Containment The PSTF air test data was examined for evidence of bubble pressure loading of the suppression pool wall opposite the vents._ These tests were chosen because the drywell pressure at the time of vent clearing is comparable to that-expected in a full scale Mark III and because-the vent air flow rates and associated pool dynamics would'be more representative than the large scale steam blowdown tests.
The maximum bubble pressure load on the containment observed during PSTF testing was 10 psig.
The Mark III design load.is l
I l
based on these tests.
L Any structures in the containment annulus that are I
within approximately 20 ft. of the initial suppression pool surface will experience upward loads during pool swell.
If these structures are attached to the con-l tainment wall, then the upward loads will be transmitted into that structure.
In addition, the region below the,,,,
HCU floors will experience the wetwell pressurization transient during pool swell froth flow at the HCU floor.
. ~. - -
p.
~
"i. ; '
- ' w'
.L 1'
' E.1'. 3 -
_ Loads on Structures in Suppression Pool Immediately following vent clearing and during bulk i
h, pool swell,. structures within the pool above the bottom vent elevation, can experience loads calculated using appropriateJdrag coefficient, and a pool swell velocity of-40 ft/sec.
This isLa bounding calculation of the.
maximum pool swell velocity.
Because of uncertainties.
of the flow pattern in.the suppression pool, the-40 ft/sec velocity. vector applies either upward'or outward, b
Structures in the suppression pool should be designed-conservatively for the drywell bubble pressure!and y
. pool' swell. drag.
(This' applies to small submerged ~
structures'e.g.,Hpipes.)
~
/
4.1.4 Loads on Structures at-the-Pool Surface l
Some structures have their lower surfaces either right at the suppression; pool surface or slightly submerged.
j This location means that these structuresido'not experience the'high: pool swell. impact loads. discussed
'in Section 4.1.5.
However, they experience pool swell drag loads produced by water flowing vertically past-the structures at 40 ft/sec.
J t
L4.1.5' Loads on Structures Between-the Pool Surface and the HCU Floors Equipment'and platforms located in the containment annulus region,jbetween the pool-surface and the HCU-f platform,; experience pool swell induced dynamic loads f
1 i
whose magnitude.is dependent upon both location and the geometry of the. structure.
The pool swell i
phenomenon can be considered as occurring in tv o
- phases, i.e., bulk' pool swell.followed by froth pool swell.
The pool swell dynamic loading conditions on a particular structure in the containment annulus are dependent upon the type of pool swell that the structure experiences.
In addition,to location, the e
size of the structure is also important.
Small pieces of equipment and structural items.will only influence the flow of a limited amount of water in the immediate vicinity of the structure.
Large platforms or floors, on the other hand, will completely stop the rising pool, and thus incur larger loadings.
For this reason such j
platforms and floors are located above the bulk pool I
swell zone, (e.g.,
the HCU floors).
This' subject is discussed in Section 4.1.6.
4.1.5.1 Impact Loads The PSTF air test data shows that after the pool has risen approximately 1.6 times vent submergence i.e.,
12 ft, the ligament thickness has decreased to 2 ft or less and the impact loads are then significantly reduced.
Conservative bulk pool swell impact loading of 115 psi on beams and 60 psi-for pipes are applied uniformly to any structures within 18 ft of the pool surface.
For evaluating the time at which
~.
s impact occurs at various elevations in the contain-ment annulus, the maximum observed water surface velocity of 40 ft/sec is assumed.
Bulk pool swell would start 1 sec after the LOCA.
The basis for the loading specification is the PSTF,
air test impact data.
These tests involved charging the reactor simulator with 1000 psia air and blowing down through a 4.25 inch orifice.
Fully instrumented s
targets located over the pool provided the impact data.
l Additional tests have been conducted which provide impact data for typical structures that experience bulk pool swell.
Data from these tests indicates that the i
specified design load is conservative.
It should be noted that impact loads are not specified for gratings.
The width of the grating surfaces (typically 1/4 inch) do not sustain an impact load.
This has been verified in the 1/3-scale PSTF test.
Grating drag loads are calculated.
For structures above the 18 ft elevation, the conserv-ative froth impingement load of 15 psi should be used.
Again, this impingement load is applied uniformly to,,,,,,,,_
all structures with the time.
This is also based on data generated during the PSTF air test series.
j i
1 i
f For structures between 18-and 19 feet above the s
suppression pool design loads and duration are linearly extrapolated from the values of 115 or 60 psi to 15 psi.
The influence of seismic induced submergence variations on the pool swell transient and resulting impact loads has been considered.
It has been concluded that the-effect on the magnitude of the pool swell impact load is not significant.- This conclusion is based on a consideration of the influence of submergence on swell velocity and the significant load attenuation which will result from the pool surface distortions.
The very significant margins between the specified loads and the expected loads provides confidence that any local increase in swell velocities will not result in loads in excess of design values.
I 4.1.5.2' Drag Loads In addition to the impact loads, structures that experi-ence bulk pool swell are also subject to drag loads as the pool water flows past them.
Drag loads are calcu-lated assuming a velocity of 40 ft/sec. between the pool surface and HCU floors.
4.1.6 Loads on Expansive Structures at the HCU Floor Elevation At the HCU floor elevation there are portions of.the,.,,,,
1 floor which are comprised of beams and grating and l
l other portions that are solid expansive structures.
L
-- -~,
The bottom of the steam tunnel is at approximately the same elevation.
The small structure portion (beams and grating) of the HCU floor is discussed in Section
^,-
4.1.7.
The expansive structures at this elevation, such as -
the bottom of the steam tunnel, experience an impulsive loading of 15 psi followed by an 11 psi pressure dif-ferential.
The impulsive load is due to the momentum s
of the froth which is decelerated by the structure.
The 11 psi pressure differential is based on an analysis of the transient pressure in the space between the pool surface and the HCU floor resulting from the froth flow 2
through the 1500 ft vent area at this elevation..
PSTF test results are the basis for the froth impinge-ment load of 15 psi lasting for 100 msec.
The 11 psi froth flow pressure differential lasting for 3 see is based on an analysis of the transient pressure in the space between the pool surface and the HCU floor.
The value of 11 psi is from an analysis that assumes'that L
the density of the flow through the annulus restriction is the homogeneous mixture of the top 9 ft of the 3
suppression pool (i.e., 18. 8 lb' /f t ).
This is_a m
conservative density assumption confirmed by the,PSTF_,,
1/3-scale tests which show average densities of
- 17.-
1 i
3 approximately 10~lb,/ft.
,The analytical.model used to simulate'the HCU floor flow pressure differential has been compared with test data.
These tests indicate the HCU floor pressure differential is more realistically l
in the 3 to 5 psi range.
l
)
The potential for circumferential variations in the pressure transient in the wetwell region beneath the HCU floor have been examined and on the basis of bound-ing calculations.it is concluded that the pressure variation will be less than 0.5 psid.
k 4.1.7 Loads on Small Structures at and Above the HCU Floor f
Elevation Small structures at the HCU floor elevation experience
" froth" pool swell which involves both impingement and f
i drag type forces.
PSTF air tests show that the struc--
tures experience a froth impingement load of 15 psi lasting for 100 milliseconds.
Structures must be designed for this short term dynamic impingement load.
Grating structures are not subjected to this impinge-ment load as discussed in Section 4.1.5.1.
Following the initial froth impingement there is a period of froth flow through the annulus restriction at this elevation with a pressure differential as discussed in the previous section.
t i
L
--r-.n v
y o
~wn--
i Those small structures above the HCU floor that could be exposed to pool swell froth are exposed to a drag-I load.
The drag load is determined for the geometric
~
shape of the structure using a froth density of 18.8 3 as in theLHCU floor differential pressure cal-lb /ft m
culation and the velocity of the froth at the elevation of the structure.
The velocity used is 50 ft/sec at 19-1/2 ft above the suppression pool and is decelerated by the effects of gravity.
The velocity of 50 ft/sec m
is a bound of the available data.
Pool swell is not assumed for structures located more than 30 ft above the suppression pool.
4.2 Vent clearing As the drywell pressure increases following a postulated LOCA, the water initially standing in the vent system accelerates
'into the suppression pool and the vents are cleared of water.
- The process of vent clearing affects the. maximum pressure that will be reached within the drywell.
GE has examined vent clearing performance as a part of its confirmation of the analytical model for computing drywell pressure response for postulated LOCA events.
This was done in one-third and full-scale tests.
Predicted drywell pressure responses from these tests agreed well with observed data thus confirming the adequacy of vent clearing predictive methods.
In addition vent clearing loads were obtained from the one-third and full-scale tests.
These loads are specified
?
^
t 5
for GE's standard plant in a form directly applicable to plant design and are identified in Appendix 3B.
The Mark III-test program with respect to the vent clearing phenomenon is complete, and the program provides adequate
.c r
data to assure that the Mark III containment vent clearing loads are properly defined.
I The following sections discuss.the vent clearing loads r
identified in Appendix 3B.
4.2.1 Loads on Drywell (Drywell Pressure)
During the. vent clearing process, the drywell reaches a peak calculated differential pressure of 21.8 psid.
During the subsequent vent flow phase of the blowdown, the peak pressure differential does not exceed 21.8 i
psid value even when it is assumed that pool swell results in some two-phase flow reaching the contain-ment annulus restriction at the HCU floor.
Interaction i
between pool swell and the limited number of structures at or near the pool surface does not adversely affect 1-the drywell pressure.
The calculated drywell pressure l
during the Design Basis Accident includes the HCU floor 1
pool swell interference effects.
The containment response analytical model was used to calculate l
these values.
l l
I
/
s During the blowdown process, the drywell is subjected to differential. pressures between levels'because of'
' flow restrictions.
This value varies with the' size-of the restriction, but a bounding value for a 25 per-cent restriction.is 0.5 psi.
On the basis of this calculation, it'has been concluded that differential, pressures within the drywell during the Design Basis Accident will be small and as such, need not be-i specifically included in the drywell loading specifica-tions.
4.2.2 Loads on Weir' Wall' The pressure drop at.'any point on the weir wall due to the acceleration of water during vent clearing is less than the local hydrostatic pressure.
Therefore, there is-no net outward load on the weir-wall due to vent I
clearing.
This conclusion is based on the predictions of the containment response analytical model.
Once flow of. air, steam and water. droplets has been-J established _in the vent system, there will be a static pressure reduction in the weir annulus that leads to approximately a 10 psi uniform outward pressure on the-i weir wall.
.This loading was calculated with the vent flow model.cnd for design purposes is assumed to exist-during the first 30 seconds of blowdown..
?
I f
I i
I
l -
i 4.2.3 Loads On Containment (Water Jet)
Examination of applicable PSTF' data indicates some
. evidence of a loading of'the containment wall due to tte water jet associated with the vent clearing process (e.g., less than 1 psi), as indicated by a small spike e
at 0.8 sec.
These water jet loads are negligible when compared to the subsequent air bubble pressure dis-cussed in Section 4.1.2 and are not specifically.includedi as a containment design load.
~<
4.3 Vent / Coolant Interaction (Vibratory Steam Condensation)
Chugging Following the vent clearing and pool swell transient asso-ciated with drywell air venting to the suppression pool there is a period of high steam flow through the vent system followed by reduced steam flow as the primary system high j
I energy fluid inventory is depleted.
During this phase-. the
- top row of vents are able to sustain the steam flow and the lower two rows are completely covered with water.
As the steam flow through the vents decreases to very low values, the water in the top row of vents begins to oscillate back and forth.
This action results in dynamic loads within the top vents and on the weir wall opposite the top vents.
Oscillatory pressure loadings can also occur on the drywell, suppression pool basemat, and containment.
This low-steam,,,,
flow oscillatory process, named " vent / coolant interaction" by the Intervenors, is referred to as " chugging."
v
~
-- 2 2 ' -
+
i The chugging loads described above have been evaluated in 1/3-scale and full-scale experiments as part of the Mark II test program.. For GE's standard plant the-loads are specified in a' form directly applicable to plant design in Appendix 3B.
Additional testing is ongoing.which will provide more data for evaluating steam condensation / chugging loads --
a very localized loading condition; however, the'already c
completed experiments referred to above provide a sufficient basis to select design loads for preliminary design purposes.
The following. sections discuss the vent / interaction loads identified in Appendix 3B.
4.~3.1 Loads'on the Drywell l
4.3*l.1 Condensation Loads Following Design Basis Accident' (DBA)
.Following the i'nitial pool swell. transient caused by the.
venting of drywell air to theLeontainment free space, there is a period of 1 to 5 minutes (depending upon break size and location) when the. vents can experience highLsteam mass flow rates. - Vent steady state steam 2
mass' fluxes of up:to 25 lbs/sec/ft occur as a result of either a main steam or recirculation lineJbreak.- The
~
'PSTF facility has undergone single vent steam blowdown l ~
o
,,w-.,_,.,
m.c,.
.,.[.n'
.,,.5
,,,w~,-
4
- w-
,.,,u,., _.,,,,,, -,
yn
i
[
23 -
4
~
tests'with blowdown orifices at least-twice the
~
nominal.DBA..
Some pressure oscillations with'a peak 1
amplitude of + 4 psi at.3 to 8 H have been observed j
g
', {
in the test facility'at the ven't' exits.
These speci-fled design values are not significant compared'to
-l t
the peak calculated drywe11' pressure.
4.3.1.2 Chugging Loads Following. Design' Basis Accident or r
Small Break Accident During vent chugging, drywell pressure fluctuations
.J result if significant quantities of suppression pool c
water are splashed into the drywell when the returning
~
water impacts'the weir wal1.
This can result in nega-
~
tive pressure on the drywell.
The maximum value of this load is (-) 1 psid.
Chugging is an oscillatory.
phenomenon having a' period of 2 to 5 seconds.
4. 3.'1. 3 Loads on Vents Due to Chugging Following Design Basis Accident or Small' Break Accident 7
.In addition to bulk drywell pressure fluctuations, vent chugging can lead to localized loads on the upper sur-faces within the top horizontal vents.
Pressures up to 450 psig have'been observed witIh durations in the 8 to 20 millisecond range.- Considering that drywell y
vent sleeves.are embedded in 5 ft thick concrete and- -.-........'._."
.the loads'are of extremely short duration, no specific dynamic. design requirements are imposed.
Sc
. 24 -
i 4.3.2 Loads on Weir Wall i
4.3.2.1.
Chugging Loads The upper vents will experience vent chugging follow-
'ing the reactor blowdown.
As a result, the weir wall will experience a series of inward acting loads due' to impingement of the oscillating water..
For weir wall loads, the large scale configuration of the PSTF is considered most. appropriate since the vents are full scale and the generic relationship between the On the basis.
top vent and the weir wall is correct.
of full scale PSTF-data,-the Mark ~III_ weir wall loading-specification for chugging is 15 psi applied over the It is' anticipated that projected area of the top vent.
for a full scale multi-vent Mark III, the' chugging
^
process will be random and only limited vent locations e
will experience simultaneous chugging.
However, the worst case in terms of the number and location of top vents simultaneously experiencing chugging loads is assumed.
The chugging frequency is 0.20 to 0.50 H.
4.3.2.2 condensation 7
There will be no loads induced on the weir during condensation, as shown by_ lack of transducer _ response,_,__,
I i
l in the tests.
l l'
,ny.,
4.3.3
' Loads on Containment 4.3.3.1 Condensation Loads During the condensation phase of the blowdown, there have been some pressure oscillations measured on the l
containment wall in the PSTF tests and there ~ magnitude is + 2 psi at 3 to 8 H.
These specified values are z
applied for design consideration.
1 4.3.3.2 Chugging Loads During the chugging phase of the blowdown there have been some pressure oscillations measured on the contain-ment wall in the PSTF tests.
Because of the short dura-tion (3 msec.) the load is not considered in the design specification.
4.4 Pool Stratification - Loss of Coolant Accident
- During steam condensation in the suppression pool due to the postulated LOCA, the pool water is heated in the immediate vicinity of the vents.
Most of the energy is released i
through the top vents.
As a result, the upper portion of the pool is heated more than the lower portion.
The vertical temperature gradient is known as " thermal stratification."
Low steam-flow chugging (as described in subsection 4.3) and circulation of suppression pool water.by the emergency core cooling system pumps will effectively dissipate this 1
_ ~ _. _
E'
--26~-
i thermal' gradient as the accident transient progresses.
Therefore, Lit is a.short-term effect.
Because of the turbu-lence associated'with the condensation process and th ~
presence of a large mass of cold water above the top row
't of vents, there is no concern for pool boiling or impair-ment of t e pressure suppression function.
This has been-t demonstrated 'by the Mark III test program.-
i 6
l S'
a t
This short-term ~ thermal stratification is offset by other '
]
i static thermal and hydrodynamic loading-conditions, and j
therefore it is not..specifically included as a preliminary.
design ' specification.
L k
4.5 Pool Stratification - Safety / Relief Valve Discharge Steam ' discharge, to the supprescion pool via the reactor safety / relief valvesE (SRV's) will take. place during certain l
-i operational transients.
The condensation of this steam.in i
the vicinity.of the safety / relief' valve discharge devices will cause local heating of the suppression pool water.
.j l
This " stratification" does not by itself cause significant 1
L loads on suppression pool components and structures, but' i
l it must be' considered in the design of,the safety / relief l
-valveLdischarge' devices in order to assure their acceptable 1
L performance under all anticipated conditions.
q is
~
o.
l L
L-
1 4
1 i
1 The performance of the safety / relief valve discharge device has been evaluated experimentally.by a foreign GE licensee.
Based on these test results,-the devices will perform as designed up to a local water temperature of l
212*F without unacceptable loading conditions being encoun '
tered.
The test results also showed the temperature dif-ferences between the discharge region'and other locations to be less than 9*F.
For reasons other than for safety /
relief valve discharge, the bulk suppression pool temperature will remain below 212*F.
Therefore, the quencher thermal" performance raises no concern for unacceptable thermal loading on the suppression pool ~ boundary.
' 4.6 Pressure Loads and Flow Bypass As discussed in Section 4.5, the safety / relief valve dis-charge devices have been designed and evaluated experi-
. mentally for effective, smooth condensation up to a local.
water temperature of 212*F.
It was also noted that other considerations will prevent the suppression pool temperature from reaching this value.
Thus there is no concern for significant oscillatory loads in the suppression pool as a result of steam condensation instability during a con-tinued' discharge of the safety / relief valves.
h
,u c-
,,_.,w...,
,r,--
-c.
= -
- 28'-
r
-The possibility of: steam bypassing the suppression pool lus a result' of disturbance of pool. surface (by local boiling, asymmetrical wave generation, seismic slosh, or.
other phenomena) has been qualitatively evaluated. - It is.
effectively precluded by the provision of a substantial volume of water above1the top row of vents and the pro-
. vision for draining part'of the upper-containmentLpool to j
5 augment the' suppression pool water volume.
m As described in Section 4.5, local boiling will not occur due to the large mass of cold water above the top row of vents.
Thus there'is no concern-for steam bypass due to 4
local boiling.
Asymetric wave generation.is evaluated.using full-scale test data from the Mark.III test program.
The test data
[
showed post pool swell wave peak-to-peak amplitudes of
- less than two feet.
The' plant designer'should take this parameter into account in the containment design to assure-no potential for steam bypass.
i Seismic slosh effects on the pool surface have been evaluated.
in a.three dimensional test.
In that' test, it was' concluded
~
that vent uncovering.will-not occur when subjected.to the seismic. spectra' set forth of USNRC Regulatory Guide 1.60.__,,
,....... ~.. _.
i..,.,.
s._.
i 4.7-: Inadvertent Upper Pool Dump Part of the upper containment pool is drained to the sup-
'pression pool if a signal < indicating a pipe rupture inside the drywell is present and one of the following signals is also present:
Suppression pool low water level or 30 minutes elapsed time following the pipe rupture.
The act of drain-ing the upper pool to the suppression pool is referred to as " upper pool dump."
The " Suppression Pool, Makeup System"-
is provided with sufficient redundancy and interlocks to assure that no single. active failure, including operator error, can result in inadvertent opening of both isolation valves on either dump line during a non-LOCA plant' condition.
5.
Ongoing Test Programs The ongoing Mark III confirmatory test program is being conducted at' General Electric's PSTF.
The remaining three phase test program is being performed primarily to obtain
- data for evaluation of
- 1) the localized conditions asso-ciated with the steam condensation portion of the LOCA
]
- blowdown,
- 2) suppression pool ~ thermal stratification, and
- 3) the effect of multiple vents on the blowdown transient.
The first of these test phases was begun in November 1976, and has been documented.
This test series consisted of eight shakedown and twenty runs for data in which the sup,
pression pool' was heated to initial temperatures of between 70 and 170*F, the simulated break size was varied to simulate both small and large breaks, and the blowdown of both saturated liquid and saturated vapor was tested.
In parallel with the above, test data was obtained for use in understanding the loading conditions on submerged structures located within the suppression pool.
Although the emphasis of this data acquisition was on the LOCA water jet and air bubble formation submerged structure loads, data on conden-sation/ chugging was also obtained.
m The second phase of the remaining Mark III confirmatory testing program consists of a full-scale test series with the same basi-c objectives as the above described 1/3-scale test series.
This testing has also been completed and documented.
The final test phase consists of a 1/9-scale test series in which a nine-vent array will be utilized to ev:Aluate multivent effects.
Installation of this vent con-figuration has been completed and testing is scheduled to be completed in 1979.
Final documentation of the Mark III confirmatory test program results is scheduled to be completed
(
l in the first quarter of 1980.
The results and interpretation 1
l of these tests have been and will continue to be transmitted to the NRC Staff and Applicants on a timely basis.
1
... _.......