ML18017A245

Forwards Design Assessment Rept,Revision 2. Proprietary Version Withheld (Ref 10CFR2.790)

ML18017A245
Person / Time
Site:	Susquehanna
Issue date:	06/09/1980
From:	Curtis N PENNSYLVANIA POWER & LIGHT CO.
To:	Youngblood B Office of Nuclear Reactor Regulation
Shared Package
ML17138B347	List: ML17138B346 ML18017A245
References
PLA-491, NUDOCS 8006110222
Download: ML18017A245 (325)

Text

A RKGULATO Y INFORHATION DISTRIBUTION STEM (RIDS) r "Jv ACCESSION NBR: 80064'10292 DOC DATE! 80/06/09 NOTARIZED ~ NO DOCKET ¹ Klectric Stationi Unit li 'Pennsylva

~

FACIL:50-387 Susquehanna Steam 0 53-50 388 Susquehanna Steam Electric Stationi Unit 2i Pennsylva 030389 AUTH'AME AUTHOR AFFILIATION CURT ISr N,H ~ Pennsylvania Po~er 8 Light Co.

. RKC IP,NAME RECIPIENT AFFILIATION YOUNGBLOODiB~ J' I.icensing Branch

SUBJECT:

Forwards "Design Assessment ReptiRevision 2 ' Proprietary version withheld (ref 10CFR2 ~ 790) ~

DISTRIBUTION COOK: PBOlS COPIES RECEIVED:LTR KNCL SIZEi y P.C7 TITLE! Propr i etar'y Info Re PSAR/FSAR NOTES'04KW 4 C V5 Q5 R 0 't A CC AWglf'g REC IP IKNT COP IKS RKC IP IKN T COPIES

~~~ ~~

ID CODE/NAME ID COOK/NAME LTTR ENCL ACTION: 02 PM~+

BC LTTR ENCL 1

1 1'+~ AD AO)4 0 LA A~ 1 1

0 0

INTERNAL: Ol-

~ I 1 1 03 OPERA LIC'R 1 1 W~ig 04 QAB 06 STRUC ENG BR+

1 1

1 1

05 MECH ENG BR 07 HATL, ENG BR 2 2 ~W-I<

~l 09 REAC SYS BR M 10 ANI. BR 11 CORE PERF BR 13 CONTHNT SYS4f 1

1 1

1 1

1 12 AUX SYS BR 14 ICC SYS BR 1

1 1

Mi Mj 15 PAR SYS BR W 1 1 16 AD SITE TECH 18 ACDKNT ANL 0 20 RAD ASHT BR~

1 1 19 EFLT TRT SYS 21 I8E 2 2 4

ee aI 23 KIRK'NOOD AD PLANT SYS

~Sf 1 1

1 1

1 0

AO FOR ENG AD SITE AAl YS 1

1 0

0 AD/CORE 8 CONT 1 0 DIRECTOR NRR" 0 HPA 1 0 NRC POR 1 0 OKLD 1 0 EXTERNAL: 2A ACRE sr LwR¹3 LA 'rr/AFF

+ 1g 1 0 LPDR NSIC 1

1 0

0 r" hre~irz I W <IAA AA g rET iver IiE I briber (LOW,)' Stfg 7 g

]9~~

re@ (L.+0,

@0 TOTAL NUMBER OF COPIES RKQUIRKDR- LTTR M ENCL

e PPaL

~

TW6 NORTH NINTH STREET ALLENTOWNs PA 18101 PHONEs t215) 821 5151 June 9, 1980 Mr. B. J. Youngblood, Chief Light Water Reactors Branch No. 3 Division of Proj ect Management U.S. Nuclear Regulatory Commission Washington, D.C. 20SSS SSES DOCKET NOS. 50-3878 50-388 DESIGN ASSESSMENT REPORT, REVISION 2 ER 100450 FILES 172-1, 840-2 PLA 491

Dear Mr. Youngblood:

Transmitted herewith are 40 copies of Revision 2 to the Susquehanna SES Design Assessment Report. Both Volume 1 and the Proprietary Supplement have been revised. Listed below are the major modifications.

1. Revision of Section 4.2, "LOCA Load Definition", to reflect the changes in load methodology required to comply with the October, 1978 NUREG-0487, as well as the addition of Subsection 4.2.3, "Response to NRC Criteria for Loads on Submexged Structure".

2. Update of Section 7.0, "Design Assessment".

3. Preparation of a non-proprietary and proprietary Section 9.0," SSES LOCA Steam Condensation Verification Test GKM-IIM".

4. Completion of Appendix A, "Containment Design Assessment", and Appendix E, "Reactor Building Structural Design Assessment".

S. Update of Appendix D, "Program Verification", to include verification of the KWU computer code VELPOT.

L o,

6. Rewrite of subsection 8.8.4, "Thermal performance of Quenchers". pm 8, JP

<gP q0 p5 q.

PENNSYLVANIA POWER 8 LIGHT COMPANY 8o06110 g g r}

Mr. B. J. Youngblocd June 9, 1980 Page 2 In addition, a number of editorial and syntactical sentence modifications have been included.

Pursuant to 10 CFR 2.790 and the affidavit submitted with our April 14, 1978 letter (PLA-244), we request that those pages marked proprietary be withheld from public disclosure.

Very truly yours,

,N. i'. Curtis Vice President-Engineering 5 Construction PAF:JLI PAF128:3

DOCKET EO. SIS DATE.

8A'OTE TO NRC AÃD/OR E,OCAL ?UBt'iC DOCENT ROOMS The following item submit ed with let r dated from is being withheld from oublic disc osure in acccrdance with Sect on 2.790.

PROPRIETARY LVEORCATiON K4stribction Serv- ce ' B~ch

~p,fl RK0(

(4 0 UNITED STATES ss NUCLEAR REGULATORY COMMIS&4 WASHINGTON, D. C. 20555

~*<< <<+

OBQIDL~i FOR: TERA. Corp.

FROM: US %5C/TIDC/Distribution Services Branch

SUBJECT:

Special Document Handling Requirements O 1. Please use the following special attached document.

distribution list for the n 2. -

The attached. document considerations:

requires the following special Do not send oversi.se enclosure to the HRC PDR.

n Only one oversize enclosure return for Regulatory File storage.

was received please Proprietary information - send aff'davit only to the HRC PDR Q Other: (speciiy) cc: DSB Files TMC/DSB Authori"ed Signature

PROPRIETARY SECTION 4 TABLES Nuebeu TITLE 4-1 Design Parameters Affecting SRV Loading Quencher Hole: ield Data HOGE!f Input Data Line Loads During SRV Opening 4-5 Line Loads During SRV Closing Line Loads Durinq Irregular Condensation 4-7 Total Quencher Loads During SRV Opening Total Quencher Loads During SRV Closinq 4-9 'Total Quencher Loads During Irregular Condensation 4-10 Quencher Arm Loads Du ing SRV Opening 4-11 Quenche. Arm Loads During SRV Closing 4- 1'2 Quencher Arm Loads Durinq Ir"egula-Condensation 4-13 measured Parameters Relative to Figures 4-28 to 4-30 4-14 Submerged Structure Pressure Di ference as a Function of Body Dimension 4-15 Submerqed Structure .'luitipliers 80061102>R 4p-5

i' PROPRIETARY 4.0 LOAD DZP'I';AITEO'.t A. V SAPETV RPLI:.P VALVE~SRV! DISCHARGE LOAD DEPINITIOR I

This section 'provides a qeneral discussion of the approach used for desiqn of the SSFS Safety Relief Valve system (Subsection 4.1.1) as well as the methods u ed to calculate suppression pool boundary and submerged structu"e loads. For clarity the loading conditions have been divided into two cateqories:

'a ~ SRV'ischa"qe hydrodynamic loads..exerted on the SRV system (pipe, quencher, and support) itself (Subs ec t ion 4. 1. 2)

b. 'SRV'ischarge loads on the suppression pool boundary and submerged structures (Subsection 4. 1. 3) .

4. 1..1 General Discussion o" the SSES A2oroach The SRV system used. for SS"=S has beo..n,designed based on the followinq criteria:

a. Redu tion to the maximum extent p" act icable or. ti."

'etwell water soace dyna ic pressu=es associated with SBV discharge

b. Avoidance of condensation ins-abilities associated with high mass flux SRV steam discna"ges in no'up to 2000~)

suppression pools.

'To satisfy these criteria, quenche=s have been developed

~

specifically for the Pennsylvania Power and Light Comoany (?PGL) by i<raftwerk Union (KMU). A SSZS-unique dynamic load sDecif ication ha been preoa" ed by K 3 for - his device and is

=- ~

described in Subsections 4. 1.2 and 4.1. 3. Du" inq an extensive quencher develor!ment program (Ref 1),AU has determined the deqree of influence o>> various SRV system design parameters on the dynamic Pressures which result from SRV discharqe and has concluded the followinq:

P'ool pressure amplitudes decrease with decreasing pool temperature. This is a consequence of the relationship between bubble steam content and saturation conditions.

b. Pool pressure amplitudes "decrease with increasing pool free water area. The effect of eccentric SRV discharge locations on pool pressure amplitudes is negligible. =

4P-7

r PROPRIETARY

. DHIHIT"g" C~ Pool pressure amplitudes decrease with decreasing quencher exhaust a "ea. For decreasing exhaus- areas, the energy'nput to the oscillatinq auoble-water system is spread over a lonqer time, with a corresponding decrease in excitation of the'oscillatinq system.

d The influence of SBV discharge line length over the range measured by KMU, 9 to 19m (29. 5 to 62. 3 ft), is insignificant for a constant discharge line air mass.

(4hen two sets of unit" (Enqlish and metric) are given, the f irst value is the oriqinal one; the second is an approximation provided for convenience.) Detailed information concerning effects due to long discharg".

lines with numerous bends will be obtained during the Susquehanna unit cell tests described in Chapter 8.

e. Pool pressure amplitudes decrease with decreasing expelled air mass, ie, total energy input to the system

'dec"eases with decreasing air volume. However, the ai" volume in the pipe should not be considered as an absolute quanticy of influence, but =ather as a relative ef"ect, highly deoendent upon the mass of water over which the input energy is distributed and the rate at which eno.rqy i" added to the system.

f he followinq pa"amete=s a feet poo'ressures because of .heir in" luence on S'.?V .ischa ge line clearing pressures, but are less important than those men"ioned above: valve opening time, steam mass flux, SRV

, ai~cnarq 0 line emperature, and su 'h merence.

more c "mplete listing of ajor and some localized parameters is contained in Table Q-l.

The effects of differences in physical parameters between SSES and KNU BNRs have been accounted for in the quencher design shown on Figure o-1 and in Table a-2. ~

o co"rect primarily fo; "he reduced steam mass flux per S? V and increased line air volumes, the SSES quenchers have been designed with an outlet area approximately 50 percent of thar which has been used for Ge man 3/Rs. This assures that optimum use has been made of the discharqe area effect on pressu"e amplitude reduction. Fu ther decreases in outlet area are not feasible due to the adverse ef feet on SRV bac<pressures and SRV discharge line design pressures.

The effect of SRY discharqe line length (the lonqest SSES SRV discharge line is about twice as long as tne longest line previously tested by KWU) on pressu"e amplitude will be studied; during the SSZS unique testinq program, as will the SSBS curved

/

PROPRIETARY EXHIBIT"g",

with cespect to inhibiting steam-air mixinq 'ipe,arranqement prior to and during vent clearinq.

4. 1. 1. 1 Thermal Performarice One of the keys to the KRU quencher device is its ability to condense stably (without larqe pressure amplitudes) the steam fraction in exhaust s"earn-air mixtures as well as pure steam discharges. The mo t restcictive conditions, which involve high steam mass fluxes a"..d elevated pool temperatu es, are of p"imacy importance. Disc'narqe hcle pattecns are arranged to enable an influx of pool water between adjacent rows under all operating conditions. This arranqement. ensures that immediate contaci is established between the cooler pool water and the warmer gas 0 ischa rq ed.

The optimum quencher hole pattern verified during the GKH quencher development program is used for the SSES design (Figure 4-1 and Table 4-2) . Th 10 mm dischacqe holes are sDaced 15 mm on centers'nd ace arranged in rows which are separated by 50 mm.

The 50 mm center-to-center spacing provides the pathway for supplyinq wa ter o the s:earn (see Piquce 4-2), thereby enabling the pool o b heated almos" to the boiling point without a "ise in the press>re ampl;tul.s associated wi:h SHV Discharge.

Verification of auencher =hermal pe=focmance may be found in Hef (on Piquce 5 1 3 an. page 5-34)

4. 1.2 Loads or. the SRV System due to SHV Aetna"ion The loading conditions which are d'esccibed .n the followinq subsections a~ply to the SHV piping, quencher body and arms, and quencher support.

4. 1.2. 1 SP.V Lin. Backoressure Load The maxi=urn SHV backnressure during stealy sta-e blowdown was investigated analytically for the quencher dlscnarqe device. The longest line geometry was used in the analysis. Zt was determined that the maximum SHV discharge line internal pressure is less than 550 psiq.

4. 1.2. 2 SH V System Rater Clearing P=essuce- Load This subsection summarizes the analytical techniques employed .o calculate internal pressures and vertical loads acting on "he SHV discharge pi ping as a cesult of water sluq clearing.

Safety relief valve steam flow was assumed satucated for all calculations. The KAU'omputer code HOGE.'t was used to compute the pressure. rise in an SHV discharge line thcouqhout the watec clearing phase which follows the liftinq of an SHV The code,

tt PROPRIHTAR Y Cl ))

DIHIHIT documented in Re f 1, has been verified with subscale (model) and.

in- plant tes t da ta.

The SSES'nique paramete s listed in Table 4-3.have. been used as .

input data to the HOGEif computer proqram.

The flow resistance. coefficient for quenchers .which had-been optimized to parameters unique to KNU-designed plants was found to be = 1.5. richen calculatinq the SSES-unique flow resistance particular .conside"ation was gi ven to the SSZS- 'oefficients, unique quencher. Due to dif erent parameters (compared to KRU plants), approximately one-half the discharge area of earlier KdU desiqns was requi ed.

Usinq an area reduction factor of 0.6, the effective discharge area of the Susquehanna guen'cher is calculated as:

Aeff q 0. 5 Aqeom = 0. 522 m~ (5.617 ft ~)

Since the cross-sectional a ea of an,SRV- discha qe line is:

0.073 m~

the area ratio becomes:

A>f cq= 0 72 AD The HOGZ I code relates the quencher flow resistance coefficient,

(, to the square of the flow velocity inside the SBV discharge lino, necessitating the calculation of a velocity ratio between the quencher discha qe and the pipe flow velocities.

An = 1 = 1 39 MD Aeff q 0.72 where:

flow velocity For quenchers typical of those used in KMU designed plants, this ratio is equal to one. 4 As a.significant po"tion of the pressu=e reduction mechanism is related to the quencher discharge area, an appropriate resistance

4 9

p p

PBOPBI-TABY Sl'p coefficient was used for the Susquehanna quencrer, based on a

.value which had been previously verified for KWU plants.

Consistent with the HOGZN code methodology, the SSES-unique value was calculated by multiplying the KWU value by the square of the SSES velocity ratio.

M~~ (1 39) ~ 1.93 or approximately 2.

W~D Hence, the Susquehanna quencher flow resistance coefficient is:

gSSES = 2 x 1. 5 = 3 where "1. 5 = 4 for KWV plants The followinq clearinq pressures were calculated for .the longest and the shortest. SRV discharqe lines, respectively, based on the

-KWU HOGEM'analysis:

Lenceth of Biecl.n~t' line Calculated clearing pre-suxe 48.3 m (158. 5 ft) 22.7 bar (314. 5 psig) =

34 9 m (1145 ft) 27. 1 bar (378. 3 psiq)

'The clearing "essure ti"..e histories a e shown on Figure 4-3.

Su bsequeri t to water clearirq, tne 'r."erna'ressure changes to the steady stz te s sam flow condition.'or calculatina t'e vertical load imposed on the quencher due <<o the directional chanqe in flow velocity -ithin the quencher (vertical SRV discharqe,line to horizonta'l quenc.".er arms), a conserv'tive esis" ance coefficient, g = 0, was used (=a<<her than the value E = 3.0 described in the pa agraphs above).

The followina vertical loads actinq on 'an SBV discharqe line result f=om a chanqe in direction of the water leg durinq water cl'earinq:

T.ength of the SBV discharge line Vertical load 48 3 m 490 kN (110. 2 kips) 34.9 m 620 kH (139. 4 kips)

The time histories of these vertical loads (with = 0) are shown on Fiqu=e 4-4.

4P-11

IL 1

PROPRIETARY 4 1 2 3 SRV Disch~a". e Line Loads Durinq water sluq clearing, the dif ferent pipe runs of .the SRV line are subjected to dynamic loads due to flow changes within the pipe (p'essure and momentum changes). The piping analysis contained in Section 5. 5 includes these. loads. Figure 4-5 represents the vertical load on the last pipe cun (ending with the quencher) . Tables 4-4,4-5 and 4-6 list the maximum loads experienced by an SRV discharqe line during SRV opening, SRV closinq and irregular condensation, respectively.

4~1.. 4 quencher Body Loads Oscillatinq. bubbles from SRV discharge into the suppression pool produce external loads on the quenchers. An operating quencher is affected by bubbles caused. by its own discharge as well as by

'bubble from adjacent quenchers. Xt has been shown experimentally (Ref -23) that the maximum external loading condition on an individual quencher occurs durinq operation of the quenche itself. .he operation of one or more adjacent quenchers does riot pcoduce increased loads. ">>xternal loads on quenchers which are not ope atinq ace evaluated using loading conditions desccired in Subsection 4.1.3'ccording to theic location in the pool.

The loads acting on the quenche: body a"e shown on Figure 4-6.

Tables 4-7I 4-8 and 4-9 list th>>a" imiim loads experienced by an SSKS quencher during SR'1 opening, SRV closing and ir=egular condensa tion, resoectively. The load time histocies ace referenced in the same tables. Seven "hous'.nd valve openings, seven thousand valve closings and one zillion irceqular condensaticn load cycles have been assumed.

4.1.2.5~~uenchec Arm Loads

,The loads actinq on eacn auenchec a=m are shown on Figure 4- 14.

Tables 4-10, 4-11 and 4-12 list the maximum loads experienced by an SS"-S quenche" arm durinq SRV openinq', SRV closing and irreqular condensation, respectively. 1he load time histo"ies are refecenced in the same tables. Seven thousand valve openinqs, seven thou=and valve closings and one -million irregular condensaticn load cycles have been assumed.

4 1.2.6 quencher Suyoo"t Loads The quencher supports have been designed for the following loads:

Loads acting on the quenche" due to SRV discharge as discussed in Subsections 4. 1. 2.4 and 4. 1.2. 5.

t PROPBI" TABBY ExHIBIT "A"

b. Loads from'he SBV discharqe 'line

c. Loads from flow deflection within the d'ischarge line Loads due to oscillating discharge bubbles

4. 1 2. 7 quencher Fatigue Loads Althouqh each clearing event is followed by nearly continuous steam flow, steam condensation does not exhibit a unifo m behavior throuqhout the, entire range of steam mass flow rates and wetwell water temperatur s. The various regions of condensation behavior are shown on Figure 4-22. The quencher experiences maximum hydrodynamic and thermal fatigue loads during.

discontinuous flow or irregular condensation (transition region, Fiqure 4-22) . The irregular condensation loads from Table 4-9 are used for fatigue considerations. One million total stress cycles (associa ted with the irreqular condensation are assumed for the analysis.

4.1.3 Loads on Suppression Pool St=uctu"es due to SBV Act >a tion This subsec-ion d. sc=ibes loads on wetted por" ions o the suppress'on ool boundarv and submerged structures. Subsections

4. 1.3. 1 th=o )gh 4. 1. 3. 3 aiv the circumferential pressure distributions on th~ suppression pool boundaries "or the various SHV actuation cases. The vertical pressure distribution on the boundaries is, discussed in Subsection 4.1.3.4. Subsection

4. 1. 3.5 gives the pressure ti".e histories used ror the analysis.

4. 1..3. 'ymmetric Loadina Con'it'on+SBV All}

The assumption that all qas bubbles ar'inq fro SRV discharge oscillate in phase with the same s rena:h (hiqhest possible) leads to ..he wo st loading case as i~scribe.l for "he normalized condition on Figure <<-23. For the entire region (vase."at, containment wetted wall, and pedestal .wetted wall), the most restrictive pressure time histories as described in Subsection 4.1.3.5 have been used for the analysis to ensure conservatism.

In the lower region the amplitude multiplier has been chosen to be consistent with the analysis presented in Subsection 4. 1. 3.5, while in the upper region the same multiplier decreases linearly to zero at the water surface shown on Fiqu"e 4-24.

4. 1. 3. 2 As~mmetric Load in~Condition

. The most restrictive asymmetrical loading condition occurs when a qroup of adjacent valves is operating. he analysis was made for, the case in <<hich three adjacent valves are operatinq. The .

normalized pressure distribution is shown on Figure 4-25.

/

IÃ

,J

PROP BI""T ARK EXHIBIT "P" The pressure distribution was defined circumferentially for a 180o segment. On both sides of a 90~ range with a constant pressure level the pressure decreases linearly to zero over 450.

On the other 180~ segment of the pool, the pressures were assumed to be zero The vertical pressure distribution was assumed to be the same as for the symmetrical case. The multiplier described in Subsection 4 1.3 1 i.s also applied to this case.

4.1 3 3 1 Si~n1e Valva Actuation Loading Condition Asymmetric, loadinq also occurs when a single valve actuates. The normalized pressure "distribution for this case is shown on Fiqure 4--26. The pressure level in the circumferential di" ection remains constant over a range of 300 and, on both sides of this ranqe, decreases linearly ove" 47.50 to 20 percent of the maximum value. Outside of this reqion, the pressure equals 20 percent of the maximuar pressure valve. For comparison, a pressure decrease related to the law 1/R is shown on the sane figure. The pressure distribution in the radial direction is also included in Figure 0-26

4. 1.3.3 Automatic Depressurization System (ADS) Loading Condition

'Assuming hat ~he,six ADS valves ("o" locat'on, see Fiqu=e 1-4) are actinq in phase, there is no great difference between th sym=e"-ic and ~'-e ADS loading conditions. .=. gu=e 4-27 depicts the normalized pressure distribution used for this case.

4. 1.3.4 7ertic=l P"essure Di=tribu ion Once th . @as bubbles have been expelled f ro.a = quencher, they coa'le-.ce and "he resulting bubble aqqlomeration rises due to buovancy effects while oscillatinq. Because of the free su face presence, pr -su"es on containment an" pe estal walls near the water su. faces are lower than the pressures on the basemat. Zor such configurations the observed vertic ~ velocity co"ponen" is in the order of 2 m/sec. However, the oubhl. oscillation is nearly d'ampe'd out after approxi ately 1 second as can be seen on Fiqures 4-.28 to 4-30.: Therefore, the assun d pressure decrease with elevation as shown on FIgure 4-24 -is conservative.

4. 1.3. 5 pressure Time Histories The definition of SBV loads on suppression pool we ted boundari'es and sub.=erqed internals can be limited ta loads resulting from qas bubble oscillation following vent clea=ing, as:hese loads have been shown to be bounding when compared to those associated with the other phases of SRV discharge (Ref 3) . This section contains a Discussion of 'ndividual pressure time histories as well as spatial effects

EXHIBIT "4" Immediately followinq the lifting of an SRV, a mixture of steam and air is discharqed into the suppression pool. The pressure time histories experienced by the suppression pool wetted boundaries and submerged structu"es differ with respect to amplitude frequency and dampinq for each actuation event (Ref

21) .

~

To obtain a bounding loading condition for SSHS containment analysis, conservatism with respect to frequency, dampinq, and pressure amplitude is required. The resulting loads are applied to the containment in accordance with the spatial described in Subsections 4.1.3.1 through 4.1.3.4.~

pressure'distributions In order to obtain a valid frequency spread, measured traces from previous K'2U full scale testing programs were selected and, analyzed. Approximately 200 runs from various Kraftwerk Union BWR power plants were available. From these, three traces were chosen from the Brunsbuttel non-,nuclear hot. functional testing proqram for use in SSES design verification (for conservatism, subsequent actuation cases have been used). The three traces are =

shown in Fiqu=es 4-28, 4-29, and 4-30 and the test conditions are described in Reference 21. ~amor paraclete s a"e lis=ed in Table 4-13.

During the Brunsbut-el :e=t 'ng proqra', al prossu=es were measured at al'1 wall cositi on adjacent to the opera"inq quencher.

Distance to "'."e nearest 'ct ua ting Gu ncher arm was approxl 'a teiy ~

-1 m (3.28 f") . The .":.ea u=e p"ossu" e t "ares a "e there f ore exoected to inclu'e all "a" e clea i~a /water~et} and ai= bubble oscilla t ion e 6":ec ts.

'She t "aces used were selected not.only fo their frequ'ency variation bu" also "or their relatively large p=essure amplitudes of 0.5 to 0.8 bar (7. 25 to 11. 6 psia) . Figure 4-28 contains the nighest pressure a..plitude ever measured during in-plant testing folio"inq the water slug clearing phase of a KiU quencher equ ip ped sa". ~t y re ief system. ~ i!i ' the oscillation shown in Piqure 4-28 is damped out rapidly, the othe two traces exhibit less damping. A comparison between Fiqures 4-29 and 4-30 indicates that, peak pressure amplitudes can be experienced at different times.

Figures 4-31 to 4-33 contain power spectral density functions for the initial 0. 6 sec. of tne measured pressure traces. For pu"poses of comparison it shouli he mentioned that the traces contain variation" in ordinate scaling. In all cases an air

<bubble oscillaticn f requency be" ween 6 and 8 cps is dominant.

Althouq the pressure amplitude of run 435 has the highest magnitude (refer to Figure 4-28), the maqnitude of the power spectral density of the dominant. bubble f"equency is small (Ziqure 4-31) when compared to the two other cases (Figures 4-32 4P-15

~

'l I'>

N'

PROPRIETARY EXH IBIT "g" and 4-33). rIhen the rapid damping of the oscillation as shown in Figure 4-28 is taken into account, a unique bubble oscillation can be presumed to have occurred.

In adDition to the most important firs" 0.6 seconds of each trace, Fiqures 4-34 to 4-36 show the power spectral density functions Qurinq longer periods of the same traces. The bubble remains the dominant frequency even though the pressure 'requency amplitudes are in practice damped out before the analy'zer trace ends. This- prevailinq f equency shows that the traces do not contain qeometrical effects such as eiqenfrequencies of the structure. Therefore, the pressure time histories are used as pure forcinq functions. Zt should be added that the pressure transducers used were fastened to a stiff sandwich wall. structure to minimi e interaction effects.

In order to obtain a conservative frequency content, 'the variation in air mass between the Susquehanna SRV discharge lin s and those used foc Brunsbuttel were taken into consideration.

The lonqest SS"=S dischacqe line has a conservatively estimated enclosed air vol u e 0<<3. m~ (109. 5 f t~) (ref ec to Table 1-3) 1 while the Bcunsbuttel di.-eh~roe. lines have an enclosed aic volume of 1 45 "~ (v1.2 f" ~) (ce er to Ref 2) The ra<<io of these volumes is 2. 14 to 1.

' s s u m 1. ilg a soherical ai" bubble, the aic bubble frequency is

3. D v e r s e 1 y 00c tl cna 1 t0 the c1be root 0 f <<he air vo lu2e ra 10 p rotor<<iona as can b e seen in Ref 4. r a flat bubble with a constant eros" sectiona 1 area is assumed, the air bubble f"equency will oe invecsel / p 1 to the square c cot o= the volume ratio a s can be s een in Ref 5. This analysis as uncs the real bubble shape to b between 'ese two limit. s, anD -the resultinq frequen" v sNift to be be<<ween <<he two models'rediction of the air-volu me ratio propoctionality.

In order to obtain a cons rvatiye frequency content, the three traces (K'iqures 4-28 to 4-30) which were used as normalized forcinq functi'orls wece expanded in time by a <<zczoc l.8 (an expansion) and reduced in time by a factor 0. 9 (a- contraction) .

NIthin a given fre'quency range one of the three traces affects an individual location in the containment structure more adverse'.y than the others.

The Susquehanna SZS quenchecs were designed to compensate for the fact that some of the Susquehanna parameters we e different from these of the Brunsbuttel -pl. nt. To ad just o- 'love" values of steam mass flux per SRV, and ror the.greater initial erlclosed aic mass,. the exit area of the Susquehanna quencne= was reduced o approximately one half of that of existinq KMU power plants. Any

~

f urther reduction in quencner discharqe area, regardless of its desirability, is unfeasible due to limitations imposed on SRV 4P-16

5

~ ~+*~4 +&

PROPR1FTARY EXHIBIT "A" discharge line internal prssures as well as SRV backpc ssuces.

Based on the experience obtained during the su'bscale testinq phase of KNU's quencher develop ent program (cexer to Ref 1),

is unlikely the maximum SS"=S pressure ampli-.udes will ever exceed it a normalized value of 1.5 when applied to the Bcunsbuttel pressure amplitudes. Therefore, this evaluation is based on a conservative normalized value of 1.5; this valu.; will be verified durinq t ~ ~ unxq t coll testing program .. which is explaine Chapter 8, nd has been used in co,;unction with pressure-t suppression e .. (Figures 4-28 through 4->0) for the wall, pedestal, and basemat adequacy p

assessments.

1.3.7 Loads on Submerged Structures due to SRV Actuation The normalized pressure time histories presented on Figures 4-28, 4-29, and 4-30 (refer to Subsection 4. l. 3. 5) are also used for the analysis of loads on submerqed structures. The vertical pressure distribution of Figure 4-24 is adopted. The loads are calculated usinq the pcessu e values and the submerged s ructure projected area. The computed loads vere assumed to be acting in the: lateral direction except foc the downcomec bracing and stiffener ring loads. the'owncomec The dovnco.".e" b"acing loa's a"e assumed to be ac ing in lateral an" vertical directions simul:aneously. .he la eral load is calc>lated usinq the reduced pressure value according to Figure 4-24. The vertical loaD is calculated usinq the ull pressure value. The downcomec rinq plate loads ace assumed to be acting in the v rtical di "ection. This vec ical load is also calculated usinq the full pressure value.

Similar to "he load= on the suppression pool wetted walls, a, F~~

multipliec was adopted when applying the noc=alized pressu"e time histories to account for dif <<ecences between SSFS and Bcuns~uttel quenche"s. The value of -he multipliec was taken differen" ly dependinq on the size (diameter) o= t' submerged stcuc uce.

Discussicn- pertaining to the choice of this multipliec a=e provided below.

For the case of a single spherical oscillating gas bubble, the pressure amplitudes relative to the suc"oundinq wate- pressure can be calculated by the simple relation:

Pressure differential attenuation =

where Ro = bubble radius 4P-17

lllsl 4 'Iss4WL rsaa OETKIL X Cassssaalsasa 0ETAIL Y Illsl

+4 CKsa I,IaaiIIII i a (sa s IIC ss4saaasal <<TCcsaIII tsFI

~

I'I arrC

-e-- o- I- """"

sa'II 4, IIICSIWC ra CVSC T

e sCV ~ )illa Id Ia IIOITaa Section A-A 4-~I 74 Jl'saalarsaas Ag aSrrSCsaA

+CasINCIsaf afaa

/IC ~ II t$ 5 Naaaf 5

/ l I

~

I.ta o.Si I MT~MZ-l

~ I I ~ 5 J ~ Ia ~ ssssss r I I

-- ~ 104 I 5 ~ ass sa (sa'I TOTAL NO. OF HOLES 44 IIOaf5 II ~ 4 4aaaa asraaal Ca?SI HOLESr

• Lrr, s alfacasrasaf Ix dl HOI.ES ~ Ildd HOLES l

Lll.v'ItiollView C

Cra D

r.

m Caa R~ C)

Z Z c+

c0 m m U 2 Caa Ch III Z

o c) Caa Cal Cll c z A m m ZZI-m

-I 0 m IH O Q R fll C)

-C m gQl 'Cl Plan View 0

Ch 0

PROPRIETARY EXHIBIT: "A".

TABLE 4-7 TOTAL~U='.)CHER LOADS DURTi'IG SRV OPENING~ 1>

Haximum 7Load Value Direction Time Histo~r internal over 27 bars See Figure 4-7 pressure (377 psiq)

External load 44 kn<2> Simultaneously in See Figure 4-8 h

(9891 lb) the horizontal and vertical quencher planes Hater deflection 620 kn Vertical See Figure 4-5 load inside the (139,376 lb) quencher Torque 40 knm En horizontal. See Figure 4-9 (29, 501 quencher plane f t-lb)

External load due to bubble oscillation (See Subsect ion 4. 1. 2. 4)

<1> . For the case of a sliding joint in the discha"qe line close to the quencher (Fiqure 4-10), tne pressure inside the pipe acts as an external force. This cas is shown in, Fiqure 4. 11.

,<2> Effects of asymmetric hole arrangement are included.

PROPRIETARY EXHIBIT "A" TABLE 8

~TOTAL URNCHRR LOADS DURING SRV CLOS1NG Maximum Load Value Direction Time History External Load 4.5 kn (1012 Simultaneo usly lb) horizontal and iver ti-n -. he See Pigure 4-12 cal quencher planes Torqu~ 6 knm Xn horizontal quencher See Figure .4- '1 2

(4425 plane ft-lb)

PROPRIETARY EXHIBIT."g".

TABLE 4-9 1 TOTAL QUENCHER LOADS DURING IRREGULAR CONDENSATION Maximum Load Value Direction Time History External load 17.5 kn Simultaneously in See Figure 4-13 3

(3934 lb) the ho izontal and, vertical quencher planes Torque 19 knm In horizontal See Figur e 4-13 (14,013 quenche plane

.Ct-lb)

1 PROPRIETARY EXHIBIT "A" TABLE 4-1 1

~UENCHER ARS LOAQS DURING SRV CLOSING

'oad

-'Haximum Value Direction Time Histor v internal over 22 bars See Figure 4-18 pressure (304 psiq)

External load 4.5 Kn Simultaneously

{1012 lb) in the horizontal and vertica1 planes See Pigure 4-19 Bending moment 3 Knm Simultaneously See Figure 4-19 on melding seam (2213 in the horizonta1.

at intersection tt-lb) and vertical planes between quencner a"m and quencher ball Therma 1 load - 2190C See Figure 4-18 (Internal temper- =

(426~-:)

atu"- )

f, 11

EXHlHIT "g" PBOPREETABY TABLE 4-12 QUENCHER ARM LOADS DURIMG IRREGULAR COMDENSAIIOM maximum Load Value Direct.ion Time History Enternal pressure 3. 0 bars See Figure 4-20 (28.8 psig)

External load 14.5 Kn Simultaneously See Figure 4-21 (6638 in the horizontal ft-lb) and vertical direction Bending moment, 9 Knm Simultaneously See Figure 4-21 on welding seam (6638 in the horizons.al at intersection 0 t-lb) and vertical plane between quenche:

arm and quencher ball Therma l load 1330C S e Figure 4-20 (Entl na emu (271. 4~F) ature)

PROPRIETARY CHAPTER 8 SSES QUENCHER VERIFICATION TEST T ABLE OF CONT ENTS 8 1 INTRODUCTION 8 1 1 Purpose of Tests 8.1.2 Test Concept 8 1.2 1 Single Cell Approach 8.1 2.1.1 Single Cell Theory 8 1.2 1.2 Application of Single Cell Approach 8 1 2.2 Simulation of SSES Parameters'rimary 8 1.2 2 1 System Pressure 8.1 2.2.2 Safety Belie f Valve (SBV) 8 1.2.2.3 Discharge Line 8.1 2 2-4 Vacuum Breakers 8.1-2.2 5 Quencher BEV 1, 3r79

8 2 TEST FACILITY AND INSTRUMENTATION 8.2.1 Test Facility 8 2.1.1 Mechanical Set-Up 8.2.1 1 1 Steam Boiler 8.2.1 1 2 Steam Accumulator 8 .2.1 1 3 Steam Line and Buffer Tank 8 2 1 1.4 Sa fety/Relic f Valve (SR V) 8.2 1.1 5 Discharge Line and Quencher 8.2 1 1 6 Test Tank 8 2 2 Instrumentation 8.2. 2 1 General Description 8 2 2.2 Instrumentation Identification 8 2.2 3 Operating Instrumentation 8.2.2 3 1 Display on Control Console 8 2 2.3 2 Acquisition by Computer 8.2.2.4 Test Instrumentation 8.2 2.4. 1 Measuring Points 8 2.2 4 2 Set-Up of Measuring Instruments 8 2.2.5 Visual Recording REV li 3/79 8-2

8 3 TEST PARAiiETEHS AND MATRIX 8 3 I Vent Clearing Tests 8 3.2 Cond ensa tion Tests REV 1, 3/79 8-3

8 4 'ZEST RESULTS 8 4.1 Vent Clearing Test Results 8 4 1.1 Test Pa ra me ters 8 4 1 2 Behavior of the SRV and System Press'ures 8.4 1 3 Dynamic Pressure Loads on the Pool Boundaries 8 4 1.4 Loads on the Quencher and Bottom Support 8 4 2 Steam Condensation Test Results 8.4 2.1 Test Parameters 8.4. 2. 2 Presentation of Test Results 8 4 2 2 1 Survey of Observed Condensation Phases 8.4.2 2 1. 1 Blowdown at Low Water Temperature 8.4 2. 2.1 2 Blowdown at High Mater Temperature 8 4 2.2.2 Statistical Evaluation of the Dynamic Pressure Loads on the Pool Boundaries 8 4 2 2.2 1 Dependence of Dynamic Bottom and Wall Pressures on System Pressure and Water Temperature 8 4 2 2 2 2 Occurrence Frequency Distributions of the Dynamic Bottom and Mall Pressures 8 4 2 2 2 3 Statistical Characteristics of the Dynamic Bottom and Mall Pressures 8 4 2 2 3 Temperature Variations in the Mater Region of the Test Tank 8 4 2 2.4 Mater Level in the Discharge Line When Opening and After Closing the SRV 8 4 3 Checking and Cal.ibr ation o f t he Measuring Instrumentation 8 4 4 Analysis of Measurement Errors 8 4 5 Repetition Tests and Reproducibility of the Results REV l, 3/79 8-4

8 5 DATA ANALYSIS AND VERIFICATION OF LOAD SPECIFICATION 8.5. 1 Evaluation of Test Tank Effects on Boundary Pressure Measurements 8.5. 1 1 Effects of Free Water Surface and Rigid Walls 8.5 1 2 Method of Images 8 5 1 3 The Test Stand as a Single Cell 8 5 1.4 Spatial Distribution of Pressure in the Test Tank 8.5. 1.5 Investigation of the Influence of Movable Walls on the Measurement Results (Fluid-Structure Interaction) 8 5 1 5 1 General Remarks 8 5 1 5 2 Experimental Investigation of the Tank's Natural Oscillations 8 5 1.5.3 Experimental Investigation of the Tank's Response to Vent clearing Loads 8 5 1 5.4 Theoretical Investigations and Model Calculations of the Influence of FSI 8 5.1.5.4 1 Computation Models 8 5 1.5 4 2 Model Parameters and Input for Calculations Without FSI (Rigid Tank) 8.5 1 5 4 3 Model Parameters and Input for Calculations With FSI 8 5.1.5 4 4 Results of the FSI Calculations 8.5.2 Verification of SRV System Load Specification Due to SRV Actuation 8 5.2 1 Pressures During the Vent Clearing Process 8 5 2 1 1 Vent Clearing Pressures for the Long Line 8.5 2 1 2 Vent Clearing Pressures for the Short line REV. l .3/79 8-5

8 5 2 1.3 Transposition of the Measurement Values to SSES and Comparison with the Design Specification 8 5 2.2 Pressures During the Stationary Condensation of Steam 8 5 2 2 1 Long Line 8 5 2 2 2 Short Line 8.5 2 2.3 Transposition of the Measurement Values to SSES and Comparison with the Design Specification 8.5 2.3 External Loads on the Quencher and Bottom Support 8.5.2 3 1 Vertical Force

8. 5. 2. 3. 1. 1 Measurement of the Vertical Force 85231.2 Measured Vertical Forces 852312.1 Long Line 8 5.2 3 1.2.2 Short Line 8 5 2 3 1.3 Transposition of the Measurement Values to SSES 8523131 Long Line 8 5 2.3 1,.3.2 Short Line 8.5.2.3 1 3-3 Summary 85232 Torsional Moment 852321 Measurement of the Torsional Moment 8.5. 2 .3 -2.2 Measured Torsional Moments 8 5 2 3 2.2 1 long Line 8 5 2 3 2.2 2 Short Line 852323 Transposition of the Measurement Values to 'SSES 85-23 3 Bending Momen ts at the Quencher Arms

8. 5. 2. 3. 3 1 Measurement of the Bending Moments R EV 1, 3/79 8-6

8.5. 2-3 3.2 Measured Bending Moments

8. 5. 2. 3. 3. 3 Transposition of the Measurement Results In to the Weld 8.5.2 3 3.4 Specified Static Equivalent Loads 8 5 2.3 3.5 E val uati.on of the Mea su remen t Results

8. 5-2.3. 4 Bending Moments at the Bottom Support 8 5 2 3 4 1 Measurement of the Bending Moments 8.5 2 3 4.2 Measured Bending Moments 8.5.2.3 4 3 Specified Static Equivalent Load 8.5 2 3 4 4 Evaluation of the Measurement Results 8.5.2 3 5 Forces on the Quencher 8-5 2-3.6 Influence of an Adjacent Quencher

~ 8-5.2 3 7 Loads on the Quencher During Steam Con densa tion 8 5 2 3 7 1 Manifestion Forms of Intermittent Condensation in the Karlstein Tests 852372 Illustration of the Measurement Values 8.5.2 3 7.3 Evaluation of the Measurement Results for the. Quencher Arm 8 5 2 3 7.4 Evaluation of the Measurement Results for the Bottom Support 8.5-2. 3 7 5 Evaluation of the Measured Torsional Moments 8 5 2 3 7 6 Evaluation of the Measured Maximum Moments at the Quencher Arm During Intermittent Condensation 8 5.3 Verification of Suppression Pool Boundary Load Specification Due to SRV Actuation 8 5 3 1 Evaluation of the Local Effects Seen at Pressure Transducer P5.5 8 5 3 2 Veri. f icat ion of t he S pecif ied Pressure Amplitudes and Vertical Pressure Profiles after Vent Clearing REV l~ 3/79 8-7

8 5 3 2 1 Overpressures 8 5 3 2 1.1 Vertical Pressure Profile 8 5.3 2 1 2 Vertical Pressure Profile Including Local Effects at P5.5 8 5 3 2 2 Underpressur es 8 5 3 2 2.1 Vertical Pressure Profile 8 5.3 3 Verification of the Pressure Time Histories Used for the SSES Containment Analysis 8 5.3.3. 1 Tranposition Method for the Oscillation Frequency 8 5 3 3 1.1 Calculation of Measured Oscillation Frequencies 8.5.3 3. 1 1. 1 PPGL Tests at Karlstein 853311.2 GKM Model Quencher Tests'KB 8.5.3.3. 1.1. 3 Hot Tests 8 5.3 3 1.1 4 Conclusion from the Frequency Calculations 8.5.3. 3 2 Multipliers for Conversion of the Bubble Frequencies From the Test Stand to SSES

8. 5. 3. 3. 3 Transposition Method for the Pressure Amplitudes 8 5 3 3.3.1 PPGL Quencher Tests at Karlstein 8.5 3 3 3 2 KMU Quencher Tests in the Model Test Stand in Karlstein 8.5 3.3.3 3 Analytical Calculations 8 5 3 3.3 4 Influence of Backpressure on the Pressure Amplitudes 8 5 3 3 4 Verification of Design Specification 8 5 3 3.4 1 Frequency Analyses of Selected Tests 8 5 3.3 4.2 Shif ting of the PSD's in the Transposition From the Test Stand to SSES REV. li 3/79 8-8

8 5 3.3 4.2 l Frequency Shift 8 5.3 3 4.2 2 Amplitude Stretching 85 3 3.43 Symmetrical Load Case (Simultaneous Blowdown of all 16 SRV's) 8 5.3.3.4 4 Unsymmetrical Load Case (Blowdown Via One SR V) 8 5 3 3 4 5 Unsymmetrical Load Case (Blowdown Via Three Ad jacent SRV')

8 5.3 3.4.6 Automatic Depressurization System (ADS)

Load Case 853347 Summary 8 5 3.3.5 Evaluation of the Measured Pressure Oscillations During Condensation 8.5 3 3.5 1 The Quencher is Cleared Continually 853 3.5 2 The Quencher is Not Cleared Continually 8 5.3 3.5 3 Condensation in the Blowdown Pipe and Thru the Sliding Joint 8 5 3.3 5 4 Transportation of the Measurement Results to SSES 8 5 4 Pool Mixing During SRV Actuation and Thermal Performance of the Quencher 8 5 4.1 Introduction 8.5. 4 2 'Equation of Motion of the Rotating Pool 8 5.4 3 Determination of the Flow Resistances 8 5 4 4 Determination of the Force Moving the Pool 8 5.4.5 Working Equations for the Rotating Pool of SSES 8 5 4.6 Estimate of the Heating of the Suppression Chamber Water 8.5.4.7 Experimental Proofs 8 5 4.7 1 Model Tank Tests 8 5 4 7 2 KKB Test During the Nuclear Commissioning REV l, 3/79 8-9

8. 5. 4. 7- 3 GKM Half Scale Quencher Condensation Test 8.5 4 8 Summary 8.5. 5 Verification of the Submerged Structures Load Specification Due to SRV Actuation 8.5. 5. 1 Loads on the Vent Pipe 8 5 5 1 1 Measurement of the Loads 8 5 5 1 2 Measured Bending Moments 8 5 5 1 3 Extrapolation of the Measurement Results and Comparison with the Specified Value 8.5 5.2 Influence of Expelled Hater During Vent Clearing 8553 Summary R EV l, 3f'79 8-10

SECTION 8.0 FIGURES Number Tit le 8-1 Mathematical Desscription of a Single Cell Configuration with Solid Walls; Solid Bottom and Free Water Surface 8-2 Eguivalence of a Single Cell Configuration and a Parallel Bubble Field Oscillating in Phase 8-3 Geometric Single Cell Partition of the Suppression Pool 8-4 Test Stand Schematic Diagram 8-5 Long Discharge Line Configuration 8-6 Short Discharge Line Configuration 8-7 Karlstein Test Tank Plan Viev Typical Vent Clearing Instrumentation 8-8 Karlstein Test Tank C-D Viev Typical Vent Clearing Instrumentation 8-9 Karlstein Test Tank A-B Viev Typical Vent Clearing Instrumentation 8-10 Karlstein Test Tank Plan Viev Typical Condensation Test Instrumentation 8-11 Karlstein Test Tank C-D View Typical Condensation Test Instrumentation 8-12 Karlstein Test Tank A-B View Typical Condensation Test Instrumentation 8-13 T-Quencher Showing Typical Vent Clearing Instrumentation 8-14 T-Quencher Shoving Typical Condensation Test Instrumentation 8-15 Test Matrig for Vent Clearing Test 8-16 Location of Test Group No. 1 in the Operation Field 8-17 Location of Test Group No. 2 in the Operation Field 8-18 Location of Test Group No. 3 in the Operation Field 8-19 Location'f Test Group No. 4 in the Operation Field REV 1, 3/79 8-11

8-2 0 Location of Test Group No. 5 in the Operation Field 8-2 1 Location of Test Group No. 6 in the Operation Field 8-22 Location of Condensation Tests in the Operation Field 8-23 Valve Opening Time Versus Accumulator Pressure Long Pipe Vent Clearing Tests 8-24 Valve Opening Time Versus Accumulator Pressure Short Pipe Vent Clearing Tests 8-25 Vent Clearing Pressure Versus System Pressure Long Line Vent Clearing Tests 8-26 Vent Clearing Pressure Versus System Pressure Short Line Vent Clearing Tests 8-27 Peak Positive Wall and Bottom Pressures Versus System Pressure Long .Line, Clean Conditions, Cold Pool 8-28 Peak Positive Mall and Bottom Pressures Versus System Pressure Short Line Clean Conditions, Cold Pool 8-29 Peak Positive Wall and Bottom Pressues Versus System Pressure Long Line Real Conditions, Cold Pool 8-3 0 Peak Positive Wall and Bottom Pressures Versus System Pressure Short Line, Real Conditions, Cold Pool 8-31 Peak Positive Mall and Bottom Presssures Versus System Pressure Long Line, Clean Conditions, Heated pool 8-32 Peak Positive Wall and Bottom Pressures Versus System Pressure - Short Line, Clean Conditions, Heated Pool 8-3 3 Peak Positive Wall and Bottom Pressures Versus System Pressure Long Line, Real Conditions, Heated Pool 8-34 Peak Positive Mall and Bottom Pressures Versus System Pressure Short Line, Real Conditions, Heated Pool 8-35 Peak Positive Mall and Bottom Pressures Versus Valve Actuation Long Pipe Test 14 8-36 Peak Positive Wall and Bottom Pressures Versus Valve Actuation Long Pipe Test 5 8-37 Peak Positive Mall and Bottom Pressures Versus Valve Actuation Long Pipe Tests 4 and 4R 8-3 8 Peak Positive Wall and Bottom Presures Versus Valve Actuation Long Pipe Tests 15 and 15R R EV 1, 3/79 8-12

8-3'9 Peak Positive Mall and Bottom Pressure Versus Valve Actuation Short Pipe Tests 19 and 19R 8-4 0 Peak Positive Mall and Bottom Pressures Versus Valve Actuation Short Pipe Tests 20 and 20R 8-41 Visicorder Trace P5 1-P5.10 Test 4 1.1 8-4 2 . Visicorder Trace P5.1-P5.10 Test 4R. l. 1 8-4 3 Visicorder Trace P5.1-P5.10 Test 4.1.6 8-44 Visicorder Trace P5.1-P5.10 Test 11.1 8-45 Visicorder Trace P5.1-P5.10 Test 12.1 8-46 Visicorder Trace P5.1-P5.10 Test 15.1.1 8-4 7 Visicorder Trace P5.1-P5.10 Test 15. Rl. 1 8-48 Visicorder Trace P5.1-P5.10 Test 19.1.1 8-4 9 Visicorder Trace P5.1-P5.10 Test 19.R2.1 8-5 0 Visicorder Trace P5.1-P5.10 Test 19.R2.2 8-51 Visicorder Trace P5.1-P5.10 Test 19. R2. 3 8-52 Visicorder Trace P5.1-P5.10 Test 19.R2.4 8-53 Uisicorder Trace P5. 1- P5. 10 Test 1 9. R2. 5 8-54 Visicorder Trace P5.1-P5.10 Test 19.R2 6 8-55 Visic order Trace P5 1-P 5 10 Tes t 1 9. R2. 7 8-56 Visicorder Trace P5.1-P5.10 Test 19 R2.8 8-57 Visicorder Trace P5.1-P5.10 Test 19. R2 9 8-58 Visicorder Trace P5.1-P5.10 Test 19.32.10 8-59 Visicorder Trace P5.1-P5.10 Test 20.1.1 8-6 0 Visicorder Trace P5.1-P5.10 Test 20.Rl.l 8-61 Visicorder Trace P5.1-P5.10 Test 20.R1.10 8-62 Visicorder Trace P5.1-P5.10 Test 21.1 8-63 Visicorder Trace P5.1-P5.10 Test 21. 2 8-64 Visicorder Trace P5.1-P5.10 Test 25. 1 8-65 Visicorder Trace P5.1-P5.10 Test 25. R2 REV lg 3/79 8-13

8-66 Maximum Resultant Bending Moment at Quencher Arm 1 ,

Long Pipe Vent Clearing Tests 8-67 Maximum Resultant Bending Moment at Quencher Arm 2 Long Pipe Vent Clearing Tests 8-6 8 Maximum Resultant Bending Moment at Quencher Arm 1 Short Pipe Vent Clearing Tests 8-69 Maximum Resultant Bending Moment at Quencher Arm 2 Short Pipe Vent Clearing Tests 8-7 0 Maximum Resultant 'Bending Moment at the Quencher Support Long Pipe Vent Clearing Tests 8-71 Maximum Resultant Bending Moment at the Quencher Support Short Pipe Vent Clearing Tests 8-72 Observed Condensation Phases During Tests 8-7 3 Typical Visicorder Trace of Stationary Operation of Quencher Test 33. 2-10 Seconds after Start 8-74 Typical Visicorder Trace of Stationary Operation of Quencher Test 35.1 Seconds after Start 8-75 Visicorder Trace Shoving Intermittent Operation of the Quencher Test 36.1 Sys em Pressure 6.2 1.0 bar Pool Water Temp 26~C 300C 8-76 Visicorder Trace Shoving Excerpt from Intermittent Operation of Quencher Test 36.1 280 Seconds after Start 8-77 Visicorder Trace Shoving Single Event Out of Intermittent Condensation Test 36. 1 8-7 8 Typical Visicorder Trace of Stationary Operation of Quencher Test 37.2 13 Seconds after Start 8-79 Typical Visicorder Trace of Stationary Operation of Quencher Test 39.1 10 Seconds after Start 8-80 Visicorder Trace Shoving Intermittent Operation of Quencher Test 40. 1 System Pressure 2. 5 bar Pool Water Temp. 89~C 91~C 8-81 Dynamic Bottom Pressures during the Blowdown Along the Upper and Lover Boundary of the Operation Field

'-82 Dynamic Wall Pressures During the Blowdown Along the Upper and Lover Boundary of the Operation Field R EV. 1, 3/79 8-14

8-83 Occurrence Frequency Distribution Positive and Negative Dynamic Amplitudes for the Condensation Tests Pool Temp. 22~C-300C 8-84 Occurrence Frequency Distribution Positive and Negative Dynamic Amplitudes for the Condensation Tests Pool Tem p. 59~C 91>C 8-85 Occurrence Frequency Distribution Positive and Negative Pressure Amplitude for Condensation Tests Pool Temp.

22 C 30 C 8-86 Occurrence Frequency Distribution Positive and Negative Pressure Amplitude for Condensation Tests Pool Temp 59 C 91 C 8-87 Nean Values of the Bottom Dynamic Pressures During the Blowdowns Along the Upper and Lower Boundary of the Operation Field 8-88 Mean Values of the Wall Dynamic Pressures During the Blowdowns Along the Upper and Lower Boundary of the Operation Field 8-89 Water Temperature Time Histories On Pool Wall Condensation Test 33.2 8-90 Water Temperature Time Histories On Pool Wall Condensation Test 35.1 8-91 Mater Temperature Time Histories On Pool Mall Condensation Test 37. 2 8-92 Rater Temperature Time Histories On Pool Mall Condensation Test 39.1 8-93 Water Temperature Time History On Quencher Arm 1 Condensation Test 33.2 8-94 Rater Temperature Time History on Quencher Arm 1 Condensation Test 35.1 8-95 Mater Temperature Time History on Quencher Arm 1 Condensation Test 37 2 8-96 Mater Temperature Time History on .Quencher Arm 1 Condensation Test-39.1 8-97 Calibration of Sensors and Registration Instruments 8-9 8 Intervals for Calibration Checks and Adjustments of Instrumenta tion 8-99 Calibration System REV. 1>> 3/79

8-100 Calibration Results Deviations from Nominal Value-P5. 1 P5-10 8-101 Mater Level in Discharge Line Test 15.1 8-102 Mater Level in Discharge Line Test 20.1 8-103 Water Level in Discharge Line Test 32 8-104 Effects of Free Surface and Rigid Tank Walls on Dynamic Fluid Pressure 8-105 Method of Images 8-106 SSES Smallest Unit Cell and the Karlstein Test Tank 8-107 Pressure Profiles for Different Bubble Locations 8-1 08 Pressure Profile for a One and Four Bubble Arrangement 8-1 09 Comparison of Measured and Calculated Normalized Pressure Profiles 8-110 Comparison of P ressure Prof iles Ca lcula ted for the Karlstein Test Tank and the SSES Suppression Pool 8-111 Comparison of Calculated and Specified Pressure Profiles 8-112 Tank Arrangement Showing Instrumentation and Explosive Charge Locations for Measuring Tank Reponse 8-113 Configuration of Explosive Container Used to Generate Underwater Pressure Impulse 8-114 Typical Tank Response Due to Pressure Impulse 8-115 Frequency Analysis of Gage MA2 8-116 Frequency Analysis of Gage MA7 8-117 Frequency Analysis of Gage MA8 8-118 Frequency Analysis of Gage P5.10 8-119 Displacement Correlations for 13 Hz Eigenmode 8-12 0 Displacement for the 13 Hz Eigenmode 8-121 Test Tank Arrangement for Shakedown Tests 8-122 Tank Displacements and .Pressure Trace During Shakedown Test 08.1 8-123 Frequency Analysis of Gage MA2 Shakedown Test 08.1 REV li 3/79 8-16

8-124 Frequency Analysis of Gage WA7 Shakedown Test 08.1 8-125 Frequency Analysis of Gage WA8 Shakedown Test 08 1 8-126 Frequency Analysis of P5 10 Shakedown Test 08.1 8-1 27 Air Nass Plow used for KOVlBl Computer Code Calculations 8-128 Unit Wall Displacement of 13 Hz Node Used in KOVlB1 Computer Code Calculations 8-129 Boundry Pressure Distribution Calculated for Unit Displacement of 13 Hz Node 8-13 0 Wall Presure Calculation with KOVlB1 Computer Code 8-131 Effects of FSZ on Bubble Prequency 8-132 Typical Pressure Trace in SRV Discharge Line Test 4.1.4 8-133 Typical Pressure Trace in SRV Discharge Line Test

20. Rl 7 8-134 Pressure in Steam Line before SRV Versus Pressure in Buffer Tank at Value Opening 8-135 Pressure in Discharge Line Versus Reactor Pressure at Vent Clearing P4.1 Long Line Tests 8-136 Pressure in Discharge Line Versus Reactor Pressure at Vent Clearing P4 4 Long Line Tests 8-1 37 Pressure in Discharge Line Versus Reactor Pressure at Vent Clearing P4.1 Short Line Tests 8-138 Pressure in Discharge Line Versus Reactor Pressure at Vent Clearing P4.4 Short Line Tests 8-139 Vent Clearing Pressure Versus Valve Opening Time 8-140 Steady State Pressure Versus Reactor Pressure - P4.1 Long Line Tests 8-141 Steady State Pressure Versus Reactor Pressure P4.4 Long Line Tests 8-l42 Steady State Pressure Versus Reactor Pressure P4. 1 Short Line Tests 8-143 Steady State Pressure Versus Reactor Pressure P4.4 Short Line Tests 8-1 44 Steady State Pressures at Different Locations Along the Discharge Line Extrapolated to 88 Bar Reactor Pressure R EV 1, 3/79 8-17

8-145 Typical Trace for Vertical Load Long Line Tests 8-146 . Vertical Load Versus Clearing Pressure Long Line Tests 8-147 Vertical Load Versus Vent Clearing Pressure Short Line

-Tests 8-148 Typical Trace for Torque on Bottom Support Long Line Test 8-1 49 Bottom Support Torque Versus Vent Clearing Pressure Long Line Tests 8-15 0 Bottom Support Torque Versus Vent Clearing Pressure Short l.ine Tests 8-151 Typical Trace for Bending Moments on Quencher Arms Long Line Tests 8-1 52 Resultant Quencher Arm Bending Moment Versus Vent Clearing Pressure Short Line Tests 8-153 Frequency Distribution of Maximum Resultant'Bending Moment on Quencher Arms and at Weld Seam 8-154 Resultant Bottom Support Bending Moment Versus Vent Clearing Pressure Short Line Tests 8-155 Frequency Distribution of Maximum Resultant Bending Moment on Bottom Support SG 4.5 4 6 8-156 Frequency Distribution of Maximum Resultant Bending Moments On Quencher Arms gt Strain Gages Intermittent Condensation 8-157 Frequency Distribution of Maximum Resultant Bending Moment at Weld Seam on Quencher Arm Intermittent Condensation 8-158 Frequency Distribution of Maximum Resultant Bending Moments at Bottom Support Intermittent Condensation 0.5 m below Quencher Center 8-159 Power Spectral Densities Test ll. 1 P5.5 8-16 0 Power Spectral Densities Test ll 1 P5 2 8-161 Power Spectral Densitie's Test 4.1. 6 P5.5 8-162 Power Spectral Densities Test 4.1. 6 P5.2 8-163 Power Spectral Densities Test 20.R1.10 P5 5 8-164 Power Spectral Densities Test 20- R1.10 P5.2 R EV 1 3/79 8-18

8-165 Maximum Specified Vertical Pressure Profile and Measured Maximum Values Overpressures 8-1 66 Maximum Specified Vertical Pressure Profile and Measured Maximum Values Considering. Local Effects 8-167 Maximum Specified Vertical Pressure Profile and Measured Maximum Values Under Pressures 8-168 Karlstein Tests Comparison of Measured and Calculated Bubble Frequency 0$ Humidity 8-169 Karlstein Tests Comparison of Measured and Calculated Bubble Frequency 100% Humidity 8-170 GKM Tests Comparison of Measured and Calculated Bubble Frequency 8-171 GKM Tests Comparison of Measured and Calculated Bubble Frequency Over pressure 8-1 72 KKB In-Plant Tests Comparison of Measured and Calculated Bubble Frequency 8-173 SSES Calculated Bubble Frequencies 8-174 Multipliers for Conversion of Bubble Frequencies the Karlstein Test to SSES 8-175 Overpressure Multiplier for Conversion of Bubble Frequencies 8-176 Normalized Amplitude Spectrum Versus Bubble Frequency-Karlstein Tests 8-1 77 Karlstein Model Tests Influence of Mater Surface on Pressure Amplitude 8-178 GKM Tests Influence of Overpressure on Bubble Pressure 8-179 PSD of Karlstein Tests 11.1 and 12.1 P5.10 8-180 PSD of Karlstein Tests 4.1.1 and 4.1.6 PS.10 8-181 PSD of Karlstein Tests 21.1 and 21.2 P5.10 8-182 PSD of Test 20.R1 10 P5.4 8-183 PSD's of Test 11.1 P5.2, 5.4 and 5.10 8-1 84 PSD Comparison Test 20 Rl 1 and Design Specification 8-185 PDS Comparison Test 4.1.1 and Design Specification REV lg 3/79 8-19

8-186 PSD Comparison Test 20.R1.10 and Design Specification 8-187 PSD Comparison Test 21 1 and Design Specification 8-188 PSD Comparison Tests 21.2 and 25.R2 and Design Specification 8-1 89 PSD Comparison Test 0.1.6 and Design Specification 8-190 IBA Drywell and Wetvell Pressure History 8-191 PSD Comparison Test 11.1 and Design Specification 8-192 Typical Cross Section of SSES Suppression Pool 8-193 Revised Quencher Arrangement 8-1 94 Velocity of Rotating Pool for One Actuating Valve in Outer Row 8-1 95 Water Motion of the Acceler ated Pool 8-196 Test Stand for Measuring Thrust 8-1 97 Measured Temperature Distribution in the KKB Suppression Pool 8-1 98 Resultant Bending Moment on Dummy Vent Versus Reactor Pressure 8-199 Resultant Bending Moment on Dummy Vent Versus Clearing Pressure 0~

8-2 00 Resultant Bending Moment on Dummy Vent Versus Pressure Amplitude at P5.7

'J 8-2 01 Specified Pressure Distribution on Dummy Vent 8-2 02 Typical Visicorder Trace for Bending Moment on Dummy Vent REV li 3/79 8-20

PROPRIETARY SECTION 8 TABLES Number Title 8 1 Typical Operating Instrumentation 8.2 Typical Vent Clearing Test Instrumentation 8.3 Typical Condensation Test Instrumentation 8.4 Parameters at Test Start Long Pipe Vent Clearing Test Series 8 5 Parameters at Test Start Short Pipe Vent Clearing Test Series 8.6 Parameters at Test Start Condensation Test Series 8.7 Behavior of the SRV and System Pressures Long Pipe Vent Clearing Test Series 8.8 Behavior of the SRV and System Pressures Short Pipe Vent Clearing Test Series 8.9 Peak Dynamic Pressures on the Pool Boundary During Vent Clearing Long Pipe Vent Clearing Tests 8.10 Peak Dynamic Pressures on the Pool Boundary During Vent Clearing Short Pipe Vent Clearing Tests 8.11 Maximum, Strains, Moments and Vertical Loads on the Quencher Arms and Support During Vent Clearing Long Pipe Tests 8.12 Maximum Strains, Moments and Vertical Load on the Quencher Arms and Support During Vent Clearing Short Pipe Tests

8. 13 System Pressures and Pool Water Temperatures of the Condensation Tests 8 10 Peak Dynamic Pressures Amplitudes During the Different Condensation Phases 8.15 Statistical Characteristics of the Bottom Dynamic Pressures (P5. 2)

8. 16 Statistical Characteristics of the Wall Dynamic P ressures (P5. 10)

8. 17 Repetition Tests Comparison of Recorded Valves 8.18 Repetition Tests Mean Values and Deviations REV. 1, 3/79 8-21

PROPRIETARY 8 0 SSE~SUENCHER VERIFICATION TEST 8 1 I NTRODUCTION

8. l. 1 P u~rose o f the Tests The optimized quencher design for SSES and the load specification on the wetted boundaries of the suppression pool, on the submer'ged structures and on the pressure relief system, are based on parametric model test studies and full scale inplant test results from a similar quencher design. The load specifications for the SSES quencher are described in detail in Section 4.1. In order to verify these load specifications and further verify the quencher's steam condensing characteristics, full scale single cell tests were conducted at the Kraftwerk Union laboratories in Karlstein, West Germany.

8.1.2 Test Concept The concepts used to design and perform the tests were:

1) Use of a conservatively defined single cell

2) The close simulation of the main sa fety relief valve system parameters 8 1.2.1 Unit Cell Approach 8 1.2 1.1 Single Cell Theory For a gas bubble oscillation in a free water space, the water mass coupled to the bubble is alternately accelerated and decelerated During this process the overpressure and underpressure amplitudes decrease with increasing distance from the bubble. When a solid wall is placed near the oscillating bubble, the water acceleration is restricted in the direction of the wall and the decrease in pressure amplitude in the direction of the wall is less. This effect can be expressed mathematically by replacing the. bubble by a potential source and accounting for the wall by the method of images. The effects of the real source and the image source are added for each point of the flow field.

For the case in which a bubble is enclosed in a narrow water space, closely surrounded by solid walls and a solid bottom with a free water surface at the top, the water space below the bubble is for all practical purposes unmoved. Only the water volume above the bubble is free to oscillate.. Consequently, the pressure gradient in the lower water space is nearly zero, while the pressure amplitude above the bubble decreases with increasing proximity to the water surface. The pressure amplitudes are zero at the water surface and the method of images applies.

REV. 1, 3/79 8-22

PROPRIETARY Analytically, the case in which a planar field of uniform strength bubbles are all oscillating in phase is the same as the case in which solid walls exist between each of the individual bubbles. The single cell test configuration used at Karlstein simulates this extremely conservative case of parallel bubbles oscillating in phase with the same source strength. A description of the equivalence of the single cell configurations, using the method of images, is contained in Figures 8. 1 and 8. 2.

For a more detailed evaluation of the Karlstein test tank single cell, see Section 8. 5.1.

8 1.2 1.2 A~lication of Single Cell ~Aroach The submergence of the quencher in the test tank is equal to the highest value in the plant. As to the water cross-section area the single cell theory described above is used Figure 8.3 shows a geometrical partition of water space. The water cross-section areas related to the different quenchers are listed below:

Average Water Related Water Surface Surface Quencher A 31 47 mz (338 62 ftz) 21. 4 mz (230- 26 ftz)

Quencher B 31 47 mz 21 4 mz Quencher C 31. 47 mz 31.3 mz {336.79 ftz)

Quencher D 31 47 mz 42 m z (451 92 ftz)

Quencher E 31 47 52 31~3 mz Quencher F 31 ~ 47 AI 2 31' mz Quenc her G 31.47 mz 42 mz Quencher H 31. 47 mz 31m 3 mz Quencher J 31.47 mz 31~3 mz Quencher K 31 47 mz 42 mz Quencher L 31 47 mz 31~3 mz Quencher 31 47 mz 31a3 mz Quencher N 31.47 mz 21 4 mz Quencher P 31.47 mz 21.4 mz Quencher R 31. 47 mz 31~3 mz Quencher S 31.47 mz 42 mz REV 1, 3/79 8-23

PROPRI ETAR Y The smallest water surface (approximately 21.4 m~) is simulated in the tests. Therefore, the dynamic pressure amplitudes at the walls and the bottom are measured under conservative boundary condi tions.

8.1.2 2 Simulation of SSES Parameters The fo'llowing section provides a description of those parameters that were simulated in the Karlstein test facility These parameters are typical of most MK II plants. For more detail on the test facility see Section 8.2 8 l. 2 2. 1 Pri ma~rSgs tern Pressure The reactor operating pressure for SSES is approximately 1000 psig (69 bar) while the highest pressure set point for any SSES Safety Relief valve is 1205 psig (83 bar), which is close to the highest primary pressure that can be simualted in the Karlstein test facility (82 bar) . This allowed the test simulation to very closely match the range of initial primary system pressures that .

can be expected in the operating plant.

8.1 2 2 2 Saf et~Relief Val'vegSRVJ In order to match the characteristics of the Safety Relief Valve, an original Crosby SRV, shipped directly from the plant site, was installed in the test stand and used in all tests.

8 1 2 2 3 Discharge J.ine In order to cover the range of discharge 'line lengths and therefore air volumes that exist in SSES, two vent clearing test series were run; one with a discharge line that simulates the longest SSES discharge line {48 m) and one that simulates the shortest SSES discharge line (35 m) . In addition, the number of bends in each line, the i'nner diameter of the main part of the (303.9 mm), and the inner diameter of the last vertical run

'ine to the quencher (288 9 mm) are closely simulated to that which exists in the SSES plant. (schedule 40 pipe and schedule 80 pipe, respectively). In addition a 24 ft. submergence, corresponding to the highest water level in the suppression pool, was used for all tests 8.1 2.2 4 Vacuum Breakers In order to closely simulate the effects of vacuum breaker operation on the tests, two six-inch diameter Crosby vacuum breakers were shipped to Germany and installed in the test stand at the same relative location as planned for the SSFS plant.

REV li 3/79 8-24

PROPRIETARY 8 1.2 2.5 Ouencher A full size prototype of the quencher installed in the SSES plant was installed in the test facility and used for all tests.

Figure 8.13 shows the quencher with instrumentation for vent clearing tests while figure 8.14 shows the quencher with instrumentation for the condensation tests.

8-2 TEST FACILITY AND INSTRUHENTATION 8.2 1 Test Facility 8.2.1.1 Mechanical Set-up The test configuration as constructed is typically illustrated diagrammatically in Figure 8.4. The test stand configuration can be divided into:

the steam boiler, the steam accumulator, the steam line before the SRV and the buffer tank, the SR V, the discharge line between the SRV and the water pool with the quencher as pipe termination, and the large tank as water pool.

8.2 1 1 1 Steam boiler The steam boiler is an oil-fired, once-through, forced-flow boiler with an output of approximately 20 HW at a maximum steam pressure of 170 bar (2499 psig) and a maximum steam temperature of 520~ C (968~ F) . The .boiler is designed for a closed operating mode in normal operation. A fraction of the boiler's output is recovered from the condensate via the high-pressure cooler. When there is an open loop {i e., lost condensate), the output is reduced. The steam flow available in this mode is approximately 8 to 9 kg/s (17.6 to 19.8 ibm/s) . The lost condensate results in a time limitation on continuous output.

The feedwater supply of the boiler is about 20 m3 (705 ft~).

Once that amount is used up, further steam supply as continuous output is possible only up to the output of the feedwater conditioning system. That amounts to 5 m~/h (176 ft3/h). For longer test periods it is necessary to interrupt operation for 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> in order to refill the feedwater storage tank.

8 2.1.1.2 Steam Accumulator As described in 8.2.1.1.1 the amounts of steam supplied continuously by the boiler are too small to test an SRV.

REV. 1, 3/79 8-2 5

PROPRIETARY To provide a way to test valves at flow rates of up to approximately 22 kg/s (484 ibm/s), a valve test facility was built using the boiler plant and a pressure vessel connected to it. This vessel is charged with a steam/water mixture by the boiler and is used as a steam accumulator. From this steam accumulator, higher steam flow rates can be delivered for a short period of time The dimensions of the pressure vessel are 1.5 m diameter and 12 m high, which results in an accumulator volume of approximately 22 m3.

Adapted to the required steam output, the accumulator is filled with saturated water and saturated steam at the specified ratio.

The steam is drawn downward through a standpipe. The high steam, flow to be extracted transiently from the accumulator results in a rapid decrease of pressure and temperature. For strength reasons, the temperature difference between the inside and outside of the accumulator vessel must not exceed a certain value. This limits the maximum pressure drop and thus the available test time.

8 2.1.1.3 Steam Line and Buffer Tank The connection between the steam accumulator and the valve test stand consists of an ND 250 pipe line. This line contains isolating devices for emergency isolation and a measurement section constructed as a Venturi nozzle. The existing eguipmeat provide for a direct horizontal connection of the valve being studied. This corresponds to the design of the SRVs used in German BMR plants and to their arrangement at the end of a tap line coming from the main steam line.

The steam supply line was rebuilt to match the design features of SSES. The previously described pipe line now ends in a T-piece.

Xn order to simulate the SSES main steam line and to keep the steam supply flow to the valve as uniform as possible, a buffer tank having a volume of 5.2 m3 was connected to the second horizontal outlet of the T-piece.

The vertical outlet of the above-described T-piece leads to the valve.

8.2.1 1 4 Safet~Relief Valve QSRVQ The SRV used in the tests is the actual version being used for SSES. These valves are arranged vertically, have a steam inlet from below and an outlet to the side. As described in 8. 2 1 1.3, the steam supply line was rebuilt in such a way that the same arrangement was possible in the test stand. The valve was mounted on the T-piece, using the same connection dimensions as in the actual plants.

REV 1, 3/79 8-2 6

PROP RI ETAR Y Operation of the valve during the tests requires the connection of power supply lines, control lines and measurement lines. The existing equipment at the valve test facility was used to satisfy most of those requirements. Some modifications became necessary in order to adapt to the construction of the valve. The SRVs in German BMR plants are operated by an electrically actuated pilot valve with its own operating medium. In contrast, the SS'>>S valve used in the test was opened pneumatically. Accordingly, the compressed-air connection was rebuilt so that the opening conditions in the actual plant could be simulated in the test stand.

8.2~1 1.5 Discharge Line and Quencher The SRV described .in 8.2.1.1.4 discharges on the exhaust-steam side into a pipe which represents the SRV discharge line . The length of the SRV discharge line and the number of bends are different for the 16 SRV's for SSES. Two line lengths were used for the tests, corresponding to the longest and shortest lengths of the SRV discharge lines in the plant. Isometric drawings of the two discharge lines are shown in Figure 8.5 (long line) and Figure 8.6 (short line) .

Pipe supports and vibration dampers were mounted at the required places. These places were not identical to the corresponding ones in the plant, because the mounting situations and especially the concrete construction of the plant cannot be-simulated directly in the test facility.

To prevent the buildup of a large underpressure in the pipe, two actual vacuum breakers were installed in a vertical part of the pipe line, as in the plant.

The quencher forms the termination of the SRV discharge line (see Figure 8 ll). The steam is conducted into the water through a large number of holes having a diameter of 10 mm. The design of the quencher is described in detail in Section 4.1.

bottom support is provided to hold the quencher in place in the A

test tank It connects the quencher rigidly to the bottom of the tank and is constructed in such a way as to make it possible to measure the loads exerted on the quencher due to vent clearing processes and steam condensation. The sliding joint provided between the quencher and the discharge line in the plant is simulated in the test stand hgdraulicallg by a corresponding annular gap 8 2 1 1.6 Test Tank For SSES, the exhaust steam from the relief valves is conducted into the suppression pool and is condensed there. Xn the test facility, a section of that pool is simulated by a stiffened REV- lg 3/79 8-27

PROPRIETARY steel tank (see Figures 8. 7, 8 8, 8 9) . In the plant, the suppression pool can be subdi'vided conceptually into suhspaces, each of which is associated with a steam supply line (see Figure 8.3). In order to adapt the conditions in the test tank to the dimensions of the smallest geometrical single sell, concrete shaped blocks were inserted into the test tank. The concrete shaped blocks are clearly illustrated in Figure 8.7. The exposed cross-sectional area of the water space is 7.2 m x 3.15 m = 22.7 m~. It corresponds conservatively to the smallest individual cell in the pla nt.

Illuminating devices and viewing ports made possible the direct observation and also photographic recording of the underwater processes.

8.2.2 Instrumentation Instrumentation is provided for controlling the test procedure, determining the prescribed measurement quantities, and recording the m.

8 2.2 1 General Description The instrumentation used in the Karlstein test facility consists of operating instrumentation and test instrumentation. Operating instrumentation assures the control of the test facility and its environment correlation. The test instrumentation records the load data which is used to verify the conservatism in the design loads as specified for the SSES in section 4.1 of this Design Assessment Report.

Details on the operating instrumentation are given in Section 8.2.2.3. A detailed description of the test instrumentation can be found in Section 8.2.2 4 8.2.2 2 Instrumentation Identification For identification, the measuring sensors are designated according to a system of letters and figures. The first one or two characters are letters which identify the type of instrument:

P Pressure Transducer T Temperature Sensor (Thermocouple)

F Flow Rate Measurements L Rater Level Measurements DG Displacement Gage SG Strain Gage ILP Electrical Impulse Signal Level Probe REV. l, 3/79 8-28

PROPRI ETAR Y These letters are followed by a number which characterizes the location within the test facility where the instrument is situated. The facility was divided into sections as follows:

Section 1 contains the steam supply, including the accumulator {only transducers of the test stand instrumentation system are contained in this section) .

Section 2 contains the steam line up to the safety relief valve and includes the buffer tank.

Section 3 contains the safety relief valve.

Section 4 contains the discharge line and quencher.

Section 5 contains the test tank.

The sensor designation is completed by adding a decimal point and a sequential number. For example, "P5.6" means: the number 6 pressure transducer in the test tank.

Additional abbreviations used are as follows:

DPS Data Processing System CTC Coated Thermocouple DCA Direct Current Amplifier CFA Carrier Frequency Amplifier HT-SG High Temperature Strain Gage SRV Safety Relief Valve PG Pressure Gage RTD Resistor Temperature Detector 8 2.2.3 Operating 1nstrumentation The operating instrumentation is provided for measurement of parameters in relation to the steam accumulator, the steam lines and the SRV's A total of 30 sensors can be recorded by a process computer which is part of the operating instrumentation system. The data are stored on a magnetic disk and can be printed out The recording frequency of the process computer was adapted to align with the instrumentation chanriels, covering a range from 0.5 Hz, for those sensors where only small transients are to be expected, up to about 200 Hz for the sensors where higher frequency signals are expected (e.g. for pipe vibrations)

The operating instrumentation comprises the measuring devices used to monitor and control the system and also the data acquisition devices needed for that purpose. Typical measuring locations for the tests are illustrated in Figure 8.4 and listed in Table 8.1.

BEV 1, 3/79 8-29

PROPRIETARY According to the type of acquisition and display, the measurement sensors can be classified into two groups= "Display on Control Console" and "Acquisition by Computer".

8 2.2.3.1 Disp~la'n Control Console To enable the operating personnel to control the test equipment, a number of quantities which characterize the operating condition of the system are displayed continuously.

In particular, they are:

Water level in : Steam accumulator, steam line, buffer tank, discharge line, test tank Pressure in Steam accumulator, buffer tank, control line, discharge line Temperature in : Steam accumulator, buffer tank, discharge line, test tank a

8 2 2.3.2 A~cuisition ~b Computer Most of the data sensors comprising the operating instrumentation are interrogated by a computer at prescribed time intervals before, during and after the te. t. The values are stored on a d isk.

The data are printed out at programmed intervals. At an interrogaticn f requency of 200 Hz, the capacity of the storage device is sufficient for a recording time of 2 minutes.

The following measurement values are interrogated:

Water level Steam accumulator, buffer tank discharge line, test tank Pressure Steam accumulator, buffer tank, steam line, control line, discharge line Temperature Before SRV, after SRV, surface of SRV, discharge line, test tank Vibrations Steam line before SRV, discharge line Valve travel SRV, vacuum breakers Switching time Electrical energization of SRV REV 1 g 3/79 8-3 0

PROPRIETARY 8 2 2.4 Test, Instrumentation Mesurement values used to verify the test tasks are determined by the test instrumentation. It is necessary to include here a few typical measuring points that are already used for monitoring purposes in the operating instrumentation on the pipes .and SRV.

Since most of these processes are of a high-frequency nature, the data is acquired in analog form by means of carrier-frequency measuring amplifiers and dc amplifiers on analog magnetic tape, and to a large extent also on visicorders. The visicorder traces allow an initial review and a pre-evaluation of the test data.

8 2.2 4.1 Measuring Points Measurements are made of the pressure on the steam line before the SRV; valve actuation and valve travel; pressure variation in the discharge line at four points between the SRV and quencher; temperature in the discharge line at three points between the SRV and quencher; water level in the discharge line before the quencher inlet at four positions for the long line and five positions for the short line; bending, axial and torsional strains on the bottom support; bending strains on the quencher; bending strain on a dummy vent pipe; temperature distribution in the test tank; temperature distribution at the quencher for the condensation test; wall pressures and bot tom pressures in the test tank.

Typical measurement points for the vent clearing tests are illustrated in Figures 8.7, 8.8, 8.9 and listed in Table 8.2.

Typical measurement points for the condensation tests are illustrated in Figures 8. 10, 8.11, 8.12 and listed in Table 8. 3.

8.2.2 4.2 Set~u of Measuring Instruments All instrumentation is channelled to one central station situated in the control room of the laboratory.

Each instrumentation channel consists of the individual sensor, connecting cable, amplifier (carrier frequency amplifier or direct current amplifier), attenuator; and are recorded on magnetic tapes and visicorders, most channels being in parallel on both systems. Three magnetic tape recorders and three visicorders were used in the control room. Each unit allows the recording of 12 channels and, in addition, a time reference signal and a physical correlation trace.

REV 1, 3/79 8-31

PROPRIETARY The sensors are connected by shielded cable to the amplifiers vhich are located near the recorders in the control room. For the strain gages, displacement gages and pressure transducers, carrier frequency amplifiers vere used which allow a frequency resolution of up to 1 KHz. For temperature measurements, direct current amplifiers (10 Hz) vere used together vith a 10 Hz lov pass filter.

8 2.2 5 Visual-Recording Three high-speed cameras vere used to film the processes in the pool during the blowdovn through the quencher. KMU uses a "HYCAM 120 m~'or that purpose. Tvo LOCAM cameras (model 51-0003) were being made available by the Standford Research Institute (SRI)

The positioning of the cameras was as follovs:

HYCAM camera in front of one bull's eye at quencher height; LOCAM camera 1 in front of one bull's eye at a tank height of approximately 4m; LOCAM camera 2 on the service platform above the tank at a height of approximately 9 m.

A correlation between the moving pictures and the data recordings on the Visicorder and magnetic tape vas accomplished by means of a timing mark on the fi'lms.

8 3 TEST PARAMETERS AND MATRIX 8.3 1 Vent Clearing Tests The test matrix for the vent clearing tests is presented in Figure 8.15. This figure shows the test number and parameter conditions used for each test.

.The number of basic tests was 25. These 25 tests were split into 5 groups of tests where by each group covered a set of test parameters. Tests numbered 26 to 32 were additional tests vhich were not required to verify the quencher design but which could prove useful in evaluating the performance of the safety relief system. Tests number 27, 28, 30 and 31 were to investigate shorter than normal SRV opening times, but, as valve opening times vere found to be quite fast, these tests were not added to the required tests. Tests number 26 and 32, with one locked vacuum breaker, were included into the test matrix. The results shoved the effect of the locked vacuum breaker to be minimal so test number 29 was not added.

R EV 1, 3/79 8-3 2

PROPRIETARY The allocation of each test group within the operation range of the safety relief system is shown in Figures 8.16 to 8.21 by test points.

Base parameters in Group 1 (Figure 8 16) are long discharge line

'length, normal discharge line air temperature, normal initial water level inside the discharge line and normal valve opening time. Each of the following groups vary one or more of these Group 1 base parameters; Group 2 (Figure 8 17) uses a low initial water level inside the SR V pipe; Group 3 (Figure 8.18) uses a high discharge line temperature; Group 4 (Figure 8.19) uses a short discharge line length and Group 5 (Figure 8.20) uses a short discharge line length and a high discharge line temperature.

Each of the basic 25 tests was comprised of two or more valve actuations where by only the first actuation is made at t,he specified conditions of the discharge line (so-called clean condition) . Any other actuation was made at the prevailing discharge line temperature and water level (so-called Real Condition) . In the case of only two actuations at a test point the time interval between the actuations was approximately 10 minutes. In the case of multiple actuations at a test point the time intervals between actuations were varied as follows:

For test points 4, 5, 14, 15 the time between successive actuations was l. 5/5/15/30/60/120 seconds, accounting for seven valve actuations.

For test points 19 and 20 the time between successive actuations was 1 5/5/15/30/60/120/5/15/600 seconds, accounting for ten valve actuations.

For vent clearing tests with only two SRV actuations, the hold-open time for the SRV was 2 seconds while for the multiple value actuation tests the hold-open time was 1. 5 seconds.

'Five test points were repeated, these were test points 4, 15, 19 and 25. Repeat tests at a designated test point are indicated '0 with a letter R in the test number i.e. Test number 20.Rl.l is the first value actuation of the repeat test at test point

20. 1 1.

A compilation of actual parameters at the start of each test is tabulated in Table 8.4 for the long pipe test series and Table 8 5 for the short pipe test series.

8.3 2 Condensation Tests In order to further verify the steam condensation capabilities of the quencher device and provide specific information regarding its steam condensation capabilities for the safety relief system l

REV. 1, 3/79 8-33

PROPHI ETAR Y operation range a series of eight extended blowdown tests were performed. These tests are designated as test numbers 33 to 40.

Each test was performed with the short discharge line configuration as described in section 8.2.1. 1.5 and with an initial discharge line temperature of approximately 90~C.

The location of the initial system conditions for each test point is plotted on the safety relief system operation range in Figure 8 22 In order to initiate each test the SRV was actuated as was done in the vent clearing tests. The valve then remained open -until the system pressure reached the predesignated value for that test. At this time the valve was closed and the test was completed. The total allowable pressure drop in the accumulator tank for each initial system pressure dictated the duration of each bio wd own.

A compilation of actual parameters at the start of each test point in the condensation tests matrix is tabulated in Table 8.6.

8 4 TEST RESULTS This section provides a compilation of the test results for the vent clearing and steam condensation tests conducted at the Kraftwerk Union laboratories in Karlstein, West Germany in order to verify the load specification and steam condensing characteristics of the quencher design for the Susquehanna Steam Electric Station. Included in this section is information about the boundary conditions at the beginning of each test, of the behavior of the SRV, primary system pressures, the'esults dynamic pressure loads on the pool boundaries and their primary freguency and the loads on the quencher and bottom support This information is provided in the form of tables, figures and actual visicorder recordings.

8 4 1 Vent Cleari~n Test Results Nineteen tests with a total of 67 vent clearing processes were performed with the long discharge line in the period from May 8, 1978 to June 7, 1978 and 13 tests with a total of 58 vent clearing processes were performed with the short discharge line in the period from June 27, 1978 to July 7, 1978.

8 4 1.1 Test Parameters The most important of the parameters being investigated was described in Section 8.3. A detailed list of test parameters for each valve actuation is given for the long discharge line tests in Table 8.4 and for the short discharge line tests in Table 8.5.

This includes R EV. 1, 3/79 8-34

PROPRIETARY type of test length of discharge line accumulator pressure water temperature in the test tank water level in discharge line air temperature in discharge line The accumulator pressure P1.1A and the buffer tank pressure P2.6A are the determinative values for the system pressure at the start of each test. The values were read by computer just prior to the start of the test. In addition these pressures were stored continuously on magnetic tape. If a long period passed between the last computer reading and the actual test start then the initial values for the accumulator pressure were taken from the corresponding computer plots The initial accumulator pressures were also read from those plots for the multiple valve actuation tests.

For accumulator pressures below 30 bar (435 psi), measuring point P2.5 was used to determine the system pressure, since measuring points Pl. 1A and P2.6A were outside the measuring range.

The water temperature at the start of the test was taken either from the computer listings or, in the multiple valve actuation tests, from the computer plots Due to the inertia of the Barton cell, the measurement value for water level in the discharge line (measuring point L4.1) in the multiple actuation tests, especially for the 2nd, 3rd and if applicable, the 8th actuation, must be disregarded or considered only as an indicative value.

The temperature in the discharge line at the start of each test was taken from the computer listings or the computer plots for the multiple actuation tests 8 4 1.2 Behavior of the SRV and System Pressures To evaluate the valve behavior, the valve opening time, t0 , was determined from the recorded valve lift variation for all tests.

This involves the time from the beginning of valve opening until attainment of the steadystate lift (see sketch below) . These opening times are listed, for the long discharge line tests, in Table 8.7 and, for the short discharge line tests, in Table 8.8.

The associated steady state lifts are also indicated. A plot of the measured valve opening times as a function of accumulator pressure at the start of each test is shown in Figure 8.23 for the long discharge line tests and Figure 8.24 for the short discharge line tests.

The so-called vent clearing times tpz are also given in Tables 8.7 and 8 8 This is the time from the beginning of valve R EV. 1, 3/79 8-.35

PROPRIETARY opening until the instant 'of maximum pressure at measuring point P4.4 in the discharge line. (see sketch below) t s valve lift vent clearing pressure pressure before quencher 1'wo values are indicated in Tables 8.7 and 8.8 for system pressures measured in:

buffer tank P2. 6 bef ore the SR V P2. 5 in the discharge line P4. 1 to P4.4 These two values are the pressure at the vent clearing time (vent clearing pressure) and the pressure approximately 1.5 seconds after the start of test (steady pressure)

The initial parameters of relevance for the classification of-tests are indicated in the row headings.

The vent clearing pressure in the discharge line before the quencher inlet (measuring point P4.4) is plotted versus system pressure (measuring point P2.6) under Clean Conditions in Figure 8.25 for the long discharge line tests and in Figure 8.26 for the short discharge line tests. See Section 8.5.2.1 for a discussion of the vent clearing pressures and their dependence on reactor pressure.

8 4.1.3 Dynamic Pressure Loads on the Pool Boundaries As read off the Visicorder traces, the peak positive and peak negative pressure amplitudes during vent clearing for measuring points P5. 1-P5.3 (bottom pressures) and P5. 4-P5. 10 {wall pressures) are compiled in Table 8.9 for the long discharge line tests and in Table 8. 10 for the short discharge line tests. In a'ddition, approximate values for the predominate frequency of the pressure oscillations are indicated. These frequencies were read from the visicorder traces.

Figures 8.27 and 8.28 show the measured peak positive pressure amplitudes at the tank bottom directly beneath the quencher (P5.2) and on the concrete wall at the quencher's mid-height R EV 1, 3/79 8-36

PROPRIETARY (P5.10) as a function of system pressure for the long discharge line and short discharge line tests. The test points plotted are all Clean Condition tests with cold water in the test tank

{approximately 25~ C) and discharge line cold (approximately 50~

C) (Long discharge line tests 1.1, 2.1, 3. 1, 4.1 1, 4.81.1 and 32.1 and short discharge line tests 16.1, 17 1, J.8.1, 19.1.1 and 19.R1.1)

As a comparison Figures 8.29 and 8.30 represent corresponding measuring points for tests performed under Real Condition (Long discharge line tests l. 2, 2. 2, 3. 2,.10.4 and 32. 2 and short discharge line tets 16. 2, 17. 2 and 18. 2) . As can be seen the pressure amplitudes are slightly higher for the Clean Condition tests and no significant change with system pressure is observed.

Figures 8.31 and 8.32 show the measured peak positive pressure amplitudes at measuring points P5.2 and P5.10 for Clean Condition tests with heated water {45 C 80~CO in the test tank for the long discharge line tests and short discharge line tests respectively. (Long discharge line tests 5. 1. 1, 6.1, 7 1, 8.1, 9.1, 15 1.1 and 15.R1.1 and short discharge line tests 20.1.1, 20.Rl.l, 22.1, 23.1, 24.1). Again, as a comparison, Figures 8.33 and 8.34 represent corresponding measuring pcints for tests performed under Real Conditions (Long discharge line tests 6.2, 7.2, 8.2, 9 2, 11. 2 and 12. 2 and short discharge line tests 20.R1.7, 22.2, 23.2 and 24.2) In contrast to the tests with cold water in the test tank, the pressure amplitudes are slightly higher for the Real Condition tests, but as with the cold water tests, no significant change with system pressure is observed.

Figures 8. 35 to 8.40 show the measured peak positive pressure amplitudes at measuring points P5.2 and P5. 10 for a number of multiple valve actuation tests plotted against the corresponding valve actuation.

Figures 8.41 to 8 65 show the first second of visicorder pressures traces (for the pool boundary pressures, P5. 1-P5.10) from various tests.

8~4 1 4 Loads On The quencher and Bottom Support The bending strains on the two arms of the quencher and at the bottom support were each measured in two mutually perpendicular directions. The resultant bending strains and bending moments were calculated from these individual values. The strain-versus-time varia tions stored on magnetic tape were read for the maximum resultant during vent clearing. A high-pass filter having a cutoff frequency of 2 Hz was inserted in order to rule out any falsification of the evaluation due to slow drifting of the zero point The upper frequency limit was at 400 Hz due to the m echa nica 1 co nd iti on s.

REV 1~ 3/79 8-37

PROPRIETARY The maximum resultant bending strains determined in this manner and the bending moments calculated from them are compiled in Tables 8.11 and 8.12 for the long and short discharge line tests respectively. To clarify the direction distribution of the resulting bending moments on the quencher arms, the components of the maximum resultant bending moments are depicted in polar coordinates in Figures 8.66 and 8.67 for the long discharge line tests and Figures 8.68 and 8.69 for the short discharge line tests.

As shown the resultant bending moments on the quencher arms occur principally in the vertical direction Figures 8 70 and 8.71 for the long and short discharge line tests show a corresponding distribution of the maximum resultant bending moments at the bottom support.

Tables 8.11 and 8.12 also indicate the maximum torsional strains and torsional moments measured at the bottom support and the maximum vertical strains and vertical forces measured at the bottom support during vent clearing. This data is based on as evaluation of the visicorder traces.

8 4 2 Steam Condensation Test Results Eight condensation tests with the short discharge line were performed in the period from July 18, 1978 to July. 21, 1978.

8.4.2 1 Test Parameters The most important of the parameters being -investigated was described in Section 8.3. A detailed list of test parameters is given in Table 8.6. Compiled in that Table are the parameters at the beginning of the tests, such as:

type of test length of discharge line accumulator pressure water temperature in test tank water level in discharge line water level in test tank air temperature in discharge line The accumulator pressure Pl.lA and buffer tank pressure P2.6A are the determinative values for the system pressure at the start of each test. The values were read by computer just prior to the start of the test. In addition, these pressures were stored continuously on tape but only up to 360 seconds after the start of tests 36.1 and 40.1. This was dictated by the limited storage capacity of the operating instrumentation computer's magnetic disk. This data was continuously stored on the visicorder traces and the test instrumentation magnetic tapes.

REV 1 i 3/79 8-38

PROPRI ETAR Y For accumulator pressures below 30 bar (435 psi), measuring point P2.5 was, used to determine the system pressure, since measuring points Pl. 1A and P2.6A were outside the measuring range.

The water temperature at the start of a test was taken from the computer listings and at the end of a test from the computer plots The values for the water levels and air temperatures in the discharge line at the start of a test were taken from the computer listings.

Table 8.13 shows the relation between the test step, test number, and ranges of pressure and water temperature as they actually occurred.

8 4 2.2 Presentation of Test Results First we will present a survey of the observed condensation phases. That is followed by a presentation of the dynamic pressure amplitudes in the water region of the test tank.

Finally the temperature variations in the water region are described.

8.4.2 2 1 Surve~ of Observed Condensation Phases In the operation field of the quencher as given by the test matrix, the observed condensation phases are indicated in Figure 8.71 for blowdowns along the upper and lower boundary lines of the operation field.

8 4 2 2.1. 1 Blowdown at low Mater Te~m erature For the blowdown along the lower boundary line, the following condensation phases were observed for the tested pressure range:

Absolute system Condensation Phase Tests Pressure in Bar 70- 2 5 Stationary 33.2, 34.1, 35.1, and initial section of 36.1 2 5- 2 Intermittent Middle section of 35.1 2 1 In the pipe (1) End section o f 36. 1 (1) It should be noted here that at the beginning portion of the steam flow has emerged through of this phase the annular gap a

above the quencher inlet. As noted in Section 8.2.1.1.5, REV 1, 3/79 8-39

PROPRIETARY this annular gap simulates hydraulically the sliding the quencher installed at SSES.

fit of Figure 8.73 shows a typical example of the measurement traces obtained with the bottom and wall pressure sensors for stationary operation of the quencher in the upper pressure range (test 33.2) . Figure 8 74 shows a typical example of the lower (test 35.1) . High-frequency pressure oscillations occur pressure'ange with very low amplitude, and without any fixed frequency.

To illustrate the intermittent operation, the variation of the bottom and wall pressures and two pipe pressures throughout the entire duration of test 36 1 is shown in an extremely time-compressed form in Figure 8. 75. The intermittent condensation phase is clearly recognizable in the middle section of the test.

Figure 8.76 shows a more time-expanded excerpt from that phase.

Supplementarily, Figure 8 77 shows a typical powerful individual event in an extremely time-expanded form. The high-frequency pressure peaks superimposed on the low-frequency sinusoidal pressure pulsations are clearly discernible in both Figures 8 75 and 8.76.

For the phase of condensation in the pipe, the test traces exhibit negligibly low amplitudes, which are close to the resolution limit of the measuring chain. Therefore, no example of such a trace is shown.

8.4.2 2 1 2 -Blowdown at High Water Te~m erature For blowdown along the upper boundary line, the phases described in 8.4.2.2.1.1 were observed in practically the same pressure ranges. However, the appearance of the pressure oscillations differs to some extent from that of the pressure oscillations at low water temperature.

First, here is the observed relation between pressure range and condensation phase:

Absolute system Condensation phase Tests Pressure in Bar

>70 4. Stationary 37 2e 38 ls 39 le and initial section of 40 1 4 2 Intermittent Middle section of 40-1 2 1 In the pipe<>> End section of 40.1 REV 1, 3/79 8-40

PROPRIETARY (1) It should be noted here that at the beginning portion of the steam flow has emerged through of this phase the annular a

gap above the quencher inlet. As noted in Section 8.2.1.1.5, this annular gap simulates hydraulically the sliding fit of the quencher installed at SSES.

For stationary operation in the upper range of pressure, Figure 8.78 shows a typical example for test 37.2. 1he lower range of pressure for this phase is represented by an example from test 39 1 (Figure 8.79) . There are also higher-f requency pressure oscillations with low and very low amplitude, respectively, and without any fixed frequency.

A typical example of intermitten~to eration is shown in Figure 8.80 by an excerpt from test 40.1. Compared to this phase at low water temperature (see especially Figure 8.76), a distinct attenuation of the strength of the pressure pulsations is observable at high water temperature. Superimposed high-frequency pressure peaks do not occur.

For the phase of condensation in the prie, the. test traces exhibit negligibly low amplitudes even at extremely high water temperature of more than 90oC.

8.4.2.2.2 Statistical Evaluation of the ~Dnamic Pressure Loads on the Pool Boundaries As described in Section 8.4.2 2. 1, the steam condensation does not have any uniform form throughout the entire range of system pressure and water temperature.

To now be able to quantify the distribution of dynamic pressure amplitudes during a blowdown from 70 bar to approximately 1 bar, the recordings from a representative bottom pressure sensor and wall pressure sensor for all the tests were statistically evaluated. This also allowed us to investigate the influence of system pressure and water temperature on the dynamic pressure a mplitudes.

B.a 2.2 2.1 Dependence of ~Dnamic Bottom and Ilail Pressures on System P re ssure and Mater Temperature The pressure-time histories stored on magnetic tape for pressure sensors P5. 2 (bottom pressure) and P5. 10 (wall pressure) were each read for maximum value at uniform time intervals. A high-pass filter with a frequency cutoff of 2 Hz and a low-pass filter with a frequency cutoff of 500 Hz were inserted into the circuit.

In'his manner, a falsification of the evaluation due to slow R EV 1, 3/79 8-41

PROP HI ET AR Y drifting of the zero point or due to electrical interference was largely excluded.

Por tests 33. 2, 34.1, 35 1, 37.2, 38.1 and 39.1,.a uniform interval of 1 second was chosen because of the relatively short test duration of a maximum of 64 seconds in test 39.1. In tests 36.1 and 40.1 with test durations of over 800 seconds, the uniform interval was 4 seconds. In these two tests, the phases of stationary and intermittent condensation and condensation in the pipe were covered separately at the same time. No error was introduced into the evaluation by the different choice of intervals, since the maximum values were covered in each case The extreme values determined for the positive and negative dynamic pressure amplitudes at the bottom and on the wall are plotted versus the transient variation of the system pressure in Figures 8 81 and 8.82. Due to the large number of extreme values, a selection was made with the aim of considering only the higher values.

The top half of the Figure shows the measured maximum pressure amplitudes for the blowdown at higher and high water temperature along the upper boundary line of the operation field. The bottom half shows them for the blowdown at low water temperature along the lower boundary line.

A similar illustration for the measured maximum wall pressure amplitudes is given in Figure 8.82.

The peak bottom-pressure and wall-pressure loads measured during the individual condensation phases are indicated as a function of water temperature in Table 8.14. Prom these peak values, we can ascertain a slight decrease of the pressure level with a hot pool for the stationary and intermittent condensation phases. Por the phase of condensation in the pipe, of course, there are practically no diffe'rences in the pressure levels for cold and hot pool-8 4. 2 2.2. 2 Occurrenc~efre uence Distributions of the~Dnanic Bottom and Mall Pressures In parallel with the determination of extreme values as described in Section 8.4.2.2.2. 1 all positive and negative peak values between the zero passages of the pressure-vs.-time variations were determined. in each time interval and classified according to magnitude.

This counting method, known as <<peak count between zero passages" or "mean crossing peak count method>>, avoids the inclusion and consequential overassessment of small intermediate oscillations.

Only the absolute maxima between two zero passages're included in the count.

REV li 3/79 8-42

PROP RI ETAR X The count result supplies the class occurrence frequency distribution at once. Positive and negative peak values were treated separately. Any error in the count results by the noise level on the magnetic tapes was largely eliminated by means of a

. prescribed amplitude suppression of 10 mV = 0. 015 bar.

A uniform class interval of 0.025 bar was chosen for the histograms. In that way, the histograms of the individual tests were able to be combined into an overall distribution for blowdowns with cold and hot pool. The histograms of the positive and negative amplitudes of the dynamic bottom pressures at measuring point P5. 2 are illustrated in Figures 8.83 and 8 84 for blowdowns with cold and hot water, respectively. Analogous historgrams for the wall pressures at measuring point P 5.10 are shown in Figures 8.85 and 8 86.

8.4.2.2.2.3 Statistical Characteristics of the Dynamic Bottom and Mall Pressures Influences of test parameters can be read of f from the statistically determined mean values, since those values are obviously much more typical than the magnitudes of individual and very rare maximum values. The mean values were determined by the group value methods using the following equation:

k Zn .P P i~1 K

illZn<

where PG = mean value; Fi class mean value; n = class frequency.

The group value method was also used for the combining of the individual histograms of a blowdown to get the mutual freqeuncy distributions. Those mean values are indicated in Figures 8. 83 to 8. 86 In general, the trends are supported by the maximum values. The unavoidable scatter of the maximum values is allowed for by forming the average value of the 10 highest amplitudes in each test. Due to the small number, they were determined by the single-value method:

N ZP PE

= i~1 N

where REV li 3/79 8-4 3

PROPRIETARY P

E

= mean value; P.= single extreme value; N = number of extreme values Tables 8 15 and 8.16 provide an overview of the abovementioned most important. statistical characteristics of the pressure-time histories at the bottom and at the wall, respectively for tests 33.2 to 40.1. Indicated are:

maximum value relative to the entire test, mean value relative to the entire test, lower limit value of the 10 highest values, mean value of the 10 highest values.

• Beside the data concerning the system pressures and water temperatures, the condensation phases are also listed. In tests 36.1 and 40.1, the phases of stationary and intermittent condensation and condensation in the pipe were treated separately.'igures 8.87 and 8.88 show plots of the mean values relative to the entire test or test section and the mean values of the 10 highest values, as functions of system pressure.

The mean values of the bottom and wall pressures are slightly higher for the blowdown with a cold pool. This trend, already alluded to in Section 8.4.2.2.2.1 on the basis of the absolute extreme values, is therefore verified statistically., The level of the mean values from the 10 highest values is higher by a

-factor of approximately 3-4 than the level of the mean values relative to the entire test or test section.

8.4 2 2 3 Te~merature Variations in the Rater Region of the Test Tank Four tests were selected to illustrate the temperature variations in the water region of the test tank:

test 33.2 for high system pressure and cold pool, test 35.1 for low system pressure and cold pool, test 37.2 for high system pressure and hot pool, test 39.1 for low system pressure and hot pool.

Figures 8.89 to 8 92 show the vertical temperature distribution obtained from the measuring points T 5.5, T 5.2, T 5.3 and T 5.4 arranged above one another on the concrete wall In each case, the measured temperatures are scattered about a mean curve. The scatter is greatest f or measuring point T 5 2 (approximate max.

a8o C). That measuring point is at the height of the quencher arm and is impinged upon directly by the sidewards directed flow impulse. The scatter is least for measuring point T 5.4 (approximate max. a50 C). The scatter can be explained by the h ig h degree o f turbulence in the pool. .

REV ~ 1 i 3/79 8-44

PROPRI ETAR Y Figures 8.93 to 8.96 show the temperature variations at quencher arm 1 for the same tests. At measuring point T 5.8 located in the middle of the hole array (see f igure 8. 14) a distinct temperature increase of approximately 15-200C, on the average, was recorded relative to the pool temperature. In contrast, the temperatures at the upper edge of the hole array (T5.9) and at the upper edge of the guencher arm (T5.10) are somewhat lower than the pool temperature at T5.1 due to a sufficient "cold water supply". This is an indication of the good circulation of water near the guencher. This .confirmed the expected condensation behavior of the quencher as related to the layout of the hole array. (See Sec tion 4.1.1. 1) .

8 4.2.2 4 Water Level in the Discharge Line When Opening and After Closi~n the SRV In the tests with the long discharge line, the water level in the pipe was measured by the "Level Probes" LP 4.1 thru LP 4.4 at four positions, one above another.

In the tests with the short discharge line, this instrumentaiton was extended by the measuring point LP 4.5 above the measuring point LP 4.4; see Figure 8.8 The measurement signals from these Level Probes were recorded on visicorders and magnetic tape.

A Barton cell, measuring point L 4.1 in Figure 8.4, was used to set and measure the water level in the d ischarge line bef ore test start. The reading of that measuring point was interrogated by the computer before and during the test and was stored The indications of the Level Probes and also the indications of the Barton cell were used to depict the time variation of the water .level in the discharge line. It must be taken into consideration that the response speed of the Barton cell is too slow for the rapid changes of the water level during vent clearing and after the closing of the SRV. The measuring point was used essentially to determine the steady-state. water levels in the discharge line.

Figures 8 101 and 8.102 show two typical examples of the variation of the water level in the pipe for the interval test 15.1 with the long discharge line and 20.1 with the short discharge line. It was found that in two instances in interval test 15.1 (Figure 8.101), the water column briefly exceeded the external water level, but fell back immediately. These two test points represent the maximum water column rise measured in the vent clearing tests.

In the interval test 20.1, the water column did not reach the level of the external water surface in any instance after closing of the SRV. The maximum water level rise was generally found, in all tests, to occur after the third valve actuation.

R EV 1, 3/79 8-45

PROPRIETARY To evaluate the effect of vacuum breaker operation on the water column ref lood following vent clearing; Test 32, with one locked vacuum breaker and a time interval of 3 seconds between the closing of the valve after the first actuation and the next actuation, was included. Figure 8-105 shows the variation of the movement of the water column in Test 32. As can be seen no adverse effects were recorded.

8.4 3 Checking and Calibration of the Neasuring Instrumentation The calibration and the electrical and physical checking of all sensors before, during and after the tests were performed in accordance with the Test and Calibration Specifications.

Fig. 8 97 shows diagrammatically the physical calibration of the sensors, the setting and calibration of the amplifiers and recording instruments, and the quality inspection of the sensors.

Pig. 8.98 shows the time intervals stiplated for the checks and calibrations in the Test and Calibration Specifications. Fig.

8.99 clarifies the chain of the calibration system from the national standards of the Physikalisch-Technische Bundesanstalt (PTB) to the measuring instruments.

The pressure sensors P 5.1 thru P 5.10 used in the tests were fully operable until the end of the tests. The lowest insulation resistance of 1.2 x 10~ 0 measured at P 5.1 after the tests can be classified as "good". The pipe pressure sensor P 4.1 failed on 31 Nay 1978 It was replaced by a new sensor for the With this new sensor P 4. 1 ~ the lowest subsequent tests.

insulation resistance for the group of pipe pressure sensors after the tests was 3 x 10~~ , which was very good There were no failures for the strain gauges SG 4 1 thru SG 4. 8, SG 5.1 and SG 5.2 Here also, a very good insulation resistance level was recorded with a lowest value of 3 x 10~ u at SG 4.6 after the tests.

J.ikewise, none of the temperature measuring pionts T 5.1 thru T 5.10 failed. The lowest insulation resistance of 1.3 x 10~~ was sufficiently high.

8 4 4 Analysis of Neasurement Errors Based on information from the manufacturers of the measuring instruments, KWU s own investigations, and taking into consideration the experience accumulated in similar test projects, the maximum measurement errors for the individual sensors can be indicated as follows:

R EV. 1, 3/79 8-4 6

PROPRIETARY Pressure sensors P 5.1 thru P 5.10 Linearity error of the sensor Error 2.5% of measured value in range of 0 to 2 bar 2. 5'5 Reproduction error of the sensor

0. 2% of 5 bar 0.01 bar Error of the measuring amplifier 0 5%

Error of the balancing unit and recorder 0. 5%

Max. total error x [ 0.01 bar + 3. 5% of the measurement value]

Pressure sensors P 4.1 thru P 4.5 Error Linearity error of the sensor 0.5% of measured value in range of 0 to 20 bar 0.5%

Reproduction error of the sensor 0.1% of 35 bar 0. 035 bar Error of the measuring amplifier 0 5%

Error of the balancing unit and recorder 0 5%

Max. total error a [0.035 bar + 1.5% of the measurement value]

P ressure sensors P 2. 3 and P 2 5 Error Linearity error of the sensor 1% of measured value in range of 0 to 40 bar Reproduction error of the sensor

0. 1% of 14 0 ba r. 0.14 bar Error of the measuring amplifier 0. 5%

Error of the balancing unit and recorder 0 5%

Max. total error a [0.14 bar + 2% of the measurement value]

Strain gauges SG 4.1 thru SG 4 8~ SG 5.1~ Error and SG 5. 2 Tolerance of the guage factor Influence of temperature on the guage factor R EV 1, 3/79 8-47

PROP RI ETAR Y Error of the measuring amplif ier 0 5%

Error of the balancing unit an recorder 0. 5%

Max. total error a 5% of the measurement value Temperature measur~in ~pints T 5.1 thru T 5.10 Error of the sensor loc Error of the measuring amplifier 0 5%

Error of the balancing unit and recorder 0. 5S Max. total error x [ l~C + 1$ of the measurement value]

A fter the f irst tests on May 10, 1978 and a f ter conclusion of the tests on August 2, 1978, additional physical checks of the pressure sensors in the water region were performed by incremental lowering of the water level in the test tank. The max. deviations from the nominal value were approximately +0.01 and -0.02 bar. Fig. 8.100 illustrates a frequency distribution of these deviations combined from both checks and for all presure sensors. It shows a typical Gaussian distribution.

In order to record the high-frequency processes correctly in frequency and amplitude, the data was acquired in analog form on magnetic tape. Por a sensor eigenfrequency of approximately 30 kHz, the dynamic range was limited not by the sensors but rather by the carrier-frequency measuring amplifiers located further on in the circuit. The frequency cutoff of the measuring amplifiers was at 1.5 kHz and that of the magnetic tape recorders was at 2.5 kHz. The frequency cutoffs of the visicorders were determined by the utilized galvanometers These frequency cutoffs are approximately 1 kHz. The frequency response of each individual galvanometer was checked prior to the tests.

8.4 5 Repetition Tests and ~Re roducibility of the Results To verify the reproducibility of the measurement results, a repetition of 5 tests was specified in the Test Matrix. Based on a preliminary assessment of the results after conclusion of the test series with the long and short discharge lines, the following tests were repeated (as mentioned previously):

Long line:

4.1 through 4.Rl Interval tests 15.1 through 15.Rl Interval tests Short line:

REV 1, 3/79 8-4 8

PROPRIETARY 19.1 through 19.R2 Interval tests 20 1 through 20.R1 Interval tests 25.1 through 25.R2 Single Actuation tests In addition to the relevant initial conditions, Table 8.17 also gives the measured vent clearing pressure (measuring point P 4.4),

max. dyn. bottom pressures (measuring point P 5.2) i mar dyn. wall pressures {measuring point P 5.10) and frequencies of the pressure oscillations for the first SRV actuation in each of the repetition tests

(>>Clean Conditions tests").

A comparison of the above-cited values for the repetition tests associated with each other demonstrates the good reproducibility under Clean Conditions. The maximum deviations from the mean value for each pair of repetition tests are (see Table 8. 18):

for the vent clearing pressure 10. 75 bar or a6%

for the bottom and wall pressures a0.05 bar or a7X for the frequency of the pressure oscillations 10 5 Hz or a7%

The mean deviations from the mean value of each pair of repetition tests, averaged for all 5 pairs o f tests, are:

for the vent clearing pressure 10.37 bar or 13K for the bottom and wall pressures 10 02 bar or a6%

for the frequency of the press oscillations RO 2 Hz Or X5%

Figures 8.37 and 8.38 illustrates the max. dynamic pressures in the pool duri ng the vent clearing for the multiple valve actuation repetition tests with the long line. Figures 8.39 and 8 40 shows the same thing for the multiple actuation repetition tests with the short line In comparison with the first SRV actuations under Clean Conditions, some larger deviations are exhibited here in the tests under Real Conditions (2nd to 7th and 10th SRV actuations). The reason for these deviations is that the initial conditions differ significantly from each other.

The visicorder traces for each "clean condition" actuation at a repetition test point is provided:

Tests 4.1.1 and 4.Rl.l Figures 8-41 and 8-42 REV 1i 3/79 8-4 9

PROPRIETARY Tests 15.1.1 and 15 Rl 1 Figures 8-46 and 8-47 Tests 19.1.1 and 19.R2.1 Figures 8-48 and 8-49 Tests 20.1.1 and 20.Rl.l Figures 8-59 and 8-60 Tests l

25. and 25 R2 Figures 8-64 and 8-65 A visual comparison of the traces from each repitition test also shows good reproducibility.

Accordingly, it can be said that:

If the initial conditions of the tests are set in a controlled manner (Clean Conditions), then the test results are reproducible.

If the initial conditions correspond to the randomly prevailing operating states (Real Conditions), then the measurement values lie in a larger scatter range.

8 5 DATA ANALYSIS AND VERIFICATION OF LOAD SPECIFICATION 8.5.1 Evaluation of Test Tank Effects on Boundary Pressure Neasurements In this Section, ve present theoretical and experimental investigations which show that the Karlstein test tank represents a good simulation of the hydraulic conditions of the SSES suppression pool. Me are concerned primarily with the effects exerted on the processes in the vater by the existing boundary surfaces such as the water surface, tank bottom, movable or immovable tank walls. The results of the investigation facilitate the evaluation and transposition cf the boundary loads measured in the tests to SSES.

8 5.1.1 Effects of Free Rater Surface and R~iid Walls The effects of the free water surface and the rigid walls of the tank on the fluid pressure will be explained first by means of the examples illustrated in Figure 8-104. The top half of the Figure shows the velocity potential and flow field of a spherical bubble subjected to overpressure or underpressure in an infinitely extended, incompressible fluid. The potential field is described by a simple 1/r law (Reference 35). If, for example, the same bubble is located in a cylindrical rigid tank which is partially filled with fluid, then the potential field and flov f ield have a visibly different appearance (Figure 8-100, bottom) . The differences .in the nonstationary fluid pressure, which is proportional to the velocity potential for sufficiently lov flow velocity (pressure field = potential field; see Reference 4 for example), are clearly evident in the pressure profiles on the right side of the Figure 8-104. The free water surface constrains the pressure to zero, vhile the cylindrical wall causes an increaseingly more poverful pressure rise with REV 1, 3/79 8-50

PROPRIETARY increasing depth. The narrower the tank, the greater is the pressure rise. The calculations rela ting to Figure 8-104 were performed by the finite-elements method {Reference 34) for a tank diameter of 3 m and a water depth of 6 m. The bubble was 2.8 m deep and 0.8 m in diameter.

Besides the pressure field, there is also an effect on the water mass which is effectively entrained by the bubble during pulsation motions (pressure oscillations) and thus also the oscillation frequency. In the case shown in Figure 8-104, the bubble in the tank has a larger coupled mass than in the infinitely extended medium. This is manifested by the fact that the pulsation frequency of the bubble is correspondingly lower i

(see Sect on 8. 5. 3. 2) .

8.5 1 2 Method of Images The method of images is an important aid which makes it possible to clearly understand the hydraulic actions of the water surface and rigid walls and to calculate them quantitatively in a simple way (Reference 35) . Zt is based on the fact that the influence of a plane rigid wall on the flow field of a hydrodynamic point source can be represented by a superposition of the flow field without the wall (infinitely extended fluid) and the flow field of an image source of identical sign and identical strength located behind the wall (Fig. 8-105) . The same holds for a plane free water surface, except that the image source has the opposite s ign.

Using this method of images, the flow field of a point source in a rectangular, vessel is obtained finally by repeated application of suitable imaging operations (Figure 8-105d and Figure 8-2) .

The immediate significance of the method of images lies in the f act that a pulsa ting bubble can be conceived of as a hydrodynamic source, thus providing a simple method to calculate the pressure field. Of special importance for the performance of tests is the consequence derived by inversion of the method of images: A configuration of bubbles oscillating in parallel can be simplified in a test by surrounding one bubble with rigid walls. This will be clarified further in the following.

8 5 1.3 The Test Stand as a SincCle Cell Based on the above discussion, an oscillating bubble in a rectangular vessel is equivalent to a plane field of simultaneously oscillating bubbles {Figure 8-2). From Figure 8-2 it follows further that vessels with several bubbles are also equivalent, since between each pair of bubbles the imaging wall section can also be omitted.

REV 1, 3f'79 8-51

PROPR1 ETAR Y Application of the method of images to the transposition of a system of valves blowing down simultaneously in a plant to a test stand with a quencher leads to the cell division illustrated in Figure 8-3..

As discussed in section 8.1, the water space of the test stand was formed according to the interior single cells C, F, K and N (Figures 8-3 and 8-108), since they are the narrowest and will therefore exhibit the highest wall and bottom pressures. That can be seen by observing that, according to the imaging principle, they conser vati vely 'simulate more quenchers lying closer together than is actually the case in the SSES suppression pool 8.5 1.4 Spatial Distributions of Pressure in the Test Tank To get meaningful test results, pressure sensors have to be mounted at suitable points in the test tank. A series of theoretical investigations was performed in order to better assess their arrangement. They consisted of calculating the spatial distribution of pressure along the tank walls for various bubble configurations under water. The KRU computer code VELPOT was used for this investigation. A bubble was simulated by a point, source normalized to unit source strength.

The results are illustrated in Figures 8-107 to 8-109. Figure 8-107 shows the calculated wall pressure distribution for a bubble in three different positions near the quencher:

Case 1 Source on the tank axis, 0. 7 m above the quencher axis Case 2 Source on the tank axis, at quencher elevation Case 3 Source at center of the quencher (eccentric) .

The results show that .,the eccentric arrangement of the quencher which became necessary because of space limitations in the tank, including the corresponding positioning of the pressure sensors (black squares in Figure 8-107), results, theoretically, in slightly higher measurement values for the pressures. The next calculation (case 4, Figure 8-108) serves to answer the question as to how the bubble's form influences the pressure distribution.

To do that, the single source from case 3, f igure 8-107, was replaced by four identical sources with the same total source strength. Figures 8-108 and 8-109 show that there are no major differences. Note also the good agreement seen between the measured pressures from shakedown test 08 1 and the calculated values i'n Figure 8.109.

The model cases 3 and 4 (single bubble at center of quencher and 4-bubble arrangement) are best adapted to the test stand geometry. Since the associated pressure distributions hardly differ at all {Figure 8-109), it is demonstrated that an exact REV 1, 3/79 8-52

PROP RI ETAR Y knowledge of the air distribution under water is not necessary for a correct arrangement of the pressure sensors In order to demonstrate the conservative nature of the chosen single cell, as already explained in Section 8.5.1.3, the pressure distribution for model case 4 is compared to the distribution calculated for the Susquehanna plant in Figure 8-110. The pressure distribution in the test stand envelops the pressure distribution in the SSES. Furthermore, the pressure distribution in the test stand is enveloped by the specified distr ibut ion {F igure B-ill).

8.5.1.5 Investigation of the Influence of Savable Salle on the Measurement Results /Fluid-Structure Interacti~on 8.5.1 5 1 General Remarks In the preceding discussion, it was assumed that the single cell has rigid and immovable walls. The construction of the Karlstein

.test tank is such that the tank, despite a series of stiffening ribs (see Figures 8-10 to 8-12), still has a residual compliance.

The time-varying loads acting during the blowdown of the quencher can therefore excite the tank into oscillation due to Fluid-Structure Interaction (FSI).

Using experimental and theoretical investigations, it will be shown that influences of tank oscillations on the measured boundary loads can be neglected. The experimental investigations consisted, firstly, of measuring the tank's response to a short pressure impulse which was produced by an explosive charge detonated near the quencher (Section 8.5. 1. 5.2) . Measurements made during the start-up tests on the test stand then supplied the tank's response to the loads occurring during vent clearing (Section 8 5.1.5.3) . Taking into consideration the inpulse response, it turns out that effects of tank oscillations at the eigenfrequencies are negligible. This statement is later confirmed by calculations and also is extended to forced oscil la tion s.

8 5.1.5 2 Ex2erimental Investigation of the Tank's Natural Oscillations The experimental investigation of the tank's natural oscillations was performed with impulsive excitation by an explosive charge in the water and simultaneous measurement of the displacements of the wall and bottom sections and of the fluid pressure.

The arrangement of the charge and sensors in the tank is illustrated in Figure 8-112. The position of the charge was chosen such that the spatial load profile in the tank matches the profile of the blowdown loads as well as possible. The charge itself was a stoichiometric mixture of hydrogen and oxygen which REV 1, 3/79 8-53

PROPRI ETAR Y was ignited in a plastically deformable flat container (Figure 8-113). Eight displacement transducers (WA 1 to WA 8) were available for the displacement measurements. They were positioned with the aim of obtaining the most useful information.

The arrangement of the pressure measuring points in the water (P5.1 to P5.10, Figures 8-10 to 8-12) was the same as in the later blowdown tests. As for the evaluation of the pressure traces in Section 8.5.3, transducer P5.10 was chosen as reference pressure transducer The charge was located at different positions near the quencher as shown in figure 8-112, i.n order to obtain enveloping load profiles. A typical result is illustrated in Figure 8-114, which shows the recordings from displacement transducers WAl to MA8 and pressure transducer P5.10 for test no. 2 (charge in position 2).

The lowest occurring frequencies are below 1 Hz, but have nothing to do with the tank's response, but rather represents a shift of the zero point The lowest eigenf requency of the tank is at approximately 13 Hz and is seen clearly in the response from transducers WA2 and WA3 oscillating in phase. Both gages are seated on the box-shaped stif fening rings as shown in figure 8-112. At the wall sections between the stif f eners (WA4 and MA6) and at the bottom (WA8), the frequencies that occur are mainly between 30 and 60 Hz. The oscillations of the flat lower stiffener rings (WA1 and WA5) are less pronounced. The smallest displacements are found at the concrete sections (WA7), where some of the amplitudes are smaller hy an order of magnitude. The pressure signal from P5..10 shows distinct excursions only during the f irst 100 ms.

To be able to better evaluate the tank's frequency response, the measured time variations were Zourier analyzed and power spectra were formed. The spectra associated with the displacement transducers on the steel wall (MA2), concrete wall (MA7) and bottom (WA8) and the pressure transducer PS 10 are shown in Figures 8-115 to 8-118.

mentioned 13 Hz It turns out that the previously oscillation in the low-frequency range is of greatest importance. The associated tank deformation (eigenmode) can be derived from the point correlations shown in Figure 8-119.

There, the displacements of the displacement transducers WA2, WA3 and MA7, filtered by a bandpass filter at 13 Hz, are plotted against each other at the same times.

The fit line through the set of points has a positive slope in the top graph and a negative slope in the bottom graph.

Therefore, displacement transducer WA3 (steel wall above MA2; see Figure 8-106)-oscillates in phase with WA2, while displacement transducer WA7 (concrete wall) oscillates out of phase. This means that the 13 Hz oscillation corresponds to an ovalizing motion of the wall (see Figure 8-120).

REV 1, 3/79 8-5 4

PROPRIETARY 8.5.1.5 3 Experimental Inves~ti ation of the Tank's Re~sonse to Vent Clear~in Loads The investigations of the tank's response to vent clearing loads were performed during the test stand shakedown tests. To measure the tank's response, the choice was made to use one di,splacement transducer each on the steel wall (WA2), on the concrete wall (WA7) and on the bottom (WA8) . The instrumentation is shown in Figure 8-121.

Test 08.1 represents a typical example of the shakedown tests that were run. The measured time histories of the wall and bottom displacements and of the reference pressure P5.10 are shown in Figure 8-122. The zero-point drift mentioned above was eliminated by using a 2 Hz high-pass filter. It can be seen that both the pressure and the displacements oscillate at the same principal frequency of 5. 1 Hz. The steel wall (WA2) and bottom (QA8) move in phase. The very small movement of the concrete wall (WA7) is almost out of phase compared to the pressure P5.10.

In addition, the displacement transducer WA8 records a higher-frequency oscillation at 30 Hz. It has already begun weakly at test start, then develops strongly at about the time of the vent clearing', and then decays again about 300 ms later The physical interpretation of the 5 Hz oscillation is obvious.

The pressure oscillation is caused by the pulsation of the air bubble which is created during vent clearing. At the same time, the tank carries out forced oscillations at the frequency of the forcing force (5 Hz pulsation of the air bubble). The sometimes phase-opposed nature of the displacements of the steel wall and bottom, on the one hand, and the concrete wall, on the other hand, makes eigenmode it plays evident that the above-discussed ovalizing a dominant role.

The origin of the rapidly decaying 30 Hz oscillation seen at WA-8 at the test start is attributed to local forces transmitted through the discharge line and the quencher support during vent clearing.

'F ig ur es 8-123 to 8-125 show the power spectral densities of the displacement time histories for gages WA2, WA7 and WA8 measured during shakedown test 08 1. Figure 8-126 shows the power spectral density of the pressure time history for P5.10 measured during shakedown test 08 l. A,review of these figures shows very little influence from the 13 Hz tank eigenfrequency or from the 30 Hz local effec t seen at WA8. Figure 8-126 showing the rsults from P5 10 shows practically no influence from either of these effects.

R EV li 3/79 8-55

PROPRIETAR Y Prom this it can Karlstein test tank be concluded is rigid that for and has no all practical purposes influence on the pool the boundary pressure measurements made during the tests.

8.5.1.5 4 Theoretical Investigations.and Model Calculations of the Influence of Fluid-Structure Interaction 8 5 1.5.4.1 Computation Models The analysis described below to compute the FSI on the measured pressures in the Karlstein test tank was performed by using the KWU computer code KOVIBlA which was developed originally and used successfully for the analysis of fluid'-structure interaction in the water pool of KWU's 69 Product Line BWR Plant.

The underlying- theory follows from a uniform formulation of the mechanical processes based on potential theory and classical Lagrangean dynamics. It unifies the dynamics of the bubble and the FSI by using the results of modal analyses. In particular, the feedback effects between bubble and structure via the fluid a re i nc1 ud ed.

8.5.1 5 4 2 Model Parameters and ~In ut for Calculations Without The model parameters and input quantities for calculations of the air bubble oscillations in the rigid tank are:

air mass flow into the bubble, water temperature (= air temperature in stationary equilibrium),

hydrostatic pressure at bubble position, hydrodynamic mass parameter of the bubble, spatial pressure distribution, initial values ,(hubble radius, etc.).

The total air mass (integrated ai r mass flow), water temperature and static pressure at the bubble position are obtained from the test data. The hydrodynamic mass constant of the bubble and the spatial pressure distribution are obtained from the corresponding potential calculations (Figure 8-107, case 1) . The time variation of the air supply into the bubble was ad justed heuristically by means of systematic trial and error, in parallel with the initial values, in such a way that the calculated and measured time variations of the pressure at transducer P5.10 exhibited optimal agreement The start-up test 08.1 was used as reference test for these calculations. The air mass flow determined in this manner is illustrated in Figure 8-127 R EV li 3/79 ~ 8-56

PROPRIETARY 8.5.1 5.4 3 Nodel Parameters and Xnaut for Calculations with Pdj Just as for the determination of the air supply into the water pool, a semiempirical method is used for the structural dynamics data. They are determined on the basis of the eigenfrequency measurements described previously. Input data for the calculation are:

eigenfrequency, modal mass, modal weight, dynamic pressure distribution.

Based on the impulse response of the tank (Figures 8-115 to 8-117), it is plausible to select the oscillation mode lying at 13 Hz. That fixes the frequency. The modal mass cannot be taken directly from the experiment, but rather can be determined indirectly via the measured unit displacements of the wall. The unit wall displacement is illustrated in Figure 8-128. It is obtained from displacements at the displacement transducers by bandpass filtering at 13 Hz and plotting simultaneous values of displacement which are normalized to 1 at the water surface. The displacement direction is defined as positive if the relevant wall section moves inward. The hydrodynamic component of the modal mass, (coupled water mass) is then calculated by methods of potential theory.

The modal weight, which is equal to the integral load relative to the modal mass and averaged over the unit displacement, is based on the load distribution calculated for case 1 (Figure 8-107, centered bubble). The dynamic pressure distribution (see Figure 8-129) is obtained from the unit displacement. by means of potential calculations.

8 5.1.5.4.4 Results of the FSI calculations The results of the calculations concerning the influence of FSI are shown in Figures 8-130 and 8-131. Figure 8-130 shows the calculated time variation of the pressure at pressure transducer P5.10, first in the rigid tank (without FSI) and then in the elastic tank with the 13 Hz eigenfrequency. There is a very slight reduction in the pressure amplitudes, but it is certainly negligible in comparison to the scatter of the measurement values themselves.

As is evident from Figure 8-131, the frequency influence of FSI also can be neglected. In that Figure, the oscillation f requency of the bubble is plotted against the bubble volume. The bubble has a slightly lower frequency with FSI effects included than without.

R EV 1, 3/79 8-57

PROPRIETARY A physically clear explanation of the very slight FSI effects found in the Karlstein Test Tank can be obtained .by comparing the volumes of fluid which are moved by the oscillating wall and bottom and by the pulsating bubble . For a bubble volume {long line) of 2. 2 m~ and pressure fluctuations of x0. 4 bar {see Figure 8-126), the volume change of the bubble is approximately 1 m~

ll isent r o pica y.

In contrast to this, for displacements like those found in Figures 8-124 and 8-125 the walls and bottom use up only about 0.05 m~, which is only 5% of the water volume coming from the bubble. Therefore, due to the compliance of the tank, 95% of the water flows upward instead of 100% (rigid tank).

Thus, the result of the experimental and theoretical FSI investigations is that effects of the compliance of the Karlstein test tank walls and bottom on the pressure loads measured on the boundaries of the tank during the tests can be neglected.

8 5 2 Verification of SHV System Load Specification Due to SRV Actuation The pressures inside the SRV discharge line were measured at four measuring points: just behind the SRV at measuring point P 4.1, in the center of the blowdown pipe at measuring point P 4. 2 (measuring point P 4.5 for the short discharge line), just above the normal water level at measuring point P 4.3, and just before the inlet of the quencher at measuring point P 4.4 (see Figure 8 4).

The long and short discharge lines are illustrated in Figures 8-5 and 8-6.

The measured pressures in the discharge line are documented in Section 8.4.1.

8.5.2 1 Pressures Duri~n the Vent ClearincC Process Typical measurement traces of the pressures in the discharge line are shown in Figures 8-132 and 8-133. The vent clearing pressure is read off at P4.4. As discussed in Section 8.4 1, the vent clearing pre sure is defined as the pressure which is read off at the first pressure maximum at P4.4. A typical feature of this pressure variation is the dynamic overshoot of the pressure above the stationary value. This- phenomenon does not occur in such a pronounced manner at the other pressure transducers along the discharge line. This dynamic effect indicates that the pressure required to expel the ~ater column is greater than. the pressure necessary to bring the steam mass flow through the quencher.

The expulsion of the water column, is also clear from the different time variations at P 4.3 and P 4.4. The pressure at 8 EV li 3/79 8-5 8

PROPRIETARY measuring point P 4.3 (above the vater column) rises much more steeply than the pressure at measuring point P 4.4 (inside the vater column) The difference betveen the tvo pressuros is the pressure which is necessary for the acceleration of the vater column.

At the time of the vent clearing, the two pressures have approximately equal values. But after the vent clearing they differ again, this time due to the different pressure losses caused by flow resistances in the pipe.

8.5 2 1.1 Vent Clearing Pressures for the L~on Line The steam mass flow through the SRV is a practically linear function o f the stagnation 'pressure (reactor pressure) . Since the steam mass flow is one of the main parameters for the pressure build-up in the air region of the discharge line and thus for the acceleration of the water column, we will plot the pressures in the discharge line as a function of reactor pressure. The pressure in the buf fer tank (P2.6) and not the pressure in the steam line before the SRV is used as the reactor pressure for the tests since the pressure in the buffer tank more closely simulates the representative stagnation pressure in the reactor. (see Figure 8-134).

To describe the dependence of the vent clearing pressure on the reactor pressure, only those tests for which the initial conditions were set and thus knovn exactly were used. Those are the tests vith so-called <>clean conditions".

From Figures 8-135 and 8-136, it can be seen that the measurement results have good reproducibility for the tests with clean conditions. The pressures in the pipe increase practically linearly with reactor pressure.

The following trends can be observed:

1) A lowered water level in the discharge line results in lover pressures during the vent clearing.

2) A hot pipe results in higher pressures during the vent clearing. This is due to the smaller percentage of condensation on the pipe wall.

3) The pressure (at the time of vent clearing) behind the SBV is always higher than the vent clearing pressure close to the quencher.

The difference is attributable to the flow loss along the line.

R EV 1, 3/79 8-59

PROPRXETARY (at the time of vent clearing) behind the SRV

4) The pressure increases -with increasing reactor pressure {or increasing steam flow ra te through the relief valve) .

Besides the clean-condition tests, there is a large number of real-condition tests and interval tests. Since the initial conditions in them were random and were not varied in a controlLed manner, the measurement values are scattered over a much wider band than in the clean-condition tests. Hence, these tests are not usable for trend analyses, but may be used for verification of maximum specification values.

The measured maximum values are:

Pressure behind the SRV (at vent clearing time): 19 bar at a reactor pressure of 72 bar Vent clearing pressure before the quencher: 14.5 bar at a reactor pressure of 72 bar 8 5 2. 1 2 Vent Clearing Pressures for the Short Line Figures 8-137 and 8-138 show the measured pipe pressures plotted against reactor pressure for clean condition tests with the short discharge line. The same trends as seen with the long line are seen here.

Since the short line has a smaller air volume than the long line, while the water column to be cleared and other parameters remain the same, the pressures in the short line are higher than those in the long line.

The measured maximum values are:

Pressure behind the relief valve {at vent clearing time):

22 bar at a reactor pressure of 73 bar Vent clearing pressure before quencher:

18 bar at a reactor pressure of 73 bar.

8 5.2 1 3 Tran~s osition of the Measurement Values to SSES and Comparison with the DesicCn deci f ication The verification tests in Karlstein were run with the actual geometry of the relief system, the actual SRV, and the highest water level in the discharge -line (6. 2 m above center of quencher) that occurs for SSES.

The measured vent clearing times for that water level and a high reactor pressure (69 81 bar) was between 250 and 400 ms REV. 1, 3/79 8-6 0

P ROP3I ETAR Y For these vent clearing times, the opening time of the SR V (measured opening times: 29 - 60 ms) has no noticable effect on the vent clearing pressure (see Figure 8-139) .

Hence, in regard to the vent clearing pressure, the only variable whose maximum value for SSES was not completely covered was the reactor pressure.

The following extrapolation applies for that:

a) Pressure behind the valve at vent clearing time The Measured maximum value for the long line is 19 bar at a reactor pressure of 72 bar A Slope of 25% is seen in figure 8-135.

Extrapolating to 88 bar, the result is:

for the long line Pox = 23 bar maximum The Measured value for the short line is 22 bar at a reactor presssure of 73 bar Slope of 25% is seen in figure 8-137.

Extrapolating to 88 bar, the result is:

P ax = 26 bar for the short line The design value given in Section 4. 1.2.1 is 550 psi = 37.93 bar.

The Karlstein tests demonstrate that the design value is very conservative for the vent clearing case.

b) Vent cleari~n pressure The measured maximum value for the long .line is 14.5 bar for reactor pressure of 72 bar A Slope of 12.5% is seen in figure 8-136.

Extrapolating to a reactor pressure of 88 bar results in Pmax = 16.5 bar for the long line.

The measured maximum value for the short line is 18 bar at a reactor pressure of 73 bar.

A Slope of 12.57 is seen in figure 8-138.

Extrapolating to a reactor pressure of 88 bar results in P max = 20 bar for the short line.

The specification value given in Section 4.1.1.2 is Pmax 27 bar The Karlstein tests demonstrate that the specification value for the vent clearing pressure is very conservative.

8 5.2.2 Pressures Duri~n the Stationary Condensation of Steam About one second after the opening of the SRV, the vent clearing process is completed and the phase of sationary steam condensation begins.

In this phase, the pressures in the discharge line are determined by the steam mass flow and the flow resistance. Since the steam R EV 1, 3/79 8-61

PROPRIETARY mass. flow is proportional to the reactor pressure, here again we will investigate the dependence of the pipe pressures to the reactor pressure.

8 5.2 2 1 Lo~n Line Figures 8-140 and 8-141 show the dependence of the steady state pressure on the reactor pressure.

Me see that the relation can be represented very well by a straight line.

As a result of pipe friction, the stationary pressure behind the SRV has higher values than the pressures just before the quencher. It also exhibits a faster increase with reactor pressure.

The measured maximum values are:

17.5 bar at reactor pressure of 72 bar for the pressure behind the SRV (P4 l0 bar at reactor pressure of 70 bar 1)>

for the pressure before the inlet to the quencher (P4.4) 8 5 2.2 2 Short Line Figures 8-142 and 8-143 show the dependence of the steady state pressure on the reactcr pressure.

The behavior of the pressure before the quencher (P 4.4) is practically identical for the short line and long line. This is not surprising, since this pressure depends only on the flow resistance of the quencher.

The pressures behind the SRV are lower than those for the long line, but display the same increase, with reactor pressure.

The different flow resistances of the two discharge lines are manifested here.

To clarify this effect, the variation of the stationary pressure at the measuring points along the discharge line are plotted in Figure 8-144 for the short and long lines. The average pressures were used, i.e., the pressures were read off from the interpolation 1'ines at 88 bar (see Figures 8-190 to 8-143).

The measured maximum values for the short line are:

Pressure behind the SRV (P4.1) 16 bar at a reactor pressure of 72 bar, and 15 bar at a reactor pressure of 63 bar REV 1, 3/79 8-62

PROPRIETARY Pre sure before inlet to the quencher (P4.4) 9.5 bar at a reactor pressure of 71 bar, and 9.0 bar at a reactor pressure of 65 bar.

8 5.2.2.3 Transposition of the Measurement Values to SSFS and comparison with the Design Speci ficat ion As was the case with the vent clearing pressure, the only variable whose maximum value in the SSES was not completely covered by the test stand was the reactor pressure.

An extrapolation of the measured maximum values to a reactor pressure of 88 bar yields the following results:

a) Long L ine The measured maximum value behind the SRV is 17. 5 bar at a reactor pressure of 72 bar.

A Slope of 22% is seen in figure 8-140 Extrapolating to 88 bar, the result is:

Pm~~ 21 bar The measured maximum value before quencher inlet is 10 bar at a reactor pressure of 70 bar.

A Slope of 16% is seen in figure 8-141.

Extrapolating to 88 bar, the result is:

Pmmx = 13 bar.

b) Short Line The measured m'aximum value behind the SRV is 16 bar at a reactor pressure of 72 bar and 15 bar at a reactor pressure of 63 bar. A Slope of 22% is seen in figure 8-142.

Extrapolated to 88 bar the result is: P max

= 19.6 bar and 20.5 bar, respectively.

The measured maximum value before quencher inlet is 9.5 bar at a reactor pressure of 71 bar and 9 bar at a reactor pressure of 65 bar.

A Slope of 16% is seen in figure 8-143.

Extrapolated to 88 bar, the result is:

max=

P 12.5 bar and 13.0 bar, respectively.

It can be stated that the design value of 550 psi = 37.93 bar for the stationary pressure behind the valve is very conser vative.

8 5.2 3 External Loads on the~uencher and Bottom Sup2ort In this Section we shall discuss the measurement results which provide information about the external loads on the quencher and REV 1, 3/79 8-63

PROPRI ETAR Y bot tom su p port. The measuring points pro vid ed f or tha t p u rpose are shown in Figure 8-13, and are as follows:

SG 4 1/4 2 Bending at quencher arm 1 SG 4 3/4 Bending at quencher arm 2 SG 4 5/4 6 Bending at the bottom support SG4 7 Longitudinal strain at the bottom support SG 4 8 'orsion at the bottom support Strains were measured at all measuring points. The measured strains were used to calculate the loads which produced the strains. The loads thus calculated are static equivalent loads which contain hydraulic and also structural-dynamical eff ects.

8 5.2.3 1 Vertical Force 8.5.2 3.1 1 Measurement of the Vertical Force To measure the vertical force, two strain ga uges, SG 4. 7, were connected in such a way that they measure strains resulting from vertical forces.

The following relation exists between the load and strain:

~ A ~

E ~

e 2 F where A ~ .016 m B

5 2 F =33 ~

e kN F ~ 2.06 x 10 N/mm B

If we insert e in pm/m, we then get the vertical force in kN.

This equation was used to convert'he measured strains into vertical forces.

8.5.2.3.1.2 Measured Vertical Forces Figure 8-145 shows a typical measurement trace for the vertical force It increases rapidly during 'the expulsion of the water column and, after reaching the maximum value, returns quickly to zero.

8 5 2 3.1.2 1 Lo~n Line The vertical force exhibits a strong relationship with vent clearing pressure as shown in Figure 8-146 This holds true for all tests, even those with random initial conditions such as the real conditions and multiple actuation test.

As discussed in Section 8.5.2 1.3, the vent clearing pressure is inturn influenced by the reactor pressure, initial water column in the'discharge line, discharge line temperature, etc. and was REV 1 3/79 8-6 4

PROPRI ETAR Y extrapolated out to a maximum reactor pressure of 88 bar.

Therefore, the maximum vertical load will be extrapolated to the maximum vent clearing pressure from Section 8.5.2.1.3.

The measured maximum value for the vertical force is:

149 kN at a 12 8 bar vent clearing pressure.

8.5.2.3.1.2 2 Short Line Figure 8-147 illustrates the dependence of the vertical force on the vent clearing pressure. In principle, the same discussion as in Section 8. 5.2. 3.1. 2.1 for the long line applies here also.

The measured maximum value for the vertical force is:

192 kN at a 16 8 bar vent-clearing pressure.

The vertical forces relative to the vent clearing pressure are practically the same.

8 5.2.3.1. 3 Transposition of the Measurement Values to SSES As was discussed previously for the extrapolation of the vent clearing pressures, the measurement values for the vertical force can also be transposed directly to the plant. For verification of extreme conditions in the plant, the measurement values are extrapolated to a reactor pressure of 88 bar. The extrapolation can be performed directly via the vent clearing pressure.

8.5.2 3.1.3 1 Long Line The measured maximum value was:

149 kN at a 12. 8 bar vent-clearing pressure Slope = 13 kN/bar (Figure 8-146)

According to Section 8.5 2 1 3, the extrapolated vent-clearing pressure for the long line was 16 bar Extrapolation of the vertical force to 16 bar yields:

Fy = 190 kN max 8 5.2 3.1..3.2 Short line The measured maximum value was:

192 kN at 16 8 bar vent-clearing pressure Slope = 13 kN/bar (Figure 8-147)

According to Section 8.5.2.1.3 the extrapolated vent clearing pressure for the short line was 20 bar.

REV 1, 3/79 8-6 5

PROPRIETARY Extrapolation of the vertical force to 20 bar yields:

Fvm x In addition Figure 8-147, shows a measured value of 149 kN at a 12 bar vent-clearing pressure.

This leads to a maximum extrapolated vertical force of:

F~~y= 252 kN 8 5 2 3.1 3.3 Summary The extrapolation of the measurement results for the vertical force yields a ma'ximum value of:

F~ ~~~ =, 252 kN>>

In Figure 4-11, the specified vertical force is given as 860 kN.

Dn the basis of the measurement results, the specification value can be viewed as extremely conservative, both in the maximum value and also in the load-versus-time function.

8.5.2 3 2'ore'ional Moment 8~2~3.2 l Meas~nement of the Torsional Moment To measure the torsional moment, two strain gauges (SG 4.8 Figure 8-13) were connected in such a way that they measure strain resulting from torsional moment only.

According to Reference 41, there is a very simple relation between the torsion or shear strain and the measured strain, when the strain gauges are mounted at a 45o angle relative to the principal shear stress direction.

Qe have:

Y ms shear strain Thereforei since the strain gauges SG 4.8 were mounted at a 45o inclination to the vertical axis, ve have:

z = shear stress G

G ms shear modulus Y =2. eandr~D a 2

REV 1, 3/79 8-66

PROPBI ET A HY torsional g = moment Ip IP = polar moment of inertia bar r ~ outside radius of the twisted cylindrical Y

r G

P Qe thus obtain the relation between torsional moment and measured strax.n D

a The shear modulus is defined as G

2(1+ p) p = poisson's ratio Mith E = 2.06 x 10s N/mm~ and

'p = 0~3 Me get:

G ~ 7.9 x 10'/mme The polar moment of inertia is defined as 4

7f ~ D (1-D /D )

p 32 Therefore:

I 4.64 x 10 m Inserting the various numerical values, we get:

0.41',

Inserting E. inMm/m, this equation gives us the torsional moment in kN-m HEV lg 3/79 8-6 7

PROPRIETARY This equation was used to convert the measured strains at SG 4.8 into torsional moments.

The torsional moments obtained in this manner represent static equivalent loads.

8 5.2 3 2.2 Measured Torsional Moments Figure 8-148 shows a typical measurement trace for the torsional moments. After the end of the vent clearing process, (approximately 1 second after test start) the amplitudes of the measured torsional moments are very small compared to the maximum amplitude during the vent clearing process There is a factor of 6-7 difference between the two of them. The maximum amplitude of the torsional moment occurs much later than the expulsion of the water column.

8.5.2.3 2.2.1 L~on Line The torsional moment at the bottom support has,its origin only in unsymmetrical processes at the quencher during the vent clearing and during the transition to stationary condensation.

i Figure 8-149 shows the dependence of the torsional moment on the vent clearing pressure. Since the vent clearing pressure is a direct influencing parameter (see Section 8.5.2.3.1.2.1) we will correlate the torsional moment with that value.

The sharply pronounced scatter band is an indication that a random process is superimposed on that dependence. That is expressed by the fact that the torsional moment is brought about by random unsymmetry.

The measured maximum value of the torsional moment is:

M T maX

= 55.8 kN-m at a 14 bar vent-clearing pressure.

$ .5 2 3.2.g 2 Short Line Pigure 8-150 again shows the dependences of the torsional moment on the vent clearing pressure. In principle, the situation is the same as in the preceding Section fo- the long line.

The measured maximum value of the torsional momen't is:

39. 2 kN-m at a 18 bar vent-clearing pressure.

8.5 2 3.2.3 Tran~s osition of the Measurement Values to SSES Shen transposing the measurement results to SSES, we shall consider in a conservative manner the load carried by the discharge line, which in the test stand is connected rigidly (but REV li 3/79 8-68

PROPRIETARY not in a leaktight manner) to the quencher and bottom support by means of weld brackets(see Figure 8-13 and 8-14) in contrast to the free moving sliding joint at SSES. To do that, we make the assumption that the discharge line is fixed in a torsion resisting manner at the f irst bend above the quencher.

That results in the following picture:

Discharge Line Quencher Bottom support

/ / / /

The torsional moment N~ acts at the quencher. The torsional moment N~~ was measured. at the bottom support. The discharge line carries the torsional moment M><.

Therefore:

+M 2 Prom the equality of the rotation, we get:

"T V Y

G IpG Therefore:

Tl 1 1 = "T2 '2 '2 G ~ Ipl G ~ Ip2 Tl = P1 ~ 2 2 T2 P2 1 1 Me have the following dimensions:

r la = 0.1775 m r 2a = 0.162 m r1g 0.125 m 2g ~ 0.3.445 m 0.45 m ~ 11.313 m 1 2 REV 1, 3//79 8-69

PROP RI ETA R Y Therefore:

4.64 -4 4

. 10 m Z 4.0,. 10

" m4 Therefore:

Tl 4.64 0.162 11.313 26.6 M 4 0'1775 0'45 2

26.6 T Tl 'T2 Tl (1 + 1 )

26.6 M 1.0376 M 1

Thus, the load transmitted to the discharge line is less than 4g of that transmitted to the bottom support.

If, without taking into first use Pigures 8-149 consideration the discharge line, we and 8-150 as the basis for an extrapolation of the measured maximum values to maximum vent-clearing pressure for the corresponding discharge line, then we get t he following maximum val ues:

a) long line Mr, ~= 59 8 kN-m b) short .line Mr) ~ax = 43. 2 kN-m If we this then now consider the torsion carried by the discharge value is increased to a maximum of:

line, "ri ~x= 62 kN-m The torsional moment specified in 4.1.2.6 for the guencher support was 40 kN-m to be applied as a step function A torsional moment step function applied to an undamped one mass REV l, 3/79 8-7 0

PROPRIETARY oscillator (quencher acting as inertial mass and bottom support as a torsional spring) corresponds to a maximum response of:

M<< = 2(40) kN-m = 80 kN-m Since the maximum torsional moment derived from the Karlstein tests is M< = 62 kN-m, the specification is conservative.

/

8-5.2.3 3 Bean~in ncnente at the guenchec Aten 8 5.2.3 3.1 Measurement of the Bending Moments In the Karlstein tests, the bending moments vere measured in the horizontal plane (parallel to the tank's bottom) and also in the vertical plane, at both of the quencner arms.

To accomplish that, two strain gauges each vere connected in such.

a way that they measured unsymmetrical strains resulting f rom normal stresses (unsymmetrical component) . The following strain gauges were mounted for that purpose (see Figure 8-13:

SG 4. 1) Moments in vertical direction SG 4 3)

SG 4.2) Moments in horizontal directon SG 4 4)

The strain gauges were mounted approximately 150 mm from the weld between the quencher arm and the central ball.

The section modulus of one quencher arm is:

W

+ D 3 (1- )

a

'D ~ 0.4064 m a

D = 0.3744 m Qe have:

a =cE=M/W M =cEW This leads to the equation betveen quantities: M = 0.38-c 8-7 1

PROPRIETARY This gives the bending moment in kN-m, if c is inserted in p m/m.

With this equation, all the measured bending strains were converted into bending moments. The bending moments thus calculated are static equivalent loads.

8.5.2 3.3.2 Measured Bending Moments Figure 8-151 shows a typical measurement trace of the measured bending moments at the quencher arms. We see clearly that the maximum values occur much later than the clearing of the quenc her.

The evaluation of the individual bending moments relates to the total resultant bending moment, i e., the bending moment which actually loads the. quencher arm. The resultant bending moment is obtained by using the relationship:

2 M

x'es

~'gM + M y z The bending moments Mg are read off at SG 4.2 and 4;4. The bending moments Mz are read off at SG 4.1 and 4 3 The resultant bending moments exhibit no deterministic dependence on the vent clearing pressure, as sho vn .in Figure 8-152. Therefore, the resultant bending moments on the quencher arms must be considered as statistical values.

The measured maximum value of the re ultant bending moment is 63 k N-m.

e 8.5 2.3 3.3 Transposition of the Measurement Results into the Weld In Section 4 l. 2. 5, the bending moments in the weld were specified. In the Karlstein test stand, the strain gauges were mounted about 150 mm from the weld in ordernot to measure localized stresses due to the weld and the intersection between the ball central body and the quencher arm. Available experience indicates that this distance is sufficient to measure a stress profile which is independent of shape factors.

From the specified force and moment (Table 4-10), we obtain for the distance between the weld and the force producing the bending moment:

lp ~ 19 ~ Oo 655 29 By treating the quencher arm as a cantilever beam, we obtain for the maximum stress and thus for the maximum bending moment:

=

g0.655 M g (0.655-0.15)

.REV~ 1>> 3/79 8-72

PROPRIETAR Y M = bending moment in the veld B max M

= measured bending moment B max Therefore:

= 1.297 M B max B meas Thus, based on the measured maximum resultant bending moment of 62 KN-m (see Section 8 5 2.3. 3.2), we obtain the following maximum bending moment in the weld:

'aximum resultant bending moment: 81 kN-m 8 5 2.3 3.4 ~Secified Static Eguivalent Loads As already noted above, the measured bending moments are to be considered as static eguivalent loads In Section 4.1. 2. 5 Table 4-10, two'ontributions were specified with respect to the bending moment in the weld:

a) a step function having a step height of 19 kN-m b) a maximum differential pressure vhich, according to Section 4.1.3.7, is 0 8 bar from KKB trace No. 35 with a 0.5 multiplier. This results in a maximum differential p res sure o f 0. 4 ba r.

The contribution of the differential pressure is to be viewed statically, since, according to Section 4 1 3.5, the freguency of the differential pressure is approximately 6 Hz. The bending eigenfreguency of the guencher arm is on the order of 100 Hz.

The contribution of the differential pressure to the bending moment in the weld is thus:

11.4 kN-m The contribution of the step funcion is to be viewed dynamically.

Therefore, the same considerations are applicable as those made for the torsional moments in Section 8.5. 2. 3.2.3. Accordingly, we have the following static eguivalent loads:

Component in one Direction Contribution from step function = 2 X 19 = 38 KN-m Contribution from differential pressure = 11.4 KN-m Total = 49.4 KN-m REV li 3/79'-73

PROPRI ETAR Y Resultant Moment Contribution from step function = 38 x ~2 = 53.7 KN-m Contribution from differential pressure = 11.4 KN-m Total = 65.1 KN-m 8.5.2 3.3. 5 Fvaluation of the Measurement Results As already mentioned in Section 8. 5.2 3.3.2, the bending mo'ments on the quencher arm are to be treated as statistical values.

Figure 8-153 shows the frequency distribution of the measured maximum bending moments in each tests and the resulting frequency disrihution of the values transposed in to the weld.

The frequency distributions are based on the peak maximum value of each individual test, which were measured either at SG 4.1/4.2 or at SG 4. 3/4.4.

The specified static equivalent loads (see Section 8.5.2.3.3.4 are introduced for 7000 responses of the relief valve.

Therefore, the loads are to be evaluated in a fatigue analysis.

It follows from Figure 8-153 that the mean value of the measured maximum values transposed into the weld is 35 kN-m.

Except for three cases, the specified resultant bending moments also cover the maximum measured values. The quencher is being evaluated for these measured maximum values.

It should be noted that both the specified stationary internal quencher pressure of 22.0 bar and the resulting thermal load of 219~C were found to be very conservative when compared to the maximum extrapolated values of 13.0 bar and the resulting saturated steam temperature of 195~C measured during the tests.

(Section 8. 5. 2. 2 3) .

8.5.2 3.4 Bending Moments at the Bottom S~ugort 8 5 2.3.4.1 Measurement of the Bending. Moments To measure the bending moments a the bottom support, two strain gauges capable of measuring the bending strains were mounted. In, the measurement arrangement, the bending strains could be measured in two mutually perpendicular directions (see Figure 8-

13) . The strains for moments about the x-axis were measured with the strain gauges SG 4.5. The strains for moments about the y-axis were measured with the strain gauge SG 4 6.

REV lg 3/79 8-7 4

PHOPRI ETAR Y The section modulus of the bottom support is:

4 W = D 32 a 3

(1-D 4

)

a W ~ 1.307 x 10 -3 m 3

We have a ~E ~

@~M/W This leads to the equation:

M ~ 0.27 ~

c This equation gives the bending moment in kN-m, in pm/m.

if c is inserted This equation was used to convert all measured bending strains of the bottom support into bending moments. The bending moments thus calculated are static equivalent loads.

8.5 2.3.4.2 Measured Bending Moments In Figure 8-151, the bending moments at the bottom support can be seen under the traces of the bending moments at the quencher a rms.

The maximum values occur at a later time than the vent clearing.

But they occur at the same time as the maximum values of the bending strains at the quencher arms. The maximum strain resulting from torsion does not occur at the time of the maximum bending strain (see Figure 8-151, SG 4.8) .

The evaluation of the bending moments relates to the resultant bending moment, i.e., the bending moment which actually loads the bottom support. The resultant bending moment is obtained by interconnecting the actual load-versus-time functions of the individual components through the relation:

The bending moments Mz are read off at SG 4.5 and the bending momen ts M at SG 4. 6 The maximum resultant bending moment was 54.5 kN-m The resultant bending moments display no dependence on the vent clearing presure, as shown in Figure 8-154. Hence, the same conclusions that were drawn for the bending moments at the quencher arms are applicable here, also.

BEV. 1, 3/79 8- 75

PBOPRI ETAB Y 8.5.2 3 4.3 Specified Static Equivalent Load.

As already mentioned, the measured bending moments are to be viewed as static equivalent loads.

The bending moments at the bottom support are introduced through the quencher.

Section 4.1 2.4 and Table 4-7 specify a transverse force of 44 kN on the quencher was used as step function.

In addition, a maximum differential pressure of 0.4 bar on the quencher was specified. The contribution resulting from the differential pressure is to be viewed as a statically acting load. It amounts to 48 kN.

Note: The discharge line and the bottom support were not considered here. The presssure difference was formulated only over the projected area of the quencher.

The specif ication then yields the following transverse forces on the quencher:

Contribution from step fun'ction = 2 x 44 = 88 kN Contribution from differential pressure = 48 kN Total = 136 kN Strain gauges SG 4.5 and SG 4.6 were mounted approximately 0.5 m below the center of the quencher. Transposed to this location, the specification yields:

68 kN-m 8 5.2 3. 4 4 Evaluation of the Measurement Results Figure 8-155 shows the frequency distribution of the measured maximum bending moments at the bottom support. The measured maximum values are also covered by the specification.

Thus, the Karlstein tests have demonstrated that the specified transverse forces on the quencher can be viewed as very conservative.

8 5 2 3 5 Forces on the Quencher In the Karlstein Quencher Tests, only bending moments were able to be determined for the quencher itself. In Section 4.1.2, forces and moments on the quencher were specified. The specified momen ts were calculated from the f orces. The measured moments are w ithin the specif ication. Theref ore, we can concl ude that the forces are also verified.

REV li 3/79 8-76

PROPRIETARY 8 5. 2 3. 6 Influence of an A dgacent Quencher During the clearing of the quencher, strong turbulences and eddies of the expelled and ambient water develop around the discharging quencher. In particular, after the vent clearing the quencher is surrounded by a large number of air bubbles which represent a locally compressible volume in the water. This state, which forms around the discharging q<<niche>i prevents effects from the blowdown of an adjacent quencher from penetrating to the quencher under consideration.

It is therefore understandable that, in within the Brunsbuttel and Philippsburg the KMU in plant tests nuclear power plants, no increase of the load on the quencher and bottom support was found for the response of several quenchers in comparison to the response of one quencher (Reference 6) .

An effect of a load on one quencher due to the firing of an adjacent quencher is to be observed only when the adjacent quencher blows down alone.

In that case, a detailed evaluation was made for the Brunsbuttel blowdown tests (Reference 38).

The result of the investigation was that the measured loads are enveloped by a pressure difference of 0.2 bar applied over the adjacent internal structures in the pool at the quencher level, i.e., also over the quencher.

A maximum pressure difference of 0.4 bar over the quencher arms was specified for SSES. The vent clearing pressures and dynamic pressures in the water pool obtained for SSES from the Karlstein tests are of the same order of magnitude as the corresponding measurement results in Brunsbuttel.

Therefore, the specified differential pressure of 0.4 bar over the quencher arms can be viewed as conservatively enveloping.

8.5 2 3 7 Loads on the Quencher During Steam Condensation The maximum mechanical and thermal loads on the quencher during the condensation phase occur during the phase of intermittent condensation. In Section 4.1.2.7, the loads resulting from intermittent condensation were taken as the basis for the fatigue design of the quencher.

The evaluation of the loads on the quencher during steam condensation in the Karlstein tests therefore relates primarily to the phase of intermittent condensation.

R EV 1, 3/79 8-77

PROPRIETARY 8.5.2.3 7 1 Manifestation /orms of Intermittent Condensation in the Karlstein Tests As discussed in Section 8.1. 3, the condensation tests vere performed along the lower and upper boundary lines of the operation field for water temperatures <30~C and also for water temperatures >590C. In both regions, the intermittent condensation phase occurs for very low reactor pressures (approximately between 2 and 4 bar). In Section 8 4.2 it shown that the maximum values for the dynamic pressures in the is vater region occur .during intermittent condensation in cold vater The same is true also for the loads on the quencher.

Fori the evaluation and comparison with the specification, we use the measurement values of the bending moments at the quencher during the intermittent condensation in the cold pool. The measurement values are documented in Section 8.4.2.

8 5.2 3.7.2 Illustration of the Measurement Values The time duration of the intermittent condensation in the cold pool was about 100 seconds. The total number of condensation events at the quencher was 52. The maximum measurement values occurred in the vertical direction at SG 4.3.

The frequency distribution of the resultant bending moments (SG 4.'3/4.4) at the quencher arm is show in Figure 8-156. The mean value of the maximum measurement values of each event is 11.8 kN-

m. The maximum measured value was 66.5 kN-m.

The f requency distribution of the resultant bending momen ts (SG 4.5/4.6) at the bottom support is shown in f igure 8-158.

The mean value of the measuremen t values is 8 9 kN-m. The maximum value was approximately 30 kN-m The measured maximum value of the torsional moment during the inter mit te nt cond ensa tion is 6. 2 k N-m.

8.5 2 3 7 3 Evaluation of the Measurement Results for the Ouencher Arm Figure 8-157 shows the frequency distribution of the resultant bending moments, which were transformed from the measuring point into the weld (see Section 8.5.2.3.3.3. The mean value of these bending moments is 15.2 kN-m. The maximum value is 86 kN-m.

The measured bending moments represent static equivalent loads.,

In Section 4. 1.2.7 and Table 4-12, a value of 25.4 kN-m was specified .for the equivalent load for the resultant bending moment in the weld during intermittent condensation. The loads specified are formulated for an occurrence frequency of 106.

REV. li 3/79 8-7 8

PROPRIETARY In the fatigue analysis, the mechanical loads represent only one load component. Another part of the fatigue loading is produced by the alternating thermal loading. The assumption made in the specification was 106 temperature steps from 35~C to 133~C and from 133~C to 35oC.

The low-frequency oscillations of the pipe's internal pressure measured at P4.4 are used as a basis for the measured temperature alternation. The saturated-steam temperatures are then correlated with those pressures.

The pressure oscillations have an oscillation frequency of about 0.5 Hz and a maximum amplitude of 0 5 bar overpressure = approx.

2 bar absolute pressure. This pressure lies below the specified value of 3 bar.

The measured maximum pressure of 2 bar corresponds to a saturated-steam temperature of 120~C. Assuming that the inflowing water in SSES is at a temperature cf at least 35~C, then the temperature step is 85 C.

A temperature step of 98~C is assumed in the specification, so that there is a reserve of 13~C.

The measurement values forming the basis for the evaluation and comparison with the specification were observed only during the phase of intermittent condensation with cold water in the test tank.

As with the boundary pressures in the test tank (Section 8.4. 2),

the loads on the quencher were considerably lower during the intermittent condensation phase with warm water than during intermittent condensation with cold water. The measured maximum bending moment during this condensation phase was (1 kN-m relative to the weld seam.

In addition, KMU in plant tests in the Brunsbuttel nuclear power plant showed that, for a pool water temperature of approximately 35~ and above, intermittent condensation loads on a quencher were smaller. This indicates that the region wh'ere intermittent condensation loads of any consequence can be expected is limited to that of very low pool temperatures (approximately 25~C) and very low steam mass flows and that heating of the pool a small amount results in a reduction in loading 8.5 2.3.7.4 Evaluation of the Measurement Results for the Bottom S~uport An impulsively acting transverse force of 17.5 kN was specified on the quencher for intermittent condensation.

REV l~ 3/79 8-7 9

PROPRIETARY The distance from the middle of the quencher to the measuring point for the bending moments at the bottom support is 0. 5 m, so that the specified bending moment with respect to the bottom support is:

(17 5 kN x 2) x 0.5 m = 17.5 KNm (static equivalent load)

The maximum resultant bending moment from the tests is approximately l

30 KN-m.

8 5.2 3 7 5 Evaluation of the Measured Torsional Moments An impulsively acting torsional moment of 19 kN-m was specified for the intermittent condensation.

This step function yields a torsional moment of:

38 kN-m as the static equivalent load The specified torsional moments conservatively envelop the measured maximum value of 6.2 kN-m.

8 5 2.3.7 6 Evaluation of the Measured Maximum Moments at the Quencher Arm during Intermittent Condensation A maximum resultant bending moment of 66 5 kN-m at the quencher arm was measured in the intermittent condensation phase, which results in a moment of 86 kN-m in the weld. The measured maximum values of the resultant bending moments at the quencher arm during intermittent condensation are on the order of magnitude of the measured maximum vlaues during the vent clearing phase (Sect ion 8. 5. 2. 3. 3. 2) .

For the vent clearing, a temperature difference of 184~C was specified. For the intermittent condensation, a temperature difference of 98~C was specified.

The total stresses loading the quencher arm are composed of mechanical and thermal stresses. The thermal stresses are distinctly larger than the mechanical stresses.

The maximum resultant bending moment at the quencher arm for intermittent condensation exceed the value specified for the vent clearing by about 40%., However, the associated temperature jump is only about half as large as for the vent clearing.

REV -

1, 3/79 8-80

PROPRI ETAR Y 8 5 3 Verification of Suppression Pool Boundary Load Specification Due to SRV Actuation In Section 4 l. 3, three pressure time histories are specified as the basis for the containment analysis due to SRV actuation. The three traces vere taken from a large number of bottom pressure time histories from various KKB in plant tests.

The evaluation of the pressure oscillation measurements in the Karlstein vent clearing tests will therefore concentrate on demonstrating that the pressure time histories specified are enveloping.

Accordingly, analysis and assessment. of the individual measured pressure time histories is restricted to a minimum.

8.5.3.1 Evaluation of the Local Effects Seen at Pressure Transducer P5.5 As shown in Figures 8-10 to 8-12, the pressure transducer P5. 5 is mounted on the concrete wall opposite the middle of the hole array on the quencher arm.

About 0.25 seconds after expulsion of the water column, P5.5, in comparison with the other pressure transducers, exhibits high-frequency positive pressure peaks which are not observed at the neighboring pressure transducers. This effect is from the local t urbu1ences.

These high frequency pressure peaks have a small energy content so that their range of action is limited to the immediate vicinity of the pressure transducer.

following Table should make this clear In this Table, the

'he ratio of the measured pressure amplitudes of the neighboring pressure transducers {P5. 10 and P5.4) to the pressure maximum at P5.5 is indicated for all tests vhich exhibited a maximum pressure amplitude > 1 bar at pressure transducer P5.5.

REV 1, 3/79 8-81

PROPRIETARY p 5.lO PS'+

Test P5 4 P5. 10 P5. 5 P5. 4/5 5 P5. 10/P5. 5 (ba r) (bar) (ba r) 4 1 6 0,6 0,55 1,0 0,6 0,55 5.1.7 0,45 0,4 li0 0,45 0,4 10 R1.7 0,73 0,55 1,7 0,43 0,32 20 Rl 9 0,45 0,4 . lr0 0,45 0,4 20 Rl 10 1,0 0,65 1,73 0,58 0,38 25 1 0,55 0,6 1,0 0~55 0,6 25 R2 0,85 0 8 1,55 0,55 0,52 From this Table we can see that the measurement value has decayed by half at about 1 m from the measuring point P5. 5.

The comparison measurement points P5.4 and P5.10 are in the region of origination of the air bubble oscillation, so that no attenuation effect due to distance effects could occur at that measuring point Therefore, the sharp decrease of the pressure amplitude which is measured nevertheless shows clearly that the pressure measured at pressure transducer P5. 5 is limited to its local vicinity.

As further verification that this effect is limited to the area around pressure transducer P5.5, a comparison is made between the power spectral densities from P 5.5 and the bottom pressure transducer P 5. 2.

REV. 1,, 3/79 8-8 2

PROPRIETARY The following tests vere selected:

Test 11.1 This test exhibited the highest power spectrum at the dominant frequency Test 4.1.6 This test exhibited the highest pressure amplitude at P5.5 for the long discharge line Test 20.R1.10 This test exhibited the highest pressure amplitude at P5 5 for the short discharge line.

The comparison can be summarized as follovs:

At the dominant frequency, the power densities are the same magnitude for the pressure oscillations at the bottom pressure transducer P5.2 and at. pressure transducer P5. 5.

The differences at the higher frequencies is significant. For tests 4.1.6 and 20.R1.10 the frequency spectrum of P5.5 exhibits significantly higher power densities at higher frequencies than the corresponding frequency spectrum at pressure transducer P5 2.

This significant factor is not noted for the frequency spectrum of test 11.1 (see Figures 8-159 and 8-160). In that test, the difference between the maximum pressure amplitudes for pressure transducers P5.5 and P5. 2 was 0 13 bar. The pressure ratio is P5 5/P5 2 = 0 '8/0 65 = 1 2.

In test 4. 1.6, the difference in the power densities at the higher frequencies is already more strongly evident (see Figures 8-161 and 8-162). In that test, the difference between the maximum pressure amplitudes for P5.5 and P5. 2 was 0.5 bar. The pressure ratio is P5. 5/P5 2 = 1/0.5 = 2.

The difference in the power densities at the higher frequencies is quite strongly pronounced in tests 20. Rl. 10 (see Figures 8-163 and 8-164) . The difference in the maximum pressure amplitudes for P5.5 and P5.2 was 1.1 bar in that test. The pressure ratio is P5.5/P5.2 = 1.73/0. 63 = 2.75.

The pressure differences or pressure ratios are not discernible in the power spectra for the dominant frequencies, but are at the higher frequencies From that we can conclude that the pressure oscillation which was measured at pressure transducer P5- 5 has approximately the same amplitude at the dominant frequency as the pressure oscillations vhich were measured elsewhere in the vicinity of the quencher, e g., at P5.2 In addition, higher frequency pressure oscillation components having a high amplitude are occasionally superimposed on the f undamental oscillation in the pressure oscillations at P5 5.

The higher frequency components, vhich occur at pressure R EV 1, 3/79 8-83

PROPRIETARY transducer P5.5, decay rapidly in time and space, so that the effect of the high frequency pressure oscillations remains limited to the immediate vicinity of measuring location P5.5 Therefore, as stated before, the measurement results for the dynamic pressures at P5.5 represent local events having no global effect on the containment.

We will therefore not consider the positive pressure measurements at P5.5 when verifying the design specification for the overall containment analysis the results from this gage are included for the verification of the loadings on the columns.

8.5.3.2 Verification of the~Secified Pressure A~mlitudes and Vertical Pressure Profiles after Vent Clearing The measured peak pressure amplitudes for the 125 vent clearing tests are tabulated in Tables 8.9 and 8 10. Section 8.4.1 also presents a number of Figures (8.27 to 8.34) which show that the pressure amplitudes measured in the tests had no significant dependence on the initial reactor pressure. Therefore, no modification to the measured pressures will be made to account for differences in the reactor pressure between SSES and the Karlstein test stand. In addition, as explained in the previous section, the positive pressure measurements a P5.5 will not be considered when verifying the design specification for the overall containment analysis.

8 5 3 2.1 Overpressures The maximum over pressure amplitude measured on the boundary of the Karlstein test tank was 1.0 bar That pressure was measured at the concrete wall (p5.4) in test 20.R1.10. A maximum pressure amplitude of l. 2 bar is speci fied in section 4.1. 3 (KKB Pressure Trace No. 35 with the 1.5 multiplier) . The maximum specified overpressure amplitude of 1.2 bar .evelops the measured maximum overpresure amplitude of 1 0 bar.

8.5.3 2.1.1 Vertical Pressure Profile It can be assumed that the maximum dynamic pressure vill occur in a sphere which surrounds the quencher and has approximately the radius of a quencher arm, (5'-0") .

At some distance from accordance with a distance it, the maximum value will be attenuated in law. For an infinite water space, the 1/R law is applicable for the decrease of the pressure with distance from the source. That law applies in all directions, i.e , in the vertical direction also. The validity of the 1/R law is based on the assumption of a stationary (i.e., fixed position) oscillating bubble in the infinite water space. That ideal case does not hold for the clearing of the relief system.

Already shortly after the expulsion of the air-steam mixture, BEV 1, 3/79 8-8 4

PROPRIETARY small air particles move to the surface of the pool because of buoyancy. Even more important, however, is the fact that the water surface and the tank boundary surfaces influence the distance law and that the pressure amplitude must vanish at the water surface itself.

Accordingly, a pressure profile in the vertical direction is specified in Section 4.1. 3.4 providing for a constant 'pressure at 6e0 (1 83 m) above the suppression pools bottom and, starting at

<hat height, a,linear decrease of pressure u p to the water sur face.

Figure 8-165 shows that the maximum specified pressure distribution very conservatively envelops the measured maximum pressure amplitudes. The conservativeness becomes clearly evident amplitude if, of based on the measured maximum value of wall pressure 1 bar at pressure transducer P5.4, we assume a linear decrease of pressure from that measur ing point to the water surface. That assumed linear pressure decrease (depicted in Figure 8-165 by a dashed line) also envelops the maximum pressure amplitudes measured in the vertical direction. In comparison with the assumed linear pressure decrease and the specified pressure distribution, the conservativeness of the specif icat ion becomes obvious.

~

2 Vertical Pressure Profile Iuclu ainu local

'-5-3-~2..

Ef facts at P5. 5 For the evaluation of the unpertubed pressure distribution in the vertical direciton, the measuring point P5.5 was omitted, even though it lies in a direct line with the pressure transducers P5.4, P5.6 and P5.7. Because of the local effect for P5.5, a separate analysis shall be performed here.

That analysis starts with an estimation of the vertical zone of influence associated with the pressure peak measured at P5.5.

The lateral holes in the quencher arms extend over an angle range of 72~ on each side. The holes are drilled radially, so that in first approximation we can assume a source flow of the emerging fluid. The high-frequency pressure peak at P5.5 occurs at a much later time than the vent clearing. It can be supposed that at that time there is a steam-air mixture flowing out of the quencher. The steam-air jets emerging from the holes have a high degree of turbulence. Thus, the edges are very soon mixed with the surrounding water. Furthermore, the emerging steam is condensed immediately and the expelled air is cooled down quickly, so that the expelled compact volume is reduced rapidly.

Therefore to estimate the range of action, source flow acts over a mean angle range of 8 = e/2 = 720/2 it is assumed that the 36o. The total range of action is then REV li 3/79 8- 85

PROPRI ETA R Y b = x tan 36~ x = 1.575 m {distance from centerline of b =1.14 m quencher arm to concrete wall)

This range of action of '1.14 m is divided into equal parts above and below the measuring point P5. 5, so that we obtain a range of action of 10.57 m relative to the measurement location Based on this range of actin the measured vertical pressure distribution considering the local effect is compared with the specified pressure distribution in Figure 8-166. The base points of the pressure elevation at P5.5 were placed on the straight line of the linear pressure drop symmetrically with respect to the quencher's center plane.

From Figure 8-166 it can be seen that the maximum specified pressure distribution results in a larger resultant force on the containment boundary and columns than does the measured pressure distribution including consideration of the local effect This means that the overall specified pressure distributrion in the vertical direction also envelopes the local pressure elevation at p5. 5.

8. 5 3 2 2 Unde~r ress ur es The maximum underpressure amplitude measured on the boundary of Karlstein test tank was -0.68 bar. That pressure was measured a the concrete wall. {P5.10) in test 25. R2. A maximum underpressure amplitude of -0.56 bar is specified in Section 4.1.3 {KKB Pressure Trace No. 76 with the 1.5 multiplier).

The'next largest underpressure recorded during test 25.R2 was

-0 50 bar.

The next largest underpressure recorded anywhere during the vent clearing tests was -0 58 bar at P5.2 in test 25.1.

Except for the two measurement values called out above all other measured underpressures were hounded .by the maximum specified value of -0. 56 har.

8.5.3 2 2.1 Vertical Pressure Profile Figure 8.167 shows a plot of the maximum specified underpressure distribution and the maximum measured underpressure values for the Karlstein tests.

It can 2S.R2, he seen that, except the maximum specified for the pressure one value at P5 10 for test distribution envelops the maximum measured pressure amplitudes.

REV. 1, 3/79 8- 86

PROPRIETARY In addition, for SSES, the most unfavorable boundary condition in this comparison is the low liquid level of 22 ft = 6.70 m in the suppression pool.

The hydrostatic pressure distribution with respect to that liquid level is indicated by a dashed line in Figure 8-167.

The comparison of the measured worst underpressure distribution vith the hydrostatic water load resulting from the worst boundary condition for this comparison (lowest water level in the suppression pool) shows that the compressive forces from the water load and the tensile forces from the underpressure distribution= maintain the equilibrium." Thus, the Karlstein tests have, in addition, demonstrated that the blowdown of the SSES relief system vith the quencher does not result in any resultant tensile forces on the steel liner, even for the worst possible superposition.

8 5.3 3 Verification of the Pressure Time Histories Used for the SSES Containment Analysis Xn order to verify that the pressure time histories used for the SSES dynamic analysis due to SRV actuation are bounding, the Power Spectral Densities (PSDs) of the specified time histories (with the appropriate amplitude increase and frequency range from Section 4. 1.3) are compared with the PSD's of the appropriate time histories recorded in the Karlstein test tank and transposed to the SSES" suppression pool.

Statements concerning the clearing of parallel quenchers are based on the unrealistic and extremely conservative assumption that the expelled, air bubbles are equally large and oscillate in phase. A quantification of that conservativeness is not given.

Me vill first discuss and transpose the oscillation verify the theory to be used to frequencies measured in the test tank to the suppression pool. Then, the appropriate multipliers for this frequency transposition will be established. A discussion is also provided for transposing the measured pressure amplitudes to the suppression pool. Finally, the actual verification is presented.

8 5.3 3.1 Tr~ans osition method for the Oscillation Frequency The theoretical basis for the transposition of the pressure time histories measured in the Karlstein tests to the SSES suppression pool is provided by the KMU computer codes VELPOT and KOVIBlA By using the test results from the PPGL quencher tests in Karlstein, the GKM quencher tests, and the non-nuclear hot tests in the Brunsbuttel nuclear power plant (KKB hot tests), we shall first confirm experimentally the correctness of the transposition BEV 1, 3/79 8-87

PROPRIETAR Y theory. That is followed by a calculation of the frequencies for the following three blowdown cases:

(1) Simultaneous blowdown of all 16 quenchers (2) Simultaneous blowdown of the 6 quenchers related to the automatic depressurization system (ADS)

(3) Blowdown of one outer quencher For each case, a comparison of the theoretically calculated frequencies with the frequencies measured in the test stand) provides a number (frequency multiplier) by which a frequency measured in the test stand must be multiplied in order to get the corresponding frequency in the SSES suppression pool A factor for the influence of the suppression pool overpressure is also determined in the same way. The corresponding measured pressure time history is transposed to the plant by dividing by this factor 8 5 3. 3.1. 1 Calculation of Measured Oscillation ~r~cr uencies 8 5.3 3. 1. 1 1 PPGL Tests at Karlstein Since it was found that Karlstein test tank has Fluid-Structure Interaction in the no significant influence on the measured pressure analysis time for a histories, rigid tank.

it is sufficient to carry out the The comparison of calculated and measured oscillation frequencies will be based on'the assumption of equal bubble volumes. The measured oscillation frequencies are taken from Tables 8.9 and 8.10 . The associated bubble volumes were calculated from the test data, using the formula:

pp-pi [P

'[p ie -cP eee 'e)

(7

]

T

~eo>

T P

sat (Tpool) ] pipe pipe f ree pipe volume (ms)

Ppipe pressure in pipe (bar) hydrostatic pressure at the quencher location (bar)

Psat saturation steam pressure (bar)

C relative humidity ( s = 1 at 1005)

Tpool water temperature (oC) pipe mean temperature in pipe (oC)

The averaging of the temperature in the pipe is performed by using the for mula i 1 E

pipe N

i where the pipe was divided into N equal sections. The temperature T in the between the measured temperatures.

i th section was obtained by interpolation REV. li 3/79 8-8 8

PROPRIETARY The comparison between the measured and calculated bubble frequency is shown in Figures 8-168 and 8-169 in which the bubble pulsation frequency is plotted versus the equilibrium volume at static pressure. Por the measurement points in Figure 8-168 was assumed that dry air was in the pipe prior to the test start, it while wet air (100% humidity) was assumed in Figure 8-169. In general, good agreement is found between the theory and measured frequency. However, we cannot overlook the fact that the measured frequencies in figure 8-168) are higher than the calculated ones, especially for small bubble volumes. This may be related to the fact that the active volume of air under water is actually smaller than the volume found for dry air from the test data. This is hinted at by the calculation of the bubble volume under the assumption of 100% humidity in the pipe. There the measurement points are closer to the calculated curve (Pigure 8.169). In order to keep the uncertainties associated with such effects as small as possible, only tests for which the initial pipe temperature was below 700C were chosen for the comparison with the theoretical case.

8.5.3.3.1.1.2 GKM Mode~l uencher Tests Another sorce used to verify the theory is offered by the GKN quencher tests (Ref 1). Since the pipe temperatures there were in the vicinity of 300C or below, uncertainties in the bubble

. volume under wa ter are distinctly smaller than in the Karlstein tests. In addition, the GKM tests were also run with backpressure in the suppression chamber, so that information derived from the computer codes for blowdown of the quencher during a loss-of-coolant accident can also be verified. The results can be found in Figures 8-.170 and 8-171. Figure 8-170 shows the calculated and measured dependence of the pulsation frequency on the bubble volume for various submergences (2 m, 4 m and 6 m) with atmospheric pressure in the suppression chamber.

The theory and measured frequency agree even better here than in the Karlstein quencher tests. This is probably due to the fact that the bubble volumes determined from the measurement values have a much smaller scatter due to the low temperatures in the pipe. The influence of backpressure on the pulsation frequency is shown in Figure 8-171. Here again, the theory is verified by the test data.

8 5.3 3 1 1 3 KKB Hot Tests In order to demonstrate the correctness of the theory for in-plant conditions also, calculations were performed for the blowdown tests with one valve in the non-nuclear hot tests in the Brunsbuttel BMR plant (Ref. 3). Pigure 8-172 shows the results.

The agreement between the calculated and measured'requency is similar to that in the Karlstein tests. The same is true for the scatter range of the measurement values. Since the pipe temperature here was at about 90OC, a larger scatter actually REY lt 3/79 8- 89

PROPRIETARY would have been expected, but did not occur because the pipe was carefully flushed with air prior to the beginning of these tests.

8.5.3.3.1.1.4 Conclusion from the Frequency Calculations The test calculations described above show that the theory

{VELPOT and KOVIB1A computer programs) describes the measured frequencies not only in one special case, but also for a broad range of geometries and backpressure:

(1) The size of the water space varies from approximately 7 m>

(GKN) to approximately 23 m~ (test tank at Karlstein) to approximately 400 m~ (suppression chamber in Brunsbuttel nuclear power plant).

(2) The quencher submergence ranged from approximately 2 m to 6 m.

(3) The bubble equilibrium. volume var ied bet ween approximately 015 m~ to 37 m~.

(4) The suppression chamber pressure varied from 1 bar to 3 bar.

(5) The water temperature in the suppression pool varied between approximatley 16~C to 800C.

Thus, the theory can be considered verified and can be used to transpose the pulsation frequencies measured in the Karlstein test stand to the SSES suppression pool.

8.5.3 3.2 Nuit~i liers for Conversion of the Bubble Frequencies from the Test Stand to SS ES Using the VELPOT and KOVIBlA computer codes, the following three blowdown cases are analyzed:

(1) 'Simultaneous blowdown of all 16 quenchers (2) Simultaneous blowdown of the quenchers A, B, G, K, M, P which are included in the ADS

{3) Blowdown of one quencher (quencher B)

The results are illustrated in Figure 8-173 which shows the pulsation f requency as a function of bubble volume (bubble in hydrostatic equilibrium). The behavior of the frequency curve for the 16-quencher case in the plant is practically the same as for the test stand (Figure 168), thereby confirming once again the suitability of the test stand geometry that was chosen. In the case of the 6 quenchers in the ADS case, the frequencies are higher due to the larger single cell corresponding to the smaller REV li 3/79 ~ 8-90

PROPRI ETA RY hydrodynamic bubble mass. They are even higher in the case of one quencher.

Based on the results shown in Figures 8-168 and .8-173, a simple formula can be given for converting from the measured bubble frequencies to these frequencies found in the plant by asking:

By what factor t"multiplier") must a bubble frequency measured in the test stand be multiplied to get a corresponding frequency in the plant'? This multiplier is plotted in Figure 8-174 versus the (measured) starting frequency. Thus, we have:

vplant =fv (vtest ).'est v

in which the muliiplier f for a given initial freguency can he read off from Pigure 8-173.

The graph in Figure 8-173 is applicable only f or cases with a pressure of 1 bar in the suppression pool air space. However, the blowdown for the ADS case during a loss-of-coolant accident is associated with a suppresson pool overpressure.

p ~>l bar An additional multiplier fpKK (pKK) is necessary for such cases, so that the frequency conversion must be written in a more general manner:

V plant

= fP (P kk) '. S (Vtest ) . V test kk The multiplier fpKK (pKK') can be taken from Figure 8-175. For a suppression as it chamber pressure of 7 bar, must be.

it has the a value of 1, The multipliers for the frequency also fix the multipliers for the oscillation period when transposing the pressure time histories measured in the test stand to the plant:

t test t l P

kk kk v test I

8 5.3 3.3 Transposition Method for the Pressure A~mlitudes As already described in detail in Section 8. 5 1, the test stand was so designed and the. pressure transducers were so arranged that the measured pressure amplitudes can be transposed to the plant without change Correspondingly, a 1: 1 transposition is made. Because of its obvious conservativeness, such a 1:1 REV 1, 3/79 8-91

PROPRIETARY aiplitude transposition offers the advantage that more exact quantitative proofs do not have to be provided. The most significant conservative features are the folloving:

(1) In blowdown case's with several quenchers, it is assumed that all bubbles are equally large and oscillate in phase.

Deviations from this assumption {such as actually occur in the plant) result only in lover pressure amplitudes.

(2) Blowdown cases with less than 16 quenchers are assigned the same pessure amplitude as the 16-quencher case. In reality, such cases have a lower amplitude due to the geometry (larger single cell) .

The conservativeness described in (1) has not yet been proven experimentally in any quencher tests, but it is already obvious from a theoretical viewpoint, since a time-shifted superposition of tvo temporal maxima always yields smaller values than an addition of the maximum values.

Concerning the conservativeness of (2), there are a number qualitative indications from the Karlstein tests themselves, from corresponding model studies at the Karlstein model test stand (Ref. 1), and from calculations with the VELPOT and KOVIB1A programs. The information obtained from all three of these investigations shall be described in the following sections.

In addition, we will also examine whether the conservative features are affected by a possible backpressure in the suppression pool air space.

8 5.3 3.3 1 PPGL~uencher Tests at Karlstein Indications of the conservativeness discussed in (2) above are obtained from the Karlstein tests on the basis of Figure 8-176 vhich illustrates the measured relationship between excitation (relative amplitude) and pressure -oscillation frequency for the Karlstein tests.

The frequency analysis for each pressure time history has at least two maxima of the power density. One power density maximum lies at low frequencies and the other at somevhat higher frequencies. There is a factor of approximately two between the tvo freqeuncies. The first peak of the power density (low frequency). is always larger than the second peak of the power density (higher frequency). Accordingly, the lov frequency is alvays designated as the dominant frequency For pressure transducer P5 10, the power densities of all analyzed tests are evaluated in Figure 8-176. Different analysis times vere selected for tests having different pressure oscillation frequencies The time vas so chosen that R EV. 1, 3/79 8- 92

PROPRIETARY approximately the same oscillation periods could alvays be evaluated.

The following analysis times were selected for the evaluation:

3 Hz Time: 0 1.8 seconds 5 Hz Time: 0 1.3 seconds 9 Hz Time: 0 0.6 seconds The area beneath the frequency spectrum vas determined and then the square root of that numerical value was taken. That results in values having the dimension >>har>>.

Those numerical values were normalized to the maximum value.

The results are then "relative pressures>> with respect to the calculated maximum pressure from the frequency spectra.

Since no dominant frequencies higher than 6. 5 Hz were measured in the Karlstein tests, the second peaks were also used to evaluate the higher frequencies. Hence, the power densities of both the dominant frequency and the next higher frequency are evaluated in Figure 8-176.

Based on an empirical evaluation, it .follows from Figure 8-176 that the pressure oscillations vith higher frequencies have smaller energy content than the pressure oscillations with lover frequencies.

Zn addition, as shown in Figure 8-169, the hubble frequency increases with decreasing hubble volume. But decreasing bubble volume with constant single-cell size means, according to the laws of similarity, the same thing as increasing the cell size with constant bubble volume Therefore, from the Karlstein test data, it can be said that the pressure amplitudes decrease with increasing cell size.

8.5.3 3.3 2 KM~U uencher Tests in the Model Test Stand in

~Ka 1stein During the development of the KWU quencher, tests vere performed.

to examine the influence of the size of the water space (specifically: free water surface) in the model test stand in Karlstein (Ref. 1). The results are illustrated in Figure 8-177, which vas taken from Refence 1. It shows directly hov the bottom pressure amplitudes decrease with increasing size of the water space (single cell) .

8. 5 3. 3. 3 3 Analytical C alculations The conservativeness described in (2) above is also confirmed from results of calculations with the VELPOT and KOVIB1A REV. 1, 3/79 8-93

PROPRIETARY programs. As for the frequency conversion, appropriate multipliers can be determined also for the conversion of the pressure amplitudes from the test stand to the plant. They depend on the influence of the water space on the stationary velocity potential (spatial pressure distribution normalized to unit source strength) and on the hydrodynamic source strength associated with the bubble dynamics. The source strength itself is dependent in turn on the pressure in the bubble,. which is determined by the interplay-of .bubble volume and air supply into the bubble. Since the air supply varies according to the different operating conditions during the blowdown, only a conservative estimate can be given within the framework of the present investigations T'e conversion .from test stand to the plant for one quencher may serve as an example here. Me obtain for the bottom pressure beneath the quencher:

P plant (1 quencher) <0.7 P test as upper value.

8.5.3 3 3.4 Influence of Backgressure on the Pressure Amplitudes As for the bubble oscillation frequency, the question of the effect of backpressure in the suppression pool air space must be investigated.

Figure 8-178 shows the bottom pressure amplitudes measured in the GKN model quencher tests for a suppression pool air space pressures of 1 and 3 bar As can be seen, the pressure amplitudes do not depend on the suppression pool air space pressure.

8.5 3 3.4 Verification of Des~i n ~S ecification In the transposition of the pressure oscillations measured in Karlstein to the SSES, the extremely conservative assumption that the same pressure time histories are acting at all quenchers simultaneously is used. Differences in the pressure time histories originating from the different discharge lines are neglected. Therefore, each measured pressure oscillation in the Karlstein vent clearing tests is a representative containment

,load for all load cases:

symmetrical load case (simultaneous response of all 16 SRV's unsymmetrical load case (response of one or three adjacent SRV's automatic depressurization in loss-of-coolant accident R EV. 1, 3/79 8-9 4

PROPRE ETAR Y A transposition of the measurement results to the plant is per fo rmed for these load cases.

The Karlstein test tank forms a conservative single cell.

Therefore, conservative enveloping pressure amplitudes were measured in that test stand. Mhen transposing the pressure oscillations from the single cell to the plant, there is an increase of the pressure oscillation frequencies as discussed in Section 8.5.3.3.2. As stated previously, the increase of the pressure oscillation frequencies is accompanied by a decrease of the amplitudes. The decrease of the amplitudes is neglected for this evaluation, The amplitudes of the measured pressure oscillations remain constant for all frequencies. That is an additional conservative feature, as already discussed in Section 8 5 3 3 3 8 5.3 3.4.1 ~foe ~nenc Analyses of Selected Tests The pressure time histories for selected Karlstein tests are illustrated in Figures 8-41 to 8-65 The freqeuncy analyses were carried out. with the Fourier Analyzer 5451 made by Hewlett Packard.

The frequency analyses were generated as power spectral densities. The frequencies at which a structure is excited into oscillation can be read off from the power spectral densities.

Freqeuncy analyses were performed for pressure transducers P5.2, P5. 4, P5. 5, and P5. 10 and for the following tests:

4el.l, 4.1 6, 12. 1,

21. 2, 25.

llel, 19 R2 7~ 20 Rl 1, 20e Rle10, 2lel~

R2 Pressure oscillations at both the wall and the bottom are considered in the freqeuncy analyses. Also considered was the frequency analysis for pressure transducer P5.5, which shows the frlocal effect The limitation of the measured f requencies of the pressure oscillations was determinative in selecting the tests to be analyzed. The tests selected were those which exhibited pressure amplitudes >0.3 bar both at low frequency and also at higher frequencies.

The frequency spectra for several Karlstein tests are illustrated in Figures 8-179 to 8-182 for pressure transducers P5.10 and p5.4.

The frequency spectra for two tests with the long discharge line and lowered water level are shown in Figure 8-179. The principal REV. 1, 3g79 8-9 5

PROPRIETARY frequency of the pressure oscillations is at 2-.3 Hz for these tests.

They are the lowest pressure oscillation frequencies that vere measured in the Karlstein tests.

Figure 8-180 shows the difference in the pressure oscillation frequencies from clean-condition tests to real-condition and/or multiple-actuation tests for the long line The pressure oscillations have a principal freqeuncy of 3.5 Hz in test 4.1.1 (clean condition) and 5 Hz in test 4.1.6 {real condition) For the short discharge line, the frequency shifts from clean to real condition are illustrated in Figure 8-181 for tests 21.1 and 21.2. The result for the short line is:

clean condition: pressure oscillation frequency 5 Hz real condition: pressure oscillation frequency 6.5 Hz The following can be said about the measured grin~ci al frequencies for the Karlstein tests:

The lowest pressure oscillation frequency vas measured in the tests with the long line and a discharge line water level lowered to 2. 5 m above the middle of the quencher.

2.0 3 Hz.

It was

2) For the clean-condition tests, pressure oscillation frequencies of 3.5 4 Hz were measured with the long discharge line.

3) For the clean-condition tests, pressure oscillation frequencies of 4.5 5 Hz were measured with the short d isch a r ge 1 in e.

4) The highest frequency for the Karlstein tests vas measured for the real-condition and/or multiple-actuation tests. The measured frequencies vere 6 6.5 Hz.

Figure 8-183 shovs frequency analyses for different pressure transducers for one test.

P 5.2 sits on the bottom beneath the middle of a quencher arm.

P 5 4 is mounted on the concrete wall at the intersection of wall and bottom.

P 5.10 sits on the concrete wall opposite the center point ball of the quencher.

of'he The frequency spectra of the pressure transducers all display a power maximum at the same frequency (3 Hz). Therefore, the R EV 1 3/79 8-96

PROPRI ETA R Y location of the measurement and the structure of the mounting position in the water region of the Karlstein test stand have no influence on the measured frequency of the pressure oscillations.

8 5.3.3 4e2 Shifting of the PSD' in the Tra~ns osition from the Test Stand to SSES The comparison of the pressure time histories measured in the Karlstein quencher tests with the pressure time histories specified in Section 4. 1 3 is accomplished by using the frequency power spectra.

The frequency spectra of the KKB traces forming the basis of the specification in Section 4.1.3 and are illustrated in Figures 4-31 to 4-33 The specified pressure oscillations have their dominant frequency in the range of 6.5 8 Hz.

To cover the pressure oscillation frequencies for SSES, the following rule for treatment of the traces was given:

The three traces should be time-expanded by a factor in the range from 0.9 to 1.8.

The pressure amplitudes should be multiplied by a factor of 1.5.

To be able to make a comparison with the measured pressure oscillations, it is necessary that the frequency spectra of the three traces be shifted in frequency and stretched in amplitude.

In this Section, we illustrate a method by which those operations on the frequency spectra can be performed.

8.5.3.3.4.2.1 Pte5uencf shift The amplitudes are preserved in the frequency shift. To ensure that, the area under the power spectrum must be held constant.

Since the analysis time range for the frequency analysis is it finite, must be made certain that the comparison involves only spectra in which approximately the same number of oscillation periods were analyzed The traces are expanded or compressed by the factor f<, while keeping the zero point fixed Let us designate the expanded or compressed frequency by the original frequency by f.

f'nd A power spectrum can always be subdivided approximately into triangles whose base is the frequency and whose altitude is the power density In the original spectrum, the area beneath a triangle is: f - f 2 1 A ~ h 2

REV li 3/79 8-9 7

PROPRI ETA R Y For the new frequency:

fl = f x fl 2

f x fp Therefore, we have for the new area:

h

Aut since A' A, h = f~ h

~r The power density of the shifted spectrum is inversely proportional to the frequency multiplier.

In this definition, the frequency multipliers are to be taken from Section 4. 1. 3. From the factor 1.8 we get fV = 1/1. 8 and from the factor 0.9 we get fV = 1/0. 9. If the frequency is reduced to half, the power density is doubled.

8 5.3.3 4.2.2 ~Am litude Stretching The following relation prevails between the amplitude of a load-vs.-time function and the power density:

a =k b f' k = correction factor 2

For the stretched amplitude, we have a' f a. The relation between power density and amplitude is preserved by the stretching, so that the same correction factor is also valid after the stretching. Therefore:

k h

b f' and thus:

a h h a h )h'= f 2 .h 8-9 8

PROP RI ET AR Y The power density ratio in the amplitude stretching is proportional to the square of'he amplitude multiplier.

8.5 3.3.4 3 Symmetrical Load Case~Simultaneous Blowdown of all 16 SR V's)

All the Karlstein clean-condition and real-condition tests are used to evaluate this load case. The multiple actuation tests are considered as irrelevant to the plant for this load case.

The one exception is the 10th blovdown test of an entire multiple actuation test with the short discharge line. Those tests are started 10, minutes after completion of the 9th blowdown test.

They are thus subject to the same conditions as the real-condition tests. Accordingly, the 10th blowdovn tests of a multiple actuation test with the short discharge line are treated as real-condition tests.

The test tank in Karlstein represents the smallest single cell with respect to the water space. That means that the maximum possible pressure amplitudes for SSES were measured.

According to Section 8.5.3.2, the measured pressure amplitudes are covered by the specification For this load case, the measured f requencies of the pressure oscillations can also be transposed directly from the Karlstein test stand to SSES (see Section 8.5 3.2) .

Thus, all the pressure time history can be transposed directly from the test stand to SSES. In order to show that the measured time histories are also enveloped by the specification, the frequency spectra of the measured pressure oscillations are compared with the frequency spectra of the specified traces.

Since the measured frequencies differ from .the frequencies of the specified traces, the spectra must be treated by the method illustrated in Section 8.5.3. 3.4 2 and brought into coincidence at the dominant frequency.

The pressure oscillations measured at pressure transducer P 5. 2 are used for this comparison, since, the pressure transducer P5.2 exhibits the highest power spectrum of all the pressure transducers that are useable for the overall loading of the containment (P5.5 is not considered see Section 8.5.3.1).

Pressure transducer P 5.2 is mounted on the bottom of the test tank, directly beneath a quencher arm. That position is also present in SSES. Therefore, this pressure transducer measures pressure oscillations having the greatest relevance to SSES.

Purthermore, the specified traces are also results of a measurement made with a .bottom pressure transducer whose location was similar to that of P5.2.

R EV 1, 3/79 8-9 9

PROPRI ETAR Y The comparison of the frequency power spectra is shown in Figures 8-184 to 8-188 We see that the frequency spectra of the KKB traces, which were frequency-shifted and amplitude-stretched as described in Section 8.5.3.3.4.2 envelop the frequency spectra of the measured pressure oscillations.

Therefore, it can be stated that:

a) the Karlstein measurement results are conservative for the load case of simultaneous clearing of all 16 quenchers

{single-cell effect);

for this load case, the pressure oscillations are enveloped

'I b) by the specification with respect to their amplitude, their frequency power spectra, 'and their spatial distribution.

8.5.3.3.4 4 Unsymmetrical Load Case slowdown Via One SRV)

For this load case, all determinative parameters, except for the -,

water surface area, were simulated in the Karlstein test stand according to their actual values for SSES.

For the load case of vent clearing with one quencher, a larger water surface area is available to the quencher in SSES than in the test in the Karlstein test stand.

Accordingly, the pressure oscillation frequencies are raised and the pressure amplitudes are lowered. In this verification, we conservatively make no allowance for the amplitude decrease with increasing water surface area.

The frequencies calculated according to Section 8.5.3.3.2 for the load case of blowdown via one SRV are compiled in the following table:

Frequency of the pressure oscillations (Hz)

Measured Frequency Plant Specified multiplier frequency.

band CLEAN CONDITION 3.5-4 1.54-1.48 5.4-5.9 REAL CONDITIONS 5 1.42 7.1 3.75-8.9 CLEAN CONDITION 1.42 7.1 4 8 0 REAL CONDITIONS 6.5 1.37 8.9 R W REV. 1, 3/79 8-100

PROPRIETARY The frequencies transposed to the plant are all enveloped by the specified frequency band.

For the load case of vent clearing of one quencher, the multiple actuation tests must also be considered (they were included under "real conditions" in the Table above)

For the load case of simultaneous blowdown of 16 quenchers, was shown that the measured power spectra are enveloped by the it specified power spectra. That statement applies for all frequency ranges. If two power spectra are brought into coincidence at one frequency and if both spectra are subjected to the same frequency shift, then there is no change in the relation of the two spectra to each other.

Therefore, the power spectra of the clean-condition and real-condition tests are also covered by the specification in the load case of vent clearing of one quencher, since, as stated above, the transposed frequencies from the test are all enveloped by the specification frequency range.

For the multiple actuation tests, test 4.1.6 is considered to be enveloping for the long discharge line, since highest pressure amplitudes.

it provided the For the short discharge line, test 20.R1. 10 (which formally can be classified as a multiple actuation test) is considered to be enveloping for the same reason.. Classified as a real-condition test, it was shown in the preceding Section that the specified traces envelop the pressure time histories for that test.

In Figure 8-189 is also it is shown that the power spectrum of test 4.1 enveloped by the specified KKB traces.

6 Even under the very conservative assumption that the pressure amplitudes measured in Karlstein can be transferred without change for the load case of vent, clearing of one quencher, the pressure time histories are enveloped by the specified traces.

8 5.3.3 4.5 Un~smmetrical Load Case slowdown via Three A~d'acent SR V ~ sg This load case is bounded by the load cases of simultaneous vent clearing of 16 quenchers and vent clearing of one quencher.

8.5.3.3.4-6 Automatic DeDzessuzizatio~as stem~A~DS Load Case In this section we discuss the load case that considers the f iring of the six quenchers associated with the ADS under LOCA conditions.

REV li 3 j'79 8-1 01

PROPRIETARY As shown in Figure 8-190, the following conditions prevail in the suppression chamber when the automatic depressurization system is actuated during IBA:

Absolute pressure in the wetwell air space, approximately 2.55 bar Pressure difference .between drywell

'and suppression chamber 0. 42 bar The Karlstein tests with lowered water level in the discharge line are used to verify the ADS case. These tests are used as they correctly simulate the discharge line as it would be with a positive pressure differential of approximately 0.42 bar in the drywell. This positive pressure differential would result in the lowering of the water level in the discharge line to the elevation of the bottom of the downcomers as was simulated for tests 10.3, ll 1, 12.1 and 13.1. Of 'those tests, the test 11.1 (enveloping in amplitude and power density) is used as the basis for the verif ication.

The amplitude-reducing influence of the larger water surface area assigned to the individual quencher in the ADS case is conservatively neglected.

Also, since earlier KMU tests proved that the backpressure in the suppression chamber has no influence on the pressure amplitudes, the measured pressure amplitudes are taken unaltered from the corresponding Karlstein tests, in which the measurements were made at a t mos phe ric pressure.

The predominant frequency in test 11.1 is at 3 Hz. According to Section 8.5.3.3.2, Figures 8-174 and 8-175, the following frequency multipliers are obtained for the ADS case for transposition of the pressure oscillations f rom test 11.1 to the plant:

Influence of the larger water surface area 1 35 Influence of the 2.55 bar backpressure 1 4 Total frequency factor 1 9 Domi.nant frequency 57 Hz Note:

The measured lowest dominant pressure oscillation frequency was measured in tests 12. 1 and 13. 1, which fall into the same category as test ll 1. Mith the total multiplier 1.9, the frequencies are raised to 3.8 Hz and thus lie within the specif ied frequency band (see Section 8. 5.3.3.5) .

The dominant frequency is within the specified frequency band REV lg 3/79 8-1 02

PROPRIETARY The comparison between the prepared trace from pressure transducer P5.2 for test 11.1 and the specif ication is shown in Figure 8-191. As for the other load cases, the comparison is made in the power spectra 'of the pressure time histories. The spectrum of test 11.1 was shifted from the dominant .frequency of 3 Hz to the dominant f reqeuncy of 5. 7 Hz while preserving the area (amplitude) .

The KKB trace of test 76 was shifted from 8 Hz to 5.7 Hz while preserving the area, and then stretched by a factor of l. 5 in amplitude. Figure 8-191 shows that the trace from the specification, treated in this manner, envelops the trace of Karlstein test 11.1 transformerd to the ADS case since the total energy represented by the area under the power spectrum curve from the specification is greater than that from the Karlstein test ll.1 8 5.3 3.4 7 Summary It has been demonstrated pressure oscillations in that the, frequency power spectrum of the the suppression chamber are enveloped by the frequency power spectrum specified in Section 4.1.3 for all load cases. Thus, the design specification provides enveloping loads also for the dynamic excitation of the SSFS containment by vent clearing of the relief system with the quencher.

8.5.3 3.5 Evaluation of the Neasured Pressure Oscillations During Condensation As discussed in Section 8.4.2, three regimes can be distinguised in the condensation process:

a) The quencher is cleared continually.

b) The quencher is not cleared continually.

c) Only the sliding joint is cleared, and the steam condenses in the discharge line.

8.5 3 3.5 1 The guencher is Cleared Continually The steam is condensed continually in the water pool outside the quencher. Calm condensation prevails for cold water and also for hot water in the blowdown tank (see Figures 8-78 and 8-79)

The measured .maximum pressure amplitude is s0.13 bar. This condensation phase was measured for reactor pressures up to about 4 bar. f The requencies of the pressure oscillations are 70-120 Hz for a cold pool and 20-45 Hz for a hot pool.

REV 1, 3/79 8-1 03

PROPRIETARY 8.5.3.3.5 2 The Quencher is not Cleared Continu~all This condensaton phase begins vhen the condensation rate outside the quencher is. greater than the steam mass flov through the line. The pressure in the quencher drops belov the hydrostatic pressure of the surrounding vater. The water penetrates into the quencher. The condensation surf ace area is thereby decreased and so is the condensation rate. The result is a pressure rise in the discharge line, so that the water that has floved in is expelled again.

The inflow of vater from the suppression chamber into the quencher and the subsequent braking and re-expulsion of the water is a nonstationary process vhich occurs periodically.

For that reason, this condensation phase is also called intermittent condensation.

The phenomenon of intermittent condensaton is dependent on the water temperature. For cold vater there is a higher rate of condensation outside the quencher, resulting in a larger generation of negative pressure inside the quencher and therefore a more vigorous flow of water into the quencher.

For a cold water pool, the profile of the dynamic pressures is similar to the profile vhich is familiar from the chugging phase of the condensation at the vent pipes; see Figure 8-76.

For heated vater in the suppression chamber, the condensation rate outside the quencher is smaller, so that the entire process takes on the form of a low-frequency pressure oscillation (See Figure 8-80)

The tests in" Karlstein yielded as maximum measurement result for the dynamic pressure: + 0.28, 0.18 bar, for a cold pool. The time between two events is about 1.0 second. For a heated pool, the measured maximum amplitude is +0. 12, 0.07, bar.

8.5 3 3.5.3 Condensation in the Discharge Line and Thru the Slidi~n Joint If the steam flov decreases further, reached in which the quencher is no a condition is finally longer cleared, but rather remains continually filled with vater. Then there is steady-state condensation of steam inside the discharge line This condensation phase proceeds very calmly and begins at reactor pressures below 2 bar.

In this condensation phase, maximum dynamic pressures of +0. 08, 0.04 bar were measured in the water pool during the Karlstein tests.

R EV l, 3/79 8- t 04

PROPRIETARY 8 5 3.3.5 4 Tr~ans osition of the Measurement Results to SSES In regard to steam condensation, the conditions of the Karlstein test stand are direclty transposable to the conditions of SSES.

On the whole, the pressure amplitudes during condensation are small compared to t'hose during vent clearing and therefore are covered by the latter.

8 5.4 Pool Mixinq Durincn SRV Actuation and Thermal Performance of the~uencher 8 5.4 1 Introduction When an SRV responds, steam is condensed in the water of the suppression pool via a quencher As this happens, the water must absorb the heat of vaporization of the steam, and so heated. When there is a long-lasting discharge of steam via a it is quencher, all the water in the suppression chamber should particpate in the heating, so as to limit the local heating in the vicinity of the discharging quencher In order to obtain good mixing of the hotter and colder water in the pool, all quenchers are positioned at a small distance from the bottom (3 6" = 1.07 m) (see Figure 8-192) ) . The water heated

~

near a quencher is specifically lighter than the colder water lying above it. Therefore, the warmer water will rise and mix with the colder water.

To obtain an additional mixing effect, the hole occupancy of the quenchers were made slightly unsymmetrical (approximately 8%) .

Mhereas the quencher arms have the same hole occupancies on the sides, only one arm of each quencher has holes on the end cap.

In that way, a unilateral thrust can be exerted on the water in the suppression pool.

In the top view of the quencher arrangement (Figure 8-193), we see that the quenchers are arranged in two graduated circles.

Along the inner graduated circle, the quencher arms all point in the circumferential direction, and the end cap with holes all point in the same circumferential direction. On the outer graduated circle, the columns would practically prevent a thrust effect if the quenchers were arranged in the same manner.

Therefore, the quenchers were directed more radially, but turned by an angle of gf = 300 in the circumferential direction from the radii. In this way, 50% of the thrust circumferential direction till (equidirectionally acts in the with the thrust of the quenchers on the inner graduated circle) . It should be noted that this new arrangement supersedes the original arrangement shown in Figure 1-4.

In the following, we shall estimate the acceleration of the water pool for the case in which one quencher on the outer graduated R EV li 3/79 8-105

PROPRIETARY circle is operated for a long period of time at a reactor pressure of 70 bar {valve failure in open position) . Then we shall present some measurement results from a test with a 4-arm quencher in the Brunsbuttel nuclear power plant and some information from the GEM Nodel quencher tests related to steam condensation with a quencher.

8 5.4 2 Equation of Notion of the Rotation Pool Zt is assumed that the water flow in the rotating pool can be considered as a straight-line channel flow due to the small curvature of the graduated circle and the low circumferential velocity.

If we place the the discharging origin of the coordinate quencher, system at the center of then the equation of motion of the rotating pool reads:

m x+c 2x'.2 Fff 5

m W

mass of water to be accelerated in the suppression chamber c sum of all flow resistances W

eff effective driving force This differential equation has the general form:

x+ax = b Substituting x = u, the differential equation takes the form:

u + au~ = b This differential equation is a special form of the Riccati d iff e re nti al equation The general solution of the differential equations reads R'ef. 53:

u(t ( n) =-.n a. b +bTanh a. b (t-K) ga. b + a. q'Tanhga'b (t-c)

The initial condition f or t = 0 reads:

0 n /a b+ b Tanh/a b -g)

/a b = a n .Tanh/a b (-g)

REV l, 3/79 8-106

PROPRIETARY This conditional equation is satisfied only if 6 and n = 0.

The initial condition then leads to the solution:

a . b u (t) = b . Tanh ga. b Since u(t) = X(t), the equation for the velocity of the rotating pool reads:

b x(t) = Taab)a. b t

)a. b For the distance covered, we have:

x (t) = pJ' (v) d v The solution reads:

X (t) = ln cosh ia. b . t 0 [ (

8 5.0 3 'etermination of the Plow Resistances The following resistances are considered:

a) Wall resistance of the channel b) Resistance for flow around the discharge lines with quenchers and bottom support c) Resistance for flow around the vent pipes d) Resistance for flow around the columns The channel has the following dimensions:

REV. li 3/79 8-107

PHOPRI ETAR Y The hydraulic diameter of the channel is:

26 ~ 822 - 8 ~ 84 Re ct 7 3 ( 2 2.8 cn 26.822 8.84 (2 7.3) +, 2 For the Reynolds number, we have:

Re  ! W. RR According to Reference 36, the kinematic viscosity for water at 40oC is v = 0.65l x 10-~ m~/s.

Xf we assume a velocity of 10-2 m/s so as 'to cover the start-up phase also, we get:

2 R + 10 x2.8 . 43x10 4

.651 x 10 The SSES suppession pool is lined with a steel liner which cannot be considered hydraulically smooth. For such large steel structures it must be assumed that the individual plates are not joined together with their edges parallel, so that the flow resistance is increased by projecting edges. We therefore conservatively assume an absolute roughness of k = 2 mm. Then we have:

K -4 7.1 x 10 dh 2.8 x 10 This corresponds to a friction coefficient of > = 0.022.

The resistance coefficient is then:

~.1m W

h 26.844 + 8.84 m' 2

~ 7f RR 56 m rw - '28)

.022 Cylindrical bodies are immersed into the water of the suppression chamber. They are the discharge. lines with guenchers, the vent pipes, and the steel columns.

REV 1, 3/79 8-108

PROPRIETARY Outside diameter Submergence Quantity m m Discharge lines 0 324 7 3 16 Vent pipes 0.61 3 35 87 Steel columns 1 06 7-3 12 For the individual structural components, we then have the following Reynolds number:

v = 0.01 m/s (see above) Re = W dm For the roughness, we assume k = 0.2 mm. Then, according to Reference 39:

Reynolds Submergence number d Discharge line with 5 x 10~ 6.17 x 10-i 22. 5 0 73 q uenc her v i.th bottom support Vent pipe 9.4 x 10~ 6. 28 x 10 5 5 0 73 Column l 63 x 10' 9x 10-+ 6.9 0 73 The resistance force is t,hen:

p 9 2 The surface area on which the wall resistance acts is:

2 4

Furthermore:

c W

6. 16 x .44 + .73 x 50 + .73 x 177.8 + .73 x 93

.2 AA 16 x 0.324 x 9.6 50m c = 238m 2

A 87 x .61 x 3.35 177.8. m AS -

12 xl.06 x 7.3 ~ 93m 2

Since the water region of the suppression chamber also contains a few structural components which vere not considered here, an additional allowance shall be made. Me choose: 2 c 300m W

R EV. 1 3/79 8-1 09

PROPRIETARY 8.5.4.4 Determination of the Force Novi~nthe Pool Forces on the water mass in the suppression pool are produced by thrust from the boreholes on one of the end caps which are present on each of the quenchers. The smallest thrust force is produced by the quenchers along the outer graduated circle, since they do not have their thrust boreholes arranged in the circumferential direction.

The quenchers along the outer graduated circle are turned by an angle 4 = 30o relative to the radial direction.

g=

50'DV The thrust force results from the impulse of the outflowing steam.

F ~ AP x A - + P x MD x A~U

= effective outlet area of quencher difference between pressure in the quencher and ambient pressure PD

= density of the outflowing steam W

D

= velocity of the outflowing steam As an effective outlet area of a quencher end cap, there is available:

A~g a< x ADU geom e

C D 0.8 (Section 8.5.2.3)

+0 geom

(~l

(

2 4 )

6.9 x 10

-3 m

2 A

DO geom 5'~2 x 10 -3 m 2

A constant reactor pressure of 70 bar is chosen for the estimate of .the effectiveness of the rotating pool.

According to Reference 37, the mass flow through the relief valve at a reactor pressure of 70 bar is:

REV lr 3/79 8-110

PROPRIETARY m = ill kg/s The resulting stagnation pressure in the quencher is:

p = ll bar and the steam pressure in the quencher holes is:

pD

= 64 bar Therefore, PD" = 3 4 kg/m> WD 462 m/s The force acting in the circumferential direction is then:

Feff F sin Feff (AP + P + Wj) ) A DU " x sin Q with Q

= 30' Therefore:

Feff 2~ O'N + lo5 KN 3 ~ 5 KN 8 5.4 5 Working Equations for the Rotation'Pool of SSES The equation of motion for the rotating pool reads:

5 '.2 m

w x + cw 2 eff This differential equation was solved in general form in Section 8 5.4-2.

To determine the mass of water which is to be moved, we must consider the internal structures which reduce the water mass. Me have:

I,4 (26.822) - (8.84) ) x .73 - x (.324) x 7.3 x 16

- x ( 61) x 3.35 x 87 - x (1.06) x 12 x 7.3]

4 4 3.5 x 10 Kg For the total resistance coefficient we have according to Section 8 5 4 3:

= 300m C<

and for the effectively acting force we have according to Section 8 5 4 4:

F = 35KN eff REV l, 3/79 8-1 11

PROPRIETARY Therefore, the eguation of motion - reads:

6 3.5 x 10 x X + 1.5 x 10 5 x X 2

3.5 x 10 or There fore, for: X+aX2 =b a = 4.3 x 10 4 Va b 6.55 x 10 t b = 9.9 x 10 The equation for the velocity of the rotating pool reads:

-1 -3 X (t) = ,1.52 x 10 Yanh 6.55 x 10 t The equation for the displacement reads:

X(t) = 23.2 1n i cosh 6.55 x 10 ti The results are illustrated in Figures 8-194 and 8-195.

8 5 4.6 Estimate of the Heati~n of the Suppression Chamber Mater The local heating of the suppression chamber vater results from the balance of the heat brought in by the condensing steam and the heat dissipated by the flowing water.

As time passes, hovever, the pool is set into motion by the impulse of .the inflowing steam and reaches a velocity such that most of the heat brought in is distributed over a larger volume of water than the assumed local volume., The difference between the local and mean water temperature decreases.

8 5.4 7 Experimental Proofs 8 5 4.7 1 Nodel Tank Tests Thrust measurements on a steam jet vere made in the Karlstein model tank in the Spring of 1973 (Ref 40).

The test set-up is illustrated in Figure 8-196 The steam pipe is connected by. a spring to the side wall of the model tank. The excursion of the spring with the steam pipe is measured by a displacement transducer.

The measurement system vas calibrated by determining the excursion of the steam pipe for a defined force.

The steam outlet opening had a diameter of 10 mm.

The mass flow density vas 600 to 630 kg/m<s.

The measured reaction forces were 20 28 N.

REV 1, 3/79 8-112

PROPRIETARY A short calculation yields:

Outlet area = 7. 8 54 x 10- ~ m~

Rest pressure before the outlet opening 4.5 bar Pressure after the outlet opening 2.6 bar Steam density (at 2.6 bar) .= 1.44 kg/m~

The resulting outlet velocity is:

W = K 1. 135 2.6 x 10

~

g ~

W 452. 7 m/s and the thrust force is:

F = (PW + hP) A ff A

e ff = 08xAgeom F ~ (1.44 x (452.72) + 1.6 x 10 ) x 0.8 x 7.854 x 10 F ~ 284N The measured values are lover than the calculated values.

The measurements have proved clearly that the impulse of the emerging steam jet becomes active as a thrust and that, vith respect to the velocity buildup of the rotating pool (and thus for the maximum local heating),

the calculated values.

it is conservatively bounded by 8 5.4 7.2 KKB Test During the Nuclear Commissioning The pressure relief system was tested during the commissioning phase of the Brunsbuttel nuclear pover plant. In one such test, a relief valve was held open for a time of about 270 seconds.

The suppression chamber cooling system vas switched on during the test. Water was drawn off in the lower part of the pool, cooled, and sprayed from pipes provided with holes and located under the top of the suppression chamber.

12 measuring points are mounted in the water region of the suppression chamber. They are arranged at three different elevations (14 m, 16.5 m, 18. 2 m) and at four different circumferential positions (5o, 75o, 195o, 245o). The water level is at a height of 18.89 m.

Figure 8-197 shows a three dimensional spatial representation of the measured temperature field in the vater just before test start (curve 1) and at 228 seconds after test start (curve 2) .

In Figure 8-197, the vertical position of the transducer is represented on the ordinate and the circumferential position on the abscissa The temperature axis points to the rear. The heating of the pool is indicated as the difference of curves 2 and 1 at three elevation positions. The mean water temperature was approximately 32. 3oC before the test and approximately 42.8oC REV. 1, 3/79 8-1 13

PROPRIETARY at 228 s later.. The maximum measured temperature was 500C, so that the maximum deviation from the mean was 7.20C.

The discharging quencher was located at 285't an elevation of 14 915 m and accelerated the water toward the left in the Figure.

Correspondingly, the.water temperature is higher above and to the left of the quencher. From that we can see the effectiveness of the quencher's arrangement near the bottom and of the unsymmetrical hole arrangement with re.,pect to uniform utilization of the heat sink of the water pool.

8 5 4 7 3 GKN Half Scale Quencher Condensation Test A series of intermediate scale (1:2) condensation tests were performed in the GKM test stand to demonstrate the high temperature performance of the guenchers(Ref. 27). Condensation tests were run on seven different versions of the quencher device. The last three versions had 10-mm diameter holes on'he quencher arms The spacing of the hole centerlines was 1.5 diameters circumferentially and 5.0 diameters axially. This hole pattern is also adopted in the actual SSES quencher design.

These tests were run at a water temperature ranging f rom 13oC to 100oC (56oF-2120F) and a steam mass flux (with respect to the hole area) range of 8 to 495 kg/m~ (1.6 to 101 ibm/ft~s). Mater temperatures as high as 1070C(225~F) were measured at certain locations in these tests.

8 5.4 8 Summary The Karlstein'uencher tests and previous GKM half scale quencher tests show clearly that smooth steam condensation can be achieved at elevated temperatures which approach the .local saturation limit.

Xn addition the calculations and KKB, in plant tests provide information which suggest that pool mixing is enhanced by steam discharge through the holes in the end caps of the quencher."

8 5.5 Verification of Submerged Structures Load ~Secification Due To SQV Actuation Section 4.1.3.7 gives the design specification for the loads on submerged structures due to SRV actuation. The basis for the specification is the three pressure time histories used for the containment analysis but instead of a constant amplitude multiplier of 1.5 various multipliers, related to the crossectional area of the object, are used. (see Table 4-15) .

The loading on the columns including the localized ef ect at P5.5 has been discussed in Section 8. 5.3.2.1.2 RE V 1, 3/79 8-114

PROPRE ETAR Y In addition the effects of air bubble oscillation loads on the quenchers have been discussed in Section 8.5.2 3.6.

The following section will discuss the loadings on the vent pipes as measured in the Karlstein test tank and provide a description of the influence for the expelled water duri'ng vent clearing.

8.5.5 1 Loads on the Vent P~ie 8 5.5 1 1 Measurement of the Loads Xn order to determine the loading of the vent pipe near a quencher, a vent pipe having the same outside diameter and wall thickness as that in SSES was installed in the Karlstein test stand and supported by typical bracing. (see Figure 8-10) .

Underneath the bracing, bending strains were measured in two mutually perpendicular planes by means of strain gauges (SG S 1 and SG S 2) (see Figures 8-11 and 8-12). The strain gauges were mounted about 100 mm below the bracing.

The outside diameter of the vent pipe is:

D =0609m and the inside diameter is:

D;=0589m Thus, the cross-sectional area is:

2 A ~ 0.0188 m and the moment of resistance is:

4 Dl -3 3 2.77 x 10 m 32 0

We have:

x W~'M ~ Gx E x W Therefore:

M ~ 2.77 x 10 -3, '0 11 '

'nd hence; M ~ 0 57c R EV. 1, 3/79 8-115

PROPRIETARY If we insert c in micrometers per obtain the bending moment in kN-m meter into this equation, we The bending moments calculated in this manner are static t

equ iv al en loads.

5-5-5.1.2 Measnned Bendi~nsaeents Figures 8-198 to 8-200 show the dependence of the measured resultant bending moments on the reactor pressure, vent clearing pressure, and. pressure oscillation amplitude that were measured near the vent pipe on the concrete wall.

Only the tests with clean conditions were used for the plot of the measured bending moments versus reactor pressure, whereas all tests in the reactor pressure range of 60-81 bar were used for the plots of the bending moment versus vent clearing pressure and pressure oscillation amplitude.

The measurements of the bending strains at the vent pipe were performed only for the tests with the long discharge line.

The measured maximum bending moment was 14.6 kN-m at a 74 bar reactor pressure and a 13.8 bar vent clearing pressure.

8 5 5 1 3 Extr~a olation of the Measurement Results and

~Com arison with the Specified Value If the measurement values are extrapolated to the extreme conditions in the plant on the basis of Figures 8-198 and 8-199, we get the following extrapolated maximum values:

16.5 kN-m with respect to an 88 bar reactor pressure, 19.0 kN-m with respect to the vent clearing pressure of 16.5 bar for the long discharge pipe, as extrapolated in Section 8. 4 for the extreme boundary conditions in the plant.

In the specification, a maximum pressure difference of 0.75 x 0.8

= 0.6 bar was specified for the vent pipe with the distribution illustrated in Figure 4-24. The pressure distribution for the vent pipe installed in the Karlstein test stand is shown in Figure 8-201 The following relation applies for the pressure at the end of a vent pipe:

~dP dP hP ss Oe4 bar

7. 3-1. 83 7. 3-3. 65 REV li 3/79 8-116

PROPRI ETAR Y At the clamping point of the vent strut, we have:

AP0 AP = 0 1 bar 7.3-1.83 7.3-6.3 The pressure distribution from the end of the vent pipe to the clamping point of the vent-pipe strut is trapezoidal.

The lever point is:

arm

0. 1 +. 0.4 2

L S

= 0.1 S

' x 2.' 65 1'59

(

of the acting force with respect to the clamping (0. 4-0.' 1) 2 )

2 3

x 2.65 For the bending moment at the clamping point we get:

M5

= ( ~2~ 2.65 x 0.6 x 1.59 ) 10 SP 63 kNm

~SP Relative to the strain gauges, we have:

MB 57 kNm SP The extrapolated maximum moment was 19 kN-m.

It is thus measurement demonstrated values and that the specification envelops the their extrapolation.

The proof that the specification envelops the measurement values and their extrapolation is based on a purely static analysis.

Such an analysis is permissible because the exciting pressure oscillations have a frequency of 4-6 Hz. However, the strain gauges indicate a natural oscillation frequency of 17-20 Hz for the vent pipe which is very close to the natural frequency of the vent assumed in SSES that

(

the 19 Hz) (see Figure 8-202).

dynamic load factor is close Hence, to it one.

can be 8 5 5 2 Influence of Expelled Water During Vent Clearinc[

A review of the high speed films and pressure traces at P5.5 from the Karlstein tests shows negligable influence of the expelled water at this gage. In addition the total penetration of the expelled vater appears to be approximately 3 feet for a 70 bar initial system pressure. Therefore, no additional loading, other than that already included in the pressure traces vill he considered.

R EV 1, 3/79 8-117

PROPRIETARY (A time correlation of a high speed film to pressure trace at P5.5 will be supplied later.}

8 5.5 3 Summary The loads measured on the dummy vent pipe are static equivalent loads, but loads which are a sum of individual components. In the specification, the transverse loads on internal structures originating from the blowdown of the relief system are

.formulated as differential pressures across the internal structures. The differential pressures have the same pressure time history as the dynamic pressures in the water region of the suppression chamber This formulation of the transverse loads on the vent pipe (more generally on the internal structures in the water region of the suppression pool) yields the enveloping static eg'uivalent load.

This was also verified by the KKB tests with the actual relief system (Ref. 38) . The maximum differential pressures calculated from the measurement results are p = 0.16 bar at the quencher arm, and p=0.11 bar at the protective pipe on the discharge line.

They are both conservatively bounded by the KKB specified value of p=0.2 bar. The KKB test results shows that there is a clear separation between the specified loads and the maximum measured loads for both the lateral and vertical loads on internals in the pool of the suppression pool.

Based on the verification of the transverse loads by the KKB tests and based on the comparison between specification and measurement for the Karlstein tests (see Section 8.5. 5.1),

be stated that the values formulated in the specification for the it can transverse loads on internal structures in the water region yield enveloping static equivalent loads.

REV 1 g 3/79 8-118

KEY: 1. Relief valve

2. Compressed air

3. HP steam line

4. Heating

5. SRV

6. Condensate e

P

,$ $ ... A,

'C I

.d' 1$ $

'4 (o!>

(C$ !P A

C 0 m C U m m

Ch Z

0 Z z z gzo$

mL m ZZr

~Dm m

n fh li 0

a! 0 Z

KEY) l. Relief valve

.))),O 2. Compressed air 17 3 ~ HP steam line

4. Heating 16

5. SRV 15 OO $ $ 6. Condensate X(4 DO ~ $

1$ 3t

)3

'4P<Pg~

$ 4.$

I 'gA 9 P

~ ~ ggl ~

I 8 ~

i .I s po

%1,,

)7 pE$

PE$

'4 )

I $$

$ e c.

Q3gy$ ~

pill

)yO C

C 0

p xm mg m

)N y g r III

~'N g o

H

e bi <<'s eyes ron Jle'e/r s+cecr co~ fttpr n Concrete sthtcttrral elements pa o a + e rn O Cl 4

,/

A n ch 0

~

4 I

~

4 0

0 O

O 0

i P5.3 0

P Owsley Ii 0 wnt ppe o <

~O P59' 0

o

~ ~

/ Broore ci D' P5.4

+Y n p55 p4 P5.6 0

nX gP P5.7 O 0 on PS.

I 0

fR '~ O r

vu RT5.3 s.c Xi on/

Ct o p 0

p 0 n ~

0 o V

B Discharge line for ttWU SRv tnnl usted for Ih+

tests in qwstionl REV 1 3/79 SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT KARLSTEIN TEST TANK PLAN VIER TYPICAL VENT CLZARIl6 INSTRU-MENTATION FIGURE '.7

LP c.5 LPr 3 I

StJ 5.(

I I

I c>G5.2 "0 't' Bracing

'0 o

0

~~ I I. 0, ~ a7 ~

0 0

0 "~a ~

I

.0 ct ~T5.c o

...- . P5.7 e 0 0

~

0 I ~ a LP42-g 0 ~ t o ~ a> ~ ~ ~

I.

0

'0 I o 0 Oummy 0 ment pt pe ct' o I ~ 0 0 J I I

c> o a Cl t>>at >'s e>es 7

0 ~

~ ~

0 pa9 o

o 0 0iJ i W J ~

Cis cctce (hat leele

!cr

(("IU 5((ar a>seat la> Ihe ah r i Lt.;,*:;

Ph: i I

>t 0

r J ~

Ii aa>estaahl h e s J

a> ~

pal i

5.5 K.S. prototype quencher P54 Pc2 P,S.3 REV I SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT KARLSTEIH TEST TAHK C-D VIEN TYHCAL VEHT CLEARING IsSSTRU-MEHTATIOH FIGURE 8.8

LPC>

LP4,4 d,d

~ 7,2m

~

6,2em SG52 o

I

~

~

SG 5.t 0

~ ~ 0

~

0 0 ~ 0. 0

~

T56 0 J PST g 'h2 LP4.2

~ ~ r 0 0

0 0 Cl 0

2.65m ~

ao

~

o 0

~ g 4 0 ~

~ p o r

T53n 0 0 0

0 P56 0 ppe Od P 5.9 0 0 Lpr I dd

~d

~ ~

o 4 J Cd

~ ~ J d

T5I I T52 PSo I P5!0 J

o 2

IO 0 J ddC rooorPJ P6i SS.ES prototype quencher REV 1 3/79 SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT KhRLSTEIN TE8T ThEK h-B VISf vrncar. err chemi; aeraU-MEm,'hTIOS F IGURE 8. 9

100 Oc 93 C= 200 F 0 Two Actuations (first clean, second real )

90 I

a

~ Mul tiple Actuations

( first clean, subsequent real )

80 O

O 70 CD CD 60 55 C =130 F 6 a 50

~ cn e <0 Operation Field a Vl ID.

<(I a

CD.

CD CD CD CD 30 CD 24 "C = 75 F 3 4 20-I I I I I 10 20 30 40 50 60 70,80 bar 90 Reactor Pressure REV 1 SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT LOCATION OF TEST CROUP N). 1 IN THE OPERATION FIELO FIGURE e y6

100 93 C= 200 F 0 Two Actuations 90 ( first clean, second real )

I a

80 CO 13 O

O 70 CD CD 60 55 C =130 F SO U 11 JD O)

CO 0

40 CII Cll Operation Field Cl CL CD CD 30 CD 24 C=7S F 10 I I I

":0 20 30 40 50 60 70 80 bar 90 Recctor Pre sure 3/79 SUSQUEHANNA STEAM ELECTRIC STATION UNITS 'I AND 2 DESIGN ASSESSMENT REPORT LOCATION OF TEST GROUP NO. 2 IN THE OPERATION FIELD FIGURE 8 17

100

'C 93 C= 200 F

~ Multiple Actuations U 90 ( first c lean, subsequent real )

I a

Ch 80 CD O a O Vl CL 70 C)

C) 60 55 C =130 F 50 L a 15 JD CA CD

~ cr'(I 40 a Operation Field a Vl CL tD C) C) 30 24 C =75 F I

g0 20 30 40 50 60 70 80 bar 90 Reacior PressUre REF 1 SUSQUEHANNA STEAM ELECTRIC STATION UNITS1 AND2 DESIGN ASSESSMENT REPORT LOCATION OF TEST GROUP IOo 3 IN THE OPERATION FIELD FIGURE

100 93C=200 F 0 Two Actvations

( first clean, second real 1 90 ~ Multiple Actvations

( first clean, subsequent real )

80 24 O a O Ih CL 70 (D 60 55 C =130 F 22 21 50 L U

Ch lD

~ oi 40 CII Operation Field t5 0 Ul GL CL lD C)

C3 ED Pl 30 CD 24 DC

~ 750F 16 17 18 19 20 I I I I I I I 10 20 30 40 50 60 70 80 bar 90 Reactor Pressure 3/V9 SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT LOCATION OF TEST GROUP NO. 4 IN THE OPKRhTION FIELD FIGURE 8. 19

100

'C 93 C= 200 F 0 Two Actuations (first clean I second real )

90 l

a 80 O a O lh CL.

70 lD C) 60 55 C =130 F L

50 L 25 a

JD Ch CD 40 CII a

Operation Field a lA CL CL C) CI C) C) 30 C) 24 C =75 F 20 10 20 30 40 50 60 70 80 bcr 90 RBQclor Pressure REV 1 SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT LOCATION OF TEST GROUP NO ~ 5 IN THE OPERATION FIELD FIGURE 8.20

100

'C 93 C= 200 F 0 One Act uatinn under clean condition 90 P

l a

6 Two Actuations (first clean, second real) 80 Cl O

O CIl CL 70 CD CD 60 55 C =130 F l

a 50 I A3 a CD CD JD CD CD

<II 40 CII a

Operation Field a Ul Vl O. ID.

C) CD 30 CD 26 /32 24 C=75 F 20-I 10 20 30 40 50 60 70 80 bar 90 Reactor Pressure aZV 1 3/79 SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT LOChTION OF TEST GROUP IO. 6 Ilf THE OPERATION FIELD FIGURE 8.21

bor 12 9

Q.

8 cn 6 a

cn 5 o 4 CD 3

2 10 20 30 40 50 60 ?0 Abs. System Pressure P 2.6 ~

80 90 bQr REV 1 3/V9 SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT VEST CLEARIRr HIESSURE VERSUS SYSTEM PRESSURE M?F LIlK V%IT CLEARING TESTS FIGURE 8.25

0 bar 14 13 12 10 9

8 Ul Vl Ol 0

cA O

6 cn 5 C

o 4 3

2 10 20 30 40 50 60 70 Abs. System Pressure P 2.6 5>>

80 90 bar 3/79 SUSQUEHANNA STEAM ELECTRIC STATION UNITS1 AND2 DESIGN ASSESSMENT REPORT VENT CLEARING PRESSURE VERSUS SYSTEM PRESSURE SHORT LINE VENT CLFAHIlS TESTS FIGURE B.p6

50 Height 0 Temperature Distribution r

1 20 before Test-Beginnig CO r

2 Temperature Distribution 228 Seconds 50 after Test-Beginnig 19 CJ e 18,2 co 0'~

18 CO ~O) cr oo 0 ga

@0 gC oc oo 17 O4C

.,o~

g 16,5 C

o c (C 16 I

)0o cl CO C i N)I C7 m

CO 0

C m

X 15 3(fC 14 ~

80'ocation 195 Quencher of the 75 0 40 120 160 200 240 280 320 360 40 80 9

p 2 yC+

2 Circumferential Location of the Condensation Chamber m3'm 22r

~Um

~uO M O MM O I I 0

2

QR7 j+~ l~+ ~q ~ ~ ge P REFACF.

his Report contains data, descriptions and anaylsis re1ative to e adequacy of the Susquehanna Steam Electric Stations design to ac mmodate loads resultinq from a safety relief, valve (SRV) disc rge and/or a loss-of-coolant accident (LOCA)./

h

I C=

~DlI T88LJ OF CONTENTS Chapter l GENERAL INFORMATION 1.1 Purpose of Report 1.2 History of Problem 1.3 Quencher Discharge Device 1 4 tlK II Supportinq 1.5 Plant Description Program 1.6 Fiqures 1.7 Tables Chapter 2 SUN'RY

2. 1 Load D ef in i t io n S umma ry 2.2 Desiqn Assessment Summary Chapter 3 SRV DISCHARGE AND LOCA TRANSIENT DESCRIPTION

3. 1 Description of Safety Relief Valve {SRV) Discharge 3.2 Description of Loss-of-Coolant Accident (LOCA)

Chapter 4 LOAD DEFINITION

4. 1 Loads from Safety Relief Valve Discharge 4.2 Loads from Loss-of-Coolant Accident 4.3 Annulus Pressurization 4.4 Fiqures 4.5 Tables Chapter 5 LOAD CONBINATIONS FOR STRUCTUR~ES PIPING~ AND EOUIPi'IENT 5.1 Concrete Containment and Reactor Building Load Combinations 5.2 STructural Steel Load Combinations

5. 3 Liner Plate Load Combinations 5 4 Dovncomer Load Combinations 5.5 Piping, Quencher, and Quencher Support Load Combinations 5.6 NSSS Load Combinations 5.7 Equipment Load Combinations 5.8 Figures 5.9 Tables Chapter 6 DESIGN CAPABILITY ASSESSMENT

6. 1 Concrete Containment and Reactor Building Capability Assessment Criteria 6.2 Structural Steel Capability Assessment Criteria 6.3 Liner Plate Capability Assessment Criteria

TABLE OF CONTENTS ~Continue~d 6.4 Downcomer Capability Assessment Criteria 6.5 Pipinq, Quencher, and Quencher Support Capability Assessment Criteria 6.6 NSSS Capability Assessment Criteria 6.7 Equipment Capability Assessment Criteria Chapter 7 DESIGN ASSESSMENT

7. 1 Assessment Methodology 7.2 Design Capability Margins 7.3 Figures Chapter 8 8~SEE OENCHER VERIVICETION TEST

8. 1 Unit Cell Approach 8.2 Simulation of SSES Parameters 8.3 Instrumentation Arrangement 8.4 Test Matrix 8.5 Analysis of Data 8.6 Figures C hapter 9 RES PONS ES TO NRC~U EST IONS 9.1 Identification of Questions Unique to SSES 9.2 Questions Unique to SSES and Responses Thereto 9.3 Figures Chapter 10 REFERENCES Appendix A CONTAINMENT DESIGN ASSESSMENT A.I Containment Structural Design Assessment A.2 Submerged Structures Desiqn Assessment Appendix B CONTAINMENT RESPONSE SPECTRA DUE TO SRV AND LOCA) LOADS Appendix C REACTOR BUILDING RESPONSE SPECTRA DUE TO SRV AND LOCA LOA DS

It

.'I

,it ) + lr I

I ii' tt'

TABLE OP CONTENTS ~Continue~d Appendix D PROGRAM VERIFICATION D.1 Poolswell Model Verification D.2 Fiqures D.3 Tables Appendix E REACTOR BUILDING STRUCTURAL DESIGN ASSESSMENT Appendix F PIPING DESIGN ASSESSNENT Appendix G NSSS DESIGN ASSESSNENT Appendix H EQUIPMENT DESIGN ASSESSHENT

CHAPTER 1 GENERAL INFORMATION TABLE OF CONTENTS

1. 1 PURPOSE 'OF REPORT 1 2 HISTORY OF PROBLEM 1 3 QUENCHER DISCHARGE DEVICE 1.4 MARK IX,SUPPORTING PROGRAM 1 5 PLANT DESCRIPTION

1. 5. 1 Primary Containment

1. 5. 1. 1 Penetrations 1.5. 1. 2 Internal Structures 1 6 FIGURES 1.7 TABLES 1-1

Number Title 1-1 Cross Section of Containment 1-2 Suppression Chamber, Par tial Plan Suppression Chamber, Section View 1-0 Quencher Distribution

CHAPTER 1 TABLES Number Title SSES X,icensing Basis "

1-2 SSES Containment'Dimensions 1-3 SSES Containment Design Parameters "-

1-3

1 0 GQNERAI INFOBHATEON 1 1 PURPOSE AND OBGANXZATXON OF BEPQRT The purpose of this report is to present evidence that the Susguehanna Steam Electric Station (SSES) design margins are adequate should the plant be subjected to the recently defined thermohydrodynamic loads which -'r'esult'rom safety relief valve (SBV) operations and/or discharges during a loss-of-coolant accident (LOCA).. in a GE.boilinq water reactor (BWB) 1-4

The criteria used for s'election of the SRV discharge device for SSES were minimization of pressure oscillation loads in the suppression pool and stable cond'ensation of steam f or the range of suppression pool temperatures over which safety relief valves can be expected to operate The options considered for satisf yinq these criteria were the rams-head tee, the quencher discharge device, and variations on these designs. Evaluation of the two principal devices indicated that the quencher offered significant advantages over the rams-head, including improved thermal performance at higher pool operatinq temperatures, as well as reduced loads.

A thermohydraulic quencher design for the safe'ty relief system of the SSZS is being engineered by Kraftwerk Onion {KMU) to satisfy the above criteria. The SSES quencher design is different from that presented in the Mark II DFFR in that it has been optimized based on parametric test studies which were conducted by KMU in order to minimize SRV discharge loads Kraftwerk Union has supplied to PPGL a package of significant desiqn and test reports pertaining to the quencher development to demonstrate design adequacy and guality of their device {refer to Table 1-1) . Mith reqard to the <<second pop<<phenomenon, KMU tests have indicated that, due to the quencher flow resistance, the water level in the SRV discharge pipe following initial discharge does not rise above the water level of the suppression pool. Refer to Subsection 0. 1.3. 6 for a further discussion.

To verify KMU's design approach a full-scale SSES unique unit ce11 test, as described in Chapter 8, is being performed by KMU for PPSL. Section 0 1 presents the analysis methods of the SRV discharqe loading 1-7

1.4 HARK II SUPPORTING PROGRAM PPGL is a member of the Mark II ovners group that vas formed in June, 1975 to define and investigate the dynamic loads due to SRV discharge and LOCA. The Nark II ovners group containment program concentrated initially on the tasks required for the licensing of the lead plants (Zimmer, LaSalle, and Shoreham) . This phase of

, work, called the short term program, is essentially complete (as of January, 1978) and a lonqer term program is undervay. The final goal of the Mark II program is to evolve a complete DPFR which vill support the plant-unique DARs submitted by each plant for its license to operate.

After qaining some understanding of the containment loads through the initial Mark II work, PPGL decided to find a qualified consultant to supplement in-house technical resources and assist in the determination of a realistic course of action for Susquehanna. In November, 1976, Stanford Research Institute, nov called Stanford Research Institute International (SRI), was selected, and an information exchange between SHI and PPGL ensued to determine what caused the greatest loads on the containment structure. After conducting a complete review of known data from the Mark II program and other knowledqeable persons and organizations, PPGX and SRI decided that the loads from main steam safety relief valve (SHV) discharge were the key 1oads to be controlled. A study of possible methods of controlling the load and a review of vhat activities were occurring in Europe led PPGL and SHI to the conclusion that an SRV discharge mitigating device {quencher) should be employed to reduce this loading on the Susquehanna containment. Although the Hark II ovners group had quencher-related tasks in their program, these tasks were not sufficiently timely to satisfy SSES-construction schedule needs.

Prom reviewinq the work done in Europe by such firms as ASEATOM, MARVIKEN, and Kraftwerk Union, PPGL discovered that all known quencher designs were based on data from Kraftwerk Union (KMU).

Thus, in March, 1977, SRI, Bechtel (the SSES Architect/Engineer) and PPGL visited KWU for discussion and tour of quencher-related facilities. In late July, 1977, PPGL employed the services of KMU to design a SSES-unique quencher device (see Section 1.3).

The definition of LOCA loads (Section 4.2) is in accordance with the Nark II program Due to the schedule restrictions f or Susquehanna. PPGL will define the thermo-hydrodynamic loads resultinq from SRV discharge usinq an approach developed by KMU.

This approach (presented in Section 4.1) differs from that of the Hark II proqram.See Table l-l for a summary of the documentation supportinq SSES licensinq.

1-8

1 5 PLANT DESCRIPTION The SSES, Units 1 and 2, is being built in Salem Township, Luzecne,County, about 5 miles noctheast of the Borouqh of Berwick. Two generating units of approximately 1,100 megawatts each are .scheduled for operation. Unit'1 for Novembec 1, 1980, and Unit 2 for May 1, 1982. General Electric is supplying'he nuclear steam supply systems; Bechtel power corporation is the architect-engineer and constructor.

The reactor building contains the major nuclear systems and equipment. The nuclear reactors for Units 1 and 2 are boiling water, direct cycle types with a rated heat output of 11.2 x 10~ Btu/hr. Each reactor supplies 13 4 x 10~ lb/hr of steam to the tandem compound, double flow tucbines.

1 5.'I P~ciman Containment a

The containment is a reinforced concrete structure consisting of a cylindrical suppression chamber beneath a truncated conical drywell. Piqure 1-1 shows the geometry'f the .containment and internal structures. The conical portion of the primary containment (drywell) encloses the reactor vessel, reactor coolant recirculation loops, and associated components 'of the reactor coolant system. The dcywell is separated from the wetwell, ie, the pressure suppression chamber and pool, by the drywell floor, .also named the diaphragm slab Major systems and components in the containment include the vent pipe system (downcomers) connecting the drywell and wetwell, isolation valves, vacuum relief system, containment cooling systems, and other service equipment. The cone and cylinder form a structurally inteqrated reinforced concrete vessel, lined with steel plate and closed at the top of the drywell with a steel domed head. The carbon steel liner plate is anchored to the concrete by stru'ctural steel members embedded in the concrete and welded to- the plate.

The entire- containment is structurally separated from the surrounding reactor building except at the base foundation slab fa reinforced concrete mat, top lined with a carbon steel liner plate) where a cold joint between the two adjoining foundation slabs is provided. The containment structure dimensions and parameters are listed in Tables 1-2 and 1-3. A detailed plant description can be found in the SSES PSAH, Section 3.8

1. 5. l. 1 Penetra tions Services and communication between the inside and outside of the containment are made possible by penetrations through the containment wall The basic types of penetrations are the drywell head, access hatches (equipment hatches, personnel lock, suppression chamber access hatches, CRD removal hatch),

electrical penetrations, and pipe penetcations. The piping 1-9

penetrations consist basically of a pipe with plate flange welded to it. The plate flange. is embedded in the concrete wall and provides an anchorage for the penetration to resist normal operating and accident pipe reaction. loads.

The internal structures consist of reinforced concrete and structural steel and have the major functions of supporting and shielding the reactor vessel, supporting the piping and eguipment, and forming the pressure suppression boundary. These structures include'he drywell floor (diaphragm slab), the reactor pedestal (a concentric cylindrical reinforced concrete shell resting on the containment base foundation slab and supportinq the reactor vessel), the reactor shield wall, the suppression chamber columns(hollow steel pipe columns supporting the diaphragm slab), the drywell platforms, the seismic trusses, the quencher supports, and the reactor steam supply system supports. See Figures 1-1 through 1-4 arid Tables 1-2 and 1-3 1- 10

106P d

y "d ~

CONTAINMENT Ct p r

C l.

I 4

s 0

'Cr 0 U p(g p

~ ~

d 136o p ~

SRV DIAPHRAGM f

'~

e,i SLAB PENETRATION ~

p<

C 03 u ~

e F ROM DIAPHRAGM ~ Cr i y SLAB PENETRATION .0> 160 O~,

SRV DIAPHRAGM Q8 SLAB PENETRATION P

0 r ~

d rg O~'EDESTAL 1

s TO QUENCHER S r ~ i I r d

NOTE: BRACING IS NOT SHOWN SUSQUEHANNA STEAM ELECTRIC STATION

.UNITS 1 AND 2 DESIGN ASSESSMENT REPORT SUPPRESSION CHAMBER PARTIAL PLAN

/

FIGURE 1-2

00 1So 3454 O. <<<< ~ ~

3150 0 0 ~

~ 454

'P P,

r,

<<P 88FT'" K"-

u 25822mm 3000 ~

~

P ~

P , <<I H Go 100FT" '

O 3048 0mm 750 2850 ~ h, 0 ~

B a'~ . A M

P '<<

0<< ~ p 2700 SOFT" 18288mm 4 a'

d

.a R P

2550 3eFT< 1050 b <<

m 0

42FT'"

F 12802mm D P~

240 V 5 r

~ P<< P 0 ~

k'35

~ C

~<< <<

0 225 <<

~

L

~

'<<<< ~ ~ e 0<<

~ 4 1954 1CS4 NOTE: INDICATES ADS-ASSOCIATED QUENCHER SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT QUENCHER DISTRIBUTION FIGURE 1R

TABLE 1-1 SSES LICENSING BASIS I. .Mark II Containment Supporting Program A. LOCh - Related Tasks Task Target Used for eeet et ~et tive t Activit e ~tee lett e Documentation SSES Licensin A.l. "4T" Phases I, II, IIX Phase I Test Report Completed NEDO/NEDE 13442-P-01 - 5/76 Yes Phase I Application 6/76 Memorandum Completed Application Memo Yes Phase II 6 III Test Report Completed NEDO/NEDE 13468-P - 12/76 Yes Phase II III Application Memorandum Completed Application Memo - 1/77 Yes h.2. Pool Swell Model Report Model Report Completed HEDO/NEDE 21544-P 12/76 Yes h.3. Impact Tests PSTF 1/3 Scale Tests Completed HEDE 13426-P 8/75 Yes Mark I 1/2 Scale Tests Completed NEDC 20989-2P 9/75 Yes h.4. Impact Model PSTF 1/3 Scale Tests Completed NEDE 13426-P - 8/75 Yes Mark I 1/2 Scale Tests Completed HEDC 20989-2P - 9/75 No EPRI 1/13 Scale Tests EPRI Report Completed EPRI NP-441 - 4/77 Yes A.5. Loads on Submerged Structures LOCA/RH Air Bubble Model 12/77 NEDE 21471 Undecided LOCA/RH Hater Jet Model 12/77 NEDE 21472 Undecided Applications Methods 12/77 NEDE 21730 Undecided Test Reports 1Q/78 Report Undecided h.6. Chugging Analysis and Testing Single Cell Report Completed HEDE 23703-P-ll/77 Yes 4T FSI Report 1/78 NEDE 23710-P Ho Multivent Model 12/77 NEDE 21669-P Ho h.7 ~ Chugging Single Vent CREARE Report 4Q/77 Report A.8. EPRI Test Evaluation EPRI - 4T Comparison Completed NEDO 21667<<8/77 Yes h.9. Multivent Subscale Testing and Analysis Facility Description and 4Q/77 Report Undecided Test Plan Test Report 1979 Final Report Undecided

Task Target Used for Number ~eetteit Activit e ~Con testee Documentation SSES Licensin A.10. Single Vent Lateral Loads Analysis Report 4Q/77 Report Undecided B. SRV Related Tasks B.l. Quencher Model DFFR Model Completed NEDO/NEDE 21061-P - 9/76 ..No Confirmatory Tests 3Q/78 Report No B.2. Ramshead Model Analysis Completed NEDO/NEDE 21061-P 9/76 No B.3. Monticello In-Plant SRV Tests Preliminary Test Report Completed NEDC 21465-P - 12/76 No Hydrodynamic Report Completed NEDC 215&1-P - 8/77 No B.4. Consecutive Actuation Transient Analysis Analytical Models 4Q/77 Report No B.5. SRV Quencher In-Plant Caorso Tests Test Plan Completed NEDM 20988 - 12/76 No Advance Test Report 1Q/78 Report No Pinal Report 4QI78 Report No B.6. Thermal Mixing Model Analytical Model 4Q/78 NEDC 23689 No Be7. SRV Water Clearing Analysis 3Q/78 Report No B.S. Quencher Air Bubble Frequency Analytical Model 4Q/77 Report No B.9v Monticello Fluid Structure Interaction (FSI) Analysis 1Q/78 Report No B.10. DFFR Ramshead Model Comparison to Monticello Data Data/Model Comparison Completed NSC-GEN 0394-10/77 No B. 11. Ramshead SRV Methodology SurnLary Analytical Methods Completed NEDO 24070-11/77 No B.12. Structural Response to SRV Discharge Analytical Report 4Q/77 Report No B. 13. Quencher Empirical Model Update Analytical Model and lQ/79 Report No Correlation

Task Target. Used for Nnet t ~Activit Activit e ~Con lotion Documentation SSES Licensin C. Miscellaneous Tasks C.l. DFFR, Rev. 3 Revision 10/78 NEDO/NEDE 21061 Revision 3 Not yet available C.2. Mass and Energy Release Report Analytical Report 3/77 GB-77-65 Yes C.3. NRC Round 1 Questions DFFR Amendment 1 Completed NEDO/NEDE 21061 Amendment 1 - 12/76 Yes DFFR Amendment 1, NEDO/NEDE 21061 Supplement 1 12/77 Amendment 1, Supplement 2 Yes C.4. Decoupling Chugging and SRV Loads On hold C.5. SRSS Justification SRSS Report Completed NEDO/NEDE 24010 - 7/77 Yes C.6. NRC Round 2 Questions DFFR Amendment 2 Completed NEDO/NEDE 21061 Yes Amendment 2 6/77 DFFR Amendment 2, Supplement 1 Completed NEDO/NEDE 21061 Amendment 2, Supplement 1 - 8/77 Yes DFFR Amendment 2, NEDO/NEDE 21061 Amendment 2, .

Supplement 2 Completed Supplement 2 Yes Supplement 3 4q/77 Supplement 3 Yes C.7. Justification of "4T" Chugging Loads Justification Completed NEDO/NEDE 23617"P-8/77 Undecided Bounding Loads NEDO/NEDE 24013-P-8/77 Undecided NEDO/NEDE 24104-P-8/77 Undecided NEDO/NEDE 24015-P-8/77 Undecided NEDO/NEDE 24016-P'-8/77 Undecided NEDO/NEDE 24017-P-8/77 Undecided NEDO/NEDE 23627-P-8/77 Undecided C.g. FSI Effects in Mark II Evaluation of FSI Effects lg/78 Report Undecided Containments C.9. Monitor World Tests Monitoring World Pressure Reports (Quarterly) No Suppression Tests II. 1NU Tests and Reports (supplied to PP&L)

Document Used for Number Title Status Documentation SSES Licensin Formation and oscillation of a spherical gas Completed AEG - Report 2241 Yes bubble Analytical model for clarification of pressure pulsation in the wetwell after vent cleaning Co pleted AEG - Report 2208

Document Used for Number Title Status Documentation SSES Licensin Report 2593 3~ Tests on mixed condensation with model quenchers Completed KWV Yes 4, Condensation and vent clearing tests at GKM with quenchers Completed KWV Report 2594 Yes

5. Concept and design of the pressure relief Report 2703 Yes system with quenchers Completed KWV Report 2796 Yes

6. KKB vent clearing with quencher Completed KWV

7. Tests on condensation with quenchers when submergence of quencher arms is shallow Completed KWV Report 2840 Yes

8. KKB Concept and task of pressure relief system Completed KWV . Report 2871 Yes

9. Experimental approach to vent clearing in a model tank Completed KMV - Report 3129 Yes

10. KKB Specification of blowdown tests during non-nuclear hot.functional test - Rev. I dated October 4, 1974 Completed KWU/V 822 Report Yes Anticipated data for blowdown tests with

- pressure relief system during the non-nuclear hot functional test at nuclear power station Brunsbuttel (KKB) Completed KWU Report 3141 Yes

12. Results of the non-nuclear hot functional tests with the pressure relief system in the nuclear power station Brunsbuttel Completed KWU Report 3267 Yes

13. Analysis of the loads measured on the pressure relief system during the non-nuclear hot, Report 3346 functional test at KKB Completed KWU Yes

14. KKB Listing of test parameters and important test data of the non-nuclear hot functional tests with the pressure relief system Completed KWU - Working Report Yes R 521/40/77

15. KKB Specification of additional tests for testing of the pressure relief valves during the nuclear start-up, Rev. 1 Completed KWU/V 822 TA Yes

16. KKB Results from nuclear start-up testing of pressure relief system Comp le ted KNJ Working Report Yes R 142-136/76 17 'uclear Power Station Phillipsburg - Unit 1 Hot Functional Test: Specification of pressure relief valve tests as well as emergency cooling and wetwell cooling systems Completed KWU/V 822/RF 13 Yes

Document Used for Number Title Status Documentation SSES Licensin

18. Results of the non-nuclear hot functional tests with the pressure relief system in Working Report the nuclear power station Phillipsburg Completed KWU Yes R 142-38/77

19. KKPI - Listing of test parameters and important test data of the non-nuclear hot functional tests with the pressure relief system Completed KWU Working Report Yes R 521/41/77

20. Air oscillations during vent clearing with Report 2327 single and double pipes Completed AEG Yes 616715/cak

TABLE 1-3 SSES CONTAINMENT DESIGN PAKQKTERS A. Dr ell and Su ression Chamber ~Dr ell Su ression Chamber

1. Internal Design Pressure 53 psig 45 psig

2. External Design Pressure 5 psid 5 psid

3. Drywell Floor Design Differential Pressure Upward 28 psid Downward 28 psid

4. Design Temperature 340'F 220'F

5. Drywell Free Volume (Minimum) 3 239,337 ft3 (including vents) (Normal) 239,593 ft3 (Maximum) 239,850 ft .

6. Suppression Chamber Free (Minimum) 148,590 ft3 Volume (Normal) 153,860 ft3 (Maximum) 159,130 ft

7. Suppression Chamber Water Volume (Minimum) 122,410 ft3 (Normal)

(Maximum) 131,550 ft

8. Pool Cross-Section Area Gross (Outside Pedestal) 5379 ft Total Gross (Including Pedestal Water Area) 5679 ft Free (Outside Pedestal) 5065 ft Total Free 5365 ft

CHAPTER 2

SUMMARY

TABLE OP CONTENTS LOAD DEFINITION

SUMMARY

2 1.1 SRU Load Definition Summary

2. 1.2 LOCA I.oad Definition Summary DESIGN ASSESSMENT

SUMMARY

2.2.1 Containment Structure and Reactor Building Assessment Summary 2.2.1. 1 Containment Structure Assessment Summary.

2. 2. 1. 2 Reactor Building Assessment Summary 2 2 2 Containment Submerged Structures Assessment Summary 2 2 3 Piping Systems Asessment Summary

2 0

SUMMARY

This Design Assessment Report contains the SSES adequacy evaluation for dynamic loads due to LOCA and SRV discharge.

2-2

2 1 LOAD DEFINITION SUMN ARY 2.1.1 SRV Load Definition S~nmmar Hydrodynamic loads resulting from SRV actuation fall into tvo distinct categories: loads on the SRV system itself (the discharge line and the discharge quencher device), and the air clearing loads on the suppression pool walls and submerged structures.

Loads on the SRV system during SRV actuation include loads on, the SRV piping due to effects of steady backpressure, transient vater slug clearing, and SRV line temperature. Determination of loading on the quencher body, arms, and support is based on transients resulting from valve opening (water clearing and air clearing), valve closing, and operation of an adjacent quencher.

Air clearing loads are examined for four loading cases:

symmetric (all valve) SRV actuation, asymmetric SBV actuation, single SRV actuation, and Automatic Depressurization System {ADS) actuation. Dynamic forcing functions for loading of the containment walls, pedestal, basemat, and submerged structures are developed using t'echniques developed in Section 4. 1. Loads on the SRV system due to SRV actuation are discussed .in Subsection 4. 1. 2, and loads on suppression pool structures due to SRV actuation are discussed in Subsection 4. 1.3. A full scale, unit cell test program is heing employed to verify SSES unique SRV loading as described in Chapter 8.

2.1 2 LOCA Load Definition Summary The spectrum of LOCA-induced loads on the SSES containment structure is characterized by LOCA loads associated with poolsvell, condensation oscillation and chugging loads, as well as long term LOCA loads.

The LOCA loads associated vith poolsvell result from short duration transients and include downcomer clearing loads, w'ater jet loads, poolsvell impact and drag loads, pool fallback drag loads, poolswell air bubble loads, and loads due to dryvell and vetwell temperature and pressure transients. Techniques used to evaluate these loads are described in Subsection 4.2 1.

Condensation oscillations result f rom mixed flow {air/steam) and pure steam flov effects in the suppression pool. Chugging loads result from lov mass flux pure steam condensation. The load definitions for these phenomena are contained in Subsection 4.2.2.

Long term LOCA loads result from those vetvell and dryvell pressure and temperature transients which are associated with design basis accidents (DBA), intermediate accidents (IBA), and small break accidents (SBA). Their load definitions are contained in Subsection 4.2.3.

Structures directly affected by LOCA loads include the drywell walls and floor, wetwell walls, RPV pedestal, basemat, liner plate, columns, downcomers, downcomer bracing system, guenchers, and wetwell piping. Their loading conditions are described in Subsection 4 2 4.

2-4

2 2 DESIGN ASSESSMENT SUNK ARY Design assessment of the SSES structures and components is achieved by analyzing the response of the structures and components to the load combinations explained in Chapter 5. In Chapter 7, predicted stresses and responses (from the loads defined in Chapter 4 and combined as described in Chapter 5) are compared vith the applicable code allovable values identified in Chapter 6; the SSES design vill be assessed as adequate by virtue of design capabilities exceeding the stresses or responses resulting from SRV discharge or J.OCA loads.

2. 2. 1 Containment Structure and Reactor Building Assessment Sum~mar 2.2. 1.1 Containment Structure Assessment Summary The primary containment valls, base slab, diaphragm slab, reactor pedestal, and reactor shield are analyzed for the effects of SRV and LOCA in accordance vith Table 5-1. The ANSYS finite element program is used for the dynamic analysis of structures.

Response spectra curves are developed at various locations within the containment>> structure.to assess the adequacy of, components.,

~

Stress resultants due to dynamic loads are combined with other 1oads in accordance with Table 5-1 to 'evaluate rebar and concrete stresses.'esign 'safety margins will are defined by comparing the actual concrete and rebar stresses at critical sections vith the code,allowable values.

2.2.'1.2 Reactor Building Assessment Summary The reactor building is assessed for the effects of SRV and LOCA loads in accordance wi th Table 5-1.

Containment basemat acceleration time histories are used to investigate the reactor building response to the SRV and LOCA loads. Response spectra curves at various reactor building elevations are used to assess the adequacy of components in the reactor building.

2. 2. 2 Containment Submerged Structures Assessment Summary Design assessment of the suppression chamber columns and dovncomer pipes is being perf ormed. Based upon an approximate, equivalent static analysis carried out to date, strengthening of these structures should not be required. This conclusion vill be confirmed when the dynamic analysis is complete Preliminary results from the dynamic analysis of the suppression pool liner plate indicate that no structural modifications are required This conclusion will be confirmed when the final analysis is complete.

2-5

The original downcomer bracing has been redesigned with pipe sections to minimize bracinq drag loads due to poolsvell and fallback. The revised bracing system is designed using a simplified equivalent static approach.

Containment and reactor building piping systems are being designed to withstand the effects of LOCA and SRV induced dynamic loads. The load combinations for piping are defined in Table 6.1 of Ref. 10.

2-6

3 1 DESCRIPTXOM OF SAFETY RELIEF VALVE DISCHARGE Susquehanna Units 1 and 2 are equipped with a safety relief system which condenses reactor steam in a suppression chamber pool. By this arrangement, reactor, steam is conducted to the wetwell. via fast acting safety relief valves and quencher equipped discharge lines. .This section discusses the causes of SRV discharge, describes the SRV discharge proce'ss, and identifies the resultant SRV discharge actuation cases 3 1.1 Causes of S V Discha~r e During certain reactor operating transients, the SRVs may be actuated (by pressure, by electrical signal, or by operator action) for rapid relief of pressure in the reactor pressure vessel. The following reactor operating transients have, been identified as those which may result in SRV actuation:

\

a. Turbine qenerator trip (with bypass or without)

b. Hain s tea m line isolation val ve (NSI V) closure

c. Loss of condenser vacuum

d. Feedwater controller failure

e. Pressure regulator failure open

f. Generator load rejection (with and without bypass)

q. Loss of ac power h Loss of feedwater flow Trip of two recirculation pumps Recirculation flow control failure decreasing flow

k. Inadvertent safety relief valve open ing A detailed description of these transients is provided in Section 15.2 of the FSAR

3. 1.2 Description of the SRV Discharge Phenomena and SRV

~t.aadin Cases Before an individual safety relief valve opens, the water level in the discharge line is approximately equal to the water level in the pool As a valve opens, steam flows into the discharge line air space between the valve and the water column and mixes with the air (see detailed evaluation in Chapter 3 of Ref 1, pages 6-12 through 6-14) . Since the downstream portion of the discharge line contains a water slug and does not allow an 3-3

immediate steam discharge into the pool the pressure inside

~ -the line 'increases. The increased pres'sure expels the water slug from the SRV discharge line and quencher. The magnitude of the water clearinq pressure is primarily influenced by the steam flow rate through the valve, the deqree to which entering steam is condensed along the discharge line;,walls, the volume of the discharge line airspace, and the length of the water slug to be accelerated.

The clearing of water is followed by an expulsion of the enclosed air-steam volume. The exhausted gas forms an oscillating system with the surrounding water, where the gas acts as the spring and the water acts as the mass. This oscillating system is the source of short term air clearing loads.

While the air-steam mixture oscillates in the pool it because of buoyancy and eventually breaks through the pool water rises surface at which time air clearing loads cease. When all 'the air leaves the safety relief system, steam flows into the suppression pool through'he quencher holes and condenses. The SSES quencher design assures stable condensation even with elevated pool water temperature.

The SBV actuation cases resulting from the transients listed in Subsection 3. 1.1 are classified, as being one of the following cases:

a. Symmetric (all valve, or AOT) discharge

b. Asymmetric discharge, including single valve discharge

c. Automatic Depressurization System (ADS) discharqe Also considered in the containment design is the effect of subsequent SRV actuations (second-pop), discussed in Subsection

4. 1 6.

The symmetric discharge case (otherwise termed the all-valve, or AOT, case) is classified as the type of SRV discharge that would follow rapid isolation of the vessel from the turbine such as turbine trip, closure of all MSXVs, loss of condenser vacuum, etc. As pressure builds up following isolation of the 'vessel, the SBVs actuate sequentially according to the pressure set points of the valves. This may or may not result in actuation of all the SRVs, but for conservatism in loading considerations all valves are assumed to actuate Refer to Subsection 4. 1.3. for 1 discussion of the loads resulting from this all-valve case.

Asymmetric discharqe is defined as the firing of the SRVs for the

~

three ad'jacent quencher devices which results in the greatest asymmetric pressure loadinq on the containment. This situation is hypot'hesized when, following a reactor'cram and isolation of the Vessel, decay heat raises vessel pressure so that low set point valves actuate. Xf, during this time of discharge of decay heat energy, manual actuation of the two other adjacent SRVs that 3-4

comprise the asymmetric case is assumed, this actuation would result in the maximum symmetric pressure 'load on the containment.

Subsection 4. 1.3.2 gives a discussion of the loads resulting from the asymmetric discharqe case.

The single valve discharge case is classified as the firing of the SRV which qives the single largest hydrodynamic load.

Transients that could potentially initiate such a case are an inadvertent SRV discharqe or Design Basis Accident (DBA). Refer to Subsection 3. 2. 3 for a discussiori of the latter possibility Subsection 4. 1.3.2. provides a discussion of t'e loads resulting 1

from the single valve case.

The ADS discharge is defined as the simultaneous actuation of the six SRVs associated with the ADS. See Pigure 1-4 for the location of the quencher devices associated with the ADS valves.

The ADS is assumed to actuate durinq an Intermediate Break Accident (IBA) or Small Break Accident '(SBA). If an ADS discharge is hypothesized'oincident to an IBA or SBA (described in Subsections 3.2.2 and 3.2. 1, respectively), the effects of an increased suppression pool temperature (resulting from steam condensation during the LOCA transient) and increased suppression chamber pressure (resultinq from clearing of the dryvell air into the pool durinq the transient) are considered in the calculation of pressure loadings for the ADS discharge case. See Subsection

4. 1.3.'3 for further discussion of the loads resulting from the ADS case.

I 3-5

3 2 DESCRIPTION OF LOSS-OF-COOLANT ACCIDENT'his event involves the postulation of a spectrum of piping breaks inside the containment varying in size type, and location of the break. For the analysis of hydrodynamic loadings on the containment, the postulated LOCA event is identified as .a Small Break Accident (SBA), an Intermediate Break Accident (IBA), or aDesign Basis Accident (DBA).

3 2. 1 Small Break Acci ent SB~A This subsection discusses the. containment transient associated with small primary system blowdowns. The primary system ruptures ia this category are those ruptures that will not result in reactor depressurization from either loss of reactor coolant or automatic operation of the ECCS equipment, ie, those ruptures with a break size .less than 0 1 sq ft The followinq sequence of events is assumed to occur With the reactor and containmeat operating at the maximum normal conditions, a small break occurs that allows blowdown of reactor steam or water to the drywell. The resulting pressure increase in the drywell leads to a hiqh drywell pressure signal that scrams the reactor and activates the containment isolation system. The drywell pressure continues to increase at a rate dependent upon the size of the steam leak. The pressure increase lowers the water level in the-downcomers. At. this time, air and steam enter the suppression pool at a rate dependent upon the size of the leak. Once all the drywell air is carried over to the suppression chamber, pressurization of the suppression chamber ceases and the system reaches an equilibrium condition.

The 'drywell contains only superheated steam, and continued blowdown of reactor steam condenses in the suppression pool. The principal loadinq condition in this case is the gradually increasinq pressure in the drywell and suppression pool chamber and the loads related to the condensation of steam at the end of the vents.

3.2.2 Intermediate B eak Accident IBA This subsection discusses the containment transient associated with intermediate primary system blowdowns. This classification covers breaks for which the blowdown will result in limited reactor depressurization and operation of the ECCS, ie, the break size is equal to or slightly qreater than 0. 1 sq ft.

Following the break, the drywell pressure increases at approximately 1.0 psi/sec. This drywell pressure transient is sufficiently slow so that the dynamic effect of the water in the vents is negligible and the vents will clear when the drywell-to-suppression chamber differential pressure is equal to the hydrostatic pressure corresponding to the vent submergence. The 3-6

CHAPTER 4 LOAD DEFINITION TABLE OF CONTENTS 4.1 SAFETY RELIEF VALVE (SRV) DISCHARGE LOAD DEFINITION 4 2 LOSS-OF-COOLANT ACCIDENT (LOCA) LOAD DEFINITION 4 2.1 LOCA Loads Associated with Poolswell 4 2 1 1 Metwell/Drywell Pressures during Poolswell 4 2 1 2 Pools well I m pac t I.oad 4 2 1 3 Poolswell Drag Load 4 2.1 4 Downcomer Clearing Ioads 4 2.1.5 Downcomer Water Jet Load 4.2. 1.6 Poolswell Air Bubble Load 4.2. 1. 7 Poolswell Fallback Load 4.2. 2 Condensation Oscillations and Chugging Loads 4.2.2. 1 Condensation Oscillation Load Definition 4 2 2.2 Chuggi ng Load Definition 4 2.3 Long Term I.OCA Load Definition 4.2 3.1 Design Basis Accident (DBA) Transients

4. 2. 3. 2 Intermediate Break Accident (IBA) Transients 4 2 3-3 Small Break Accident (SBA) Transients 4 2.4 LOCA .Loading Histories for SSES Containment Components 4.2. 4 1 LOCA Loads on the Containment Mall and Pedestal 4 2.4 2 LOCA LOads on the Basemat and Liner Plate 4.2. 4.3 LOCA Loads on the Drywell and Drywell Floor 4 2 4 4 LOCA Loads on the Columns 4.2.4.5 LOCA Loads on the Downcomers 4 2.4 6 LOCA Loads on the downcomer Bracing 4-1

4.2 4 7 LOCA Loads on Metwell Piping 4 3 ANNULUS PRESSURIZATION 4 4 FIGURES 4 5 TABLES 4-2

CHAPTER 4 P~XGU R S Mum~be Title 4-1 These figures are proprietary and are, found in the through proprietary supplement to this DAR.

4-37 k 4-3 8 SSES Short Term Suppression Pool Height 4-39 SSES Short Term Wetwell Pressure 4-40 SSES Pool Surface Velocity vs Elevation 4-41 Basemat SSES Water Clearing Jet 4-42 SSES Jet Impingement Area (Water Clearing) 4-43 SSES Poolswell Air Bubble Pressure 4-44 Air Bubble Pressure on Suppression Pool Walls 4-45 Symmetric and Asymmetric Spatial 'Loading Specification 4-46 SSES Drywell Pressure Response to DBA LOCA 4-47 SSES Wetwell Pressure Response to DBA LOCA 4-48 SSES Suppression Pool Temperature Response to DBA LOCA 4-49 SSES Drywell Temperature Response to DBA LOCA 4-50 SSES Suppression Pool Temperature Response to IBA 4-51 Typical Nark II Containment Response to the IBA 4-52 Typical Nark II Containment Response to the SBA 4-53 SSES Components Affected by LOCA Loads 4-54 SSES Components Affected by LOCA Loads 4-55 LOCA Loading History. for the SSES Containment Wall and Pedestal Local Loading History for the SSES Basemat and Liner Plate 4-57 LOCA Loading History for the SSES Drywell and Drywell Floor 4-3

4-58 LOCA Loading History for the SSES Columns 4-59 LOCA Loading History for the SSES Dovncomers 4-60 LOCA Loading History for the SSES Dovncomer Bracing System 4-61 LOCA Loading History for SSES Qetvell Piping 4-62 This figure is proprietary 4-4

CHAPTER 4 TABLES Num~be Title 4-1 These tables are proprietary and are found through in the proprietary supplement to this DAR 4-15 4-16 LOCA Loads Associated with Poolsvell 4-17 SSES Dryvell Pressure 4-1 8 SSES Plant Unique Poolsvell Code Input Data 4-19 Input Data for SSES LOCA Transients 4-2 0 Component LOCA Load Chart for SSES 4-21 Hetvell Piping LOCA Loading SItuations

0 LOAD DEPXNTTION

4. 1 SAFETY RELIEF VALVE SB~VDISCRARGE LOAD DEFINITION See the Proprietary Supplement- for this section.

4-6

4 2 LOCA LOAD DEFINITION Subsections 4.2.1, 4.2.2 definition of loads resulting and 4.2.3 vill discuss the numerical from a LOCA in the SSES containment. The LOCA loads are divided into three groups.

V (1) Short term LOCA loads associated with poolsvell (Subsection 4. 2. 1)

(2) Condensation oscillations and chugging loads (Subsection 4. 2.2)

(3) Long term LOCA loads (Subsection 4. 2. 3) .

The application of these loads to the various components and structures in the SSES containment is discussed in Subsection 4.2.4.

4 2 1 LOCA LOADS ASSOCIATED WITH POOLSWELL A description of the LOCA/Poolswell transient has been given in Section 3.2 of this Design Assessment Report. The LOCA loads associated vith poolsvell are-listed in Table 4-16. The appropriate Mark XI generic document from which SSES plant unique loads are calculated is also shown in Table 4-16. A discussion of these loa'ds and their SSES unique values follows.

The drywell pressure transient used for the poolswell portion of the LOCA transient (< 2.0 seconds) is given in Table IV-D-3 of Ref 7. A portion of this table is reproduced herein as Table 4-

17. This drywell pressure transient includes the blovdown effects of pipe inventory and reactor subcooling and is the highest possible drywell pressure case for poolsvell.

The short term poolsvell wetwell pressure transient resulting from this 'dr@well pressure transient is calculated by applying the poolswell model contained in Ref 8. The equations and assumptions in the poolsvell model were coded into a Bechtel computer program and verified against the Class 1, 2 and 3 test cases contained in Ref 9. This verification is documented in Appendix D to this report. Other inputs used for the calculation of the SSES plant unigue poolswell transient are shown in Table 4-18. The short term suppression pool surface elevation and corresponding wetwell pressure transient calculated with the.

poolsvell code are shown in Pigures 4-38 and 4-39 respectively.

The short term wetvell pressure peak is 56 1 psia {41.4 psig).

The (drywell minus wetvell) pressure differential is also plotted on this curve. The minimum A P occurring durinq poolswell is

-9. 2 psid at 0. 893 seconds after vent clearing(1. 58 seconds after the break occurs) 4-7

4.2 1 2 Poolswell Zm act Loa'd Any structure located between the .initial suppression pool surface (el. 672 ') and the peak poolswell height (el 690', see f iqure 4-38) is sub ject to the poolswell water impact load There are only minor structures (such as miscellaneous wetwell pipinq) in this portion of the SSES wetwall. This load is calculated as specified in Ref 10, S'ubsection 4 4.6. A SSES plant-unique velocity ys elevation curve has been qenerated with the poolswell model (Figure 4-40). I< is used in conjunction with impact pressure vs velocity curves for various size and shape components (Ref 10, Figures 4-34, 4-35 and 4-36) to develop a peak impact pressure at the componen-t's elevation. The pe'ak impact pressure is combined with a generalized impact pressure time history curve (Ref 10, Figure 4-37) to specify the structural load. All structures, subject to poolswell impact loads in the SSES containment are classified as >>small structures>>.

2.1. 3 Poolaeell D~aa Load The poolswell drag load applies to any structure located between the elevation of the vent exit (el. 660') and the peak pool swell

,heiqht (el. 690') . The load is calculated'for all components in the region based upon the maximum pool surface velocity (29.35 fps), regardless of elevation. The drag load pressure is calculated from Ref 10, Equation 4-24 using Vf = 29.35 fps for the velocity and p f = 62.4 1bmjft~. for the density of water, P =(1/2)CD p f Vf~ (4-1)

P (psi) = 5.8 C '4-2)

The appropriate drag coefficient for the structure involved is selected from Ref 10, Figure 4-29. The pool swell drag load is applied in either the horizontal or vertical direction (Subsection 4. 4 5. 2 of Ref 10). forceps For the case of a component oriented vertically with its axis parallel to the velocity of the pool surface,* the skin friction coefficient, AC~, used in Ref 10, Subsection 4 4. 8 is applied in place of CD. I'his method would apply, for example, to the vertical loads on downcomers, columns, or safety relief lines in the wetwell. Usinq C f = 0.0023, the vertical drag on a vertically oriented component is recalculated using Equation 4-26 of Ref 10.

F V

(lbf) = 0.0133Af (in~) . (4-3)

Here Af is the skin friction area (wetted surface area) subject to,the vertical drag force.

LOCA loads on the downcomer bracing are described in Subsection 4 2 4 6

Vertical loads on the downcomers during downcomer'learing can be estimated by using a drag load formula similar to Equation 4-3.

In this case the vent clearing velocity is 60 fps (Ref 10, Subsection 4.4.5.1) and Af is the wetted inside area of the downcomer, conservatively calculated to .be Af = {12 ft} (m) (2 ft) = 75~4 ft<

From Equation 4-3 the vertical clearing load on the downcomer for SSES is, P

V

= 0.6 kips.

'his is of similar magnitude to the vertical thrust load of 0.7 kips on the downcomer durinq steam blowdown (Ref 10, Subsection 4. 2. 3) .

Lateral loads on the downcomers during clearing are estimated from Ref 11, Table 3-4 to be less than 3 kips.

4 2 1 5 Downcomer Hater Jet Load The water clearinq jet load is calculated based on the approach developed in the design guides {Befs 12 and 13). This load is experience as a drag load by structures located within the )et cone beneath the downcomers and as a,jet impingement load by the basemat . The jet impingement load on the basemat is calculated from Ref 10, Equation 4-25, p. 4-43.

P g= pf A vf2 (4- 4)

Here p is the density of water {taken to be 62.4 ibm/ft~), A is the total jet impingement area and v is the attenuated water velocity corresponding to the maximum vent clearing get velocity (Ref 10) . Figures 4-41 and 4-42 show elevation and plan views of the SSES downcomers and their associated jet cones. The radius of the jet cone at the basemat is 2.69 ft. and the total area intercepted by the 87 downcomers in the SSES wetwell is 1978 ft~.

As seen in Figure 4-42 there is no significant overlap of adjacent jets on the basemat.

The vent clearing velocity of 60 fps is attenuated by a factor of 0.68 usinq the method described in Ref. 10,'ubsection 4.4.5. 1 to yield a value of 40. 8 fps at the basemat. The jet impingement pressure is calculated from Ref 10, Equation 4-26, p. 4-43 to be

= v I pf P

=

P I 22m 4 ps'.~

Using the value for A of 1978 ft~ for the SSES design the total downcomer water jet impingement load on the basemat is 4-9

F = 2848.3 kips.

This load acts vertically downward on the basemat from the time the break occurs until 'the dovncomers have cleared, at 0.6863 sec (Ref 7) .

4. 2. 1. 6 Poolswell Air Bubble Load The poolsvell air bubble pressure load'as containment walls is it applies to the described in Ref 10, Subsection 4.4.5.3.

This load is viewed as an increase in the hydrostatic pressure on the suppression pool walls belov the vent exit plane and is caused by the air bubble which has been purged from the drywell in the initial stages of the LOCA. The air bubble pressure transient calculated with the poolsvell model (described in Subsection 4.2. 1.2) is shovn in Figure 4-43. Figure 4-44 shovs the normalized total. pressure distribution (hydrostatic plus air bubble) to he applied to the containment as a result of this load. The pressure on the wetvell walls between the vent exit and the water surface contains a linear decrease to 0.0 psig at the water surface (Ref 10, Subsection it applies 4.4.5.3).'his load as to submerged structures is described in Refs 13 and 14.

4.2. 1.7 Poolswell Fallback Load The poolswell fallback 3.oad is a drag load vhich applies to all structures between the peak poolswell height (el. 690') and the vent exit (el. 660'). This load is calculated for components in this region using the analysis of Subsection 4.4.5 4 of Ref 10 Since the vertical structures are parallel to the fallback flow, they are subjected to negligible fallback loads. (For a fallback velocity of 30 fps the load is significantly less than 1 kip).

The downcomer bracing structure at elevation 668'-0" is, however, perpendicular to the fallback flow and vill undergo a fallback load applied vertically dovnvard. The fallback drag velocity is calculated using the equation on page 4-45 of Ref 10.

VFB= '.82 (8 ) </~ (4-6)

For the SSES design, the maximum downcomer submergence, Ho, is 12 feet so the fallback velocity is 34.05 fps. The drag pressure due to this velocity is calculated from Ref 10, Equation 4-24. to be PFB (psi) = 7 8 (4-7) where Cp is the appropriate drag coeff icient f or the structure being loaded.

4-10

Pallback loads are calculated using Refs 12 and 13.

2.2 Condensation oscillations a~ad chu in'oads Condensation oscillation and chugginq loads follow the poolswell loads in time. There are basically three loads in this time period, i.e., from about 4 to 60 seconds after the break.

Condensation oscillation is broken down into two phenomena, a mixed flow regieme and a steam flow regieme. The mixed flow reqieme is a relatively high mass flux phenomenon vhich occurs during the final period of air purging from the drywell to the vetvell. Thus, the mixed flow through the dovncomer vents contains some air as well as steam. The steam flov portion of the condensation oscillation phenomena occurs after all the air has been carried over to the vetwell and a relatively high mass flux of pure steam flow is established.

Chugging is a pulsating condensation phenomenon which can occur either. folloving the intermediate mass flux phase of a LOCA, or during the class of smaller postulated pipe breaks that result in steam flow through the vent system into the suppression pool A necessary condition for chugging to occur is that pure steam

.flows from the LOCA vents. Chuqqing imparts a loading condition to the suppression pool boundary and all submerged structures.

4.2.2. 1 Condensation Oscillation Load Definition The load specification for the mixed and steam flow phases of condensation oscillation is taken from Appendix A to Ref 20.

The mixed flov portion of the condensation oscillation load is specified as a sinusoidal load at the containment's critical frequencies, between 2 and 7 Hz vith an amplitude of x 1.75 psi.

This load is to be applied uniformly to the vetted portion of the suppression pool boundary below the vent exit with a linear attenuation to the free surface o'f the suppression pool. The duration of this load is from 4 to 15 seconds after the break has occurred.

The steam flow portion of the condensation oscillation load is specified as a sinusoidal load at the containment's critical freguencies betveen 2 and 7 Hz vith an amplitude of a 5.0 psi.

The load is to be applied uniformly to,the wetted portion of the:

suppression pool boundary below the vent exit with a linear .

attenuation to the suppression pool free surface. Also a sinusoidal load of amplitude a 0.5 psi is applied uniformly to the drywell boundary at critical frequencies between 2 and 7 Hz.

The duration of both the dryvell and suppression pool steam flov condensation oscillation load is the time period from 15 to 25 seconds followinq the initial .break.

Condensation 'oscillation loads on submerged structures are calculated usinq Refs 12 and 13.

4-11

The pool boundary chugging load is specified in Ref 15 Tvo loadinq conditions are described: symmetric and asymmetric.

The symmetric loadinq condition is specified as +4.8 psig/-

4.0 psig and is to be applied uniformly around the entire pool boundary as shovn.in Figure 4-45 (extracted from Ref 15).

e The asymmetric loading condition has a specified maximum positive/negative pressure of +20 psig/-14 psiq and has the circumfezential spatial distribution depicted in Figure 4-45.

Chugqing loads on submerged structures vill be evaluated when the desiqn guide dealing vith these loads is completed The chugging load imparted to the downcomer will be specified when the appropriate dynamic forcing function becomes available 4-2 3 LONG TERM LOCA LOAD DEFINITION The loss-of-coolant accident causes pressure and temperature transients in the drywell and wetvell due to mass and energy released from the line break. The dryvell and wetvell pressure and temperature time histories are required to establish the structural loading conditions i.n the conta'inment because they are the basis for other containment hydrodynamic phenomena. The response must be determined for a range of parameters such as leak size, reactor pressure and containment init'ial conditions.

The results of this analysis are documented in Ref 7.

The DBA LOCA for is conservatively estimated to SSES 3.53 ft~ break of the recirculation line (Ref 7). 'hebe a SSES plant unique inputs for this analysis are shown in Table 4-19. Drywell and vetvell pressure responses are shovn in Figures 4-46 and 4-47 (extracted from Ref 7) These transient descriptions do not, hovever, contain the effects of reactor subcooling Suppression pool temperature response is shovn in Figure 4-48 (Ref 7~). This transient description also does not contain the effect of reactor subcoolinq. Dry'veil temperature zesponse is shown in Figure 4-49 and similarly does not contain the effects of pipe inventory or reactor 'subcooling

4. 2.3. 2 Intermediate Break accident~IBA) Transients The worst-case intermediate break for the Mark XI plants is a main steam line break on the order of 0. 05 to 0.1 f t~. At this time plant unique IBA,data .for SSES is available only for the suppression pool temperature response to a 0. 05 ft> break (Ref 7). This data is shown in Piqure 4-50. Drywell temperature and wetwell and dryvell pressures for the SSES XBA are estimated from curves for a typical Mark IZ containment shown in Figure 4-51 (extr acted from Ref 10) 4-12

At this time plant-unique SBA data for SSES is not available.

The wetwell and drywell pressure and temperature transients 'for a typical Nark II containment are used to estimate SSES containment response to these accidents. These curves are shown in Figure 4-52 {extracted from Ref 10).

4 2 4 LOCA LOADING HISTORIES FOR SSES CONTAINNENT CONPONENTS The various components directly affected by LOCA loads are shown schematically in Figures 4-53 and 4-54. These components may in turn load other components as they respond to the LOCA loads.

For example, lateral loads on t'e downcomer vents produce minor reaction loads in the dryvell floor from which the:downcomer's are supported. The reaction load in the drywell floor is an indirect load resulting from the LOCA and is defined by the appropriate structural model of the downcomer/drywell floor system Only the direct loadinq situations are described explicitly here. Table 4-20 is a LOCA load chart for SSES. This chart shows which LOCA loads directly affect the various structures in the SSES containment design Details of the loading time histories are discussed in the folloving subsections..

4 2 4. 1 LOCA Loads on the Containment Wall and Pedestal Figure 4-55 shows the LOCA loading history for the SSES containment wall and the RPV pedestal. The wetvell pressure loads apply to the unwetted elevations in the wetwell; the appropriate hydrostatic pressure addition is made for loads on the wetted elevations. Condensation oscillation and chugging loads are applied to the wetted elevations in the vetwell only.

The poolswell air bubble load applies to the vetwell bo'undaries as shown in Figure 4-44.

4 2.4.2 LOCA Loads on the Basemat and Liner Plate Figure 4-56 shows the LOCA loading history for the SSES basemat and liner plate. Wetwell pressures are applied to the wetted and unwetted portions of the liner plate as discussed in Subsection

4. 2.4. 1. The downcomer water jet impacts the basemat liner plate as does the poolsvell air bubble load. Chugging and condensation oscillation loads are applied to the wetted portion of the liner plate.

4 2.4.3 LOCA Loads on tahe Dc well and D~rwell Floor Fiqure 4-57 shows the LOCA loading history for the SSES drywell and drywell floor. The drywell floor undergoes a vertically applied, continuously varying differential pressure, the upward component of which is especially prominent during poolswell vhen the vetwell air space is highly compressed 4-13

Figure 4-58 shows the LOCA loading history for the SSES columns.

Poolswell drag and fallback loads are very minor since the column surface is oriented parallel to the pool swell and fallback velocities. The poolswell air bubble, condensation oscillations and chugqinq will provide loads on the submerged {wetted) portion of the columns.

4.2.4.5 LOCA Loads on the Downcomers Fiqure 4-59 shows the LOCA loading history .for the SSES downcomers. The downcomer clearing load is a lateral load applied at the downcomer exit {in the same manner as the chugging lateral load) plus a vertical thrust load. Poolswell drag and fallback loads are very minor since the downcomer surfaces are oriented parallel to the pool swell and fallback velocities. The poolswell air bubble load is applied to the submerged portion of the downcomer as are the chugging and condensation oscillation loads.

4.2.4.6 LOCA Loads on the Downcomer Bracing Figure 4-60 shows the LOCA loadinq history for the SSES downcomer bracinq system. This system is not subject to impact loads since it is submerged at elevation 668's a submerged structure is subject to poolswell drag, fallback and air bubble loads. it Condensation oscillations and chugging at the vent exit will also load the bracing system both through downcomer reaction {indirect load) and directly through the hydrodynamic loading in the suppression pool. I

4. 2.4 7 LOCA Loads on Metwell Piping Figure 4-61 shows the LOCA loading history for piping in the SSES wetwell. Since the wetwell piping occurs at a variety of elevations in the SSES wetwell, sections may be completely submerqed, partially submerged, or initially uncovered. piping may occur parallel to poolswell and fallback velocities as with the main steam safety relief piping. For, these reasons there are a number of potential loadinq situations which arise as shown in Table 4-21..Zn addition, the poolswell air bubble load applies to,the submerged portion of the wetwell piping as do the condensation oscillation and chugging loads.

'-14

4 3 ANNULUS PRESSURIZATION The RPV shield annulus has the recirculation pumps suction lines passing through it {for location in containment see Figure 1-1).

The mass and energy release rates from-a postulated recirculation line break constitute the most severe transient in the reactor

, shield annulus. Therefore, this pipe break is selected for analyzing loading of the shield wall and the reactor pressure vessel support skirt for pipe breaks inside the annulus The reactor shield annulus differential pressure analysis and analytical techniques are presented in Appendices 6A and 6B of the SSES Final Safety Analysis Report (FSAB) 4-15

17.71 ft 0.883 sec 15

'2 J 0:

w CO

~

10 Z

0 I L,

5 I K Cg fL w D X T30766.1 10477 0.0 0.25 0.50. 0.75 1.0 TIME AFTER'VENT CLEARING (SEC)

SUSQUEHANNA STEAN QLIKCTRIC STATION UNITS 1 ANQ 2 DESIGN ASSESSMENT REPORT SSES SHORT TERM SUPPRESSION POOL SURFACE HEIGHT FIGURE 4-38

WETWELL PRESSURE (PSIA)

O O O

O o

0Q QO V Ol 4 CP rm

(

m z oUl 0

III R

Q O

Q o

C 0 m Q Ill P m X o tO (0 M Q Z CO D

z Z 4l fll pC> CO

~

A 0

K I

I O m~

CO m o Q

Co CO

~ o m CO o o4 O fll H ZJ gm z+amcZ (DRYWELL WETWELL) QP (PSID)

C Xf rv fl tll m 0 O CO.

O z

f P

~y h

'1 l

f, H

k '

DOWNCOME R B.O. VENT PIPE E L. 660'-0" PEDESTAL " i DIAPHRAGM SLAB SUPPORT COLUMN 12'4" EL. 648'-0<'ASEMAT SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2:.

DESIGN ASSESSMENT REPORT BASEMAT SSES WATER CLEARING JET FIGURE 441

t II A

A -, I 1

W A

~"

tt

/-

/

I

/ I j l

I I

//

COLUMNS

//

/8 X /

/ ql I

/

/ /

t /

(/

/ j l

/

/ h l z ~ ~

~

f JET IMPINGEMENT AREA (22.73 SQ. FT./VENT)

++I 4P (FROM DOWNCOMER WATER CLEARING) 4 ~

v e ~ ~

~ g 4

~

~

~ e.O ~

4

~

~

~y

~ 5 ~

g I

~ I CONTAINMENTWALL

~ ~

I

~ ~

~ r ~ ~ ~

SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT SSES JET IMPINGEMENT AREA (WATER CLEARING)

FIGURE 442

~I P.

I A

A l'

'I

\'

~ \ C

~ r E

CONTAINMEAT COLUMN PEDESTAL WALL 0.0 PSIG 0.0 PSIG EL. 672'-0" PB (t)+ PB (t) +

HYDRO. HYDRO. EL. 660'0" STATIC STATIC EL.648 0 PB (t) + HYDROSTATIC PB (t) + HYDROSTATIC PB (t) + HYDROSTATIC BASEMAT SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT AIR BUBBLE PRESSURE ON SUPPRESSION POOL WALLS FIGURE 444

30 DRYWELL 20 Cll CO LQ sc WETWELL c

10

-1OO 10 102 103 , 104 105 TIME (seel (a) CONTAINMENTPRESSURE RESPONSE FOR INTERMEDIATE BREAK AREA 300 200 DRYWELL ss:

D I

IZ 100 I

1OO 101 10 103 1O4 105 TIME (sec)

(b) DRYWELL TEMPERATURE RESPONSE FOR INTERMEDIATE BREAK AREA SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT TYPICAL MARK II CONTAINMENT RESPONSE TO THE IBA FIGURE 4-51

CONTAINMENT WALL 0 o 0 K oo 0 000 0 000

~ 04 oo

~0: ac, A0 o 0 "oo oo 0

OR oy.

o 00 0 000~

DOWN COMERS COLUMNS, 0

0'Ie q o

o ~oo 0 oo oo ooo o WETWELL PIPING NOTE:

DOWNCOMER BRACING IS ONLY PARTIALLYSHOWN IN THE INTEREST OF CLARITY.

LETTERS INDICATE SRV QUENCHERS SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT SSES COMPONENTS AF-FECTEO BY LOCA LOADS FIGURE 4-S3

~

~ '\

~

r

~,

II N I

I B.O. SLAB

~ 1.: ~

~,

EL. 700'- 3" B.O. HYDROGEN VACUUM BREAKER RECOMBINER El'. 692'-1" E L. 691'-0" I I T.O. PLATFORM E L. 691'-0" MAXIMUMPOOL SWELL EL. 690'-0" MAXIMUMPOOL SWELL HEIGHT ~ 1.5 X MAX C~

VENT SUBMERGENCE

~ 18'-0'

~

HIGH WATER LEVEL EL. 672'.0"

'RACING NORM WATER I

671'-0" LEVEL'L.

EL. 668'-0" MAXIMUM VENT SUBMERGENCE 0ti B.O. VENT PIPE EL. 660'-0" SLAB WETWELL 12'IAPHRAGM SUPPORT COLUMN PIPING 0II 3t 6tt T.O. SLAB EL. 648'-0" SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT

. SSES.COMPONENTS AFFECTED BY LOCA LOADS FIGURE 4-54

TABLE 4-16 LOCA LOADS ASSOCIATED MITH POOLSMFLL Load Reference

1. Hetwell/Dr ywell Pressures Ref 7, Table IV-D-3; during Poolswell Ref 10, Subsec-t ion 4.,4. 1. 5

2. Poolswell Impact Loads Ref 10, Subsec-tion 4.4.6 ~

3. Poolswell Drag Loads Ref 10, Subsections 4 4 5 2, 4.4.7, 4 4 8 4.. Downcomer Clearing Loads Ref 10, Subsection

4. 3. 1, Reference 11, Subsection 3. 3. 1. 2

5. Downcomer Hater Jet Load Ref 10, Sub-section 4.4.5.1

6. Poolswell Air Bubble Load Ref 10, Sub-sect ion 4. 4 5. 3

7. Poolswell Fallback Load Ref 10, Sub-section 4.4.5.4

T A B.LE 4- 18 SSES PLANT UNIQUE POOLSWZLL CODE INPUT D'ATA Downcomer Area (each) 2.96 ft~

Suppression Pool Free Surface Area 5065 03 ft~

Maximum Downcnmer Submerqence 12.00 Downcomer Overall Loss Coefficient 2. 5 Number of Downcomers 87 Initial Metwell Pressure 15.45 psia Hetwell Free Ai Volume 149,000 f t~

Vent Clearing Time 0.6863 sec Pool Velocity at Vent Clearing 3. 0 ft/sec Initial Drywell Temperature 135oF Initial Drywell Relative Humidity 0 20

0 TABLE 4-19 INPUT DATA FOR SSES LOCA TRANSIENTS Drywell free air volume 239,600 fthm (includinq ventsj Metwell free air volume 149r000 ft>

Naximum downcomer submerqence 12.0 ft Downcomer flow area (total) 256.7 ft Downcomer loss coefficient 2.5

'Initial drywe11 pressure 15. 45 psia Initial wetwell pressure 15. V5 psi a Initial drywell humidity 205 Initial pool temperature 900P Estimated DBA break size 3 53 ft~

Number of vents 87 Initial mass of steam in vessel 24,500 1bm Initial mass of saturated water in 674,000 1bm vessel Minimum suppression pool mass 7.6x10~ ibm Initial vessel pressure 1, 055 psia Vessel 6 internals mass 2,940,300 ibm Vessel 6 internals overall heat 484.9 Btu/secoF transfer coefficient Vessel and internals specific heat 0 123 Dt u/1 bm F Initial control rod drive flow 10 83 ibm/sec Initia1 steam flow to'ain turbine 3931. 5 1bm/sec RCIC 6 HPCI (HPCS) flow initiation 489.5 in level, distance from vessel "0"

Tahle 4-19~Conti nuedg RCIC 6 HPCI (HPCS) flow shutoff 564. 0 in level (normal water level), distance from vessel "0<<

Rated RCIC flow rate to vessel 83.4 ibm/sec Rated HPCI (HPCS) flow rate to vessel 695 .ibm/sec RCIC shutoff pressure 165 psia HPCI (HPCS) shutoff pressure 165 psia Condensa te storage tank entha1py 48 Btu/ibm CRD enthalpy 48 Btu/ibm Initial power level 3. 23x10~ Btu/sec Peedwater enthalpy 78 Btu/ibm Cleanup system flow 36.94 ibm/sec Cleanup system return enthalpy 413. 2 Btu/ibm Initial vessel fluid enthalpy 573. 1 Btu/1 bm RHR heat exchanqer >>K~~ in pool 306 Btu/sec ~F cooling mode RHR heat exchanger steam flow in 25 lbs/sec condensing mode RHR heat exchanger flow in pool 1390 lbs/sec coolinq mode RHR heat exchanger outlet enthalpy 108 Btu/lhm in condensinq mode Service water temperature 90 ~F

CHAPTER 5 LOAD COMBINATIONS FOR STRUCTURES'IPING'ND EQUIPMENT TABLE OF CONTENTS 5 1 CONCRETE .CONTAINMENT AND REACTOR BUILDING LOAD COMBINATIONS 5 2 STRUCTURAL STEEL LOAD COMBINATIONS 5 3 LINER PLATE LOAD COMBINATIONS 5 4 DOMNCOMER LOAD COMBINATIONS PIPING'UENCHER'ND QUENCHER SUPPORT LOAD COMB'IN AT IONS 5.5.1 Load Considerations for Piping Inside the 'Dryvell 5.5.2 Load Considerations .for Piping Inside the Metwell 5.5.3 Quencher and Quencher Support Load Considerations

5. 5. 4 Load Considerations for Piping in the Reactor Building 5 ' NSSS LOAD COMBINATIONS 5 7 EQUIPMENT LOAD COMBINATIONS 5 8 FIGURES 5 9 TABLES 5-1

CHAPTER 5 F IGURES

~umber Title 5-1 Piping Stress Diagrams and Tables 5-2 Piping Stress Diagrams and Tables 5-3 Piping Stress Diagrams and Tables 5-4 Piping Stress Diagrams and Tables 5-2

CHAPTER 5 TABLES Title 5-1 Load Combinations for Containment and Reactor Buildinq Concrete Structures Considering Hydrodynamic Loads 5-2 Load Combinations and Allowable Stre'sses for Structural Steel Components 5-3 Load Combinations and Allowable Stresses for Downcomers 5-3

5 0 LQRU ~CQBINILTIO~S FOR STRUCTURES PIPING IND EQUIPNENT To verify the adequacy of mechanical and structural design, it is necessary first to define the load combinations to which structures, piping, and equipment may be subjected. In addition to the loads due to pressure, weight, thermal expansion, seismic, and fluid transients, hydrodynamic loads resulting from LOCA and SRV discharge are considered in the design of structures, piping, and equipment in the drywell and suppression pool. This chapter specifies how the LOCA and SRV discharge hydrodynamic loads will he combined with the other loading conditions. Zor the load combinations discussed in this chapter, seismic and hydrodynamic responses are combined hy the methods specified in Ref. 10 Subsection 5.2.2 and Ref 10 Section'.3.

5-4

5 3 LINER PLATE LOAD COMBINATIONS The liner plate and anchorage system are designed for the load combinatioms listed in Table 5-1 except that all load factors are taken as unity.

5-7

~

5. 4 DIIRHCO~NR LOAN COMBINATIONS I.oad combinations for the downcomers are given in Table load combinations are based on the load combinations given 5-3.'hese in Table 6-1 o f Be f 10.

5-8

5 5 PTPIN~G~UEQCH~E~ A~ND UgNCHER SUPPORT LOAD COMBINATIONS T.OCA loads considered on piping systems include poolswell impact loads, poolswell drag loads, downcomer water jet loads, poolswell air bubble loads,,fallback drag loads, condensation oscillation loads, chuqginq loads,'nd inertial loading due to acceleration of the containment structure produced by LOCA loads. Loads due to SRV discharge on piping systems include water clearing loads, air clearing .loads, fluid transient loads on SRV discharge piping, reaction forces at the quencher, and inertial loading due to the accleration of the containment structure produced by SRV discharge loads.

The load combinatioas and the acceptance criteria for piping systems are given in Tab1e 6-1 of Ref 10.

5 5.1 Load Considerations for Pi~in@ Inside the Drywell Piping systems inside %he drywell are subjected to inertial loadinq due to the acceleratioa of the .containment produced by LOCA and SRV discharge loads in the wetwell. 'he SRV discharge piping in the drywell is also subjected to fluid transient forces due to SRV discharqe.

5 5.2 Load Considerations for ~pi incn'nside the Metwell All piping in the wetwell is subject to the inertial loading due to LOCA and SRV discharge.

Drag and impact loads due to LOCA and SRV discharge on individual pipes in the wetwell depend on the physical location of the pipinq. Other SRV discharge and LOCA loads applicable to piping

-in the wetwell are discussed in the paragraphs that follow.

Piping systems located below the suppression chamber water level are shown on Figures 5- and 5-2. These lines are located 1

outside of the jet impingement cone of the downcomer. In addition to the inertial loads'hese piping systems are subject to air bubble loads, condensation oscillation ldads, and chugging loads due to LOCA and SRV operation. The SRV piping, quencher, and quencher support are also subject to fluid transient forces due to SRV discharge.

Piping systems within the poolswell volume are shown cn Figures 5-2, 5-3 and 5-4. All horizontal runs of these pipes are above the suppression chamber water level. The following loads, in addition to inertial loads, act on these systems:

a. 'The horizontal runs of pipe below elevation 690',

experience poolswell impact , poolswell drag, and fallback draq loads.

5-9

b. The vertical, portions of pipe in the water below elevation 690'- experience" poolswell drag and fallback drag loads.

~55.3 nenc~hr and quencher sn~ort Load considerations The quencher and quencher supports are subjected to the'ollowinq hydrodynamic loads in addition to the pressure, weight, thermal, and seismic loads:

a. Unbalanced loads on the quencher due to SRV water clearinq and air clearing transients, irregular condensation, and steady state blowdown

b. Drag loads due to SRV discharge and LOCA

c. SRV pipinq end loads

d. Inertial loading due to the acceleration of the containment produced by SRV discharge and LOCA.

5. 5.4 Load Considerations for Piping in the the Reactor Building The effects of the inertial loading due to acceleration of the containment produced by SRV discharqe and LOCA loads will be evaluated for this piping.

5-1 0

5 6 NSSS LOAD COMBINATIONS To be provided later.

5-11

5 7 EQUXPMZNT L'OAD CONBINATZONS

'1 Load combinations for safety-related equipment located within the reactor building and containment will be assessed and described in a revision to this 'Design Assessment Report (" Safety-related" is defined in Table 1.8-1 of the PSAR) .

5- 12

F I G. NO. LINE NO. QTY SYSTEM EL A 12"-GBC-101 M.S.R.V. DISCHARGE PIPING CS1'-0" R.F.C.M.E.L.

12"-G BC-101 10 S.J.B.D.P.N.G.K.A.H. CS1'4I" SI.EEVE PENETRATION

~ 4 . ~

I E L 704'-0" I '1. 0.

~ ~

I I

I EL 694'-0" TWO DIR HORZ 84 ANCHOR EL 694'4"

! TORSIONAL REST.

HEIGHT EL, 680' HIGH WATER LEVEL EL 672'4" s

0 ' a, L 668'-0" TWO DIR HORZ REST.

o.'AEL A

EL 649'4I" TWO DIR HORZ 64 TORSIONAL REST.

FIGURE A FIGURE B BOTTOM SUPPRESSION POOL EL 040'4I" SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT PIPING STRESS DIAG 1'A@8 AND TABLES FIGURE 6-3"

FIG. REST.

QTY LINE NO. SYSTEM EL A El. B El. C RAD Y DIM. X EL NO.

697'-0" 6"-GBB-120 RHR MS'-6 1/2" M6'4" 897'4" 12'4 3/6" 1S 6/8" 696'4" Q RPV RAD Y ELC 24 VERTICAL5 i AXIAL REST.

ELB

~

'ERTICAL REST.

POOLSWELL EL 690'4" EL A I

HIGH WATER EL 672'A" DIM. X EL 048'4I" a%

~ ~

4 a SUSQUEHANNA STEAM ELECTRIC STATION UNITt 1 AND 2 DESIGN ASSESSMENT REPORT=

IMPING STRESS DIAGRAMS AND TABLES'IGURE

$4

TABLE 5-1 LOAD COMBINATIONS FOR COHTAINMEHT AND REACTOR BUILDIHG CONCRETE STRUCTURES {CONSIDERING Load. Single Equation Condition D L P 0

T 0

R 0 SS p

A R

V SRV<>> AOT ADS ASYM Valve LOCA~>>

Normal 1

w/o Temp 141710 1 5 X(tl Normal 2

w/Temp. 1013101010 1 3 3 Normal Sev. Env. 1 0 1 0 1 0 1 0 1 0 1 25 1 25 4 Abnormal 1.0 1.0 1. 25 1 0 1 0 1 25 X X 4a Abnormal 1.0 1.0 1251010 1 0 5 Abnormal Sev. Env, 1.0 1.0 1010 X I Sa Abnormal "Sev Env 1010 1 1 1 0 1 0 1 0 Normal 6

Ext. Env. 1010101010 1 0 1 0 7 Abnormal Ext. Env. 1010 0 1 0 1.0 1.0 1.0 1.0 X I 7a Abnormal Ext. Env. 1010 1 0 1010101010 Load Descri tion D -" Dead Loads E 0

= Operating-Basis Earthquake L . = Live Loads ESSS

= Safe Shutdown Earthguake PB = SBA or IBA (LOCA) Pressure Load T = Operating Temperature Loads P = DBA (LOCA) Pressure Load 0 A R = Operating Pipe Reactions T = Pipe Break Temperature Load 0 A P. = Operating Pressure Loads R = Pipe Break Temperatures Reaction Loads 0 A SRV = Safety Relief Valve Loads R = Reaction and jet forces associated with the pipe break

~No e

1) X indicates applicability for the designated load combination.

2) por the coluans designated AOT, ADS, ASYH, and Single Valve, only one of the four possible colunns nay be included in the load conbination for any one equation. For exanple, in equation 1 either AOT or ASYN nay be considered with the other loads but not both AOT and ASY5 sinultaneously.

3) LOCA chugging and condensation oscillation loads will be included in a subsequent revision to this table

TABLE 5-2 LOAD COMBINATIONS AND ALLOWABLE STRESSES FOR STEEL STRUCTURAL COMPONENTS {Suppression Chamber Columns, Douncoaer B~racin ~and Reactor Building Structural Steel)

Stress

~cCuation . Condition Load Combination ,'Limit Normal D+L+SRV F w/o Temp.

Normal D+L+T +SRV 0

w/Temp Normal/ D+L+T +E+SRV 0

1 5 F S

Severe Normal/ D+L+T +E'+SRV 0

1.5 F S

Extreme Abnormal D+L+P+ (T +T ) +R (Note 1)

+SRV 6 Ab no r mal/ D+L+P+ (T +T ) +R+E (Note 1)

Severe +SRV Abnormal/ D+L+P+ (To +Ta) + R+ E (Note 1)

Extreme +SRV Note 1: In no case shall the allowable stress exceed 0.90F in bendinq, 0.85F in axial tension or compression, 5nd 0.50F> in shear. Where the design is governed by requirements of stability (local or lateral buckling),

the actual stress shall not exceed 1. 5FS.

TABLE 5-2~Continuedg Notations:

Fq = Allowable stress according to the AISC, "Specification for the Design, Fabrication, and Erection of Structural Steel for Buildings", dated 1969, Part 1 0 Dead load Live load Tp Thermal effects during normal operating conditions including temperature gradients and equipment and pipe reactions.

Ta Added thermal effects (over and above op'crating thermal effects) which occur during a design accident.

Design Basis Accident pressure load Local force or pressure on structure due to postulated pipe rupture including the effects of steam/water get impingement, pipe .whip, 'and pipe reaction.

E Load due to Operating Basis Earthquake.

Load due to Safe Shutdown Earthquake.

SHV Safety relief valve loads.

Fy Ninumum specified yield stength

TABLE 5-3, LOAD COMBINATIONS AND ALLOMABLE STRESSES FOR DOMNCOMERS Primary Stress

~Euatgon Condition Load Combination Limit Upset D+P +SRV 0 ALL 1SS m Emerqency D+.P 0

SR V ALL+E 2.25 S Emergency D'PSBA 'SR ADS'E'BA 2 25 S m

Faulted D ~P 0

+S RV ALL+E 3 S m

Faulted D+PIBA +SRVADS+E+ m IEA Faulted 0+PS/A [or m PjBA)

(or BA) 7 Faulted D+PA+E'+DBA 1

3 S Fa ul ted D+P A+ E' DBA 2

3 S Notat ion s:

S m

Maximum allowable stress according to Table I-10. 1, Ref 28.

D Dead weight of the downcomer P

0 Pressure differential between drywell and suppression chamber during normal operating, condition.

SBA Pressure differential between drywell and suppression chamber durinq SBA.

IBA Pressure differential between drywell and suppression chamber durinq IBA.

'A Pressure differential between drywell and suppression chamber during DBA.

SRV Dynamic lateral pressure and inertia load due to ALL the discharge of all 16 sa fety relief valves sequentially.

SRV Dynamic lateral pressure and inertia load due to ADS the discharge of all 6 ADS safety relief valves simultaneously.

Load due to Operating Basis Earthquake

El Load due to Sa e Shutdown Earthquake SBA Chuqqing loads due to SBA as follows:

. 1. Horizontal load at bottom of downcomer, and

2. Horizontal and vertical inertial loads.

ZBA Chuqqing loads due to IBA as follows:

1. Horizontal load at bottom of downcomer, and

2. Horizontal and vertical inertial loads.

DBA 1

Vertical loads due to:

/

1. Viscous and pressure forces exerted by the flowing steam, and

2. Inertial load due to DBA DEA 2

Chuqqing loads due to DBA as follows:

1. Horizontal load at 'bottom of downcomer. and

2. Horizontal and vertical inertial loads.

CHAPTER 6 DESIGN CAPABILITY 'ASSESSMENT CRITERIA TABLE OF CONTENTS 6 1 CONCRETE CONTAINMENT AND REACTOR BUILDING CAPABILITY ASSESSMENT CRITERIA 6.1 Containment Structure Capability Assessment Criteria 6.1.2 Reactor Building Capability Assessment Criteria.

6 2 STRUCTURAL STEEL CAPABILITY ASSESSMENT CRITERIA 6 3 LINER P'LATE CAPABILITY ASSESSMENT CRITERIA 6 ' DOWNCOMER CAPABILITY ASSESSMENT CRIT ERI A 6 5 PIPE NGi QUENCHER AND QUENCHER SUPPORT CAPABILITY ASSESSMENT CRITERIA 6 6 NSSS CAPABILITY ASSESSMENT CRITERIA 6 7 EQUIPMENT CAPABILITY ASSESSMENT CRIT ERIA 6-1

6 0 DESIGN CAPABILITY ASSESSMENT CRITERIA The criteria by which the design capability is determined are discussed in. this chapter Design of the SSES is assessed as adequate when the design capability of the structures, piping, and equipment is greater than the loads (including LOCA and SRV discharge) to which the structures, piping, and eguipment are subjected. Loading combinations are discussed in Chapter 5. The margins by which design capabilities exceed these loadings are discussed in Chapter 7, Design Assessment.

6-2

6 3 LENE PLATE CAPABILITY ASSESSMENT CRITERIA The strains in the liner plate and anchorage system (welds and anchors) from self-limiting loads such as dead load, creep,.

shrinkage, and thermal effects are limited to the allowable values specified in Table CC-3720-1 of Ref 29, and the displacements of the liner anchorage are limited to the displacement values of Table CC-3730-'I of Ref 29.

Primary membrane stresses in the liner plate and anchorage system (welds and anchors) from mechanical loads such as SRV discharge and chugging are checked according to Subsection NE-3221. 1 of Ref 28. Primary plus secondary membrane plus bending stresses are checked according to Subsection NE-3222.2 of the same code.

Zatigue strength evaluation is based on Subsection NE-3222. 4.

Allowable design stress intensity values, design fatigue curves, and material properties used conform to Subsection NA, Appendix of Ref 28.

The capacity of the liner plate anchorage is limited by concrete pull-out to the service load allowables of concrete as specified in Ref 30.

6-5

The allowable stresses for the dovncomers are given in Table 5-3.

These allowable stresses are in accordance with Ref 28; Subsection NE. As permitted by, Subsection NE-1120 for MC components, the downcomers are analyzed in accordance vith Subsection NB-3650 of Ref 28; however, the lover allovable stresses, Sm~ from Table I-10. 1 for NC components are" used vhen performing the analysis.

6 5 PIPING QUENCHER AND QUENCHER SUPPORT CAPABILITY ASSESSNENT CRITERIA Piping in the containment and reactor building is analyze'd in accordance with Ref 28 Subsections NB3600, NC3600, and ND3600 for the loading described in Section 5.5.

The quencher is designed in accordance with Ref 28, Subsection NC3200,for loading discussed in Subsection 5.5.3. The, quencher support is designed in accordance with Subsection NF3000- of Ref 2 8.

6-7

To be provided later.

6-8

6 7 E UIPMENT CAPABILITY ASSESSMENT CRITERIA Assessment criteria for safety-related equipment subject to LOCA and SRV discharge loading which is located within the containment and reactor building will be described in a revision to this Design Assessment Report (" Safety-related" .is defined in Table 1.8-1 of the PSAH) 4

CHAPTER 7 DESIGN ASSESSMENT TABI E OP CONTENTS 7 1 ASSESSMENT METHODOLOGY 7.1. 1 Containment and Reactor Building Assessment Methodology 7 1 1 1 Containment Structure Assessment Methodology 7.1.1 2 Reactor Building Assessment Methodology 7.1 2 Structural Steel Assessment Methodology 7-1 21 Su ppression Chamber .Columns Assessment Me thodolog y 7 1.2 2 Dovncomer Bracing Assessment Methodlogy 7'. 1. 3 Liner Plate Assessment Methodlogy 7.1. 4 Downcomer Assessment Methodology 7 1.S' Piping and SRV Systems Assessment Methodology 1 6 NSSS Assessment Methodlogy 7 1 7 Eguipment Assessment Methodology 7 2 DESIGN CAPABILITY MARGINS 7 3 PIGURES 7-1

CHAPTER 7 PIGURES

'Nu

~be Title 7-1 . Geometry Plot of Containment Structure Model 7-2 Equivalent Model Damping Ratio vs. Model 'Frequency Structural Stiffness Proportional Damping 7-3 Finite Element Model of Column Liner Plate Loads - Normal Condition 1

7-4 7-5 Liner Plate Loads Abnormal Condition 7-6 Downcomer Analytical Model 7-2

7 0 DESIGN 'ASSESSNENT J.oads on SSES structures, piping, and equipment are defined in Chapter 4. The methods by which these loads are combined are discussed in Chapter 5 The criteria for establishing design capability are stated in Chapter 6.

This chapter describes the assessment of the adequacy. of the SSES desiqn by comparing design capabilities with the loadings to

. which structures, piping, and components are subjected and demonstrating the extent of the design margin. The first section

~

,of this chapter discusses the methodology by which design

'capability and loads are compared. The second section indicates

'the results of these comparisons.

7-3

7 1 ASSESSMENT METHODOLOGY 7.1 1 Containment and Reactor Buildin Assessment Methodolo~

'The dynamic analysis for the structural response of the contaiament and internal structures due to the SRV discharge loads and LOCA related loads is performed using the finite element method. The ANSYS finite element computer program is chosen for the transient dynamic analysis. Figure 7-1 shows the ANSYS finite element model.

Plat shell elements are used to model the reinforced-concrete containment structure and the reactor vessel. Pipe elements are used to model the columns supportinq the diaphragm slab. The

. soil structure interaction is taken into consideration by modelling'he soil using a series of discrete springs and dampers

.in three dir'ections as shown on Figure 7-1. These discrete sprinqs and dampers are specified based on the formulae for lumped parameter foundations found in Ref. 32.

The ANSYS program uses stiffness-proportional-damping, implying a structural damping matrix in the following form:

{C) = g {K) ~

where C Damping Matrix a sti ffness-proportional damping constant K Stiffness Matrix

, Fiqure 7-2 shows the equivalent modal damping ratio versus the modal frequency for structural stiffness-proportional-damping A value of g equaling 0 00063 is used in the ANSYS model which corresponds to a structural modal damping of approximately 4 percent of critical at 20 Hz.

Two computer programs have been developed, one as a preprocessor and the other as a postprocessor to the ANSYS computer program.

The preprocessor transforms the pressure forcing functions acting on the suppression pool walls, base mat, and pedestal into a

~

concentrated force actiag at the associated nodes of the ANSYS model. The postprocessor calculates the acceleration time history from the displacement time history obtained by ANSYS and scans for the maximum. displacements and accelerations.

Acceleration time histories, maximum structural displacements, accelerations, and broadened acceleration response spectra at selected nodes and directions are developed for the analysis of the piping, equipment, and NSSS systems. Response spectra curves are developed for all the previously mentioned SRV discharge and LOCA loads.

7-4

The response spectra are furnished for four different spectral damping values ie, 0.5,, 1, 2, and 5 percent of criti'cal. Each spectrum has been broadened to account for the uncertainties in the structural modeling techniques and material properties; All spectral accelerations are expressed in units of g (the gravitational constant).

Appendix B contains examples of the broadened response" spectra curves developed for the different loading cases of SRV discharge and LOCA related loads. (The pressure time history sho~n on 4-29 is used as the basis for the examples given ) 'iqure The ANSYS program (stress pass) is also used to compute the forces and moments due to the SRV discharge and LOCA related loads. These forces and moments are then combined with the nonhydrodynamic loads in accordance with Table 5-1. Material

. stresses at the critical design sections in the primary containme'nt and internal concrete structures are analyied using

. the CECAP computer proqram (Refer to Appendix A to FSAR Section 3.Q). Concrete cracking is considered in the analysis of reinforced concrete sections.

The construction of the SSES reactor building is such that no direct coupling with the containment occurs. A 2 in. separation

)oint is kept between the containment structure and the reactor buildinq at all points where the two structures abut, except at the base slabs where a cold joint exists. This arrangement minimizes the transfer of any direct dynamic response to the reactor building from the containment, where the SRV discharge, and LOCA hydrodynamic loads originate.

The average horizontal and vertical base accelerations from the containment dynamic analysis are computed and used as input motions on the reactor building foundations. This results in two horizontal motions and one vertical motion. The input motions are used in the form of acceleration time histories at the base slab. Reactor buildinq seismic models (horizontal north-south and east-vest and vertical), as shown on PSAR Figures 3.7-9 throuqh 3.7-11 and explained in detail in Subsection 3. 7.2. 1b of the FSAR, are used in the structural response analysis due to SRV discharge and LOCA loads.

Appendix C provides eXamples of the broadened response spectra curves for the reactor building due to SRV discharge loads for the abnormal operating transient (AOT) case at selected locations. The pressure time history shown in Piqure 4-29 is used as the basis for the examples given). The response spectra curves are developed for use in the design of piping and NSSS systems. The response spectra are furnished for four different spectral dampinq values, ie, 0.5,,1, 2, and 5 percent of critical. Each spectrum has been broadened to account for the 7-5

uncertainties in the structural modelling techniques and material propeties. Al'l spectral accelerations are expressed in units of q (the gravitational constant) . The forces and moments due to SRV discharge and LOCA loads are combined with the non-hydrodynamic loads in accordance with Table 5-1.

7.-1 2 S ructural Steel Assessment Nethodolo~

4 7 1.2 1 Su ression Chamber Columns Assessment Nethodol~o The assessment methods used for non-hydrodynamic loads such as dead, live, pressure, temperature, seismic, and pipe, rupture loads are described in the FSAR, Section 3.8..3.4 5.

For the analysis of the columns for hydrodynamic loads, the AHSYS computer program is used A typical column is modelled as a

, fixed-ended beam as shown on Figure 7-3. The total length of the column is'ivided into beam finite elements )oined at node points. An effective water mass due to submerqence is considered. Dynamic horizontal forces are applied to the column at the node points below water level. Time-varying forces and moments in the column are calculated for each finite element.

These results are combined with those for non-hydrodynamic loads to determine the t'otal forces and moments in the column.

Axial loads are produced in the bracing due to lateral loading on the downcomers. See Subsection 7. 1.4 for a description of the analysis of the downcomers for lateral loads. To determine the maximum axial load in the bracing, lateral loads are assumed to occur on all downcomers within a 90 degree influence zone in

.ei.ther the radial or tanqential directions. Bracing for the 16 SRV discharge pipes is included with the downcomer bracing. A sliding support is provided at the connection of the bracing to the discharge pipe to allow the discharge pipe to move vertically without producing a reaction load on the bracing. Since these lateral loads on the downcomers due to seismic and hydrodynamic loads are randomly oriented, various combinations of load directions are considered in order to determine the maximum axial load in the bracinq.

Xn addition to the axial load, there are lateral pressures applied along the lengthof the bracing members due to direct hydrodynamic loading Since the bracing members are of varying lengths, several different lengths of bracing members are considered for the analysis. Stresses in the downcomer bracing due to equivalent static lateral pressures are calculated using classical beam theory equations. Stresses in the downcomer bracinq due to dynamic lateral pressures are calculated using the ANSYS computer program.. The total length of the bracing member is divided into beam finite elements joined at node points. An effective water mass due to submergence will be considered.

Dynamic lateral forces are applied to the bracing at the node

points., Time-varying forces and moments in the bracing member f

are cal cul a ted for each inite element. Maximum stresses are calculated from these results using classical beam theory equations.

7. 1. 3 Liner Plate Assessment~ethodolocCy PSAR Subsection 3.8.1 provides a description of the. liner plate and anchorage system for the containment.

The analysis of the liner plate and anchorages for non-hydrodynamic loads is in accordance with Ref 18 For the analysis of the liner plate and anchorages-for hydrodynamic suction pressure loads, the load on the liner is the net negative pressure load. The net negative pressure load equals the dynamic negative pressure viue to SRV actuation or LOCA chugging minus the static positive pressure due to hydrostatic pressure or LOCA.. Pigures 7-4 and 7-5 describe the loads on the base mat and suppression chamber eall liner plate for the normal and abno'rmal load combinations respectively.

Por the normal condition, the hydrostatic pressure on the base mat is 10.4 psi and the maximum negative pressure due to the actuation of all SRV's is 7.8 psi.

The distribution of these pressures on the suppression chamber wall is shown in Figure .7-4.

Por the abnormal condition, the total positive pressure on the basemat is 35.4 psi which consists of 10.4 psi from hydrostatic pressure plus 25.0 from LOCA (small or intermediate break accident) . The total maximum negative pressure on the base mat is 21.8 psi due to the asymmetric chugging load. The maximum negative pressures from SRV actuation and chugging are combined for conservatism. It is recognized that the probability of these too phenomena producing peak, negative pressures at the same time is very los The distribution of pressures on the suppression chamber wall is shown in Figure 7-5.

Since the negative pressure is more than balanced by the positive pressure, the liner- plate does not experience any net negative pressure. Therefore, there are no flexural stresses induced in the'iner plate.

7.1.4 Dovncomer Assessment Methodolo~

Stresses in the dovncomer pipes due to static loads, such as dead

@eight and pressure, are calculated using classical eguations.

Stresses in the'dosncomer pipes due to inertial loads caused by seismic and hydrodynamic loads are calculated using the response spectrum method. The ANSYS computer program is used to solve for 7>> 7

the mode shapes and frequencies of the dovncomers and the dovncomer bracing. A group of dovncomer pipes and bracing members is represented by a lumped mass model. The inertia effect of the water surrounding the submerged portion of the dovncomers is approximated by the addition of an effective vater mass. The mass of water inside the downcomers is included in the model for all dynamic loadings except LOCA. For the LOCA conditions, the water has been vented from the dovncomers and therefore it is not included in the model.

The ANSYS computer program is used to calculate the stresses in the downcomer pipes due to hydrodynamic lateral loads. A typical dovncomer pipe is modelled as shown in Figure 7-6. Point A at the top of the dovncomer is restqained to'epresent the fixity of the downcomer at the dryvell floor. Point B is laterally .

restrai'ned to represent th'e lateral support furnished by the downcomer bracing. The total length of the downcomer is divided into beam finite elements )oined at node points. Dynamic horizontal-forces are applied to the downcomer at the node points below water level. Time-varying forces and moments in the dovncomer are calculated for each finite element. Maximum.

stresses are calculated from these results using classical. beam theory equations.

The piping and SRV systems vill be analyzed for the loads discussed in Section 5.5 using Bechtel computer programs HE101 and ME632. These programs are described in FSAR Section 3.9-Static and dynamic analyses of the piping and SRV systems are performed as described in the paragraphs below.

Static analysis techniques are used to determine the stresses due to steady state loads and/or dynamic loads having equivalent static loads. The drag and impact loads are applied as equivalent static loads.

Response spectra at the piping anchors are obtained from the dynamic analysis of the containment subjected to LOCA and SRV loading. Piping systems are then analyzed for these response spectra following the method described in Ref 19.

Time history dynamic analysis of the SRV discharge piping subjected to fluid transient forces in the pipe due to relief valve opening is performed using Bechtel computer code HE632.

7 1. 6 SSS Assessment Nethodol~og To be provided later

~71.7. si mes m sesseemt Ne~hodolocpy Analysis methodologies for saf ety-related equipment within the containment and reactor building subject to LOCA and SBV discharge loading will be described in a revision to this DAR

(<<Safety related<< is defined in Table 1.8-1 of the FSAR) 7-9

7 2 DESIGN CAPABZLXTY MARGINS Stresses at the critical sections for each of the ab'ove structures, piping, and eguipment vill be evaluated for all the loadinq combinations presented in Chapter 5.

'The results of the structural assessment of the containment and submerged structures vill be summarized in Appendix A.-'Figure A-2 shows the design sections in the basemat, containment walls, reactor pedestal, and the diaphragm slab considered in the structural assessment) The tables of Appendix A at present give the calculated design margins for load combination Eguation 1 of Table 5-1 which applies to the previously mentioned structural>

components. Similar tables vill he included in a future revision of this report in order to present the full assessment of the design capability margin for all the other load combinations.

The reinforcinq steel and concrete guality control test results shov that material strengths are higher than the minimum specified values used in computing these margins. This conservatism, along with the overload factors in the load combinations given in Table 5-1 and the material understrength factors built into the allovable stress criteria, results in actual safety margins qreater that those given in the tables of App'endix A.

villresults The of the be summarized structural assessment of the reactor building in Appendix E.

The results of the analysis of the piping systems vill be summarized in Appendix F in the form of tables. tables vill provide the maximum stress for the critical These load combination, the allowable stress, and the design margin The results of the assessment of the Nuclear Steam Supply System (NSSS) vill be summarized in Appendix G.

The results of the Appendix H.

assessment of equipment vill he summarized in 7-10

NOTE:

RPV X - AXIS IS IN PLANT EW AND Y - AXIS IN PLANT NS DIRECTION RPV SHIELD CONTAINMENT RPV PEDESTAL SUSQUEHANNA STEAM ELECTRIC STATION

'UNITS 1 AND 2 DESIGN ASSESSMENT REPORT .

GEOMEYRY PLOT OF CONTAINMENTSTRUCTURE MODEL FIGURE 7-1

DRYWELL FL.

Og p.~ v

~

rri vr j5 s Q-DECK 3lt 42"P STEEL PIPE COL.

FINITE ELEMENTS L ~52'-3" NODE POINTS HIGH WATER LEVEL 24'-0" BASEMAT 4.j (V

MODEL dpiI g 0

io 4 iC SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT FINITE ELEMENT MODEL OF COLUMN FIGURE 7-3

PEDESTAL CONTAINMENT WALL HYDROSTATIC 24'10.4 psi

+10.4 psi.

BASE MAT SR VIII 18'7.8 psi

-7.8 ps<

TOTAL 18'2.6 psi

+2.6 psi SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT LINER PLATE PRESSURES NORMAL CONDITION FIGURE 7C

POSITIVE PEDESTAL CONTAINMENT WALL HYDROSTATIC

+10.4 PSI

+10.4 PS I BASE MAT WETWE LL PRESSURE DUE TO SBA OR IBAD

+25 PSI NOTE:

WETWELL PRESSURE DUE TO DBA= 34 PSI +25PSI TOTAL

+35,4 PSI

+35.4 PSI SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT LINER PLATE PRESSURES ABNORMALCONDITION FIGURE 7-5 SHEET 1 OF 3

NOTE RPV X ~ AXIS IS IN PLANT EW AND Y - AXIS IN PLANT NS DIRECTION RPV SHIELD CONTAINMENT RPV PEDESTAL SUSQUEHANNA STEAM EI.ECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT 3-D CONTAINMENT FINITE ELEMENT MODEL (ANSYS MODEL) fIGURE 7

DAMPING RATIO P~0.00063 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 FREQUENCY

]0 20 40 SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT EQUIVALENTMODAL DAMPING RATIO VS MODAL FREQ. FOR STRUCTURAL STIFFNESS PROPORTIONAL DAMPING FIGURE 7-2

NEGATIVE PEDESTAL CONTAINMENT WALL

<<VADS

-7.8 PSI

-7.8 PSI BASE MAT 12'HUGGINGASyM

-14 PSI

-14 PSI TOTAL

-19,2 PSI

-21.8 PSI

-21.8 PS I SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT LINER PLATE PRESSURES ABNORMALCONDITION FIGURE 7-5 'SHEET 2 OF 3

PEDESTAL CONTAINMENT WALL

+11 PSI

+13.6 PSI BASE MAT TOTAL = POSITIVE + NEGATIVE SUSQUEHANNA STEAM ELECTRIC STATION

'NITS 1 AND 2 DESIGN ASSESSMENT REPORT LINER PLATE PRESSURES ABNORMALCONDITION FIGURE 7-5 SHEET3OF3

SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT MNINCOMER ANALYTICALMODEL FIGURE 7%

CHAPTER 9 RESPONSES TO NRC QU ESTIONS TABLE OF CONTENTS 9 1 ,IDENTIFICATION OP QUESTIONS UNIQUE TO SSES 4

9 2 QUESTIONS UNIQUE TO SSES AND RESPONSES THERETO, 9.3 FIGURES

This chapter vill provide responses to those Nuclear Regulatory Commission (HRC) questions whichhave been designated by Ref 10 (as amended) to be found in the plant-unique Design Assessment Report and to those questions for vhich the response in Ref 10 is inapplicable. The NRC questions for vhich responses vill be provided are j.dentified in Section 9.1, and detailed responses to the questions are found in Section -9.2.

9-2

1 IDENTIFICATION ~OF UESTIONS UNIQUE TO SSES The below listed questions address concerns unique to SSES.

These questions are answered in detail in Section 9.2.

NRC~ueetiau Number Question Topic M020 26 Primary and Secondary LOCA Loads N020.27 Inventory Effects on Blowdown N020.44 Poolswell Waves and SeismiC Slosh N020 55 SRV Loads on Submerged Structures M020 58 (1) i (2), (3) Plant Unique Poolswell Calculations M020. 59 (1), (3), (4) Downcomer Lateral Braces MP20. 60 Wetwell Pressure History M020 61 Poolswell Inside Pedestal M130 1 Pressure Loading Due to SRV Discharge M130 2 Load Combination History M 130 4 Soil Nodellinq N130 5 Liner and Anchorage Nathematical Model N 130. 6 Containment Structural Model-Asymmetric Loads N130. 12 SRV Structural Response 9-3

~9 2 UESTIONS UNIQUE TO SSES AND RESPONSES THERETO

~UESTION M020 26 The DFFR presents a description of a number of LOCA related hydrodynamic loads without differentiating betveen primary and secondary loads. Provide this differentiation betveen the primary and secondary LOCA-related hydrodynamic loads". We recognize that this differentiation may va'ry from plant to plant.

We vould designate as a primary load any load that has or result in a design modification in any Nark IX containment since vill the pool dynamic concerns were identified in our April 1975 generic letters.

~ESPONSE NO20 26 The table below shows the LOCA-related hydrodynamic loads on the SSES con'tainment. Those loads which have resulted in containment desiqn modifications are designated as "Primary Loads " These primary loads result from the poolswell transient.

Dryvell floor uplift pressures during the wetwell compression phase of poolswell lead to the decision to increase the SSES drywell floor design safety margin for uplift pressures by relocating drywell floor shear ties.

Poolsvell impact, drag, and fallback loads resulted in the relocation of equipment in the SSES wetwell to a position above the peak poolsvell height. Furthermore, the downcomer bracing system vas redesigned.

All other LOCA-related hydrodynamic 1oads are designated as "Secondary Loads" since no design modification has resulted from their presence.

LOCA Load "Primarv Load>> Secondary Load'R

1. Wetwell/Drywell Pressures XC1)

(During Poolswell)

2. Poolswell Impact Load x<>>

3. Poolsvell Draq Load x~>>

4. Downcomer Clearing Load

5. Downcomer Jet Load

6. Poolsvell Air Bubble Load

7. Poolswell Fallback .Load xc+ >

8. Mixed Flov Condensation Oscillation Load 9-4

9. Pure Steam Condensation Oscillation Load

10. Chugging.

11. Wetwell/Drywell Pressure and Temperature during DBA LOCA (Long Term)

12. Wetwell/Drywell Pressure and Temperature during XBA LOCA (Lonq Term)

13. Wetwell/Drywell Pressure and Temperature during SBA LOCA (Lonq Term)

Footnote's:

(3,) Shear ties changed in drywell floor.

(2) Equipment moved in wetwell.

(3) Equipment moved in wetwell. Bracing system redesign.

(4) Equipment moved in wetwell.

QUESTION M020 27 The calculated drywell pressure transient typically assumes that the mass flow rate from the recirculation system or steamline is equal to the steady-state critical flow rate based on the critical flow area of the jet pump nozzle or steamline orifice.

However, for approximately the first second after the break opening, the rate of mass flow from the break will be greater than the steady-state value. It has been estimated that for a Nark I containment this effect results in a temporary increase in the drywell pressurization rate of about 20 percent above the value based solely on the steady-state critical flow rate. The drywell pressure transient used for the LOCA pool dynamic load evaluation, for each Nark II plant, should include this initially higher blowdown rate due to the additional fluid inventory in the recirculation line.

RESPONSE M020 27 The drywell pressure transients have been recalculated by GE (Ref

7) with the additional blowdown flow rate produced by the inventory effects included in the analysis. The LOCA loads presented in Section 4. 2 have been calculated using these recalculated drywell pressure transients. Specifically, the drywell pressure transient resulting from the DBA LOCA including the effects of pipe inventory has been used as input to the poolswell model.

UESTI N 020 44 Table 5- 1 and Figures 5-1 through 5- 16 in the DFPR provide a listing of the loads and the load combinations to be included in the assessment of specific Mark II plants. This table and these figures do'not include loads resulting from pool swell waves followinq the pool swell process or seismic slosh. Me require that an evaluation of these loads be provided for the Mark containment design.

II RESPONSE M020 44 This information will be supplied in a subsequent revision to this DAR gUESTZON N020.55 The computational method described in DFFR Section 3 4 'for calculating SRV loads on submerged structures is not acceptable.

It is our position that the Mark II containment applications should commit to one of the following two approaches:

(1) Design the submerged structures for the full SRV pressure loads acting on one side of the structures; the pressure attenuation law described in Section 3.4.1 of NEDO-21061 for the ramshead and Section A10.3."1 of NEDO-11314-08 for the quencher can be applied for calculating the pressure loads.

(2) Follow the resolution of GESSAH-238 NJ on this issue.

The applicant for GESSAR-238 NI has proposed a method presented in the GE report, "Unsteady Drag on Submerged Structures," which is attached to the letter dated March 24, 1976 from G.L. Gyorey to R.L. Tedesco. This report is actively, under rev,iew.

RESPONSE M020 55 Loads on submerged structures due to SRV actuation. are discussed in Subsection 4.1.3.7.

~OESTZON M020.58 Relatinq to the pool. swell calculations, we require the following information for each Mark II plant:

(1) Provide a description of and justify all deviations from the DPPR pool swell model. Identify the party responsible for conducting the pool swell calculations (ie, GE or the AGE). Provide the program input and results of bench mark calculations to qualify the pool swell computer proqra,m.

9-6

(2) Provide the pool swell model input including ail initial and boundary conditions. Show that the mbd01 inject represents conservative values with resPect to obtaining maximum pool swell loads. In the case of calculated input, (ie, drywell pressure response, vent clearing time), the calculational methods should. be described and justified. Xn addition, the party responsible for the calculation (ie, GE or the AGE) should be identified.

(3) Pool swell calculations should be conducted for each Nark II plant The following pool swell results should be provided in graphic form for each plant,:

(a) Pool surface position versus time (b) Pool surface velocity versus time (c) Pool surface velocity versus position (d) Pressure of the suppression pool air slug and the wetwell air versus time.

RESPONSE N020 58 A specific response to this question can be found in Subsection 4.2.1.1. Verification of the SSES poolswell model is provided in Appendix Section D.l (2) Input and discussion of the poolswell mod@1 input can be found in Tables 4-17, 4-18, and Section 4. 2. 1. 1.

(3) The requested graphic results of the SSES poolswell calculation can be found in Figures 4-38, 4-39, 4-40, and 4-43.

QUESTION M020 59 Xn the 4T test report NEDE-13442P-01 Section 3.3 the statement is made that for the various Nark II plants a wide diversity exists in the type and location of lateral bracing between downcomers and that the bracing in the 4T tests was designed to minimize the interference with upward flow. Provide the following information for each Nark II plant:

A description of the downcomer lateral bracing system.

This description should include the bracing dimensions, method of attachment to the downcomers and walls, elevation and location relative to the pool surface. A sketch of the bracing system should be provided.

(3) The basis for calculating the impact or drag load on the bracing system or downcomer flanges. Thd magnitude and duration of impact or drag forces on the bracing system or downcomer flanges should also be provided.

9-7

(4) An assessment of the effect of downcomer flanges on vent lateral loads.

RESPONSE N020.59

.A downcomer bracing system is furnished to resist lateral loads on the downcomers. The original downcomer bracing was designed to resist seismic inertia loads. A revised downcomer bracing system has been designed to

.resist hydrodynamic loads as well as seismic inertia loads. The revised, bracing system consists of horizontal 6 in. diameter steel pipes spanning between the downcomers and embeds in the suppressi'on chamber wall or the RPV pedestal. The pattern of bracing members forms a horizontal truss as shown on Figure 9-1.

The bracing members are bolted or welded to the downcomers and embeds in the suppression chamber wall as shown on Figure 9-2. The bracing system is located 8 from the bottom end of the downcomer which is .3 ft below ft the normal water level.

(3) The basis for calculating the impact or drag loads on the downcomer bracing system (el. 668') and downcomer stiffener rinqs (el. 668'nd el. 682') is given in Section 4 2. The magnitude and duration of impact'r drag forces on the bracing system and downcomer stiffener rings is also qiven in Section 4.2 .

(4) This item is not applicable to the SSES design.

QUESTION M020 60 In the 4T test report NEDE-13442P-01 Section 5.4.3.2 the statement is made that an underpressure does occur with respect to the hydrostatic pressure prior to the chug. However, the pressurization of the air space above the pool is such that the overall pressure is still positive at all times during the chug.

Me require that each Nark,II plant provide sufficient information reqardinq the boundary underpressure, the hydrostatic pressure, the air space and the SRV load pressure to confirm this statement or alternatively provide a bounding calculation applicable to all Nark II plants.

RESPONSE N020. 60 This information -will be supplied in a subsequent revision to this DAB.

~UMNTZON M020.6 1 Siqniticant variations exist in the Nark II plants with regard to the desiqn of the wetwell structures in the region enclosed by the reactor pedestal These variations occur in the areas of {1) concrete backfill of the pedestal, {2) placement of downcomers, (3) wetwell air space volumes, and (4) location of the diaphragm 9-8

relative to the pool surface. In addition to variation between plants, for a given plant, variations exist in some of these areas within a given plant. As a result, for a given plant, significant differences in the pool swell phenomena can occur in these two region s. Me will. reguire that each plant pro vide a separate evaluation of pool swell phenomena and loads inside of the reactor pedestal.

RESPONSE N020. 61 The SSES pedestal and vetwell area is shown on Figures 1-1 and 9.3. Due to the absence of dovncomers in the pedestal interior, no pool swell would be expected in this region. There are 12 holes in the pedestal, hovever, eight of which would allow the flow of water from the suppression pool to the pedestal during a LOCA. Some dovncomers are near the pedestal flow holes, leading to the possibility that air could be blown through the pedestal holes, which would lead to a greater pedestal pool swell than vould be experienced by incompressible vater flov alone. One would expect the pedestal pool swell to be much reduced from the suppression pool swell due to its relative separation from the suppression pool and the lack of direct charging from dcwncomer vents. Indeed, l/13.3 scale model tests of the SSES pedestal desiqn conducted at the Stanford Research Institute under the sponsorship of EPRI show that the pedestal pool swell is less t'han 20 percent of the pool swell in the suppression pool (Ref 31) . There is no piping or equipment inside the SSES pedestal and, since the pedestal pool swell is very small, the only load involved due to pedestal pool swell vould be a small ~ P across the pedestal due to different water levels between the suppression pool and the pedestal. This load is considered in the design of the SSES pedestal 9~USTION N 130 1 Provide in Section 5 a description of. the pressure loadings on the containment wall, pedestal wall, base mat, and other structural elements in the suppression pool, due to the various combinations of SRV discharges, including the time function and profile for each combination. If this information is not generic, each affected utility should submit the information as described above.

RESPONSE N 130- 1 Chapter 4 describes the pressure loadings and time histories due to SRV discharge and 'other hydrodynamic loads.

~UESTIOH N130 2 In DFFR Section 5. 2 it is sta ted that the load combination histories are presented in the form of bar charts as shovn on Figures 5-1 through 5-16. It is not indicated hov these load 9-9

combination histories are used. In particular, it whether only loads represented by concurrent bars will be is not clear combined and properties of it the should be noted that depending on the dynamic structures and the rise time and durati6n of the loads, a structure may respond to two or more given loads at the same time even though these loads occur at. different times.

Also, although condensation oscillations are depicted as bars on the bar chart's, the procedure for the analysis of structures due to these loads has not been presented. Accordingly, the description of the method should include consideration "of such conditions. Also, for condensation oscillation loads and for SRV oscillatory loads, include low cycle fatigue analysis.

RESPONSE M130 2 The loads will be combined according to Tabl'es 5-1 and 5-2 of this DAR to assess the containment'tructural components Chapters 5.and 7 explain the load combination methods used in containment analysis. The structural analysis procedur'e to account for condensation oscillation load will be presented in a subsequent revision to this DAR.

QUESTION M130 4 Through the use of figures, describe in detail the soil modelling as indicated in DPFR Subsection 5.4.3 and describe the'olid f inite elements which you intend to use for the soil.

RESPONSE M 130 4 Soil modelling is explained in Subsection 7.1.1.1 and Ziqure 7-1.

QUESTZOE 5130.5 Describe the mathematical model which you will use for the 'liner and the anchorage system in the analysis as described in. DPFR Subsection 5.6.3.

RESPONSE M130 5 The mathematical model which will be used for analysis of the liner and the anchorage for hydrodynamic suction pressures is described in Subsection 7. 1.3 QUESTION 6130. 6 In DPPR Subsection 5.1.1.1 it was stated that the SRV discharge could cause axisymmetric or asymmetric loads on the containment.

In Subsection 5.4.1 an axisymmetric finite element computer program is recommended tor dynamic analysis of structures due to SRV loads, and no mention is made of the analysis for asymmetric loads. Describe the structural analysis procedure used to consider asymmetric pool dynamic loads on structures and through 9-10 0

the use of figures, describe in more detail the structural model which you intend to use.

RESPONSE 8130. t3 The dynamic analyses and models used are explained in Chapter 7 gUPSTZON N 130 12 Reference is made in DFFR Subsection 5. 4. 3 to studies of structural response to SRV load. Provide citations for this reference and where such studies are not readily available, copies are requested.

RESPONSE M130. 12 Studies mentioned in DFFR Subsection 5.4.3 are the results of analysis completed for a specific plant at the time of writing of the DFFR. Reference to the studies was intended to indicate the need for considering strain dependent soil properties. For. the SSES analysis, Ref 32 is used to determine the soil consta<Lts in the analysis 9-1 1

g0 CONTAINMENTWALL c.b + >i Sy

-i j+o 0 0 0 0

0 0 0 0 MSRV DISCHARGE PIPE (TYPICAL) 0 DOWNCOMER (TYPICAL)

COLUMN 0 0 (TYPICAL)

BRACING MEMBER (TYPICAL) rt Ie e

'y e 4 r .~ gt o4 ~

RPV PEDESTAL SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT DOWNCOMER BRACING SYSTEM FIGURE 9-'I

DIAPHRAGM SLAB DOWNCOME R

~ ~

o io'

~

O 3-1" 5 H.S. BOLTS 6" $ PIPE 1 1/4" Q 1/2" It TOP & BOTTOM.,

DOWNCOME R DETAIL 1

-.a EMBED/

HIGH WATER 6" j5 PIPE LEVEL g O,P EL 672'4" 4

rei4

>o 0

~

o BRACING 5

EL 668'4" DETAIL CONTAINMENT WALL BASEMAT b'+~ igO

)

r+

SUSQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT-DOWNCOME R BRACING DETAILS FIGURE 9-2

DOWNCOMERS

~

~

'I HIGH WATER LEVEL E L. 672'-0" PEDESTAL .

HOLES 12'USQUEHANNA STEAM ELECTRIC STATION UNITS 1 AND 2 DESIGN ASSESSMENT REPORT SPATIAL RELATIONSHIP OF DOWNCOMERS AND PEDESTAL HOLES FIGURE 9-3

10 0 REFERENCES Dr. N. Becker and Dr. E. Koch, <<KKB-Vent Clearing with the Perforated-Pipe Quencher" (translated by Ad-Exp Watertown, Nassachusetts),, KWU/E3-2796, Kraft wez'k Union, October 1973.

I 2\ Dr. M- Becker and Dr. E. Koch, "Construction and Design of the Relief System with Perforated-Pipe Quencher>>

(translated by Ad- Ex), E3/E2-2703, Kra f twerk Uniori, July 1 973.

3. Dr. N. Becker, <<Results of the Non-Nuclear Hot Tests with the Relief System in the Brunsbuttel Nuclear Power Plant" (translated by Ad-Ex), KWU/R113-3267, Kraftwerk Union,

, December 1974.

4. Dr. H. Weisshaupl, <<Formation and Osci1lations of a Spherical Gas Bubble Under Water<< (translated by Ad-Ex), AEG-Telefunken Report No. 2241, Kraftwerk Union, December 1972

5. Dr. H Weisshaupl and Schall, "Calculation Model to Pressure Oscillations in the Suppression Chamber Clarify'he After Vent Clearing>> (translated by Ad-Ex),

Report No. 2208, Kraftwerk Union, March 1972.

AEG-'elefunken

6. Dr. M. Becker, Feist and M. Burro, "Analysis of the Loads Measured on the Relief System During the Non-Nuclear Hot Test in KKB<<(translated by Ad-Ex), R 113/R 213/R 314/R 521-3346'raf twerk Unions April 1975-

7. Letter, J. W. Mi1lard to N. J. Lidl, "Susquehanna 1 6 2:

Mass and Energy Release for Suppression Pool Temperature Analysis during Safety Relief Valve and LOCA Transients," GB-77-65, March 14, 1977.

8. R. J. Ernst and M. G. Ward, "Nark II Pressure Suppression Containment Systems: An Analytical Model of the Pool Swell Phenomenon,<< NEDE-21544P, General Electric Co.,

December 1976.

9. Letter, F. C. Rally to Nark II Technical Steering Committee Members, "Pool Swell Nodel Test Cases," MKII-301-'E, August 22, .1977.

10. >>Dynamic Forcing Functions Information report (DFFR),<< Rev.

2, NED0-21061, General Electric Co. and Sarqent and Lundy Engineers, September 1976.

11. T. Y. Fukushima, et al, "Test Results Employed by GE for Containment and Vertica1 Vent Loads, <<NEDE-21078-P, BWR Table 3-4, General Electric Co., October 1975.

10- 1

12. F. J. Moody, Analytical Model for Liquid Jet Properties for Predicting Forces on Rigid Submerged Structures, AEDE-21472, General Electric Co., (to be published) .

13. R. J. Ernst, et al., Mark II Pressure Suppression Containment Systems: Loads on Submerged Structures An application Memorandum, NEDE-2)730, General Electric Co., (to be pub Lished) .

14. F. J..Moody, Analytical Model for Estimating Drag Forces on Rigid Submerged Structures Caused by LOCA and Safety Relief Valve Ramshead Air Discharges, NEDE-21471, General Electric Co., (to be published)

15. Mark II Phase I, and Report to 4T M.

Tests Applications Memorandum, Letter R 'utler (NRC) from J. F Quirk (GE),

June 14, 1976.

16. M. J. Bilanin, et. al., Mark II Lead Plant Topical Report:

Pool Boundary and Main Vent Chugging Loads Justification, NEDE-23617P, July 1977.

17. Warmeatlas (Heat Transfer Data), VDI (Society of German Engineers), Dusseldorf, 1974.

1'8. T. E. Johnson, et al., <<Containment Building Liner Plhtb Design Report,>> BC-TOP-1, Bechtel Corporation; San Francisco, December 1972.

19. <<Sei'smic Analysis of Piping Systems," BP-TOP"1, Rev 2, Bechtel Power Corporation, San Francisco, Januhry 1975.

20. Letter, J. R. Martin to Mark II Owners Group and TSC, MK XI-

,'50-E,

Subject:

Condensation Oscillation Excerpts to Applications Memorandum, July 1, 1977

21. D. Hof fman and E Schmid, >>Brunsbuttel Nuclear Power Plant List of Test Parameters and Most Important Measurement Results of the Non-Nuclear Hot Tests with the Pres ure Relief System<< (translated by Ad-Ex), R 52 1/40/77, Kraftwerk Union, August 1977

22. D Gobel, "Results of the Non-Nuclear Hot Tests with the Relief System in the Philippsburq Nuclear Power plant<<

(translated by Ad-Ex), R 142- 38/77, Kraftwerk Union, March 1977.

23. D. Hoffman and E. Schmid, >>Philippsburg I Nuclear Power Plant List of Test Parameters and Most Important Measurement Results of the Non-Nuclear Hot Tests with the Pressure Relief System" (translated by Ad-Ex), R 521/41/77, Kraftwerk Union, August 1977.

10- 2

24. Klans-D. Werner, >>Experimental Studies of Vent Cleaving in the Nodel Test Stand<<(translated by Ad-Ex) ', KWU/R 521-3129, Kraftwerk Union, July 1975.

25. D. Gobel, >>KKB Nuclear Start-Up Results of the Tests with the Pressure Relief System>> (translated by Ad-Ex), R 142-136/76, Kraftwerk Union, September 1976.

4

26. D. Hof fman and Dr. K Melchior, <<Condensation and Vent Clearing Tests in GKN with Perforated Pipes" (translated by Ad-Ex), KWU/E3-2594, Kraftwerk Union, Nay 1973.

27. GE Drawing 761E579, Bechtel No. 8856-N1-B11-89

28. ASNE Boiler 1, 1974 and Pressure Vessel Code, Section III, Division

29. ASNE Boiler 2, 1974 and Pressure Vessel Code, Section III, Division 30 ACI 318-71

31. R.L.Kiang and B.J. Grossi, <<Dynamic Nodelling of a Nark IZ Pressure Suppression System," EPRI-NP-441, Palo Alto~

April 1977.

32. "Seismic Analyses of Structures and Equipment for Nuclear Power P la n ts, <<BC- TOP-4A, Bech tel Power Corporation, November 1974 10- 3

APP ENDIX A CONTAINMENT DESIGN ASSESSMENT TABLE OP CONTENTS A 1 CONTAINMENT STRUCTURAL DESIGN ASSESSMENT A 2 CONTAINMENT SUBMERGED STRUCTURES DESIGN ASSESSMENT A 3 FIGURES

APP ENDIX A FIGURES Number. Title Concrete and Reinforcement Stress Elements A-2 Typical Section Shoving Section Location Reinforced Bar Arrangement Containment margins-Dr yvell Mall Containment margins Shield Mall and RPV Pedestal A-6 Containment Margins Metvell Mall Containment Nargins RPV Pedestal A-8 Containment Margins Base Slab Containment Nargins-Diaphragm-Slab A-2

APPENDIX A Containment Design Assessment This appendix indicates the containment elements and cross-sections where stresses are to be determined and contains a tabulation of. the predicted stresses, allowed stressed, phd design margins for each loading combination considered. The structural assessment of the containment is covered i>i Section A. 1; the submerged structures are assessed in Section A.2.

A 1 CONTAINMENT STRUCTURAL DESIGN ASSESSMENT Typical examples of this material are included in the de.port (Figures A-1 through A-9); a complete Section A.1 will be included in a future revision to this report A 2 CONTAIN'MENT SUBMERGED STRUCTURES DESIGN ASS ESSM ENT To be included ia a future revision to this report.

A-3

CECAP OUTPUT LOAD COMBINATION EQN. 1=1.4D+1.5 SRV(ASYM)

STRESSES IN KSI STRUC.

ANSYS INSIDE FACE REBAR4 OUTSIDE FACE REBAR 4 PRINCIPAL TURAL SECTION SHEAR ELEMENT CONC.

COMPO NUMBER TIES NUMBER STRESS NENT VERT. HOOP VERT. HOOP SPIRAL I SPIRAL 2 86 -0.017 -0.067 -0.123 0.123 0.027 -0.027 -0.233 -0.039 103 -0.099 -0.054 ~ 0.145 0.052 -0.018 ~ 0.076 4.103 -0.026 231 ~ 0.264 ~

0.017 ~ 0.373 0.080 ~ 0.126 ~

0.166 .4I.127 -0.063 0

C 311 -0.350 0.409 -0.480 0.618 0.097 0.044 -0.140 -0.076 m

m rn z

z z+

m pc Ch 315 ~ 0.636 0.570 -0.595 0.586 0.007 0.015 0.304 -0.090 cn~+

m m

~ ~

en~+

cn

<+m mz~ "ALLOWABLEREINFORCING STEEL STRESS = 54 KSI ZOm0

+~

m A

g7 co 0

z

CECAP OUTPUT LOAD'COMBINATIONEQN. 1=1.4D+1.5 SRV(ASYM)

STRESSES IN KSI ANSVS INSIOE FACE REBAR% OUTSIDE FACE REBAR% PRINCIPAL TURAL SECTION SHEAR ELEMENT CONC.

• NUMBER TIES NUMBER VERT. HOOP VERT. HOOP SPIRAL 1 SPIRAL 2 STRESS NENT 165 12 0 13 Z

4) Ch c z cD I- 362 m

rp m

p

~ m CO L m zZ Z

r2 Lc+

r<m >zen m+m CO Co K 15 pg m+m 2 Z "ALLOWABLEREINFORCING STEEL STRESS=54 KSI 0 ~0mA I m

mB 0 02 m cn CO

"-I O

Z

CECAP OUTPUT LOAD COMBINATIGN EQN. -1 ='1.4D+1.5 SFIV(ASYM)

STRESSES IN-KSI STRUC.

ANSyS INSIDE FACE REBAR" OUTSIDE FACE REBAR" SECTION PRINCIPAL ELEMENT SHEAR COMPO NUMBER TIES CONC.

-NUMBER NENT VERT. HOOP VERT, HOOP SPIRAL I SPIRAL 2 STRESS 441 -0.99 6.11 3.60 1.39 1.36 -0.080 ~ 0.145 455 -0.94 3.76 ~

0.95 2.52 0.99 0.59 -0.140 -0.140 473 ~ 0.92 2.91 ~

0.87 2.08 0.62 0.59 -0.078 ~

0.131 0

0 C 475 10 -1.23 6.10 ~

0.703 1.69 1.65 4.052 -0.191 m

0 m hl Z n zz I

z ac+

gzc, 495 -1.24 4.09 ~ 0.83 3.11 1.12 1.17 ~ 0.32 ~ 0.19 Ill Ill m-Im CO CO rZ I

I m>m ZZ

~0m I rraz ill 0

Z O XI O

"ALLOWABLEREINFORCING STEEL STRESS=54 KSI M

zO

CECAP OUTPUT LOAD COMBINATION EQN. 1=1.4D+1.5 SRV(ASYM)

STRESSES IN KSI STRUC- OUTSIDE FACE REBAR" ANSYS INSIDE FACE REBAR% PRINCIPAL TURAL SECTION SHEAR ELEMENT CONC.

COMPO. NUMBER TIES NUMBER VERT. HOOP VERT. HOOP SPIRAL 1 SPIRAL 2 STRESS NENT 484 16 1.71 0.472 -3.15 0.687 0.0 0.0 1.68 ~ 0.552 550 17 ~ 0.953 0.926 -1.78 2.57 0.0 0.0 0.160 -0.264 l-595 18 -1.12 0.306 -1.73 0.48 0.0 0.0 -0.030 ~ 0.257 cc:

606 19 -1.07 -0.038 -2.19 0.238 0.0 0.0 0.234 ~ 0.337 0

O C m m 0 cn X z a zz z 20 -1.28 -0.031 -2.17 0.196 0.0 0.0 0.281 -0.330 x

az cn +

Z c Cg Py ill fll mm "ALLOWABLEREINFORCING STEEL STRESS=54 KSI mz CO 0+

lll m~m Z Z g IOmA g 0 xl g7 m

~

Q 1l z 0 A O

Z

CECAP OUTPUT LOAD COMBINATION EQN. 1 =1.4D+1.5 SRV(ASYM)

STRESSES IN KSI STRUC.

ANSYS TOP FACE REBAR~~i BOTTOM FACE REBAR""" PRINCIPAL TURAL SECTION SHEAR ELEMENT CONC.

COMPO. NUMBER TIES NUMBER RADIAL STRESS NENT TANG ENTIAL RADIAL TANGENTIAL 551 30 8.51 2.10 2.43 0.92 0.501 -0.129 26 0.431 1.28 -0.26 -0.13 4.277 4.082 Q

29'.55 2.41 1.42 0.82 0.095 -0.075 c D m O 0 m m 0 M Z

710 27 0.554 0.680 3.11 1.96 0.200 -0.102

+'Z pc+

ez~

en+ m gm cn~ ~

r~

CO m+m z .727 5.89 0.54 0.02'I 0.481 4.105

~ Z 0 c m

m2 X~ O m

Q Z A CO i NORTH - SOUTH BARS 4%

EAST- WFST BARS

""~ ALLOWABLEREINFORCING STEEL STRESS=54 KSI 0

z

CECAP OUTPUT LOAD COMBINATION EQN. 1 =1.4D+1.5 SRV(ASYM1 STRESSES IN KSI STRUC.

ANSYS TOP FACE REBAR% BOTTOM FACE REBAR" PRINCIPAL TURAL SECTION SHEAR ELEMENT CONC.

COMPO. NUMBER TIES NUMBER RAOIAL STRESS NENT TANGENTIAL RADIAL TANGENTIAL 25 3.60 3.71 2.90 -0.321 -0.050 Cl 411 24 1.93 3.40 3.23 5.55 4.220 U

IT:

A K 440 21 3.33 1.55 2.93 1.69 4.350 4.048 C 0 CL C

ITI O O

o m cn UZ z z c. z+ 452 22 0.844 3.62 1.92 4.42 4.61 -0.073 xzK 33

~zen gWm

)O Ill G) Z Ch

<+m m z co I

E zom

+MO 470 23 0.392 3.87 2.18 4.56 4.031

~Q ITI O 'ALLOWABLEREINFORCING STEEL STRESS=54 KSI z 0 g7 CO I

0 z

APPENDIX D PROGRAM VERIPICATXON TABLE OP CONTENTS D 1 POOLSMELL 5ODEL VERIFICATION D 2 PIGURES D 3 TABLES

APPENDIX D

~FZ ORES

~uebeo D-1 Code Verification-Poolsvell Height for Class l'lant D-2 COde Verification-Poolssell Velocity for Class 1 Plant D-3 Code Verification-Poolssell Height for Class 2 Plant D-4 Code Verification-Poolsvell Velocity for Class 2 Plant D-5 Code Verification-Poolswell Height for Class 3 Plant Code Verification-Poolsvell Velocity for Class 3 Plant D-2

APP BNDIX D

~urbe ~T~te 4

D-1 Drywell Pressure Transients for the Test Cases 4

D-2 PLant Specific Parameters for the Test, Cases D-3 Comparison of Haximum Pool Smell Velocity for Classes, l, 2, and 3 Test Cases D-4 Common Assumptions for the Test Cases D-3

APPENDIX D ROG A VERI XCA ION The purpose of this appendix is to provide information which verifies the accuracy of the computer programs used in conjunction with A

SSES design assessment.

D OOLS LL- NOD L V I ICA ION This subsection demonstrates the accuracy of the SSES DAR poolswell model by comparing it with the model developed by the General Electric Company . The latter model has predicted conservatively the results of the 4T poolswell tests {Ref 8) .

To evaluate the agreement between the GE poolswell code and the poolswell code used for the SSES DAR, three test cases were selected. The test cases used were the Classes 1, 2, and 3 plants described in Ref 10 The input data for these three problems are given in Tables D-1 and D-2 (taken from Ref 9). (In the verification of the model, the boundary conditions assumed by GE in Ref 9 were used.. These assumptions are shown in Table D-4.), These data are representative of typical U.S. Nark II BWRs.

The poolswell code used in this DAR was revised until the results were in close agreement with GE's results as given in Ref 9.

Agreement was judged by examining the peak swell velocity predicted, since this is one of the most important pool swell parameters and one that is fairly sensitive to how the phenomenon modelled. The degree of agreement finally achieved between

's the poolswell 'code used in this DAR and the GE code is shown in Table D-3 where peak swell velocities are compared. Transient comparisons for Classes 1, 2, and 3 plants are shown on Figures D-1 through D-6 where the transient predictions of the two codes are shown to be essentially identical Prom the good agreement shown in the check cases, the poolswell code used in this DAR is verified to be the same as the GE code for evaluation of pool swell.

APP EN DIX E The results of analysis of the reactor building structure will be summarized in this appendix. This appendix will be provided in a future revision to this report

f, t