ML20235E598

Assessment of Shoreham Nuclear Power Station Secondary Containment Severe Accident Mitigation Capability, Final Ltr Rept

ML20235E598
Person / Time
Site:	Shoreham File:Long Island Lighting Company icon.png
Issue date:	06/26/1987
From:	Greene S OAK RIDGE NATIONAL LABORATORY
To:	NRC
Shared Package
ML20235E593	List: ML20235E590 ML20235E598
References
NUDOCS 8707130017
Download: ML20235E598 (72)

Text

AN ASSESSMENT OF THE SHOREHAM NUCLEAR POWER STATION'S SECONDARY j l

j CONTAINMENT SEVERE ACCIDENT MITIGATION CAPABILITY Sherrell R. Greene -

l 1

i 1

Boiling Water Reactor Severe Accident Technology Program Oak Ridge National Laboratory Oak Ridge, Tennessee Letter Report June 26, 1987

't e

Research sponsored by the U. S . Nuclear Regulatory Co mmi s-sion Office of Nuclear Regulat o ry Regulation under Int e r- l agency Agreement DOE 0554-0554-Al with the U.S. Department of Energy under contract DOE-AC05-840R21400 with the Martin Marietta Energy Systems , Inc.

j NOTICE This report was prepared as an account of work sponsored by an age ncy of the United States Gove rnme nt. Nei the r the United States Government nor any agency thereof, or any of their employees, makes any warranty, expressed or implied, y

or assumes any legal liability or responsibility for any j third party's use, or the results of such use, o r a ny in-formation, apparatus, product, or process disclosed in this '

report, or represents that its use by such third party would not inf ringe privately owned rights.

I i

8707130017 870630 2 DR ADOCK 0500 i

11 CONTENTS

.P,aggi,

1. INTRODUCTION ................................................. l'

2. ' SCOPE' AND LD3ITATIONS OF ASSESSMENT ......................... 2

3. UTILITY CLAIMS .............................................. 4

4.

SUMMARY

OF FINDINCS ......................................... 6-REFERENCES ...................................................... -8 Appendix A. FACTORS AFFECTING BWR SECONDARY CONTAINMENT PERFORMANCE .........................'............... '9 Appendix.B. COMPARISON '0F BROWNS FERRY, PE ACH BOTTOM,. AND SHOREHAM SECONDARY CONTAINMENT PERFORMANCE ......... 16' Appendix C. CRITIQUE OF FAI MODELING APPROACH FOR C9D, CADRF, AND CIA SEQUENCES ........................... 22 Appendix D. - ORNL SHOREHAM SECONDARY CONTAINMENT CALCULATIONS ... 29 Appendix E. ACRONYMS AND SYMBOLS ............................... L37 1

O

l

{

j

1. INTRODUCTION ]

i

]

The Long Island Lighting Company 's (LILCO)' Shoreham Nuclear L Power I Station was granted a Fa cility Operating License on July 3, 1985,.

authorizing operation and testing at or below 5 %' of. full rated power.

' Since ' that time . mos t licensing issues - (with the exception. of certain emergency planning issues) related to the granting of a full power li-cense have been resolved. In Ap ril, 1987, LILCO submitted a request to the Nuclear ' Regulatory Commission for authorization 'to increase Shore-ham's operating power level to .25% of full: rated power.1 . Se c tion -

50.47(c) : of the Code of Federal Regulations enables the NRC to grant ]

such'a request, providing the applicant demonstrates (a) the deficien-cies in the emergency plans are not significant,.or (b) that : adequate interim compensating actions have been or ' will be taken, or (c) that- '

there are other. compelling reasons to- permit plant operation.

LILC0 has prepared a probabilistic risk assessment for 25% power operation at Shoreham and submitted this PRA in support of. their re-quest. It is the utility's . contention- that "... . the risk and conse- l quences of accidents at 25% power are so greatly reduced that any remaining unresolved emergency planning issues become entirely insignif-icant".2 This low risk profile is achieved, in part, .due to the (claimed) ability of Shoreham's secondary containment (reactor building

. and refueling bay) to retain significant fractions-of the-fiscion pro-duct masses which would escape the primary ccatainment during severe accidents. i d

This report describes the results of analyses conducted at Oak .

Ridge National Laboratory to evaluate the LILCO's claims for Shoreham's secondary containment severe accident mitigation capability, and assess the degree to which these claims reflect the specific design character-istics of the Shoreham secondary containment. This assessment was con-ducted at the request of the Risk Applications Branch of the Nuclear Regulatory Commission Of fice of Nuclear Reactor Regulation.

Section 2 describes the scope and limitations . of the assessment.

LILCO's claims for Shoreham secondary containment severe accident miti-gation effectiveness are summarized in Sect. 3. .. Section 4 presents a summary of the findings of this assessment. The details of the analyses conducted as part of this evaluation are described in Appendices A through D. The definitions of acronyms and symbols f requently employed in this report are given in Appendix E. ,

O

1 2

2. SCOPE AND LIMITATIONS OF ASSESSKENT e- j 1

Due to the limited time available for. this assessment, the approach "

adopted for this- evaluation relies heavily on previous experience and insight gained from previous analyses of BWR secondary containment per- 1 f orma nce. 3 This exp eriential base was augmented by discussions ~ with ' I LILCO .and. its contractors, and a site visit in which the ' Shoreham secondary containment was thoroughly examined and extensive measurements were taken. This evaluation - consists 'of three separate, but; related, efforts.

j Task 1 consists of a review of those " key" BUR secondary ' contain-

-ment design characteristics which influence fission product and aerosol j decontamination f actors (defined as the ratio of the integrated fission <

product mass . which enters the secondary containment and the integrated mass of fission products which escape. the secondary containment), and. a comparison of Shoreham's key parameters to those. of Browns Ferry and Peach Bottom.-- plants for which detailed analyses have .been completed by ORNL. The goal of this analysis is an assessment of the reasonable-ness of the claimed secondaqr containment' ' decontamination f actors in light of previous ORNL studies. No attempt has ' or should be made to ,

determine unique sequence-dependent Shoreham ' seconda ry containme nt  !

decontamination factors (DFs) from this analysis. A s ummary of the. j Task 1 findings in presented in Sect. 4 and the details of the analysis i are described in Appendices A and B of this report. l J

, Task 2 is a review of the modeling approaches, assump tions, and in-put parameters employed by LILCO and its contractor, Fauske and Associ-ates, Inc. (FAI), for simula tion of the Shoreham secondary contain-ment.

The result of this task is an evaluation of the extent to which FAl's simulation approach is consistent with the "a s-bu ilt" Shoreham design. Inconsistencies are identified, and an indication of the quali-tative impact on calculated decontamination f actors is presented. The results of Task 2 evaluation are summarized in Se c t. 4 :and the details of the analysis are presented in Appendix C.

The final task (Task 3) in this assessment involves an independent evaluation of the probability of hydrogen deflagrations in the Shore-ham's secondary containment. A simplified multi-tell MELCOR4 model of the Shoreham secondary containment was constructed, and utilized in con-junction with Shoreham primary containment vent flow and blowdown sources (generated by S. A. Hodge with the.BWRSAR code) to evaluate this issue. The results of this analysis are summarized in Sect. 4 and the details of the analysis are presented in Appendix D.

The analyses presented in this report provide an assessment of the reasonableness of LILCO's claims for Shoreham seconda ry containme nt .

effectiveness. It must be emphasized, however, that independent verifi-

~ cation of Shoreham's secondary containment DFs (calculation of exact DFs ,

for a variety of accidents) would require significant efforts beyond those which were f easible during the brief period . available for'this li

3 work. Such an effort would begin with the development of a more de-tailed secondary . containment model than that utilized for the analysis presented in Appendix D. The resulting model would be employed for a variety of case studies in which the primary containment f ailure mode, location, and blowd own his tory are va ried to simulate the regimes expected. f rom a spectrum of risk-dominating accidents. This evaluation would include an assessment of the impact of reactor building standby ventilation system (RBSVS) and unit cooler operation on secondary con-tainment hydrogen deflagration probability and fission product retention capability.

1 1

4 l

1 e  !

i

)

l

]

4

3. UTILITY CLAIMS An asses sment of Shoreham's secondary containment decontamination factors was performed as part of the original Shoreham probabilistic risk assessment (PRA). That assessment, which utilized the MARCH / CORRAL and CONTEMPT code' series, yielded decontamination f actor estimates of 1 to 3 - for cesium and iodine.5-7 The secondary containment decontamina-tion factor reassessment conducted for the 25% power PRA employed the MAAP 3.0 code, and generated' DF estiaates in the range of 10 to 50 for the risk-dominating accident sequences.8 This section presents . a-summaqr of those accident sequences and the associated. secondary con-tainment DFs as presented in the.Shoreham 25% power PRA.

The 25% power PRA indicates that 25 detailed MAAP calculations were performed for a spectrum of- accidents including ATWS, station blackout, and large LOCA.8 The accident sequences were then grouped into six "re-lease categories", based on similarities. in the calculated fission ' pro-duct release histories. One " representative" accident sequence from each of the six release categories was then selected' for additional.

analysis. While DFs as high as 170 were calculated for some sequences '

the DFs calculated for three of the " representative" sequences range from 10 to 50. No secondary containment DF was claimed for cases C10E1, C3C, and C6A1. A brief description of the three sequences for which secondary containment DFs were claimed is given below.

Case C9D: Case C9D is a transient-initiated ATWS with MS1V closure e at time = 0. Reactor vessel water level is maintained at the top of the active fuel by various combinations of HPCI, RCIC, CRD, and LPCI,' while the SRVs vent steam to the pressure suppression pool. The primary containment is vented from the wetwell airspace when the prima ry containment pressure reaches 60 psig (due to evaporation from the pressure suppression pool). The vent' line is assumed to f ail at its junction with the RBSVS ducting (reactor building elevation 101 f t), and the resulting harsh reactor building environme ntal conditions are assumed to f ail all reactor vessel injection systems. Reactor vessel f ailure is calculated to occur at 10.4 h, at which time the drywell vent is also opened to maintain primary containment pressure. at or below 70 psig. Ten percent of the Cs1 inventory is released into the secondary containment, which has a calculated DF of approximately 10. The maximum secondary containment hydrogen concentration of 2 molar % occurs at -11 h into the accident. No hydrogen deflagrations are predicted.

Case CADRF: Case CADRF is a seismic initiated recirculation line LOCA, with a coincident drywell head failure of 3 ft 2. All reactor ves-sel injection systems are lost. . Only the refueling bay is credited for

- fission product removal,.and the RBSVS system is assumed to be unavail--

able. Ten percent of the Cs1 inventory is discharged to the refueling bay, which is calculated to have a DF of 10. The highest secondary con-

. tainment hydrogen concentration (5.6 molar %) is predicted to occur at 4.6 h into the accident, and no hydrogen deflagrations are predicted to o c cur.

1 5

-l Case CIA:. Case CIA is a station' blackout sequence (all AC power 'i

,. lost), combined with a stua open relief valve, and a f ailure to isolate the drywell equipment and floor drain lines. The RBSVS system is not available. HPCI and RCIC (both turbine driven)' are initially available, .

but HPCI is lost due to low HPCI turbine ateam . flow at: 8. 5 . min,' fol- l lowed by loss of RCIC at 45 min into the accident. . Reactor vessel '

i failure is calculated to occur at 13.9. h into the accident, by which time 30% of the Cs1 inventory has been deposited in 'the suppression pool. The . pool is- calculated to remain subcooled throughout the 50 h ' '

duration of the analysis. Approximately 2% of the.Cs1 is transported to i the. reactor' building via f ailure of ~ the equipment and: floor drain tank hardware in the basement of the reactor building. A secondary . con-tainment DF of 50 is calculated ' for this sequence. The secondary containment hydrogen concentration never ' exceeds 2 ' molar % and ~ no i hydrogen deflagrations are predicted.

A ' review of the detailed sequence descriptions reveals that,the basis for Shoreham's asserted low severe accident environmental source '

terms rests on three factors. - First, the presence of the , f our down-comers within the reactor pedestal region -is asserted to preclude any significant core / concrete reaction (due' to the assumed dumping of the. .

majority of the corium into the suppression pool). Although the quench- '

ing of this material does produce a prompt pressurization of the primary containment, the long term containment pressurization and- aerosol gen-eration f rom the core concrete . reaction does not occur as? it would-in a .

HK I facility. Thus, high primary containment pressures and suspended-aerosol (fission product) concentrations are avoided. . - Se condly , the-Shoreham emergency procedures call for.' venting f rom the wetwell air space when the primary containment pressure reaches 60 psig, followed by drywell venting if the pressure cannot be maintained below:70 psig. The' combined impact of the venting and the inpedestal downcomers is to pre-clude high drywell pressures which " pump" fission products into the secondary containment f ollowing primary containment f ailure. Finally ,

the absence of secondary containment hydrogen burns and incorporation of the resulting secondary containment DF further reduces the significance of any fission products which are released f rom the primary containment. i The purpose of this analysis is to evaluate,the third f actor noted above -

the secondary containment decontamination factor. The se-quences of interest for this assessment are, therefore, cases C9D, CADRF, and CIA (respective' Cs! and Cs0H DFs of 10, 10, and 50) . The following section presents a summary of results of the analyses con-ducted to evaluate these claimed secoMe g containment DFa. The details  ;

of each of tN analyses are presented in Appendices A through D.

e

...._..s- - - - - . . . . . .- - - . - - - -' "' ' - ' - - : " ' ' ' ' ' ' " ' " ' ' ' ' ' ' '

\

6 I i

I

4.

SUMMARY

OF FINDINGS t

l The major findings from the three analyses conducted during this investigation are summarized below. The details and supporting argu-

~

ments for the findings are contained in the Appendices to this report.

Additional findings which are judged to be of less significance than ,

those noted here are described in the summary section of each Appendix. j Secondary Containment Hydrogen Deflagrations Are Probable: The use l of "best-estimate" severe accident hydrogen generation sources results in secondary containment deflagrations for cases C9D and CADRF (the two j cases simulated for this analysis). The FAI analyses of these two I sequences f ailed to predict such burns - apparently due to MAAP's pre-diction of relatively low hydrogen generation rates during these acci-dents. The secondary containment decontamination f actors calculated by i FAI for cases C9D, CADRF, and (probably) CI A nust, therefore, be viewed as non-conservative estimates in the absence of detailed confirmatory calculations. The extent of the non-conservatism in each of the DF estimates cannot be assessed apriori, but ' could be significant. The l magnitude of this non-conservatism probably over-shadows that resulting i from other uncertainties described in Appendices B and C of this report.

RBSVS Ope ration Ca n Increase The Se ve rity Of Deflagrations and

. Reduce Se conda ry Co nt ainme nt DFs: Operation of the RBSVS can actually increase the severity of secondary containment hydrogen deflagrations by promoting a well mixed secondary containment atmosphere, resulting in a severe, global hydrogen deflagrations for cases in which at least 800 lb of hydrogen are available. RBSVS operation might also decrease the secondary containment DF for cases in which the primary containment f ails into the lower region of the reactor building, by actively trans-porting fission products from the lower regions of the building to the refueling bay (which would be the secondary containment f ailure location in most accidents).

The Shoreham Primary Containment Venting Procedure Would Result in I Immediate Failure of the Ref ueling Bay Siding: This f ailure is a result I of the inability of the RBSVS to maintain negative pressures in the I secondary containment f or source flows in excess of the RBSVS exhaust capacity (currently 1160 cfm). Since implementation of the venting procedure would result in f ailure of the primary containment vent duct and the injection of over 3000 cfm of steam and non-condensibles into the secondary containment atmosphere, failure of the refueling bay siding would immediately follow.

De cont ami na tion Fa ctors Fo r Cases C9D, CADRF, and CI A: While a variety of conservative and non-conservative modeling assumptions were made by FAI (described in Appendix C), the dominant f actors which would affect the calculated DFs are: (1) . the absence of hydrogen deflagra-

, tions, (2) the use of a non-conservative aerosol sedimentation area f or cases C9D and CIA, and (3) the use of an erroneous (high) heat sink area for case CADRF. While the effect of deflagrations would be to reduce l

'j 7

l DFs, the significance of the reductions cannot ' be de termined in the absence of detailed calculations. Likewis e, the use of more conserva -

tive sedimentation areas would be to reduce ~ the DFs for cases C9D and CIA, even if deflagrations ' are~ not 'considere d. - The use of an appro-priate heat sink ' area for case CADRF would- (in the absence of deflag-rations)' result in a decrease in the predicted DF for r % t sequence from 10 to as low as 6. While previous ORNL' calculations tor Browns Fe rry and ' Peach - Bottom support generic secondary containment decontamination f actor estimates of 10 to 40 - (f or cases in which hydrogen burns ' do occur), it is unclear that these values can be extrapolated to Shoreham. This uncertainty is' due to a variety of competing effects which are created by Shoreham's very' high secondary containment' connectivi ty.

Resolution of Uncertainties: While this assessment has attempted

~

to present a balanced evaluation of the Shoreham secondary containment-DFs calculated by ' FAI, the rather' qualitative nature of this assessment was dictated by the limited time available for the enalysis. Most (but not all) of the uncertainties outlined , above . and in .the Appendices of this report could be eliminated via a detailed confirmatory calculation effort. If assurance of significant Shoreham secondary containment fission product retention capability is required, it.is recommended that such confirmatory calculations be conducted. Such calculations should employ best-estimate hydrogen sources, multi-cell secondary containment models, and fully explore the effects of the various RBSVS ' and unit cooler operating modes on hydrogen deflagrations and aerosol behavior, a

j j

I l

1 1

l l

.I

G 'i i

REFERENCES

1. Request For Authorization To Increase Power To 25%, U.S. NRC Docket Number 50-322-OL, Long Island Lighting Co.

2. Ibid, page 5

3. Sherrell R. Greene, "The Role Of BWR MK I Secondary Containments In  ;

Severe Accident Mitigation," paper presented at the Fourteenth Water Reactor Saf ety Information Meeting, National Bureau of Stan-dards, Gaithersburg, Maryland, October 28, 1986.

4 F. E. Haskin, et al. , " Development and Status of MELCOR," SAND 8 6-21150, Sandia National Laborato ries , presented at the Fourteenth Water Re actor Safety Information Me eting, Na tional Bureau of Standards, Gaithersburg, Maryland, Octobe r,1986

5. "Probabilistic Risk Assessment - Shoreham Nuclear Power Station,"

SAI-372-83-PA-01, Science Applications Inc., June 24, 1983, pp. 3-420.

6. Ibid, Appendix D, p. 65

. 7. Ibid, Appendix K, p. 58  !

8. " Severe Accident Analysis Of The Shoreham Nuclear Power Station -  !

o 25% Powe r, " FAI/87-14, Fauske & As sociates, Inc., Ma rch 1987, I Appendix D, Table 0.2 I

l i

9 l

. Appendix A FACTORS AFFECTING BWR SECONDARY CONTAINMENT PERFORMANCE -

The purpose of this appendix is to summarize the BWR secondary con- j tainment characteristics which have the greatest impact on fission pro-duct retention and to briefly discuss the mechanism by which this impact in manifested. These key parameters are listed in Table A.1 and dis-

)

cussed below.

{

Prima ry Containment Failure Location: The first factor affecting BWR secondary containment effectiveness is the location at which the primary containment f ails (which determines the location at which pri- ,

mary containment blowdown enters the secondary containment). The fail-ure location determines both the physical state of material entering the secondary containment (single or two phase), and (together with secondary containment f ailure location) the volume of the secondary con-tainment which participates in the various fission product retention processes.

While prima ry containment f ailures can be postulated to occur in any location, LILCO has submitted an analysis by Stone & Websterl which identifies the most probable f ailure locations of the Shoreham primary containment are the wetwell airspace (130 psig), wetwell floor i

(134 psig), and the drywell head (140 psig). The first two f ailu re locations would result in blowdown into the lowest regions of the reac-

. tor building, while the drywell head f ailure would result in blowdown directly into the refueling bay.

It should be noted that a f ailure at the wetwell floor level would discharge wetwell pool water directly into the basement of the reactor building. As suming the pool were initially at its normal operating f level and a 10% mass addition due to SRV discharge from the reactor (a l very conservative assumption), the volume of water in Shoreham's wetwell (if saturated at 134 psig), would be approximately 100,500 ft3 . As s um-ing that 25% of the reactor building baseme nt volume is occupied by equipment and that none of the pool flashes (results in " upper-bound" estimate for pool depth), the maximum resulting pool depth in the reac-tor building basement would be 18 ft. Thus, the pool would rise to the 26 ft elevation in the reactor building. The static pressure head of this water would be ~7 p s i . The water would therefore begin flowing back into the wetwell when the prima ry containment pressure dropped to ,

7 psig. The equilibrium pool height (uniform depth in both wetwell and reactor building) would be ~9.7 f t (surf ace elevation 17.7 f t), which is )

below the bottom of the drywell vent downcomers. The pool water would, however, form a seal between the primary containment and the reactor )

building, resulting in a configuration similar to that of a Ma rk III I

containment system.

Appendix G of the original Shoreham PRA2 presents an analysis of  !

reactor building basement flooding due to a variety of reactor building pipe break accidents. The analysis presented there indicates that pool depths in excess of 2 ft would be expected to fail the HPCI, RCIC, and i

I 1

Table A.1. Factors af fecting BWR secondary

, containment fission product retention capability Pri. Cont. Failure Location Se c. Cont. Failure Location Sec. Cont. Volume Se c. Cont. Connectivity Sec. Cont. Heat Sink Area & Mass Se c. Cont. Sedimentation Area Se c. Cont. SGTS & RBSYS Capability Se c. Cont. Fire Protection System Spray Coverage Hydrogen Burns In Sec. Co nt.

. i l

i I

l i

i l

1 1

10 core spray system pumps, while levels in excess of 5 f t would f ail the RHR systems pumps. The control rod drive (CRD) hydraulic pumps are lo-cated on the 40 ft eleva tion, and would appear to be immune f rom the effects of flooding. Thus, early wetwell basemat failure would result in the loss of all injection systems except the CRD hydraulic pumps,

, which might continue to inject water into the reactor.

Seconda ry Containment Failure Location: Together with the primary containment f ailure location, the failure location of the secondary con-t ainme nt determines the the pathway through which fission products must flow prior to reaching the environment. The Shoreham reactor building is a reinforced concrete cylinder (outside diameter 135 f t) extending f rom the top of the concrete basemat to the polar crane rail (elevation 203 ft) some - 195 ft above. The wall is 2 ft thidt everywhere except between the fuel pool girder columns where it is 3 ' f t thick. The wall-f rom the crane rail elevation to the roof of the refueling bay (~25 f t) is steel f raming with insulated me tal siding. The roof construction consists of steel t russ e s , concrete slab on metal forms covered with insulation, and asphalt and gravel roofing. The floors are generally 18 in, thick concrete slabs spanning between reinforced concrete or steel beams. l The only significant potential pathways through the reactor build-  !

ing wall to the environment are the double doors (21 ft by 26 f t) to the

• truck bay airlock (40 f t elevation), which are used to import and remove the fuel cask, the reactor building normal ventilation system (RBNVS) intake grating (112 f t eleva tion) , the RBNVS exhaust ducting (101 f t elevation) and the 10 in. diam reactor building standby ventilation sys-tem (RBSVS) filter train exhaust duct (175 f t 9 in. eleva tion) . The truck bay appears to be an unlikely failure site, as the bay is designed as an airlock s t ructure, with two sets of doors in series. The RBNVS intake structure is housed in a separate room (behind an air-tight door) in the south side of the reactor building 112 f t elevation. The door opens into the reactor building from the room. Failure of this door appears unlikely, bu t might possibly be induced by localized hydrogen burn-induced pressure pulses. A more detailed analysis would be neces-sary to evaluate the probability of such a f ailure.

Leakage through the RBNVS and RBSVS exhaust ducting might be in-duced via localized hydrogen burns (or severe environmental conditions i associated with failure of the primary containment vent on the 101 ft floor). Failu re of the ducting appears unlikely, howe ve r, since the ductwork and exhaust fans that f o rm the recirculation system during RBSVS operation are designed to Seismic Category I criteria and the re-mainder of the system is supported in accordance with Seismic Category 1 criteria.3 The possibility of f ailure should be evaluated further, how-

~

ever, since the RBNVS ducting is located very near (a few feet) the probable site of the prima ry containment vent line f ailure. In these circumstances, the RBSVS ducting is the only barrier between the mate-rial vented f rom the primary containment and the outside environment.

11 The weakest structural element of the Shoreham secondary contain-ment is (like other BWR secondary containments) the refueling bay sid-ing. Although the Shoreham refueling bay does not have the blowout panels typical of Nbrk I plants, the utility has indicated that the refueling bay siding panels would deform and begin to leak at pressure differentials of ~0.5 psid.4 This is an advantageous f eature, because this failure location results in the longest secondary containment fission product release pathway (for most accidents), combined with an elevated (more than 180 f t above ground level) release point.

Although it appears the refueling bay would be the site of the  ;

secondary containme nt failure in most accidents, the potential for hydrogen burn-induced f ailures at other locations cannot be dismissed without additional analyses. Such an analysis should be conducted since failure at other locations might result in a more direct path from the primary containment to the environment.

Se conda ry Containne nt Volume: Se condary containme nt volume is a major performance indicator because the available reactor building and refueling bay volume determine the residence (holdup) time for fission  !

products (including noble gases). The holdup time is important for noble gases because it represents the only me chanism for mitigation (delay) of noble gas release. The residence time is particularly impor-tant for aerosols and fission product vapors, since it places a limit on the degree to which internal secondary containment phenomena (condensa-tion, aerosol settling and deposition, and fission product removal by the SGTS or RBSVS) can reduce fission product . releases to the

, envi ronment.

Se conda ry Containme nt Connectivity: Another secondary containment characteristic intimately related to total volume is connectivity.

Stated simply, connectivity is the characteristic which measures the ex-tent to which the total secondary containment volume can participate in mitigation phenomena. Connectivity is high when large, multiple flow paths exist between dif f erent regions of the secondary containment, and l is low when significant portions of the secondary containment are iso-lated from other regions due to the existence of few, small, or clustered flow paths.

Se conda ry Containment Heat Sink Area and Mass: The heat capa.:ity of the secondary containment affects both the secondary contalament pressure (which provides the driving f orce f or transport of fission pro-ducts from the seconda ry containment to the environment) and aerosol removal mechanisms within the containment. This heat capacity is deter-mined by the surface areas and masses of the structures which comprise the secondary containment and are contained within it. The heat sinks

. can be grouped into two classes (a) concrete, and (b) metal. This classification is based on the manner in which the heat sinks partic-ipate in the various heat transfer processes. Concrete structural tem-

, peratures tend to lag that of the local secondary containment atmosphere because these structures have enormous heat capacity and low thermal conductivity. Metal structures have low heat capacity, but high thermal

q 12 conductivi ties. The practical impact of this difference is that metal structures tend to follow secondary containment atmospheric temperatures much more closely than do concrete structures. This dif ference can be- j come important in terms of deposition and revolatilization of some fis-sion products.

The concrete structure is relatively easy to characterize from )

plant drawings. The metal structure is, however, ve ry difficult to I characterize, because it consists of a multitude of structures of vari-ous sizes (heat exchangers, pumps , s tai rwells , ducting, etc). In a general, larger heat sink areas and masces afford greater opportunities l for fission product vapor condensation and aerosol removal processes (diffusiophoresis and thermophoresis) to scrub fission products f rom the primary containment discharge prior to release to the environment. It should be noted, howe ve r, that diffusiophoresis and thermophoresis are generally lese significant removal mechanisms than sedimentation. For this reaeon the effect of secondary containment heat sink capacity cu containment pressure could be more significant than its role in aerosol )

removal f or most sequences. The portion of the total secondary contain-  !

ment heet sink capacity which participates in accident mitigation is a j direct f unction of building connectivity. '

Secondary Containment Sedimentation Area: The amount of sedimenta-tion area available in the secondary containment impacts the degree to which gravitational sedimentation processes can remove aerosols prior to their discharge to the environment. A lower bound estimate for sedi-mentation area available in a secondary containment is simply the total floor area available in the reactor building and refueling bay. The actual sedimentation area available in any containme nt is larger than the floor area due to the availability of deposition surfaces on equip-ment, ducting, and stairwells, etc. As in the case of heat sink capac- 1 ity, the magnitude of the " participating" sedimentation area is heavily j influenced by building connectivity.

Secondary Containment SGTS and RBSVS Capability: All BWR secondary  !

containments incorporate a system which is designed to filter [through charcoal and high ef ficiency particulate absolu te (HEPA) filters] and exhaust primary containment purge gas during normal startup and shutdown operations. These systems are also employed to process the secondary containment atmosphere during accident conditions. Most MK 1 plants utilize high capacity standby gas treatment systems (SGTS), which are of the once-through design. These sys t ems draw suction on the reactor building and refueling bay (typically 20,000 cfm total), filter the entire gas stream, and exhaust it through an elevated stack to the atmosphere. Secondary containment makeup air is provided via infiltra-tion and controlled inleakage f rom the environment.

Some MK 11 plants (Limerick) utilize low capacity SGTS in conjunc-tion with reactor enclosure recirculation sys t ems (RERS). The RERS mixes, filters, and recirculates 60,000 cfm between the reactor building and refueling bay. The independent SGTS filters and exhausts much smaller flows (3000 cfm) to the envi ronment. The Shoreham plant

]

I i

13 j

1

(

utilizes a reactor building standby ventilation system (RBSVS) which  !

draws suction on the reactor building (45,000 cfm) and discharges (without filtration) to the refueling bay. - Less than 1200 cfm of this ,

flow is filtered and exhausted to the environment. ;It is apparent that '

the maj or impact of SGTS operation' on the secondary containment l

. atmosphere is filtration and dilution. The major effect of the RERS is filtration and mixing, and the maj or impact ' of the RBSVS is ' mixing.

During severe accidents, the exhaust. capacity of the ' SGTS, RERS, of RBSVS (if operating) would have a major impact on secondary containment 'j fission product retention, since primary containment blowdown rates in  ;

excess of the system exhaust capacity can result in secondary contain-l ment pressurization and direct leakage from the secondary containment to the environment.

Se condary Containment Fire Protection System Spray- Coverage: The- i design of secondary . containment fire protection systems varies widely from plant to plant. Most fire protection systems incorporate some ]

" pre-a ction" fused-link sprinklers such as are utilized in: office buildings. The source of water for these sprinklers typically includes storage tanks and diesel driven pumps. The sprinkler heads require no external power source to function, and are designed to initiate at some pre-selected t empera ture (typically 160-170'F) in the' event of a fire..

While designed f or fire protection, these systems would actuate during j

severe accidents due to the rise in secondary containment temperatures

. associated with primary containment failure. Previous _ ORNL studies 5 have demonstrated that these sprays can significantly enhance secondary containment aerosol DFs via direct scrubbing of aerosols and cooling of

, the building atmosphere. The sprays also decrease steam concentrations in the building, which tends to f acilitate the initiation of hydrogen burns via elimination of the steam-inerting mechanism.

Hydrogen Burning In The Secondary Containment: The occurrence of a hydrogen burn requires the combination of a trigger -(spark or flame) and the approp riate atmospheric gas composition (typically 5 molar % or greater oxygen, 8 molar % or greater hydrogen, and less than 55 mole %

steam).6 The Shoreham reactor has an invessel Zi rcaloy inventory of ~105,000 lbs, which is comprised of fuel pin cladding and fuel chan-nel canisters. Complete oxidation of this Zircaloy would produce ~4600 lbs of hydrogen, which would be released to the secondary containment following primary containment f ailure (or venting). While 100% Zircaloy oxidation is unlikely, Shoreham calculations performed at ORNL by S. A.

Hodge in support of the NRC's evaluation of LILCO's submittal 7 indicate that as much as 45% of the total Zircaloy might be oxidized in some cir-  !

cumstances. Forty-five percent oxidation would result in ~2100 lbs of hydrogen - enough to produce four global deflagrations in the Shoreham reactor building, or two global deflagrations- throughout the entire

.econdary containment (reactor building and refueling bay) under the most conservative conditions.

l

' The tendency to achieve the atmospheric composition necessary for deflagration is a function of many factors. The two most important fac- .

tors are the degree of secondary containment conp artme ntalization (o r i connectivity) and the impact of the SGTS/RERS/RBSVS system operation.

Low connectivity tends to promote the occurrence of localized hydrogen i

burns, while high connectivity'tends to result in fewer, but larger and

~

more severe, burns. Tne impact of ' the larger number of small burns on .

secondary containment' fission product retention is much less severe than that of a few global burns, 'in which , higher peak pressures are devel-oped, and the entire building vents to the outside ' environment over a

. 'short time period;5. Burns .in highly. compartmentalized structures usually result in lower peak pressures thanido . global burnsL in large open volumes.5 Additionally, .the involved region in the compartmental-ized structure is vented within the reactor building, ratter than into the environment - (which is the case when large. portions of the. secondary.

- containme nt ' are involved in a single ' global deflagration eve nt) ..

SGTS/RERS/RBSVS operation may influence combustibility by (a) mixing the secondary ~ containment volume and '(b) importing air f rom outside the secondary containment. The mixing function tends. to increase the amount of hydrogen necessary to reach combustible conditions, while the. impact.

of air infiltration is sequence dependent. In some cases the impact of.

the air infiltration is to dilute the hydrogen concentration below ~ coor bustible limits, while in other cases (oxygen lean environments) the infiltration results in an increase in oxygen concentrations to combus-tible limits. l The major potential' ef fects of hydrogen burn-induced pressure and g temperature pulses are: (a) loss of secondary containment integrity due j to structural damage, and (b) . pumping fission products to the outside )

, environment via normal or accident-induced secondary containment vent j paths. The extent to which this. pumping occurs is L a' function of burn timing. For accidents in which ' the primary containment is vented f rom

, the wetwell, the majority of s econda ry containment hydrogen -deflagra .

tions might occur prior to the release of significant' quantities ~ of 3 aerosols into the secondary containment. The major impact of deflagra-tions in such cases might be to degrade secondary containment' integrity. 1 rather than to directly pump fission products to the environment. ]

I 4

)

a 1

15 APPENDIX A REFERENCES

1. LILCO Response to NRC Question 6.b

2. "Probabilistic Risk Assessment - Shoreham Nuclear Power St a tion,"

SAI-372-83-PA-01, Science Applications Inc. , June 24, 1983.

3. "Shoreham Updated Safety Analysis Report," section 9.4.2.3 4 Answer to review question 6, NRC/LILCO Meeting, Shoreham, N.Y. , May 12, 1987.

5. Sherrell R. Greene, "The Role Of BWR MK I Secondary Containments In Severe Accident Mitigation," paper presented at the Fourteent h Water Reactor Safety Information Meeting, National Bureau of Stan-dards, Gaithersburg, Maryland, October 28, 1986.

6. "MELCOR Hyd rogen Package Us e rs' Guide, Version 1.5.0," June 16, 1986

7. S. A. Hodge and R. M. Harrington, " Considerations' Regarding Certain Aspects Of Severe Accident Mi tiga tion Afforded By Operation Of Shoreham at Reduced Powe r," Le t ter Re po rt , Oak Ridge Na tional

. Laboratory, June 12, 1987 s

1 I

O e

a i

i

)

16 Appendix B i

. COMPARISON OF BROWNS FERRY, PEACH BOTTOM, AND SHOREHAM  !

SECONDARY CONTAINMENT PERFORMANCE The purpose of this section is to compare the Browns Fe r ry , Peach Bottom, and Shoreham secondary containment designs, and evaluate LILCO's claims for Shoreham secondary containment performance in light of ORNL's previous evaluations of Browns Ferry and Peach Bottom. The objective of this analysis is not to determine exact sequence -dependent Shoreham j secondary containment DFs, but rather to estimate a reasonable range of Shoreham secondary containment DFs based on design similarities and dif-ferences between Shoreham, Peach Bottom, and Browns Fe rry.

The Browns Ferry Nuclear Plant is a 3 unit BWR-4/MK I facility, owned by the Te nnessee Valley Au thority and located near At hens ,

Alabama. The three Browns Ferry units are housed in independent reactor buildings, but share a common refueling bay. The Peach Bottom Atomic Power Station is a 2 unit BWR-4/NK I f acility, owned by the Philadelphia Electric Co., and located near Lancaster, Pennsylvania. The two units at Peach Bottom are housed in independent reactor building refueling bay s t ruc tures.

1 The Browns Ferry nuclear plant has been the subject of continued f study by the ORNL Severe Accident Sequence Analysis and BWR Severe Acci- j dent Technology Programs during the past six years. During the past two

}

~

years, ORNL has developed and exercised a series of detailed, large l scale secondary containment models for Browna Fe rry Unit 1. These l models have been applied in a variety of studies designed to character- j ize secondary containment performance during severe accidents initiated j at full power.1 Particular emphasis has been given to the evaluation of l station blackout scenarios in which the standby gas treatment system is l not operational, and to evaluation of the impact of secondary contain-ment fire protection system spray actuation during severe accidents. )

More recently, a large scale Peach Bottom secondary containment model i was developed and applied to the analysis of a station blackout scenario at that plant.

The results of these studies indicate that BRR secondary contain-ment DFs of 10 to 40 appear credible for scenarios in which the primary containme nt fails into the lowest regions of the reactor building.

Current best-estimate Peach Bottom analyses predict DFs of ~15 for sta-tion blackou t cases (f rom 100% power) in which the SGTS is inoperable and the primary containment fails into the reactor building basement.

The best available estimates f or the Browns Ferry Unic 1 secondary con-tainment DF for a similar scenario range from 10 (no fire protection sprays) to 33 (fire protection system sprays operational). Investiga-tions of other sequences at Browns Ferry have produced DFs as high as i 100 for cases in which the primary containment f ails at low pressure into the basement of the reactor building, the fire protection system t

.l 17 l

sprays ' are operational, and no hyd rogen bu rns occu r 'n i the reactor.

building or refueling bay.

Table B.1 presents a summary of Brown s Fe rry , Peach Bottom, and 2 Shoreham secondary containment design characteristics. The character- l istics described in Table B.1 are secondary containment volume, heat d sink area, floor area (lower bound on aerosol sedimentation area), SGTS or. RBSVS circulation and filtration data, and fire protection spray cov-erage data. The data for Browns Ferry and Peach Bottom were developed during previous ORNL analyses, while the Shoreham data were developed by .j ORNL specifically for.the present analysis. Table B.2 presents the same  ;

data in a "per megawatt" (based on full rated power levels at all f acil- .)

ities) normalized format which is useful for identifying inconsistencies i' in basic design and performance data. Since integrated primary contain-ment blowdown energy and fission product ' sources are :(approximately)'.

linear functions of power level, the data in Table B.2 can provide use- i ful insights into the relative effectiveness of the three secondary con- l tainment designs. The potential impact of plantLdesign differences.on secondary containment decontamination f actors are discussed below.

Se condary Co ntainme nt Volume: The volume of Shoreham's secondary containment (Table B.1) is 50% smaller than that of Browns Ferry and 7%

smaller than that of Peach Bottom. The refueling bay and total volume data are presented in two forms in Table B.2. Since Browns Ferry is a .

multiple unit plant which employs a single refueling bay co service all )

units, consideration must be given to the manner in which the " shared" .}

volume and surface areas are credited. Thus the first number in each column of the Browns Ferry refueling bay and total volume rows is based ]

on the assump tion that all units in the facility share the common refueling bay volume and area equally (the normalization is based on total f acility power rather than single unit power). The second number in each Browns Ferry column is based on the assumption that all of the shared volume and surf ace area is available for mitigation of.a single  !

unit accident (the normalization is based on single unit power).

A comparison of the data in Table B.2 indicate that Shoreham has approxi-mately 32% less secondary containment volume per megawatt . than does Browns Ferry, and 26% more volume per megawatt to that of Peach Bottom.

These characteristics would tend to produce Shoreham secondary contain-ment DF's which are somewha t larger than those of Peach Bottom, bu t smaller than those of Browns Ferry.

Se condary Containment Connectivity: Ta bles B.3 through B.5 compare

~

the connectivity of the three subj ect secondary containments. While  ;

connectivity is a three dimensional. characteristic, Tables B.3 through B.5 focus on the connectivity between various levels of the reactor building rather than between various regions within a level. This f ocus l is appropriate because the point of secondary containment fission pro-duct release is usually the refueling bay, while prima ry containment blowdown generally enters the secondary containment at lower levels in

, the reactor building.

An analysis of Tables B.3 through B.5 reveals that Browns Ferry's and Peach Bottom's connectivity is relatively low, with the majority of all inter-floor communication occurring in one or two quadrants (south i

)

Table B.1 Courparison of Browns Ferry, Peach Bottom, and i Shoreham Secondary Containment Designs '

Parameter r s eac Shoreham

., p, .

Rated power (MWt) 3,293 3,293 2,436 Rx. bldg. vol. (f t ) 1,411,800 1,146,800 1,372,600 '

Refuel. bay vol. (ft ) 2,745,000a 1,096,400 717,700 ,

Total volume (ft ) 4,156,800 _2,243,200 2,090,300 l Rx. b1dg. heat sink 239,300 213,100 196,800 ]

area (ft ) 1 Refuel. bay heat sink 161,300 69,300 46,100 area (f t )

Total heat sink area 400,600 282,400 242,900 (ft )

Rx. bldg. floor area 54,700 42,400 45,700 (f t )

Refuel. bay floor area 48,700 14,700 10,700 (ft ) 1 Total floor area (f t ) 103,400 57,100 55,500 SGTS exhaust filter train 21,500 25,000 1,160 D flow (cfm)

. Pre-action fire protection 21,000 100 800 i spray coverage area (f t ) l l

, " Browns Ferry Refueling Bay services 3 units.

i b

Shoreham RBSVS (SGTS) system circulates 45,000 cfm between the reactor building and refueling bay, but filters and exhausts only 1160 cfm of this flow.

e

.m

j Table B.2' - Normalized Browns Fe rry, Peach Bo ttom, and 'l Shoreham- Secondary Containment Characteristics  !

l Parameter Shoreham j Fe y tto Number of Units 3 2- 1 Rated power (MWt/ unit) 3293 :3293 2436 Rx. b1dg. ' vol. (f t / megawatt) 429 348 563 Refuel. bay vol.- (f t / megawatt) 2788/834 D .333' 295

. Total volume- (f t / megawatt) ' 7070 /1262 d' 681 858-Rx. b1dg. heat sink area 73 65- 81 (f t / megawatt)

Refuel. bay heat sink 16aj49b 21, 19 area (f t / megawatt)

Total heat sink area- 890/122 d 86 100 (f t / megawatt)

Rx. b1dg. floor area - 17 13 18 (f t / megawatt)-

Refuel bay floor area 58/15 b 4 4 ,

(ft / megawatt) .I Total floor area (f t / megawatt) 220/32 d 17 22' I Time for SGTS or RBSVS to mix 1 - -

0.8-

,. sec. cont.

vol. (h) .g

. Time for SGTS or RBSVS to 3.2 1. 5 30.0 -!

filter 1 sec. cont. vol. (h) J

, Percentage of Rx. b1dg. 38 <1 2 -

floor area covered by sprays q

' " Parameter / ( (# units) (power / unit) ).

barameter / (single unit power).

C Refuel. bay value / ( (# units)(power / unit) .) + Rx. b1dg.

value / (power / unit) .

d Total value / (single unit power).

i 4

1 4

i Table B.3. Browns Ferry secondary ,

containment connectivity

~

l Quadrant flow area Total flow splits Path area (%)

(ft )

N E: S W l Hatchway to refuel. 357 0 0 100 0 bay 621 ft to hatchway 75 0 0 100 0 621 ft to 639 ft 278 0 50 0 50 593 ft to hatchway 75 0 0 100 0 593 ft to 621 ft 139 0 0 0 100 565 ft to Hatchway 150 0 0 100 0 565 ft to 593 ft 669 0 0 43 57 519 ft to 565 ft 345 23 23 2 52 l

i 1

l 4

'Y

Table'B.4. Peach Bottom Secondary; j Containment Conne ctivity. -]

1

- Quadrant, flow area ~ j

. . Total flow splits  !

Path' area' , '(%)-

(f t )

NE SEj SW ' NW . ,

.q 1

195 ftlto refueling: 357 0' 100!' O. O. i bay. j 165 ft to 195 ft. .357 0 100 0 'O 135 ft to 165 357 0 100' , O. 0 92.5 ft to 135 ft 93 0- 0 72- 28 1

I i

1 l

u 4

i a

l l

1 l

>9 l.

-)

i

1

.) .

Table B.S. Shoreham L Se condary Containment Connectivity Quadrant ~ flow area

. Total flow ' splits-

.)

Path ~ area (%) 1 (f t ') . .

NE SE SW' NW

'l 696 "

150 f t 9 in. to re- . 12 13 . 56 =19 fueling bay

- 112 ft 9 in. to 745 -6 22 52 20 1 150 ft 9 in.- j 78 ft 7 in. to 759 -13 17 .51 19-  :

112:ft 9 in.  ;

63'ft to 78'ft 7'in. 611 -

7 4 63- 26 40 f t to 63 f t 3276 1 ,4 7 ' 48 4 8 f t to 40 f t 277. 0 39 0 .61. .

1 l

i i

. l

.I l

i 1

imJ

18 l l

l and west in Browns Ferry, southeast in Peach Bottom). The connectivity of the Shoreham reactor building (Table B.5) is, in contrast, quite high. This characteristic would enhance Shoreham's secondary contain-ment fission product retention (relative to Browns Ferry and Peach Bot-tom) for sequences in which global hydrogen burns do not occur, since

, this enhanced conneccivity would allow larger fractions of Shoreham 's volume, heat sinks, and floor areas to participate in aerosol removal processes than would be the case at the other two plants. Sho reham 's secondary containment fission product retention could be reduced, how- ,

ever, if this high connectivity (and resultant gas mixing) leads to i global hydrogen burns.

Heat Sink Area and Mass: Sho reham's total secondary containment heat sink area is 39% less than that of Browns Ferry, and 14% less than that of Peach Bottom. Howe ve r, the data presented in Table B.2 indicate that Shoreham has more reactor building heat sink area per megawatt than do either Browns Ferry or Peach Bottom. This characteristic indicates that the importance of diffusiophoresis and thermophoresis in Shoreham's secondary containment would be similar to that in Browns Fe rry and Peach Bottom for similar accident sequences. i Floor (Sedimentation) Area: Shoreham's total secondary containment f floor area is 46% less than Browns Ferry's and 3% less than Peach Bot- A com's. Shoreham's normalized reactor building floor area is 6% larger

. than that of Browns Fe r ry , and 38% larger than that of Peach Bo t t o m.

Shoreham's normalized total secondary containme nt floor area is 31%

smaller than Browns Ferry's and 29% greater than Peach Bottom's. This

, is a significant f actor since aerosol sedimentation is normally the most important removal mechanism in the secondary containment.

SGTS/RBSVS Capability: The comparison of Shoreham's RBSVS system - I capability to that of Browns Ferry and Peach Bottom reveals some signif-icant strengths and weaknesses of the Shoreham design. Due to the large recirculating capacity of the Shoreham RBSVS system, only 0.8 h are re-quired to cortpletely circulate one secondary containment volume. This characteristic would tend to reduce the probability of secondary con-tainment hydrogen burns for sequences in which small amounts of hydrogen are produced in the reactor and primary containment, since the contain-ment would tend to be well mixed. Hyd rogen bu rns that occur would, however, tend to be global in nature (see Appendix D), and could sig-nificantly reduce the secondary containment's fission product DF.

l Another negative impact of Shoreham's rapid mixing rate is that '

fission p roduc ts f rom lower regions of the reactor building might be transported more quickly to the refueling bay (probable site of sec-ondary containment failure) at Shoreham than at the other two f acil-

,, ities. This is a particular concern at Shoreham, since the RBSVS filter capacity is small compared with its recirculation capacity. The time l required to filter and exhaust one secondary containment volume at i

• Shoreham is 30 h, compared with only 3.2 h at Browns Ferry and 1.5 h at I Peach Bottom. The second impact of the low RBSVS filter train (exhaust) flow is that primary containment blowdown events which would not result )

i i

19 in secondary containment pressurization and direct exfiltration to the environment at Browns Ferry or Peach Bottom, would result in substantial- l secondary containment pressurization and exfiltration at Shoreham.

This comparison indicates that significant improvement in Shore-

- ham's secondary containment aerosol and iodine DF could be achieved by  ;

increasing the RBSVS filter train capacity. A significant incremental '

capacity increase night be achieved by arranging to overcide the auto-matic throttling dampe rs which currently operate to maintain 1160 cfm flow even if both filter-trains are in service. It should be noted, however, that an improvement in Shoreham's RBSVS filter train capacity would have little impact f or accidents in which the RBSVS system is un-available, or for accidents in which the fission product release is totally dominated by noble gases. The combined effect of Shoreham's higher circulating rate and lower filtering rate is difficult to pre-dict. It is probable that the combined impact of these characteristics will be sequence dependent.

Fire Protection System Sprays: A comparison of Sh oreham's fire protection water spray system to that of Browns Ferry and Peach Bottom indicates that Shoreham's DF would not be enhanced by fire protection spray system operation, due to the very limited area covered by this system. This characteristic of Shoreham is very similar to Peach Bot-tom. -

e Summa ry : In summary, a coqparison of Shoreham's secondary contain-ment characteristics to those of Browns Ferry and Peach Bottom indicates

. the f ollowing: I I

1. For cases in which secondary containment hydrogen burns do not occur, Shoreham 's claimed total seconda ry containment DF of 10 (Case C9D) appears to be reasonable. This judgment is based on the similarity between Shoreham's snd Browns Fe r ry 's volume, and heat sink and sedi-mentation area characteristics.

2. Shoreham's claimed refueling bay DF of 10 (Case CADRF) appears to be higher than can be justified based on previous ORNL calculations f or Browns Fe rry and Pea ch Bo t tom, which have yielded lowe r-bound DF estimates of 10 to 14 for the entire secondary containment.

3. For cases in which secondary containment hydrogen burns do not f occur, Shoreham's claimed secondary containment DF of 50 (Case CIA) 1 appears to be s omewhat high. It should be noted, howeve r, this DF is l claimed for a station blackout sequence in which high primary contain-ment pressures are never achieved, the prima ry containment vents into the reactor building basement, and the RBSVS is not operational. The s e i

. factors, in combination with Shoreham's high connectivity, could result in high DFs for this sequence. l j

4 Based on S. A. Hodge's analysis of invessel hydrogen production and the deflagration limit analysis presented in this report, the pos-sibility of hyd rogen bu rns in the secondary containme nt cannot be I 1

l 4

s

l 1

1 20 excluded (see Appendix D). FAI analysts have indicated that deflagra-tion limits were not reached in any of the HAAP simulations performed for the 25% powe r PRA. The absence of hyd rogen burns in the FAI analysis appears to be a result of the low Zircaloy oxidation f ractions j (and resulting low hodrogen masses) typically calculated by MAAP. j

5. Sho reham's RBSVS design philosophy is significantly different than that of Browns Ferry and Peach Bottom. The design emphasizes mix-ing over dilution and filtration. Operation of the RBSVS will reduce i the probability of localized hydrogen burns, but burns that do occu r l will be of a global nature (see Appendix D) and would tend to flush I fission products f rom the secondary containment into the environment.

i

6. RBSVS operation may actually reduce secondary containment DFs I for accidents in which the prima ry containme nt fails into.the lowe s t portions of the reactor building. This is a possibility because RBSVS operation will transport fission products f rom the lower regions of the ,

reactor building to the refueling bay (which is the most probable site I f or secondary containment f ailure).

7. Shoreham's low RBSVS filter exhaust capacity renders the plant more vulnerable to pressurization f rom primary containment blowdown than q are Browns Ferry and Peach Bottom. Prima ry containment blowdown rates l as low as 1200 cfm could initiate pressurization of the secondary con- l

, tainment and leakage of fission products to the environment. This is a

{

important consideration, since the utility estimates that primary l containment venting procedure employed in most accidents will result in

, a 3000 cfm steam source to the reactor building.

I i

21-1 APPENDIX B REFERENCES

. 1

1. Sherrell R. Greene, "The Role Of BWR MK I' Secondary Containments In Seve re Accident Hitigation," paper presented at the Fourteenth - .

Water Reactor Safety Information Meeting, National Bureau of Stan-dards, Gaithersburg, Maryland, October 28, 1986. .

-l l

l l

)

l l

l

'I 8 ,

O

22 I

Appendix C CRITIQUE OF FAI MODELING APPROACH ~FOR C9D, CADRF, and CIA SEQUENCES The purpose of this section is to provide a critique of the Shore-ham secondary containment modeling approach (input data and assumptions) .,

employed in the MAAP modeling effort performed by Fauske and Associates l for LILCO. The comments in this section are based on a review of the l MAAP input data decks listed in Appendix C of the FAI report.1 An' a t temp t is made to identify discrepancies between (and among) the input decks and the actual- plant design, and to estimate the impact such discrepancies migh t have on the calculated secondary containment DF.

For the purpose of this dis cussio n, a " conservative" approach is _ one which would yield lower estimated DFs (higher fission product releases) than those which would be produced by utilizing parameter values more representative of the actual secondary containment construction.

I General Comments - Model Top ology : It should be noted that a single cell secondary containment model was utilized for ~all analyses The issue of whether or not a single cell approach performed by FAI.

l produces conservative or non-conse rva tive predictions is sequence '

dependent, and a function of f actors such as the locations of the pri-  ;

. mary and secondary containment failures, secondary containment connec- )

tivity, SGTS or RBSVS operability, and the amount of hydrogen released into the secondary containment. If the primary and secondary contain- ' '

ment f ailure locations, building connectivity, and RBSVS operation combine to produce a well mixed building (in which the majority of the volume and structures participate in the mitigation process), a single ,

cell model can produce realistic results. I For cases in which the primary and secondary containment failure locations, building connectivity (compartmentalization), and SGTS or RBSVS system operability result in isolation or bypass of portions of the secondary containment, the single cell modeling approach can produce either conservative or non-conservative estimates of building perf orm-ance. N( -conservative results may be produced for cases in which hydrogen deflagrations do not occur (because little hydrogen is avail-able), since a single cell model over-estimates the fractions of total secondary containment volume, heat sinks, and sedimentation areas which actually participate in accident mitigation. If significant hydrogen is p roduced, the single cell model can yield conservative results, since it will result in fewer, but much more severe, hydrogen deflagrations than would actually occur in the compartmentalized structure.

, As noted in Appendix B of this report, the Shoreham secondary con-tainme nt's connectivi ty is significantly higher than that of plants previously analyzed by ORNL. For cases in which the RBSVS is opera-

• tional (Case C9D) the single cell modeling approach may yield realistic results. For cases in which the RBSVS is not operational (Cases CADRF and CIA) it is not clear (in the absence of independent multi-cell

23

'l 1

calculations) that single cell models will produce realistic predictions -

for ' secondary containment performance. ORNL 's multi-cell Browns Ferry .

and Peach Bottom calculations (in which the SGTS was not operational and hydrogen deflagrations did occur) have not predicted secondary contain- '

. ment DFs lower than 10. g Ceneral Comments - Hydrogen Deflagrations: As noted in Section 3- -

of this report, the MAAP analyses perf ormed by FAI did not predict com- .]

bustible secondary containment conditions for any of the six representa - j tive ; accident sequences.

This results appears to be an artif act of MAAP's core melt modeling approach.- The hydrogen generation analysis -

performed by S. A. Hodge and the secondary containment hyd rogen deflagration study presented in Appendix D of this report' indicate that the probability of hydrogen deflagrations 'in Shoreham's secondary con-tainment (and associated burn-indicated fission product releases) was underestimated in the FAI analysis. This approach . represents a major  ;

non-conservatism in the FAI analysis, and imposes' a great deal of uncer-tainty on the.resulting secondary containment DF estimates. This uncer- ]

j tainty is compounded by the use of a single cell secondary containment i model which presents a simplistic view .of secondary containment )

performance. i l

It is probable that the degree of uncertainty in the DF estimates associated with these two issues is significantly greater than that

, resulting from the combined effect of all of the f actors . discussed below.

It should be noted, however, that previous ORNL analyses of Browns Fe rry and Peach Bottom secondary containme nt performance have consistently yielded DF estimates of 10 to 40 for situations in which multi-cell models were employed and hydrogen deflagrations did occur.

In light of these considerations, it is unclear why the FAI analyses did not result in higher DF estimates than those reported.

The following paragraphs provide a critique of the modeling approach employed by FAI for the individual accident sequences. Ta ble q C.1 presents a summary of the modeling parameters discussed below. '

Case C9D: A single cell (combined reactor building and refueling bay) model is employed, with a total cell volume of 1.56 million cubic feet. This is ~25% smaller than that calculated by ORNL, and would (in the absence of hydrogen burns) tend to result in conservative (lowe r) estimated DFs than those which would be expected from a containment with the larger volume. A minimum of 593 lbs of hydrogen would' be required '

to achieve combustible conditions in this volume. FAI has indicated that the highest hydrogen concentration achieved in this sequence is 2 molar % at 11 h into the accident - a value below the 8 molar % concen-tration required for deflagration.

~

A single heat slab with a surface area of 129,600 ft2 was em-ployed. This area is 47% smaller than that estimated by ORNL for the

~

total secondary containme nt architectural heat sink. area (includes floors, walls, ceilings, but not equipment). The use of this value could result in conservative DF estimates (due to the underestimation of

Ta ble C . I . Comparison of FAI MAAP input parameters to ORNL estimates M AP input value for

, p a ORNL estimate C9D CADRF CIA Rx. bldg. volume 1,372,600 Refuel.. bay volume 717,700 744,670 Total sec. cont. vol. 2,090,300 1,556,048 1,556,048 Rx. bldg. heat sink >196,800 surface area Refueling bay heat >46,100 129,608 sink surface area Total sec. cont. heat >242,900 129,608 129,608 sink surface area Rx. bldg. sedimentation - >45,700 area Refueling bay >10,700 14,454 sedimentation area Total sec. cont. >55,500 237,140 237,140 sedimentation area RBSVS exhaust flow rate 1,160 1,165 0 0

" Units: Volume (ft )

Area (f t )

, Flow rate (f t / min).

I l

1 1

l 24

)

diffusiophoresis and condensation) unless the resulting lower steam con-densation ' artificially prevents hydrogen deflagrations. Appendix F of l FAl's report l indicates that diffusiophoresis is not a maj or aerosol'

{

removal mechanism for these sequences. The thermal properties (thermal  ;

conductivity, specific heat, and density) of the heat sink are appro- '

priate for concrete, which is the dominate heat sink material. No steel heat sinks are simulated f or these sequences. j The sedimentation area employed by FAI for this calculation was l 237,140 ft2, which is roughly four times larger than the combined floor surface area of the reactor building and refueling bay. The available sedimentation area within a building is actually the total floor area j plus the horizontal projection of all structural surfaces (ducting, heat exchangers, stairs, pipes, etc. ) within the building. The ratio of the secondary containment volume and the total sedimentation area can be viewed as an equivalent aerosol settling height. F AI' s value's for these two parameters result in a settling height of 6.6 ft, while the j) ratio of ORNL's estimated secondary containment volume to total floor area results in a settling height of 37.7 f t.

While it is acknowledged that a settling height of 37.7 f t is un-realistically high, the use of a settling height of only 6.6 f t appears non-conservative in the absence of detailed measurements and calcula-tions for sedimentation area. Appendix F of the FAI report does not in-dicate that parametric studies of the effect of sedimentation area vari- )

ation were performed for the C9D sequence. Appendix F does present an analysis of Case CSAA (a "V" sequence which is not one of the six " rep-

. resentative" sequences) which indicates that a 50% reduction in the sedimentation area for that sequence results in only a 4% reduction in the mass of aerosols removed via sedimentation, with a coincident reduc-tion in the calculated secondary containment DF f rom 167 to 145. That analysis demonstrates that sedimentation area changes are not linearly mapped into sedimentation mass changes. FAI's use of an non-conserva-tive value for sedimentation area for tbis sequence would lead to non-conserva tive es tima tes of secondary containme nt DFs, bu t the signifi-cance data.

of this non-conservatism can not be deduced from the available l

A review of the MAAP C9D input deck employed by FAI reveals that,  !

while the RBSVS exhaust flow was simulated, the impact of RBSVS coolers and reactor building unit coolers was not incorporated in FAI's anal-ysis. These syctems would probably continue to operate for some time j f ollowing accident initiation. The RBSVS incorporates two 100% capacity cooling coils in series (rated at 106 Btu /h cooling capacity). The unit cooler subsystem employs 2 sets of 4 unit coolers, each having a rated air flow capacity of 40,000 cfm. Four of these unit coolers are rated at 5. 7 x 105 Bt u/h , and four are rated at 106 Btu /h heat removal capacity. Four unit coolers draw suction on the refueling bay and ex-haust to the refueling bay. Two unit coolers draw suction on the 40 ft

, elevation and sxhaust to the 8 f t elevation, and two unit coolers draw suction on the 8 f t elevation and exhaust to the 8 ft elevation. The

i 15 RBSVS and unit coolers are serviced by a system of 4, 50% capacity, 275 ton water chillers.

The two major impacts of RBSVS cooling coil and unit cooler opera-tion during severe accidents would be to condense large quantities of-steam f rom the secondary containment atmosphere, and to remove aerosols from the secondary containment atmosphere via direct impaction and dif-fusiophoresis across the unit cooling coils. The major impact of the steam removal is to reduce secondary containment pressure (and leakage to the environment), and, potentially, to promote secondary containment hydrogen deflagrations via a reduction in the steam inerting mechanism.

It should be noted, howeve r, that the impact of hydrogen deflagrations in these circumstances would be less severe than normal, since secondary containment suspended aerosol concentrations would be lowe r and less aerosol mass would be available to be " pumped" to the environment by the burn. Although a more detailed analysis of the impact of cooler opera-tion on secondary containment performance is needed, it appears that FAI's approach of ignoring the RBSVS coolers and unit coolers is a con-servatism in their calculation of secondary containment DFs for this sequence.

Case CADRF: A refueling bay DF of 10 was calculated by FAI for this sequence. According to Appendix G of the FAI report,1 29 kg of aerosols were retained in the refueling bay via sedimentation (25 kg),

diffusiophoresis (3 kg), and thermophoresis (1 kg).

A single cell (refueling bay only) secondary containment model is

. employed, with a volume of 744,670 ft3 which is approximately equivalent (3.8% larger) to that calculated by ORNL. Approximately 284 lbs of hy-drogen would be required to raise this volume to deflagration condi-tions. The highest estimated hydrogen concentration (5.6 molar %) was below deflagration limits and was predicted to occur at 4.6 h into the accident.

A single (steel) heat sink is employed, with a surfact area of 3 129,608 f t . This area is identical to that employed to represent the )

total secondary containment (reactor building and refueling bay) area in cases C9D and CIA, and is 2.8 times larger than the total refueling bay floor, ceiling, and wall area. The use of this value must, therefore, be regarded as an error. The impact of this error on secondary contain-ment fission product retention is somewhat uncertain. Since the second-ary containment DF is defined as the ratio of the aerosol mass (Q) injected into the secondary containme nt to the mass of aerosol which escapes to the environment, the amount of aerosol inj ected into the secondary containment in this sequence can be calculated as: l l

Q / (Q -- 29. ) = 10. i (10. )(Q) -- 290. = Q Q = 290. / 9.  ;

, Q = 32.2 kg. i

I 26=

1 If the amount of aerosols removed by condensation were reduced f rom 3 to 1 kg (conservatively accounting for the error in heat sink surface area), the total retained aerosol mass would drop to 27 kg, and the resulting secondary containment DF would be. j i

DF = 3 2. 2 / (3 2. 2 -- 27. ) = 6. 2.  ;

1 The sedimentation- area employed by FAI for .this calculation was 14,454 ft2, which is 35% larger than the refueling bay floor area, and j represents a reasonable estimate of the actual sedimentation area avail-  ;

able in the refueling bay.

Case CIA: A secondary containment DF of 50 was calculated f or this sequence. All secondary containment input parameters are identical to those employed f or Case C9D, except the RBSYS flow rate is zero (RBSVS system not operational). The discussion for case C9D is therefore also j appropriate 'f or this case.

fumna ry: The Shoreham secondary containment hydrogen burn analysis j press;n:ed in Appendix D of this report indicates that hydrogen deflagra- l tions are probable for cases C9D and CADRF. The FAI MAAP analysis of l these two sequences indicated that burns would not occur. The - reason I that deflagrations did not occur in the FAI analy ses may be due to  ;

MAAP's low hydrogen generation rates during the core degradation phase  ;

of the accident. The absence of hyd rogen deflagrations in the FAI '

analysis represents the single greatest uncertainty in their assess-

. ment. If hydrogen burns had occurred, the resultant DFs would probably be lower than those predicted by the FAI analysis. It is probable that the uncertainty associated with this issue is greater than that of all ,

the other items discussed below. l The single cell modeling approach employed by FAI may be reasonable for Cases C9D :nd CADRF, but may not be reasonable for cases such as l CIA, in which the primary containment blowdown enters the secondary con-tainment low in the reactor building, or the RBSVS system is not opera-tional. The influence of this simplification on secondary containment DFs is sequence dependent and cannot be assessed in the absence of  ;

independent calculations.

Both conservative and non-conservative assump tions were employed by FAI for the C9D analysis. It is possible that the impacts of these assumptions would be self-compensating. Therefore, with the exception of the hydrogen burn issue noted above, no clear basis exists (in the absence of independent calculations) for modification of the DF for this sequence. Independent audit calculations should include a sinulation of the RBSVS and unit coolers and a more conservative es tima te of sedi-mentation area than was employed in the FAI calculations.

, Correction of an erroneous heat sink area employed in the CADRF sequence would result in a decrease in the calculated DF from 10 to approximately 6. The occurrence of hydrogen deflagrations could reduce

27 this still further, but the magnitude' of ' the reduction cannot be deduced -

,in ~ the absence of detailed, sequence specific calculations.

The use of a. single cel1~ reactor buil' ding . ' model and a non-

~ conservative sedimentation. area in the: CIA simulation is the source of -

'. significant uncertainty in the calculated DF' f or this sequence. . .Thel use of a single. cell building 'model .is of ' particular concern in: this; ,

sequence, :since the primary; containment blowdown enters' the reactor '!

building in.the lowest possible elevation, and the RBSVS is not operat- )

ing - (mixing - the building atmosphere): during:this' sequence. A.combina-  !

tion of higher hydrogen concentrations and a multi-cell model would lend )

to hydrogen deflagrations.7 However, the combined effect of higter -

hydrogen _ concentrations, lower sedimentation areas,- and 3 a . multi-cell model on the recalculated DF' is dif ficult to . predict. - It ;is unlikely, however, that . such : calculations would' result in estimated . DF. values higher than those claimad by. FAI. -

l i

j r

'i l

J

28 i

APPENDIX C REFERENCES

1. " Severe Accident Analysis Of The Shoreham. Nuclear Power Station -

25% Power," FAI/87-14, Fauske & Associates, Inc. , March 1987.

2. S. A. Hodge and R. M. Harrington, " Considerations Regarding Certain Aspects Of Severe Accident Mitigation Afforded By Operation Of Shoreham at Reduced P owe r ," letter Repo r t, Oak Ridge National Laboratory, June 12, 1987 4

i 1

i i

I 4

s l

1 1

I

29 Appendix D

~

ORNL SHOREHAM SECONDARY CONTAINMENT-CALCULATIONS-This appendix describes a preliminary multi-cell Shoreham secondary containment MELCORI model developed at ORNL, and summarizes the results of calculations performed to assess _ the probability of secondary con-tainment hd deflagrations during the C9D: ATWS and CADRF LOCA sequences.7 rogenThe goals of this analysis are (1) to assess the potential for secondary containment hydrogen deflagrations during the . C9D and the CADRF sequences, and (2) to evaluate _ the imp act of hydrogen burns . on secondary containment integrity and thermodynamic conditions. Although the model could easily be applied to the. evaluation of Shoreham second-ary containment decontamination factors,L such evaluations' could not be conducted during the limited time available f or this analysis.

The C9D and CADRF sequences analyzed here are defined to be identi- d cal to those described in the Shoreham 25% power PRA. . Case C9D is an ATWS initiated by MSIV closure. ' The primary containment is vented f rom the wetwell airspace when the primary containment pressure reaches 60 f psig and from the drywell when the primary containment pressure exceeds )

70 psig. The vent line is assumed to f ail at its junction.with the 1 RBSVS ducting (reactor building elevation 101 ft.), and the resulting  !

harsh reactor building environmental conditions are assumed to f ail all reactor vessel injection systems._ The RBSYS circulation system is assumed to be operational throughout the entire course of the accident, but the cooling effect of the RBSVS cooling coils and the unit coolers  !

is ignored. <

Case CADRF is a seismic initiated recirculation line LOCA, with a I coincident primary containment (drywell head) f ailure. The primary con-tainme nt blows down di rectly into the refueling bay through a 3 ft2 l hole. The RBSVS and all unit coolers are assumed to be unavailable. j 1

The primary containment vent (case C9D) and blowdown- (case CADRF) - {

histories (steam, nitrogen, oxygen, and hydrogen mass flows) employed in .]

these simulations were generated by S. A. Hodge with the~ BWRSAR code,3 '

and are believed to be best-estimate sources for both sequences. The prima ry containment vent or blowd own source for each. sequence is' i injected into the appropriate region of the reactor building and result- _

ing secondary containment response is predicted by MELCOR. The follow-ing sections present descriptions of the preliminary ORNL Shoreham ]

secondary containment MELCOR model and the results of the two sequence analyses. j

. ORNL MELCOR Shoreham Secondary Containment Model: Figure D.1 is a schematic representation of the preliminary ORNL MELCOR model of the Shoreham secondary containment. The model consists of 13 cells or sub-

, compartments (Table D.1). Each of the 'six main floors ~ of the reactor building is divided into two regions - a northern. half- and a southern half. The flow areas (Table D.2) between the two halves and between I

a

l u

C W

E C-m e

u

. C O 8 4

- 1 y w. m m e m m -x a ...

g

=

C C O 3 8 m

Q s e z s E ~5 O E 6

. e 2 05

>z u

a m0 q' '

m a e<

z :c x w

C>

wa W d C T1 2* z z z z z z u&

gsa -

z z z z z z at

9 0 9 9 9 9 ET*

R % 5  % > 4 5 O >

w w

w w

w w .

=J J J J E J .

W W W W W W Q 5

m m. s g $  ;; j

=

o n e '

to O 5 e

j i

l l

Table D.1. ORNL 13 Cell MELCOR Shoreham Secondary Containment Model Cell Data y y ,, Structural Surf ace Areas (m2 )

Floor Ceiling Walls Steel North 8 ft 3361 345 329 1022 35 South 8 ft 3361 345 335 1022 35 North 40 f t 2416 329 329 734 33 South 40 ft 2416 335 55 734 26 North 63 f t 1637 329 369 482 33 South 63 ft 1637 55 307 482 100 North 78 f t 7 in 4040 369 422 983 37 South 78 ft 7 in 4040 350 397 983 35 North 112 f t 9 in 5148 422 537 1072 42:

South 112 ft 9 in 5148 397 504 1072 40

- North 150 ft 9 in 2835 354 352 682 35 South 150 ft 9 in 2835 321 327 686 32 Refueling Bay 20323 1023 1252 2036 102 I

i.

i i

l e

i e

Table D.2. ORNL 13 Cell MELCOR Shoreham Secondary Containment Model Flow Path Parameters Flow Area Path

. (m2 )

North 8 f t -- South 8 f t 128 North 8 f t -- North 40 f t 16 South 8 f t -- South 40 f t 10 North 40 f t -- South 40 f t 93 North 40 f t -- North 63 f t 15 South 40 f t - South 63 f t 289 North 63 f t -- South 63 f t 73 North 63 f t -- North 78 f t 7 in 19 South 63 f t -- South 78 f t 7 in 38 North 78 f t 7 in -- South 78 f t 7 in 176 North 78 f t 7 in - North 112 f t 9 in 23 South 78 f t 7 in -- South 112 f t 9 in 48 North 112 f t 9 in -- South 112 f t 9 in 242 North 112 f t 9 in -- North 150 f t 9 in 18 f South 112 f t 9 in -- South 150 f t 9 in 51 North 150 f t 9 in -- Refueling Bay 20 South 150 f t 9 in -- Refueling Bay 45

. Ref ueling Bay -- Environment (normal infiltration) .029 Refueling Bay -- Environment (Failure) 1. m2 @ 0.5 psid

, (linear ramp between 0.5 psid (3447 Pa) and 483. m2 @ 3.0 psid 3.0 psid (20684 Pa)]

l i

i l

)

i

31 i

floors are based on measurements taken during a site visit and analysis

. of plant drawings. All s tai rwells , hatchways, and gratings are accurately reflected in the inter-cell flow path characterization-utilized in the model.

Four heat sink s t ructures are represented in each cell: floor, i ceiling, inner and outer walls (one structure), and a " miscellaneous l s teel" s t ructure. The miscellaneous steel s tructure is intended to rep-resent structures such as piping, ducting, pumps, and gratings. The floors and ceilings are assumed to be 9 inches thick, with an insulated outer boundary, while the walls are assumed to be 1 foot thick, with an insulated outer boundary. The miscellaneous steel is assume d to be 0.25 in. thick, with an insulated outer boundary, and a surface area equivalent to 10% of the floor area in each cell except the southern 63 ft cell. The surface are of the miscellaneous steel in the southern 63 ft cell is larger than 10% of the floor area, to account for the multitude of gratings, pipe hangers, and metal support structures located in that region of the reactor building.

The RBSVS is represented in simplified manner. The RBSVS circula-tion and exhaust system is modeled as a set of (constant volumetric flow rate) flow paths between the appropriate compartments and the refueling bay. The information regarding the location of the RBSVS suction points and their associated flow rates was provided by L1LCO.4 The RBSVS exhaust train is assumed to draw suction directly from the refueling bay and exhaust to the environment. The secondary containment makeup air (infiltration) flow is provided via a constant-area leak path between the refueling bay and the environment. This infiltration path is char-acterized by a flow area and loss coefficient which will result in 1160 i cfm infiltration at the design RBSVS refueling bay -- to - environment pressure dif ferential of -0.25 psid. The RBSVS cooling coils and unit coolers are not simulated in the present model.

The primary containment blowdown sources (steam, nitrogen, oxygen, and hydrogen) are taken f rom the BWRSAR case C9D and CADRF analyses per-formed by S. A. Hodge, and are believed to represent "bes t -es t imat e" sources f or both scenarios. The primary containment vent source f or the C9D sequence is injected into the northern 78 ft. 7 inch compartment ,

which also represents the 101 ft and 95 f t mezzanine level (the 101 f t floor is believed to be the location of the primary containment vent ,

duct f ailure). The injection location for the CADRF sequence is the  !

refueling bay.  ;

I The vent flow in case C9D is assumed to enter the reactor building '

with a temperature equivalent to the wetwell atmosphere temperature -- a

~

conservative assump tion. The blowdown would actually enter the reactor building several degrees cooler than the wetwell temperature, due to the expansion (throttling) f rom primary containment pressure to secondary containment pressure. The blowdown in case CADRF is assumed to enter the refueling bay with a temperature equivalent to that of the drywell, 1 which is assumed to have a pressure of 16 psia just prior to the seismic-induced containment f ailure.

i l

I I

l

i 32

. Case C9D Simulation Results: . The . secondary containment analysis ~!

for this sequence is initiated at 90 minutes. into the accident, at the time the primary . containment is vented. (and the vent duct is assumed to l fail). Figure D.2 displays the integrated ~ steam, nitrogen, oxygen, and hydrogen vent flows into the secondary containment throughout the~30-h period of the analysis. The temperature of the incoming material is plotted in Fig. D.3. The reader should note that all.of the plotted.

results for this sequence are referenced to the time of vent failure, i rather than the actual start of the accident. The elapsed time between  !

accident initiation and. a specific event can be' derived by adding 5400 a to the times depicted in the plots.

Approxima tely 650000 lli of steam, 12400 lb of nitrogen, 500 lb of oxygen, and 1800 lb of hydrogen are injected into the secondary contain-ment over the 30 hour3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br /> duration of the analysis. Since the flow through the vent is choked, the steam. is injected in a' linear . manner over the' duration of the accident. -Virtually all of . the nitrogen (which ini-tially served to inert the primary containment) and oxygen have been vented from the primary containment during the first 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> of the accident. Throughout the course of this analysis, hydrogen produced in the reactor vessel is flowing via the SRVs into the pressure suppression pool, and f rom the suppression pool into the reactor building - via the prima ry cuatainment vent. The integrated hydrogen flow into the reactor .

building correlates directly with the core melt phenomena occurring l

. within the reactor vessel. The initiation of hydrogen vent flow corre-sponds to the onset of core oxidation. .

The accelerated rate of hydrogen j vent flow at approxima tely 28000 s is due to the collapse of second l

, radial region of the core. The increased hydrogen flow rate at 50000 s i is due to the collapse of the third radial region of the core.

4 Figures D.4-D.6 present the pressure and temperature histories of  !'

each of the 13 computational cells. in the model. The-initial decrease in the cell pressures is an artifact of the non-equilibrium initial pressures specified in the model input and should be disregarded. - The prompt pressurization effect of the vent flow exceeds the capacity of the RBSVS to maintain negative secondary containment pressure, resulting in deformation of the refueling bay wall panels and the opening of a'  ;

3.2 m2 (equivalent area) hole from the refueling bay to the .outside.

I During the next 14.6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />, the combination of lower source temperatures and steam condensation in the secondary containment result in rather  ;;

stable pressures and tempe ratures throughout the reactor building and i refueling bay.  !

Figures D.7 throcgh D.13 display the compartmental steam, nitrogen, .

l oxygen, and hydrogen mole fractions throughout the course of the analy-sis. At 52683 s, (16.1 h into the accident) the hyd rogen concentration

, in the northern 8 ft elevation compar tment reaches 8 mole %, and a-deflagration is initiated. At this time the hydrogen concentration in eve ry cell of the secondary containment is above 7.5% (due to RBSVS

• mixing and high building connectivi ty) , and the burn immediately propagates throughout the entire secondary containment. The burn lasts f or only 3 seconds, bu t produces peak reactor building pressures of

ornt. 13 coll molcor -

t,horeham cou c9d.

6.o r r- r-

'~-

b.b -

, gyg i

4 5.8 - - 11.0 l

t.6 - 10.0

- e.e

} e,

% - e.e g 1.6 -

_g

- ,.e i 3.0 -

.Q X s.s U C 3.s y s.e i n.e e.o t.b a 3.6 88

. Lg 0.b p e#'# . g,g c.e ~_,so.e "as av.o

. , , , , , , , , ao I o.o so.o so.o so.e ro.o em.o m.o so.s no.s ti.e 1103 st 1

-- int.g. r0 . w c. <

---int.o. 2. u

-- int.g. i.z .<, ore.

j ornt 13 cell meteor - uhoreham case c9d i

l w.v -r m 75.0-he

- to.s

.. - .m f m.e - me i m.. -***

f 3

- me.e IWLO - . rt a **

e am.o *

- l is. .

i u.e i h

o a

sm.s I

.i

. ..e l

4

- 150 .0 no.e -

)

im.o a A8

. go,o e

e.o r , , , , , , , ,

aa c.s so. a 4.o in a en.s to e no.s m.e m.e 90.0 se.a sia.o 3

tie.110 si

. <nt.9. .t. . ooc o.

[-- l l

l Fig. D.2. Shoreham case C9D integrated steam, nitrogen j oxygen and hydrogen sources.

7 f

1 i

j e

I 4

l 1

]a or,it 13 cett ,,,teng . choecham caso c9d i

.. , I N

I n.s

<JA.e

33d.l 42J.s ,

- M.0 em.8 W.S M.8

.A 488.3 -

. ggg,s f

e 418.0 Jul.9 w_

. (

ms ~

g4 4.0 3 55.0

.... ""5

.m 277.5

.. m 272.5

1. E 8 . , ,

e0 10.0 As.0 E .0 es.O to.a 60.9 F0.0 mi s en a sina.0 eso.e n 003.i

[- .~.o 7c7t .p ]

i I

d

.I i

s d

. 1 i

Fig. D.3. Shoreham case C9D source temperature. I i

I

'3

I l

orril 13 cell roelcor shoreham coco c9d l 1 was.o 'r r---

i.

I m.. .

i 1

e .. - '*

I I

• m. .

H . s.e j i

7m.. . ...

2 -. 9 . . , =

l n  ;

. .. o i h o* ; l

. 4 t us s

} . e.s s Sm.s

... 3 me.

0.3 e...

U ~ - ...,

\

u., n - s .-

m. . . . . . . . . . . r. **

eo is.o m.s m.e ea.o hi.e no o Fo o an.a se.s saa no.s (L I103 mi j

- _ o,8 n. et.. + _ u, n n. .t.v

--- south 8 ft. elev -&- south 63 ft. elev i j

--- resth south 4040f ft.

t. elevelev 1 J

l Ort al I3 Coll nielcor - foliort4 vim COSO Cfkl iL.u - - ~r

- w.m 15.0

17. 5 ns.o -

o ,, .m j

ns.s kl 1

() is.76 s

- 164.0 - ,

()

L "o sia.e

() is.a .

o e

(. no.o is.m

() is.75 ,

sm.a . t

() - as.so me -

~ IS. A see.o ()

l j h o

4

• { itJb

- ar - a u  :. g c.

t e - O c -- -!? ' C:

anu.e . . . . . . . . . .

.=

o.s io.e as a m.o m.o m.a m.a Po.o a.o m.s son o na o 3

(Lee i10 el sorth 8 (L. elev + r w th O ft. elev

. --* i.outh 8 ft. elev + seuth f23 f t. elev

- - - eiurth 40 (L. elev south 40 IL, elev

_-~

Fig. D.4. Shoreham case C9D reactor building pressure and temperature - elevations 8 ft, 40 ft, and 63 ft.

ornt 13 cell melcor shoreliom coso c9d O - s.o es e - []

me.o . () '**

un . t} e.e i I j 208- e3' gy I *

"" I e.a .

t e k, " ' ' g y e.s o 2

san 0- h, I) - 0.4 d I i 45 0.0 -

g 3 - c.3 m.n h

' (- " -

r.

r u.a us e - e; T. O 4

- e.s ae in.e de s.e ei.e 0.n 0.a de 0.0 0e IEn.e He.8 time (103 el l

-- r w th 70 ft, elev - D rw th 15G (L. elev

- -- south 78 f t. elev -O- south 150 ft. elev.

-- rorth (12 ft, elev

--- south 142 ft. etuv 1

ornt 13 cell meteor - shore-hom case c9d i

n.....___,.._._._,.._.

ns.a

- is.m IH.b it.re na.a . ss.a k L

c. no.o "" ,,

c.

Q

• (1 is.m e a e

! us-L > is.so ins.a ll I I is.a 104.0 l l is.m H

[: l i - 64. 5 Www.g- , big - Ahggd

• nou.o , , , , .

a.a in a m.a m.e m.o 'd.e so.e to a so.e eu.a im.o sio.o tLee (103) sun th 76 f t. elev -G- swth 150 f t. el +

, -- south '/h fi etev --O - sowtsi 150 ft et'v e

- - ru th n 2 ft. .te,

' -

• an]eth 112 I t . elev Fig. D.5. Shoreham case C9D reactor building pressure and temperature - evelations 78 f t 7 in,112 f t 9 in, and 150 f t 9 in.

oriil 13 celt. melcor - shoreham caso c9d hu.e ,

suu o - am.e a

sAs 7.

us eu.e 3

~ u.e - C

{ - sm.a -

5 5 [

4 um .e $

h . em.e u..

an.n euo.s

- Am.0

~

Pe.a - ,

l 100.4 I am.a .

o.o so.e m.s no so.s u.o an.o po.o so a m.e im.e s w.o '

u 410 3s) 1 - r-u o s I 1

ornt 13 cetL molcar - shoreham case c9d

• ipe.a -r,-

IM.El 101.M - II E

. is.m sala no.ob ms.m k sw.4 -Hm

~

ri 9 sez.w -

. H.es e f afu a h a  !

,5 sw.m '"* A a

888 4 H.75 ins .w

- nt.70 wi .a IDI .m - 14.06

, ,i-y,- L~ - H.w lim.a -

r- , , , , , , . , ,

u.o so.s ao si.e en a wo so.s so.e m.o so.e sta e e so.4 3

u no .i

(_ r.,r.ungw3 o l Fig. D.6. Shoreham case C9D refueling bay pressure and temperature.

e m

/

e s _

d ds e 9 c

o , , __

_ e s , __ d a s ,,_ .-

e c  % e m  % - d a

i N - n lc -

d e o r

o ,\ i t

fs i

.x '-- - -

_ e. te v a

, d3 0 e r

s

^

e 8

1 e

l e

o g d mi

^ -

L i

m ~

t

-: , n m

~ -~ _ d e

o

~

i t ~

_ t l ~ i s -

e

e. . '

d s c ,T' -

o p

3 s - -

e m 1

g' '~ -

o t

i m g s

e c

' i222

'l

- ,s sho, e E

f .

o r .... r

-% tttt ifff e 8888 h N'  %' e p

/ s.ss s e s e 8

e s G F h s

s e

m t

e 2

e a

e e.

e - - o m

g,$E- *g -

t a

e g n

i d

l i

e u b

r

, o

_ t s * .

g, c N _ a C - -

e r _

o r

_ _ g S - -

O , _ D .

_ 9 =

C -r .%' _

C -

R H s

._. - g N '

e

, s a

4:

l oi -

v

,' -~ 4 c

l 5 -i '

. , le m g g3 a h

0 8

1 e .

't -

f , -

e rh u v 5 ,

m ot C

~ _

i t h u -

l e -- _

_ S o s

m r

.d

-- 7 n ll: 4 s

s D

. a .

(

% _g .ht 3 -

gr 1 - ,% _-

o - L i o l

n F n

}

i e222 sh c .

- - t o'ts r ,b ttti f

r ~ ieIs

~~ ~ ti It0a 8 e - N' ~ f ..i.

e ee .

e. s s r. s s i y e.

e a e e e e e g e e -

,i UP - e o*

1

~ -

s e

e s

e l

( _,_,e _ _

,e t

s _

_ 9 C s

- , _ e

,e G e _ _ s S -

e O  %

C w_ a m

.' - _ ,n u

u ,s ~ -

n t

~ ~- e o le v _

'- -- ie F i

W t M s .m' - a S  % . - e.

s e v a

s3 0

e

^ l .

e

, t

. ,. 4 P .

,tl, e - o P e

.j .u m O v -

., i -

{

C.

~ ._ , t n

O ,

s o E

n -

r .~

,e i

_ t -

I. i le -r 1

- ~. - ,w e s 3

l C

k N* g .%

- e.

o p

m o

t -

% .w , 7sa c

\ o

^ t22G

~-x s

dI - \ shce e

t i

- * ~e re r -

\ tttt t T-s IiIf e

\' unu0 h

- h '"w. e s

ti41 te.s p

s

. s s F 6 s , 2- e o o u

e -

s e e s e e e e e m

- t

,o $a eg - a _

e g ~

n i

d l _

e i u

b u a

m r

o e

t

. c e a u

! a

- n es e -

S - ~- -

r -

C -

- e D O r * -

- rsu 9 S - - C O ..

C s" == - e. e -

a - " - iree s _

a r

i c

IH* _

'f O

t x ,,.~' - -

.s o.

7 m a

t t h .

IS -

es

e. a eh r

' - e3 rt

~

0 1 a u .

-  % t h o .

C 3

e *e,a

. e S s i s m .,.

C E

. _ i _

- t dt

- .dn

_ e n

I r _ .e 8 a

. - _ Dh lt - s .- e t -

f a , ,' , im .r go _

?s - _

3 i n e 3 l

- s w

,e e.

F t

s' , a f

l i , \' '.,-

t2a2 sh n s

r E

, \'

gs e

e tLtt 0

4 e ifff s

,\ -w, n00,10_

i4 a

\ ' ( e ns n_ .

e e e a s . , 2 s. e.

e e e s s e a a. m -

h j3y e,I

g _

/

d

• 4 9

c _

s ,

a _

g s

o p _

c '

m ,

_. _ g a s _

h e  % ,

r -

n _ g n o s -%

_ o h _ i -

s ,%' , 1e t a

% g3 0 v cc I' s' ,"[ 1 i

l e

g mie e

e l ,s. . t o _

m ,% g

~. n l  %. - o l .

i e * ,

t c \ ,~.

, i s

3 s,

- s _

1 ,

o _

, p .

l -

• , r3 . m n \

N ' t222 shon o .

r c .

o \ -s , ....

., tttL s

ffif e _

' 31'l3 r

\ \. , 6666 e _

, o6.s h _

e. s e p . s

• p 3 e s a e a

• i a

2 e

' e e.

s .

,* '..8 y *.g

~-

- t o

m a

g _

n i

d l

i u

b a

w r s

o t

, - c

.e., _ e s

a r - sa hi . - e r

'c e

1m e

D s

o 9

C c

e

.,% ~ -

is e

e

- e a

m 4 ~

- - s h v -- a e ~-

.e

e. c i

r o +

% +

p m

h '*- --roo - i a s A* e h .

g eh

- rt s

s ," t '0 1 i

ou r ,

.s , , e Tw - em e h o o r S s c - .

e i

t t -

d en e

,~- -- ,

.s . n e

- _ e 9 a t --

t m Dh e  % -- .m .tr c '

,N go 3 .

/ .

s in e F

. 1 s

/  %

,' a

,s

. t t.

\ iQ22 s as f i

r ,\ ' , - e

,s

.... 3 o ,

e t ttt ffIi 6

'y 3111

- N., s 6566

\ e e.nnn s

a s

e

. .s e

e

- e.

e - - .

'ojgi* 3a=

" l' ,

l ,

/

t  % ,,,

e 9 - _ ,

C -

e '

s ~ _ ,

o ~ 1-. .

r _

n 1

H '

le - --- - ,3 -

d - -

K ,% - - n f

S o

,N-

- ,. l a i _

~ - 3 t ,

\ ,

0 8 a r

o i

s - ,

,w m 1

e l

v e _

lc.

• s. , i t

e mt l

l s. . ,

_ n _

e + ~. , o C - ,m i _

s t 3

1 y i

s u

~- ,m i p

o de ,N n i222

\ ,

', mh a n m g

o -

, g' .... o s N-s r,, tttt ffes c _

- '\ 68t1a e 7777

\ . r _

S e. F.

• S 2 2

. mt . = . e a e e e e.

m e e s o.

---

h p

ci$[ .' -

--- s .

- - o _

m t

a _

g _

n _

i d

l i

u b

e e r

-  : o t

I

_. - .ee

. c a

9 t

s e

L O

r

,e e

D r

S v

- s 9 _

O ._ - C _

C  % . - e _

i

' ~ *n o e .

m i .- -. -- sh l

w

- s

- -- e at cu e .  % - - .e o

-s o

r ms lt '~ - a hd _

8

,N - e =

,N

.a a3 en 0 ra

\ , ,-

- 4 1

o

-s o ,

g q ,]- e *e -

e hh g , Tm m S t c , ,_

i t r d a _, s , ,

_ , -e .n o

I n

. k . ,

-3m 0

1 n ll. .- " _ -o e

. , .i C

._ n g7 e

- 's -

1. i 3 , - - - Ft 1 _ <

~

. e f

' . .m e s'

tn _ ~-

tmt n 222 8 7

-e r N' o. . . .

o ,

\' ., rm t

fftLis

t. -

)' 61la 1t

. 1777

\

( N e ne. e., .

. s e , . s , 1 a.

• m e , e e e , e o

• o - .

,* $[ .. E - ---

)

e. _

e i

s

e. _

_~

e 1

, _ n a

9 __

't -

c - ., _

o - _ e.

s - .. _ S

. a c

e.

a

.%' ~._. -

r i

_ d n w w o le ,

. e.

i t

-r ,' .e

'---~ _

y o . a l.ir  % . v .

s a te e g ,

_ d 3 l

e 0

, 3

4. l, 1

r ae . - -

o ,

[s ,e.

e e o

~ , s -

lc.

, L

( n e --. ,, o m e i -

- , ,e t i

l  %. -

l e

,e s

o c  % p

' a

' - m 3 ,' - o 1

,a e c J

l s e

n \ .

tJ22 sI nn

, ' N ..,, r e

o a

.s. tttt ffff "N

• e h 2222 l1l1 p

_ l1l1 s e e

\ \ e.

s saa* o .

a 8

e s

a a

e s s a. , e. m e e e , e - - t g 4.$f

- -- a _

- - g n -

i d

l -

i e

b u -

r o

t c

a a e

t e

r e

e. D 1

4

,  % . - ,es 9 _

9 - C _

C m . -

_ - - ,te e t

e ,

- a s

s _

- - a .

a .

t - ch C _. * - - , e.

mu t

r.

, * - - es u

u w

- - ao

, h s l

z  %

- - e

,ee.

t

- rd

'M ' -

r on tt s'

' - h a _

iL -

a is S N' -

iea, I h

- _ s D s

. "e, I I .tr .

'W i v . 1 o .

4 s'

\,

s

, .w == i 1 n _

t l . ,

e D n m ,

.- , a i -

ie

~ .

g9 l.

l i e

e o Ft c , _- rn f 3

- -- - 2 1 _

- e 1 a

a . 1

, la t. t2 2an2 s _.

s

.\ ' Me ....

t ttt o ,

\ res Iffi g N 2222 11I1

\

t e 11I1

\ e . nn e e p

=. s i a e.

e o o e e e s e - -

,* g* e - - -

1 a

e s

e d  % _, __,ean 9 ,

t c __

e ~ __ _ a a

s - _ s o __

c  %

w n

u ,% .~._ _

,s e.

e n

o l

e s -

' e. i r -- - ,e y t o

i ' a v

f s s _

a e y ' _ ,e la

- n3 l g _

01 e r * -s, 1 o 9* .s , ' 8,m m e e -

c - ,

n i

l / t e s o

m q ,ao i

'- , t t ll - , i e

c g -- , ,o e_

s o

\ i p 3 '

m 1 ' -

o .

-, ,e an c -

t s e i

r -

tsl2'. en22 e r '

s e ..,. r e

r_ .%

< ,e tttt s ffif

\' N 0l h

"' M55M t p

e I11I s

I \

g s e r .

e e*6a o .

g a s s. s. 1 2 , e. m e

e e e e e , e - "

t g J8* 1E

^

- a g _

n -

i -

d l

. i u

b r

o t -

e c .

e a n

i e r

a 1 , _- _- ,en s

D 9

9 -

C c

e

- ,e a

e s

• s s .

o. r. - ah r '

. - ct

' s.

u m .- -

,s e

mo

o. -

g

% ~- as l . -

~ h e -

g

,e.a ed w.

--- - rn h

s P si h oa q y' w

vn e.

0 Sh t

e ra"eII s l c.

e .ro p  % e e 2 n o - . , s L t 1 l , n m

~ - ._ ,

,e a Di 3 _ s l

. , 9

. l g

e a it c y .a Ff a

3 s . 0 1

.g

' _ ya e 5

. t i

i

\

+ ' ,

J .

t. h2 2. 21 1

r s' - . . .

o

- e. tttt.

g g'.,- ,o s ff(i

'  %. m!wW B j= iltI i t\ e

\ /I e .

g.

g s

a e.

e t

e e e.

e s.

e

e. i e e J. *

" e

e. . nn n_

g:.:t *I . -

O ornt 13 cell melcor - shoreham cose c9d e.o a.e -

e.a -

e.7 \

s e.s -

kg g

.a

. g \,

4*5 N.s~e ,

.g

• W ,#6*w, i e4 s N

. \ 'g s.3 'N= %,.-. I I

s. a \.

%.s

8. 5 l,

,,,,,,5 ,'**""*=*=-. . . . ..

, ** I

,,. ~--~~-........

e.e . . . . . . . . . .

a.o so.o m.a m.e to.o so.s so.e 7s.o so.s sa.e iso.e un o 3

tLeo 110 si

- refuel-boy ste

--- refuel boy h2

-- refuel-boy c,2 rofwel-boy n2 Fig..D.13 Shoreham case C9D refueling bay atmosphere composition.

2

33 i approximately 6 psid, and peak reactor building atmosphere temperatures '

of approximately 1200*F. Secondary containment burns of this magnitude might result in f ailure of ducting (such as the RBNVS exhaust ducting I

located near the - site of the primary containment vent failure) and equipment, but their impact of Shoreham's secondary containment integ-

. rity (outer walls) is unclear. High internal pressure capability (in excess of 3 psig) was not a design criteria for. the reactor building walls.5 The peak induced refueling bay pressure is approximately 1.4 paid, j resulting in additional deformation of refueling bay wall panels and an J increase in the effective refueling bay wall failure hole size to 177 m . It' should be noted that att ptote presented in this Appendix employ a one minute plot resolution (time intervat betveen plotted

{

pointe). This explaine uhy the predicted peak temperatures and pree- f sures are not accurately ref7ected in the plots. The peak induced re- )

fueling bay temperatures are in excess of 800'F. The RBSVS is assumed i to continue to operate following the bu rn. Additional burns are not predicted during the remainder of the 30 hour3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br /> calculation, due to a I combination of low hydrogen concentration and high steam concentrations within the secondary containment.

The results of this analysis indicate that the use of best-estimate hydrogen source te rms would result in secondary containment hydrogen

. deflagrations during the C9D sequence. The severity of the bu rns appears to be increased by RBSVS operation and Shoreham's high reactor building connectivity. The possibility of burn-induced f ailure of the RBNVS exhaust ducting at a location very near the site of the primary containment vent duct f ailu re, and the impact of such a f ailure on secondary containment DFs should be analyzed further. It should be noted that the effect of unit cooler operation would oe to increase seconda ry containment mixing and decrease steam concentrations within the secondary containment. Since much of this cooling capacity is  !

located in the 8 ft elevation, the effect of unit cooler operation could be to decrease the time interval prior to the first bu rn. Additional calculations should be conducted to examine the impact of RBSVS f ailure, and RBSVS cooling and unit cooler operation on deflagration phenomena.

Case CADRF Simulation Res ult s: The primary containment blowdown sources f or case CADRF are displayed in Figure D.14. The temperature 'of the source material is plotted in Fig. D.15. The initial rise in the integrated steam, nitrogen, and oxygen sources is a result of the-blowdown of the primary containment from 16 psia at the start of the accident (time = 0. 0 s) . The increase in ficw rates at approxima tely 3000 s is due to the onset of fuel, clad, and canister melting within the core. The large increase in flow rates at 11,300 s is a result of l

, the in-vessel steam spike (and associated SRV flow into the primary con-tainment) resulting f rom collapse of the second radial column of the core. Approximately 89,000 lb of steam, 14,550 lb of nitrogen, 600 lb

, of oxygen, and 1,265 lb of hydrogen are injected into the refueling bay during the 4 h accident analysis period.

s Ornl 13 Cell melCor sharehom cose cadrf.

l 7.a ,

s.s . #

- n.e l 1

e s.s . ]

6.6 - - ta.e

• s.e - ]

n l

-/

I.s ,,,,

so.e

'h O SJ-

.J* .

e.o 3.s . g .

3 g s.e - l

. s.o

{ l.5-a.s 4.8 5.1 88 2.0 j

-s * -~::"- * ~ ' ~~ ~ ~2 '~. Y *I-------.

" o.o s'.e i.e i.a 4'

..e .'.e l i? 4* da ti.e si.e si.e ti.o 5.0 44.0 l 3

tl.e (10 el int.+ n2 so ece 3 int.g. o2 vo sce j

- - - integ. h2 sa wce  ;

i

. q ornl 13 cell. melcor shoreham cose codef i l

i 4s..

I m.*

a..

l

'i m.. . d i

1

. ro.a

.y no- l

. m.. I

'h n.. .

1,

.! m.. - e s

. . a.o ,

1 h

e 11.0 ,

- as

... p . . .

- te.D o.s u i.e 24 2.. .. .. i i.. i.. a . ,; .. ii.. ii.. i;.. g., i..

Lt no 3.i l - int.g. .t. .o,

.l Fig. D.14. Shoreham case CADRF integrated steam, nitrogen, oxygen, and hydrogen sources.

e ornL 13 cell melcor shorehom cose cadrf em.e sm.o . en.o tse.s m.s -

<A.0 eaa.n - .a,

~"'*

2 me,e .

me.s _

i. em.s -

mA g.

.e  :

- m.e un o -

- a50.s I

no.e - - am.s

- ano.e ma.s -

- m.e I

3*o.a - , , , , , , , , , , , , , ,

a.o a.a 3.e 2.s e.e s.a s.a 7e e.o 4.s m.e si.e ia.o es.a so.a ala n.e ti (103 s1

[-- c. (..p. l Fig. D.15. Shoreham case CADRF source temperature.

l 34 Figures D.16-D.18 display the secondary containment compartmental pressure and temperature histories as predicted by MELCOR. The initial primary containment blowdown results in a peak refueling bay pressure of 0.76 psid, which is approximately equivalent to the highest refueling bay pressure induced by the hyd rogen burns which occur later in the

- sequence. Some refueling bay wall panels are deformed by this pressure an equivalent refueling bay wall failure opening of s50.7 pike, m 2. producing The peak refueling bay temperature induced by the blowdown is approximately 170*F. Since the primary containment f ails directly into the refueling bay, and the RBSVS and unit coolers are not operating, the only mechanism for mixing of the refueling bay and reactor building atmospheres is natural convection.

This effect is demonstrated by Figures D.19 through D.25, which depict the compa r tmental steam, nitrogen, oxygen, and hydrogen mole f ractions throughout the secondary containment. The natural circulation flows do not become well developed until late in the analysis period.

A refueling bay hyd rogen deflagration is initiated at 3898 s (1.1 h) when the local hydrogen concentration in the refueling bay reaches 8 mole %. The burn lasts for approximately 3 s, producing a peak refueling bay pressure of 0.8 psid and a peak temperature of approxima tely 910*F. The burn does not propagate 'into the . reactor building. The pressure spike deforms a few additional refueling bay

, siding panels, resulting in an increase in the equivalent refueling bay 2

failure hole size to 58.4 m. Tne peak burn-induced reactor building pressure and temperature are ~0.5 psid and 180*F, respectively.

Following this burn, natural circulation produces an increese in some of the reactor building compartmental hydrogen concentrations (to as high as 6 mole %), but the entire reactor building is steam inerted i during the period between tb two refueling bay burns. 1 A second refueling bay hydrogen deflagration is initiated at ,

11130 s (3.1 h), when the local hydrogen concentration in the refueling  !

bay (which has been above 6 mole % since the 6000 s point in the acci-  !

dent) again reaches 8 mole %. This burn produces a much lower pressure pulse within the refueling bay and reactor building than did the first burn, but produces a peak refueling bay temperature of approximately 1000'F. This burn, like the first, does not propagate into the reactor building.

The results of this analysis indicate that the use of best-estimate hydrogen source flows would result . in secondary containment hydrogen deflagrations during the CADRF sequence. The hydrogen burns are less severe that those predicted for the C9D sequence, but would induce

, failure of the metal siding in the refueling bay, providing a ve ry direct fission product transport path from the drywell to the environ-ment.

Summa ry : The results of the analyses presented in this Appendix indicate that the use of best-estimate severe accident hydrogen sources

ornL 13 cett, melcor shoreham cose codef su.s 12.8 sne . -[]

. p3 i

l' - 1m.8  !

c l }l 2a.e - - me Q. ,

w s- - ss.*

l und- . i s..

2 I

, i(%a i

. 317.5 - A ,,,,,,

h h m.s . ,

I '

is.e anas-e ... -

~

sno.e -

ru e/.

1 f . os.s l 'W$

]pa m ,,~~~~~%~~Yl

, . m.e

/ ..e e.e s'.a i.e i.e i.e i.e i.e i.e i.e i.a d.e ti.e ti.e ti.e is.e d.e is.o stee (103 el

- north 8 ft. etov -O- north 63 f t. etov

--- south 0 ft. etav --G- oovth 63 f t. etov

-*- north 40 ft. ete.

-- - south 10 f t. stew orol 13 cell melcor - shoreham cose endef

, in. ..

io..a .

is.as see.m .

ins.n - '55 g)

()

. sg.m is.a -

O se.m O

-& mam- g) [

n O -

o ina.ps . . H.m

(

O

' tea .no -

f (I

um (

{ is.n. u l O - H.m

()

108. 5 - . w,n ice .w -

ti s s H.10 464.A

{

tol .m . , [ u

• HJS e ,

o.e s'.e i.e i.e i.e i.s i.e i.e i.e i.e d.e si e d.e d.e d.a d.e m.e ]

(Lee 1103 al'

- twrth 0 ft. ete. - O - north 63 ft. tev g

--- south 8 ft. etov O south 63 ft etov <

-- north 40 f t. vtew

--- unuth 10 ft, stov j i

Fig. D.16. Shoreham case CADRF reactor building pressure '

and temperature - elevatiens 8 ft, 40 f t, and 63 ft. ,

i i

i

ornl 13 cell msLcor shoreham case codef 370.0 .

3D.e us. fF 2

. . in.e moi.

• - O '"'

.. ,. i .

i1 \ ,

- va.e

, ,, () - san.e

)

(i C

3 se.e - j1 g .

. iso.e

- i f i l

4-' (

g .e.

.... q o i

• j>

f I.i

(

l g . sm.e . -

j j aa4 - .

.; . g ;l Q , , , ,

, , , . b i U sio.e apA Ju%p f

d

. ..e sn.e<

~"

am.el. {

^**

e e s'.e i.e is i.e i.e i.e i.e i.e is a.e aA as an A a * A tie.1103 el rwr th 78 ft. elev -G- north 150 f t. elev

--- south 70 ft. stov south 150 (t. elev

-- nurth 112 ft. etov

--+ south 112 ft, eLov ornl 13 Cell melCor shoreham cose codef iou -

as.es

'8' 8 ' . og.m

' {

in.s -  !

. se.m

.e . .,.

g se.m g

_a ..

,,, se.m e 2 O k

~, O .' se.as

). ina.e .

k sei .6 - g, - se.fo

,, ()

ll .A, l te es pmi + . _rr -

pC soi s $ ' .s "*'4 %"7""**"[pg I 3'{ ~ ~ %i j $* . it..

j t iI se.m I

• '** ,e.,

e.e <> in i.e is is i.e is in i.e 3 a.. a.e as as a.e as l tie.110 en l

-O- north 150 ft, elev ')

- rmr th 70 f t, ete.  ;

-O- south 150 ft, elev

--- south 70 ft, ele.

' --- . noro, i 2 r t. et.,

• - + south 112 ft, elev I

I Fig. D.17. Shareham case CADRF reactor building pressure  ;

and temperature - eleiations 78 ft 7 in, 112 ft 9 in, and 150 ft 9 in.

l 3

ornL 13 cell melcor shoreham cosa codrf M.S

- es.O .

1 m.. .

M.0 DED.0 - ,

m.. . - m.e

~ -

e MS wo.o I

saa.s - '

• Ms e i

em.s .

- 35.0 ano.e-am.o me s.

W~ # - lnDA

• • ' o s i.e i.e i.e i.e i.e i.s i.e s'.e i.e d.e si.e si.e de is.e do is.o time (103 si j - ,..r u m3t ,og ]

- ornL 13 cell melcor shoreham cose coder 1

.I e

' . gg,g )

103. 5 -

- 34. 5 103.0u .

I im.M. . ns.m 802.49

84. 5 g "a nu.n. L

~"

( tm.co- {

[ ,

(

106.M - ,,,,3 001.te .

64.75 405. 5 -

~ 88'#

IDI.M is.M . "" 84.00

}f r ,

, 6.0 s'.a i.e i.e i.0 i.e i.e i.0 i.e i.e E 8 ei 9 6i.8 55.0 89 8 85.0 86.8 Gas 1103 e's j - r.r .n, 3 i.o, l Fig. D.18. Shoreham case CADRF refueling bay pressure and temperature.

f 1

r x -

_. n .

t -

c - _ n .

. e s

o

_. 4a .

c 1

. lu m

i

_ _ - n e

_ n, r _ _ o h

o ._ _ i t

s ._. H g. a

~ _ v

_. e r

~

_. 5~ .,. l e

t c

e o

~

~

._ n<t m ~ - n

~

o l

l '

~

-  :- i t

e .

- i c ~ - < s 3

~~ o p

1

~

~ - -

m t

n

~

- n t.a22 o c

r

~

i s'. e

- - tttt i

r'ir r _

- - a8ee e

• - , ,t.

- s*.. h p -

. :. . " '- s -

- o

.J e .

m -

~Z t a

g n

i d

l i

u _

b _

r -

o -

. t _

. c _

a _

f r e r .

e _. _ a i

F t

u _. _ r. R .

c i D

_ A e

s _

4.h. C o

c n

u n

i.

e s

a c

l w

e r

_ _ ii m

a h

o s

, _ R g.

., h e

r .

_ _ _. oh

, ht H,,_

r o

_ S u o -

l c

e ,

~ _

n ,,

9d

.s m 1 n -

m -

. a l D l

e _ -

. .h t c

gr

. 3 -

- n i o Fn 1 -

t -

._ n i

0.

,2,2 2 n t

f n - e r - ..

o _

- 8 s.

gatt g,a.

grSr ea nnon 3se- ., m _.- -

,jl i 1l

. f ,, -

g r - -

t  ;

a c -

e -

n

. s s - 7, o o  ; i c _- ,

t n _

_,_ _ ; a

_ v to _

. , e e .

_ a l r

o

_ . e

. ,l -

h _  ; e s _

, ;3 0 9 n 1

s o

r _ _

e i o ..

__ {; t i t c ,

~ i l - , ~, s e - __ -

o m  ; p

- _ m L o L

e

- ~

%, c c -

e 3 -

- . r e

1

- m t222 h t

sh o n p n - ..,. s r

o - - ,

tttt ffif o

-  ; m

-- 00o0 t y 7, 44t1 eeas a

. . . , t 1

=- g

. . . e u .

u '

- n i

ci$! *,I d

l i

u b

r o

t c

a e

r F

n a

R D

A f

- T l

e -

- C -

a , -  : e c - - .

r s e

_ ~

a a s ,s c o _

n.

c _ _ m

_ _ - . a m ,  ;

h o -. _ _ -

e b -

r

- . r .

e _ ,_ - a oh r _ _ - ht o

h s

. l s S u o

s

- v.lHo 0d r - _ ;m e

2 n _

a i

o _

t ._

c _ ~ , _ . -

t . ,_

_ :. Dh .

m e , - ._ ," .tr

- - . go

. . 1 t - i n

. t e

- -~. F c -

K-t f

3 1

n a

0 _

- t222 4

, t n

-- r.

~

s9 en r - -

~. lftttt o . -  ;

ff _

-- 0oI0

, e .

7 T. -

4tW4 r nnn

. . . . . s. .

i

. .a. " - -

,i oJ i.f -

e as f

, .ees

. - [

.M -

h c

c m \'

'e s.

p V +.

ies e

1. g e E, -a e  % .

ii s

t o *- . es n c _ ran

_. o n __ i w , _ .e. t w _, _

a ie .

.eo v e

r t o _. l i

_ e.

le e ls -

.s 3

e 0 1

,e 1

%, . -. e n r ~ -

,e. m o o ., -- _s i i

c .

t t

t e e .

., . ,6 i

n i - s s

• s. o p _

~. ,5 t.

l c

e e

~'

W# '

a, m

o c

~

3 ~

~ -,3e e 1

- ~ e r t ~ ~

e. tJ2d e4 oe e

n * ~

,a h l

~*

r .- ....

tttt p

o  %

• e ffff s

,e.s 3l31 o

• 6E66 m e ee e t e o e

e.

e e s.

e n a. ,

e. a

. e e e e e -

- g 53 y

• E -

- n e i d

l i

u b

r o

t

e. c

-e s a

-e e

f * -

,s r

-e h I -

-e F n ,  !, *

- ,te R c

i f - - D

-a A e *_- ,-n C s -

o ~., --e e c - iis s e e a n , --

,t e

c l

u , -- -

e _

e m

- - eo t a

-n , h w ., - e .

,ee le is , ~

- 3 rh ot

, - r., e 0 ,e h u

,, 11

~  ;, - S o e

r s o ,, .e

- m l

c . - J-7 i t

.d i

a 1 n w - f s

e 2 a s -

• -~

t

- - ~

~

' .e.s Dh t 2 t e ~ * .r c m "

"'A ~ te.e go

. - ~

~ - i n 3 ~ - F 1

~

~ r3s.

_ t

~ ~

~ o . f te " - e. tJ22 sI ce

,a

• - 3 ri - " ....

tttt 6 o ^ ,ee f_ e

, ,- - ifff li31

- e EG66 non e.

e e e e t , a .

e. e.

i e e e e e e a e

,a 1 8 g 3 E - - --- _

j a

i _

a#

• .n

. 'N. 3 -

" P p .

7n

~a S - 'n _

u _

s

.~ , n o

,n i t

'g a

- a l v

e

- - - 7,, 3l e f80

~

1

- e -

- - e m

-- - ', i t n.

- . o g

e i t

i

'~~ _7 s

- o s' ~

, . p

- i m o

. c

~3

. e

- t. a h 2. d e r

" -. .... e

._ 7. tttt fref h p

N '* Hee 8 s

' U.

7r77 m..s o

' 1 , '

t m

g . -

a

.i32 . a* -

g n

i d

l i

u b

r o

t c

a e

. r m

. F i $ rn R r - + - D e A d ,

b. .

- ,.n C o _.

s .

c .

'q

- . e e

- s _

m,n

_

• a s _

o .i ~~ - - u c .

h c , ,N , -

, . mt m, ,. _. au m.n h o u _ -

_ _ es t _ -

e ,

r r

o

- .. od h ., _ _.

_ - e. is hn s _ _ _

3 S a

- 0

, 1 8

.h t r ,

~ .

. ,,. - W 2 r o - t 2 o e _ ,

n _

t -

,. - D e .-

n m _ -

-._ .- . .i _

- g' _

t -- s -

h_N.

i7

. t .- - F e ~ -

E. t c '

• f 3 _-

~ -.

,3 1 8

' ~

~

-u ,ag 7 _

. t n -

~ - ,,,,

r -

- . ,t,,,g ,g o , - ,

- a,,3

_ y777 3 . _

_ . o ,, ,, ,,

. . , s s , 1 a .

. . e , _ _.

,*,:1?*

• t _

a e

e s

,e. s e f J .

f r <*

_ei .

_ ,s m ,* .

e c .

e.

ow . ,i e

s 1*

\,t -u _

s

e. -

o \ _

,a s n c i o -

s s ,e. i

- s

. i t o -

e a l

e -

. ,e v s

e h

r o - .

,s e

1, l

e s _

3 -

_. e 0 1 -

_ ,s t

_ _ n o

r ._

,rs.  ;

g o

_ i c

l e  % ._. e.

s t

i w_,s. _'

- s m ~ -

,e o p

l . .

l y _ m e w, .,.e. o _

c ,

\ - ._e .

c _

s '

_e 3 -

- .s a e

1

- tg2d r l . -- e.

sg oe e n "

,z .,,. h r

• tttt ff(f p

o , -

, e. 2222 s

, - i 33gl o g- \ _e 33gl m

s ...s t a

s e , s s. .

.e. e s e e a e

. i.

e a.

e " e.

e

_ g

,= j[ 3E ___-- i n

d l

i u

b .

r o

t c

a .

o s

e i

r v

- ,s ,s e

F -

f

• e e

R l

< -.. e D o *e . ,c i A c l

- C e.

Y .

e e

. i-4s e s k .Y -

s o \

. ,su a .

c .

ch

,-se t

a . .~ .

mu u ~

t e

ao _

le , -- ,s h s r - .

e e o - _ rd h -

.es 3 1

on s .

h a

,es Ml S

- I h

_. - - . t e , _

. e. e . r n ~ _ -

,r it 3 o e - .

. - 2 n t

e * ,es m _. D n t

l e

'r, .,

N __,_~~ s -.e i

e g9 i

c .

% .. o,,- es

- F t w ~ f 3 -

e _

1 - .s e 2 t .

- -.ao tJ2d sI oe 1 1

. n -- ....

t ttt r -

-e ffff

- o .

'ss

.t 2222 8181

- N \-- 1111 n

e v r e nnnn a

e.

e o s

e s.

e e

s e

e "o e.

,ig[

• I - - -

a r

e. -

r ,t

,n -

. r 's - * ,c i

P

[-

. , e.

a i -

c 3 5 g e.

o .n -

s 4 ._

. a 'g ,g e .e, a c , m1. 8 '\ . i n '-

o l

m oi

~

e

,V i

t a

e

~ 4

,s i v r

o

_- e s l fi ~ .- .e le e m s -

s _

- 3 t i

s'

- - 0

.e 1

._ - t p

%"6 -

e r -

. m n -

o

_ o

.r i c y, t i

t r, e t -

e .s s

n ~

• %, -., i s

'(

e e o l

l e *}

__ .s p m

g ..

o 3

c ] g e-c -m

%% .s. .

e y3 1 a L .

- t222 sh o n r -

t * -

.e.

a e h n *-r .. .

hp r,

tttt -

r ffff o

.e.i MMSM O s =

9 e.

IIII o n

1a ' e ssas t i

e. e . ' . s . s a

• e. -

a i s e

' e a a e

• e - +

3ahtie * *- g n

i d

l i -

u A -

b e r o

t c

a h e

e r m

- F R

r , D e[/

s e

__. .u ~ A t

a

.  ?.p

.e.

e e C c gM, .

a s.

. e. e e

s 1

,' I Q

g_ ,si s

a .

ch o lg V , ,su Ig ,i* t c

m ,

, ~. ,

mu ao a hs h

e

- e. e .

~ r. rd .

r o - ._

e on m h - ,s h a i

-- 03 m

_ S i s _

3 a

- ... e 0 l ._

e r 1

%e,s -

I

.tr .

r .;,. .

w*

e m 4 o o

c g _ , i t 2 n -

l

, ne **- r-e D n .

e  % s, s a

.i m

~ _ -

l l

',T

, ,u i

g9 -

e .

. Ft .

c ,- g ' , ., r f _

, re f, ) ,, .

3 e -, 0

,s 1 - e - - 5

- - t222 1

. t n

,s a 9 en

..,. +

f,[

. tttt =

r 1 ., ifff o . ,e. I11 i

V 4 *.M d

- g , I' 1 j

.e I8II nnnn s

a e ._ ' . s s  : ' o.

a

. i s e ' e .e e ' e

,2j i -

C m s

i l

1 i

e i 1

t

-)

ornt 13 cell melcor shoroham case codrf i,

I l

0.5 -

.j

..,. -'\

\ ,/ N, . . s* -u I 6-q g

o-g

,f (/ 1, i i )

j i .

e.s. I), ,/ l

. i /

1 . .. . .-

. \ e/

')

4 1

.... j

'l

. ..e ,. .............-. .  ;

i

s. . ~. .

,- a,.s.

e.. .  %,s '

r is e,a , ,

,, o, -- w , , , , , , ,

s a

e.o s.o s.a r.a e.s e.o io a an.o aa.s si.e it.s si.e is.o s.o i .e 4.0 3.s ti 1103 el

- refuel-tm3 st.

- - - refuel-tmg h2

~~ -

r.fuoL-tm3 o2 1

--- coru.t-im3 n2 .

4 1

1 l

1 l

J l

i j

)

5 Fig. D.25. Shoreham case CADRF refueling bay atmosphere composition.

35 ,

would resultI. in hydrogen deflagrations ' in the ~ Shoreham- secondary con-tainment ~ for the ' C9D and CADRF- sequences.. . The severity of these -

deflagrations appears' to be increased by operation of . the ' RBSYS.' opera .

tion. c BWRSAR/MELCOR predictions for the C9D- ATWS sequence indicate that -

a severe global burn . would : ' occur at- approximately 16. hours into.the.

b accident, ' producing peak' reactor: building pressures ;of '6 psid and tem-peratures as high . as .1200'F. . BWRSAR/MELCOR predictions for the . CADRF.

sequence indicate that . refueling bay hydrogen deflagrations would occur'.

at 1.1 ' and 3.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> into the accident, . with peak induced pressures and '

temperatures .of 0.8 psid - and 910*F, . respectively. Additional analyses should be . conducted to ass'ess: the' impact' of these deflagrations on secondary containment integrity. and fission ' product - - retention . capa-:

bility.-

l

/

)

I l

i

)

l 5

3

r T I

36  ;

1 i

APPENDIX D REFERENCES l i

• I

1. F. E. Haskin, et al. , " Development and Status of MELCOR," SAND 86- - )

2115C, Sandia National Laboratories, Presented at the Fourteenth i 5 Water Reactor Safety Information Meeting, Gaithersburg, Maryland, October, 1986 '

2. Severe Accident Analysis Of The Shoreham Nuclear Power Station - -

2 5% Power, FAI/8 7-14, Fauske & As sociates, Inc. , March,1987, Se c- -

tion 5. 1

3. S. A. Hodge and R. M. Harrington, " Considerations Regarding Certain Aspects Of Severe Accident Mitigation Af f orded By Operation Of Shoreham at Reduced Powe r", Le t te r Re po rt , Oak Ridge National Laboratory, June 12, 1987.

4. Telephone conversation with Fli Chiang, LILCO Staff, June 1,1987.

5. Telephone conversation with Lap Cheng, LILCO staf f, June 10, 1987.

i A

h l

i l

I 1

l e

l

P-  !

37 .j Appendix E

-ACRONYMS AND SYMBOLS- 1 1

1

• Alternating Current AC ATWS Anticipa'.ed Transient Without Scram

BWR Boiling Water Reactor CRD Control Rod Drive Cs1 Cesium iodide Cs0H Ce sium - hyd roxide

.DF Decontamination Factor FAI Fauske and Associates, Inc.

'HEPA High Efficiency Particulate Air HPCI High Pressure' Coolant Injection LILCO Long Island Lighting Company LOCA - Loss of Coolant Accident LPCI Low Pressure Injection System MK Mark MSIV Hain Steam Isolation Valve J -NRC Nuclear Regulatory Commission f- ORNL Oak Ridge National Laboratory PRA Probabilistic Risk Assessment RBNVS Reactor Building Normal; Ventilation System s, RBSBS Reactor Building Standby Ventilation System RCIC Reactor Core Isolation Cooling RERS Reactor Enclosure Recirculation System a SGTS Standby Gas Treatment System SRV Safety / Relief Valve-l

'l i

1 l

e i

i e

1 J