ML20093E206
| ML20093E206 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 09/30/1984 |
| From: | Beahm E, Hodge S, Carl Weber, Wichner R, Wright A OAK RIDGE NATIONAL LABORATORY |
| To: | NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| References | |
| CON-FIN-B-0452, CON-FIN-B-452 NUREG-CR-3617, ORNL-TM-9028, NUDOCS 8410120013 | |
| Download: ML20093E206 (199) | |
Text
{{#Wiki_filter:.. . ._- - . . - .-
~
NUREG/CR-3617 . I ORNL/TM-9028 a OAK RIDGE'
. NATIONAL' LABORATORY Noble Gas, lodine, and Cesium i
Tran' sport in a Postulated MARTIN MARIETTA Loss or. Decay Heat Removal Accident at Browns Ferry R. P. Wichner C. F. Weber S. A. Hodge E. C. Beahm A. L. Wright
?
i e if; , l I Prepared for the U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research Under Interagency Agreements DOE 40-551-75 and 40-552-75 g OPERATED By gg o g g3 e4093o MARTIN MARIETTA ENERGY SYSTEMS,INC. ' CR-3617 R PDR
. FOR THE UNITED STATES ,
DEPARTMENT OF ENERGY , l
, . _ . . . - . , - , - . ~ , .
~
.. e , 1 a-
..=
c
.l l
e r v - 9 , Printed in the United States of America. Available from National Technical Information Service J U.S. Department of Commerce 5285 Port Royal Road, Springfield, Virginia 22161 Available from
- GPO Sales Program Division of Technical Information and Document Control U.S. Nuclear Regulatory Commission -
Washington, D.C. 20555 This report was prepared as an account of work ' sponsored by an agency of the
. United States Government. Neither tha u nited States Govemment nor any agency thereof,' nor any of their employees, makes any warranty, express or implied or ,
- ar,sumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or
! represents that its use would not inf ringe privately owned rigt ts. Reference herein s - to any specific commercial product, process, or service by trade name, trademark,
manufacturer, or otherwise, .does not necessarily constitute or imply its endorserr. recommendation, or favoring by the United States Govemment or -
. any agency ..,areof. The views and opinions of authors empressed herein do not necessarily state or reflect those of the United States Govemment or any agency
- thereof.
l - t: hi
- o.
t f- $ I ~ g
~..
i e [.< i ( f {1
5g: NUREG/CR-3617
-0RNL/TM-9028 Dist. Category RX, 1S o
Chemical Technology Division Engineering Technology Division s NOBLE GAS, IODINE, AND -CESIUM TRANSPORT IN A POSTULATED IDSS OF DECAY HEAT REMOVAL ACCIDENT AT BROWNS FERRY R. P. Wichner S. A. Hodge C. F. Weber E. C. Beahm A. L. W'right Manuscript completed -- July 20, 1984
' Date published . Augus t 1984 e-Notice: This document contains information of a preliminary na-ture. It is subject to revision or correction and therefore does not . represent a final report.
Prepared for the U.S. Nuclear _ Regulatory. Commission Office of Nuclear Regulatory Research Under Interagency Agreements DOE ~ 40-551-75 and 40-552-75 NRC FIN No. B0452 n, Prepared by the
- OAK RIDGE NATIONAL LABORATORY Oak Ridge, Tennessee 37831
* . operated by MARTIN MARIETTA ENERGY SYSTEMS, INC.
for the U.S. DEPARTMENT OF ENERGY
, . under Contract DE-AC05-840R21400 r
111
. CONTENTS Page ABSTRACT ......................................................... I
- 1. INTRODUCTION ................................................. I 1.1 Objectives and Approach ................................. I 1.2 -Outline and Acknowledgements ............................ 2 References for Chapter 1 ..................................... 5
- 2. ACCIDENT SEQUENCE
SUMMARY
.................................... 6 2.1 Introduction ............................................
6 2.2 Bvents Prior to Containment Failure and Loss of Injection ............................................... 8 2.3 Events Following Loss of Injection and Containment Failure ................................. 10 2.4 Reactor Vessel Conditions ............................... 10 2.4.1 Predicted core heatup progression ................ 10 2.4.2 Predict.d corium debris temperature in the reactor vessel ................................... 11
.- 2.4.3 keactor vessel conditions ........................ 11 2.4.4 Containment vessel conditions .................... 12 2.5 Predicted Flowrates Between the Reactor Vessel. Wet-well, and Drywell ....................................... 12 2.6 Reactor Building and Refueling Bay Response ............. 13 2.7 Fission Product Transport Pathways ...................... 15 2.8 Standby Gas Treatment System (SGTS) Behavior ............ 16 References for Chapter 2 ..................................... 18
- 3. FISSION PRODUCT TRANSPORT ASSUMPTIONS ........................ 44 3.1 Introduction ............................................ 44 3.2 Chcnges in Iodine an.1 Cesium Transport Assumptions ...... 44 3.2.1 Esticated release from drywell rubble ............ 44 3.2.2 Organic iodide production rate ................... 46 3.2.3 Iodine volatility ................................ 48 3.2.4 Cesium and iodine chemical behavior .............. 50 References for Chapter 3 ..................................... 51
- 4. AEROSOL DRODUCTION AND TRANSPORT ............................. 54
- 4.1 Introduction ............................................ 54 4.2 Drywell Debris Bed Behavior ............................. 54 4.3 Aerosol Production Rates in the Drywell ................. 56
- 57 4.4 Aerosol Transport in the Drywell ........................
1 iv i
-l l
Page -
-l '4.5 Aerosol Transport in the Reactor Building ............... 58 4.6 - Summary of Aerosol Transport Results .................... 59 References for Chapter 4 ..................................... 61
- i l
- 5. FISSION PRODUCT TRANSPORT CALCULATIONS AND RESULTS ........... 73 5.1 Initial Nuclide Inventories ............................. 73 5.2 Control Volume Characteristics .......................... 74 l l
5.2.1 Primary system volumes ........................... 74 i 5.2.2 Containment and building control volumes ......... 75 1 5.3 Calculational Procedure ................................. 75 5.4 Aerosols in the Reactor Vessel .......................... 76 5.4.1 Production and deposition rates .................. 76 5.4.2 Aerosol effects on fission product transport ..... 77 5.5 Noble Gas Transport Results ............................. 78 5.6 Iodine Transport Results ................................ 79 5.7 Cesium Transport ........................................ 81 References for Chapter 5 ..................................... 83
- 6.
SUMMARY
AND CONCLUSIONS ...................................... 126 6.1 Summary of the Work Performed ........................... 126 6.2 Summary of Results and Conclusions ...................... 126 6.3 Principal Uncertainties ................................. 129 6.4 Significance of the Study ............................... 130 APPENDIX A: REACTOR BUILDING AND REFUELING BAY CALCULATIONS ..... 131 A.1 Introduction ................................... 131 A.2 Results for the Loss of Decay Heat Removal Accident Sequence ...................... 131 l A.3 Combustible Gas Concentrations in the Secondary Containment Atmosphere ............... 135 A.4 Secondary Containment Response Without Sprays ......................................... 135 References for Appendix A ........................... 137 APPENDIX B: DETAILS OF CALCULATIONS FOR CORE-CONCRETE DEBRIS l BED BEHAVIOR, CORE-CONCRETE AEROSOL PRODUCTION RATES, AND AEROSOL TRANSPORT IN THE DRYWELL AND ' REACTOR BUILDING .................................... 159 B.1 Introduction ................................... 159 i l
. 1
V g Page B.2 Core-Concrete Debris Bed Behavior Calcula-tions .......................................... 159 B.3 Core-Conc'ete Aerosol Production Rate ~"
-Calculations ................................... 161 B.4 QUICK Drywell Aerosol Transport Calculations ... 162 B.4.1 Fixed QUICK parameters .................. 162 B.4.2 Aerosol production-rate data ............ 162 B.4.3 Drywell leak-rate data .................. 162 B.4.4 Drywell pressure and temperature data ... . 163 B.4.5 Drywell geometric data .................. 163 B.4.6 QUICK drywell results ................... 163 B.5 QUICK Reactor Building Aerosol Transport Calculations ................................... 163 B.5.1 Reaccor building aerosol production-rate data .......................... .... 163 B.5.2 Reactor building leak-rate data ......... 164 B.5.3 Reactor building pressure and tempera-ture data ............................... 164 B.5.4 Reactor building geometric data ......... 164 B.5.5 QUICK reactor building results .......... 164 References for Appendix B ........................... 165 APPENDIX C: ACRONYMS AND SYMBOLS ................................ 191
-6 e O
4:9 _ V. 1 4 3 L - NOBLE GAS, IODINE,' AND CESIUM TRANSPORT IN A T*: ' POSTULATED LOSS OF DECAY HEAT REMOVAL ACCIDENT AT BROWNS FERRY u R. P..Wichner S. A. Hodge C. F. Weber. ( E. C. Beahm A. L. Wright ABSTRACT This report presents an analysis of, 'he movement of noble gas, . iodine, and cesium fission products within the . Mark-I
- containment - PWR reactor system represented by Browns Ferry Unit 1 during a postulated accident sequence initiated by a loss of decay heat removal capability following a scram. The event sequence. has - been analyzed separately and is documented in a companion ' report. The event a..alysis showed that this
- accident could be -brought under control by various means, but the sequence with no operator action ultimately leads to con-tainment (drywell) failure followed by loss of water from the reactor - vessel, core degradation due to ov.Jheating, and re-actor vessel failure with attendant movement of core debris onto the drywell floor.
The analysis of fission ' product transport presented in this report is based on the no-operator-action sequence and.
. _ . . provides an estimate of fission product . inventories, as a
=*' -
function of time, within 14 control volumes outside the core, with the atmosphere considered as- the final control volume in the - transport - sequence. As in the case of accident. sequences previously studied, we find small barrier for noble gas ejec-A tion to air, these gases being effectively purged from the drywell and reactor building by steam and concrete degradation gases. However, significant- decay of . krypton isotopes occurs during the long delay times involved in this sequence. In contrast, -large - degrees of holdup for iodine and cesium are projected due - to the chemical reactivity of these elements. Only about 2 x 10-4% of the initial iodine and cesium activity are predicted to be released to the atmosphere. Principal barriers for release are ' deposition on reactor vessel and con-tainment walls. A significant amount of iodine is captured in y the ~ water pool formed . in the reactor - building basement after actuation of the' fire protection system.
- 1. INTRODUCTION
,- '1.1 Objectives and Approach The authors of this report are primarily applied researchers in various areas - relating to fission product transport under LWR accident e'
P
c 2 conditions, and in general, work in an experimental environment within , the Chemical Technology Division. It is not our purpose or inclination to produce a general purpose computer code for estimation of LWR accident consequences. Instead we have the following primary goals:
- 1. To provide a conduit for applying the most current research infor-mation to the methodology for computing accident consequences, and
- 2. To place all known fission product transport phenomena, physical and chemical, in a context that allows evaluation of their impact on the magnitude of the accident consequence.
As researchers in this field we know full well the uncertainties involved in the estimates presented in this report, dealing as it does with the transport or chemically reactive materials through regions of high temperature in complex chemical mixtures and varying gaseous compositions. With this insider's knowledge, we do not claim any specified degree of accuracy for the quantitative results on iodine and cesium release to the atmosphere. Such a claim from any source at this time would be premature in view of the number of poorly known factors involved in the transport of these materials. What we do state is that given the careful and detailed analysis of this accident sequence provided in Ref. 1. 5, and by utilizing the most current information bearing on fission product transport properties, the computed release rates are as correct as possible with the information available at this time. Since this subject is still developing, several technical areas have been identified that may have significant impact on the results. The reader is referred to Sect. 6.3 for our view of the areas of principal uncertainty. Currently, we are attempting to upgrade our calculational procedures to apply new developments in these areas as they occur. 1.2 Outline and Acknowledgements This is the third in a series of fission product transport analyses for postulated accident sequences at a Mark-I type BWR. Specifically, Browns Ferry Unit 1 is used for the model and numerous reactor-specific details provided by the utility-owner, TVA, are utilized for the analy-sis of the sequence of events and fission product transport. We would like to gratefully acknowledge the cooperation of the TVA Engineering Support Offices in Chattanooga and Knoxville in providing the numerous design and control details without which these studies would be extreme-ly difficult and much less realistic. In addition, we acknowledge the aid provided by CE reactor safety personnel who have contributed helpful comments and information throughout the course of the study. This re-search is eponsored by the Containment Systems Research B ranch of the Division of Accident Evaluation of the NRC Office of Research. This report presents an analysis of the fission product transport
- attendant to the severe accident phase of a prolonged loss of decay heat removal capability following a scram at the Browu Ferry Nuclear Plant. The postulated loss of decay heat removal capability involves
3 the loss of the power conversion system and both the pressure suppres-sion pool cooling and the reactor vessel shutdown cooling modes of the residual heat removal (RHR) system. With the RHR decay heat removal capability unavailable, the reactor decay heat energy would be concen-trated in the pressure suppression pool.
'a Earlier studies in this series have dealt with a postulated station blackout eventl *1 and a small-break LOCA outside of the containment vessel
- caused . by a piping failure in the control rod drive hydraulic system following scram.l* Both of these accidents are predicted to lead to loss of core cooling, core collapse, and breach of the reactor vessel .nd containment system in the very unlikely event that operators do not intervene with appropriate corrective actions. Without effective operator action, the Loss of Decay Heat Removal Accident Sequence, which is the subject of this report, also leads to severe results with the difference that containment failure here precedes (in fact indirectly causes) loss of core cooling followed by reactor vessel failure. In all cases, the fission product transport analyses (Refs. 1.1, 1.2, and this report) are accompanied by companion reports devoted to a description of the event sequence.l.3-1.5 For the convenience of the reader, a brief description of the acci-dent sequence of events is provided in Chap. 2, where we have followed our usual practice of abstracting sequence details bearing on fission product transport from the companion report which deals exclusively with the accident sequence analysis. However, a significant parameter in this event sequence has been modified so that the summary presented in Chap. 2 differs in some of the details from that in the original re-
- port.l*5 Information obtained from Ames Laboratory af ter Ref. (1.5) was published pointed toward a smaller projected drywell overpressurization failure area [2 ft2 (o,19 .m2 ) vs 10 ft2 (0.93 m2 )]. Thus the temperatures, flows, and pressures vs time shown in Chap. 2 differ from
, those presented in Chap. 8 of Ref. (1.5) due to this alteration.
In addition, a significant aspect of the event sequence is the pro-jected behavior of the reactor building which surrounds the containment vessel (drywell and wetwell). Description of this behavior was not provided in Ref. 1.5 and appears here in Chap. 2 and in more detail in Appendix A. In Chap. 3, three modifications to the fission product transport analysis performed for the earlier studies are described. A brief dis-cussion is presented regarding some recently recognized features of cesium and iodine chemistry not incorporated in this study but which may in the future significantly impact the reactor vessel transport model for these elements. In Chap. 4 are described estimates of the amount of aerosol pro-duced in the drywell as a result of the core / concrete interaction after reactor vessel bottom head failure and the transport of these aerosols through the drywell and reactor building. A point of departure from
*The following terminology is used for the reactor containment structures: reactor vessel, containment vessel including the wetwell and drywell, and reactor building.
6 L____
I l 1 4 previous studies 'is the use of CORCON/ MOD ~ 1. for predicting rubble tem- , peratures . and VANESA for aerosol production. Although CORCON is be-If eved- to be superior to the INTER subroutine of MARCH, the early ver-sion at our . disposal probably contains some developmental aspects which i for our case may have resulted in higher than realistic rubble bed tem- l
-peratures.
- Both VANESA and CORCON are developments of Sandia National Labora- j tory :(SNL) and we gratefully acknowledge the helpful discussion with j Sandia personnel regarding these codes. We are particularly indebted to Dana Powers for the VANESA results which were run at Sandia for our 1 case.
The results of the fission product transport calculations are pro-vided in Chap. 5. The study and the calculated results are summarized in Chap. 6, which also provides a discussion of the principal uncertain-ties. . The thermalhydraulic conditions in the reactor building and re-fueling bay during the accident sequence - are discussed in Appendix A. Appendix B provides the details of_ the calculations concerning the corium-concrete reactions on the drywell floor and the - subsequent aero-sol transport in the drywell and the reactor building. The acronyms and symbols used in the report are defined in Appendix C. e O
5 References for Chap. I 1.1 R. P. Michner et al., - Station Blackout at Brovna Ferry Unit 1-Iodine and Noble Gas Distribution and Release, NUREG/CR-2182, Vol.
. 2 (August 1982).
1.2 R. P. Michner et al., SBLOCA Outside Containment at Brouna Ferry Unit 1 Vol. 2. Iodine, Cesium, and Noble Gas Distribution and Re-lease, NUREG/CR-2672, Vol. 2 (September 1983). 1.3 S. A. Mcdge et al., Station Blackout at Brcuns Ferry Unit 1-Accident Sequence Analysis, NUREC/CR-2182, ORNL/TM-455 VI (November
-1981).
1.4 S. - A. Hodge et al., SBLOCA Outside Containment at Browns Ferry Unit 1 - Accident Sequence Analysis, NUREG/CR-2672, ORNL/TM-8119/V1 (November 1982). 1.5 S. A. Hodge et a1., Loss of DHR Sequences at Browns Ferry Unit 1 - Accident Sequence Analysis, NUREG/CR-2973, ORNL/TM-8532 (May 1983). O O e
l 6
'2. ~ ACCIDENT SEQUENCE
SUMMARY
2.1 Introduction The predicted response 'of Unit 1 at the Browns Ferry Nuclear Plant. to a prolonged. post-shutdown loss of decay heat removal capability has been ' analyzed and is described . in detail in a companion document 21 to
~ this report. In this chapter, we ' abstract . information from that docu-
. ment, emphasizing the aspects that are most-pertinent to' fission product transport analyses. The general outline of the accident progression is provided here for the convenience of the reader. Nevertheless, Ref. 2.1 should be consulted for other:important aspects of the accident sequence
.such as the effect of various postulated operator actions or equipment failures on the sequence of events, a detailed description of the resid-Jual heat ' removal system and ;its operational modes, and a discussion of
.the. effect of thermal stratification in the pressure suppression pool.
The Loss ' of Decay _ Heat Removal - (DH2) accident sequence has been listed among the risk-dominant accident sequences leading to core melt in every BWR probabilistic risk assessment (PRA) conducted to date.* It is .an extremely important sequence from the standpoint of fission pro-duct transport ' because the containment. vesself is failed by overpres-surization before there is a loss of injection to the reactor vessel. However, the long -period (about 36 houre) between reactor shutdown and core uncovery provides partial compensation for the unavailability. of primary containment when fission product releases begin. . In this study, fission product transport ' calculations are performed for -the period from drywell failure until time 3100 min, an interval of about 17.25 h. The decision to terminate calculations at time 3100 min is lsomewhat arbitrary, but all major transient events of the accident sequence have occurred by this time. The following listing provides a. general outline of the sequence of events:
' Time (h) Events 0 . Initiating reactor trip followed by M91V closure and failure ' of both' the suppression pool cooling and the shutdown cooling modes of the RHR system.
0 - 34.4 Normal reactor vessel water level maintained. Monotonic heatup of pressure suppression pool - by condensation of periodic - reactor vessel SRV discharge. Primary con-
.tainment pressure and temperature increase. SGTS
*This is the "TW" sequence ' first identified in the Reactor Safety -
Study, WASH-1400. i The term containment vessel refers to the wetwell and drywell. i:
p.w _ -- . 7 e automatically initiated on high drywell pressure sig-nal. Primary containment reaches failure pressure [132 psia (0.91 MPa)] at end of period and failure occurs at juncture of cylindrical and spherical sections of dry-well. Drywell begins to depressurize into reactor build-ing, and harsh environmental conditions there cause loss of reactor vessel injection capability. 34.4 - 36.7 Without injection capability, the reactor vessel under-goes a pressurized boiloff of coolant. Core uncovery begins at the end of this period. 36.7 - 41.5 Drywell blowdown into reactor building essentially com-plete. First cladding failures occur (38.1), melting begins for control blade sheaths (39.1), fuel assembly channel boxes (39.6), and fuel (40.1). 75% of the core becomes molten and " core slump" into the lower plenum of the reactor vessel occurs at the end of this period. 41.5 - 43.2 The fallen corium boils off the water in the reactor vessel lower plenum and attacks the reactor vessel bot-ton head. The bottom hermd fails at the end of this pe-riod and the corium mass drops onto the dry concrete floor of the drywell. 43.2 - 51.7 The corium-concrete reaction phase of the study. Cal-culations are terminated as of the end of this 8. 5 h period (time 3100 min). It should be noted that the information presented in Chapter 8 of Ref. 2.1, which concerns the events of the accident sequence after core
-e uncovery, is superceded by the corresponding material presented in this report. The information presented in this report has the benefit of re-cent improvements to the ORNL version of the MARCH code. Also, the 2-ft2 (0.186 2 area assumed here for the drywell opening upon its m) failure by overg)ressurization 10 ft2 (o,929 m opening assumed is believed to be8 more in Chapter reasunable of Ref. 2.1. than the For accident sequences that involve failure of the BWR Mark I con-tainment vessel by overpressurization, it is necessary to assume the failure pressure, failure location, and failure size. A recent study at Ames Laboratory2 .2 provides invaluable assistance in the first two tasks, indicating that the failure would be expected to occur at a pres-sure of 132 psia (0.910 MPa) by yielding at the juncture of the cylin-drical and spherical sections of the drywell. The Ames Laboratory study does not address the expected failure size, but in subsequent discus-sions, its authors have indicated that it would be expected to be less than 10 f t2 (0.929 m2 ). The 2 f t2 (0.186 m2 ) opening assumed for the fission product transport analysis discussed here is slightly larger than the minimum hole size necessary to terminate the pressure increase
- in the containment vessel. The assumption of a larger break opening would accelerate the depressurization of the containment vessel but would not significantly affect the fission product transport calcula-tions since, for the Loss of Decay Heat Removal accident sequence, core
, uncovery does not occur until 2.2 h af ter drywell failure.
8 For this accident sequence, in which the containment vessel has , already failed by the time the core becomes uncovered, it is evident that consideration of the fission product removal and holdup potential of the secondary containment is of paramount importance in the analysis of the overall fission product transport.
- Section 2.2 provides a brief discussion of the sequence of events during the 34.4 h period leading up to containment failure by overpres-surization and the concomitant loss of reactor vessel injection capabil-ity. Subsequent events are described in Sect. 2.3, followed in Sect.
2.4 by a detailed discussion of the progression of core melt and, af ter core collapse, the temperature of the debris bed in the reactor vessel lower plenum. The flows between the reactor vessel, wetwell, and dry-well are described in Sect. 2.5, and Sect. 2.6 provides a discussion of the response of the secondary containment, which becomes an important part of the overall analysis immediately af ter drywell failure. The fission product transport pathways are discussed in Sect. 2.7, and very important considerations concerning the performance and endurance of the SGTS are discussed in Sect. 2.8. 2.2 Events Prior to Containment Failure and Loss of Injection A detailed discussion of the events leading to failure of the pri-mary containment in a postulated Loss of Decay Heat Removal (DHR) acci-dent sequence at Browns Ferry Unit 1 is provided in Chapter 3 of Ref. . 2.1. The reader is also referred to Appendix A of Ref. 2.1 for a de-tailed description of the RHR system and its potential for providing re-liable pressure suppression pool cooling under accident conditions. Table 2.1 provides a timetable and summary of events for the first , 34.4 h of the Loss of DHR accident sequence, during which the primary containment remains intact. The reactor vessel is isolated by the FBIV closure that accompanies the scram and is assumed to remain isolated throughout the remainder of the accident sequence. The practical impact of this is that the main condensers of the power conversion system are not available as a decay heat removal mechanism. With the reactor vessel isolated, the steam generated by decay heat is relieved into and condensed within the pressure suppression pool. Since the means for cooling the pool are by definition not available in this accident sequence, the pool temperature increases monotonically. Evaporation from the pool surface causes a concomitant increase in the primary containment pressure. After one hour, the primary containment pressure reaches 2 psig (0.115 MPa) causing automatic actuation of the standby gas treatment system. Also at the 1-h point , the pressure suppression pool temperature exceeds 120*F (322 K). Plant Emergency Operating Instructions (E01s) require the operator to depressurize the reactor vessel when this occurs, so that sufficient reserve capacity is maintained in the pres- , sure suppression pool as necessary to absorb the discharge of a subse-quent large-break LOCA without exceeding other temperature limits.
w - 9
~
,,' At the 2-h . point, the combination of high drywell pressure 'and
' operator-induced low reactor vessel pressure causes automatic initiation L of the low pressure ECCS pumps. The high-pressure (HPCI and RCIC) in-
^
-jection . systems also remain available, although the pressure suppression pool - temperature is above the -. recommended maximum for their lube oil
> -coolers.* The operator continues to - have available much more reactor vessel injection capability than he needs.
. Four hours af ter shutdown, the level of reactor decay heat has sub-sided to the point where_ all necessary reactor vessel injection can be b . supplied by the continuously operating control rod drive hydraulic sys-
- tem, which provides cooling water to the reactor control blades. At the 8.6-h point, the operator must act to reduce the control blade cooling flow to avoid excessive water levels in the reactor vessel.
Although . they would not normally be needed, it is worthwhile to note - the ' occurrence of loss of availability of the high-pressure injec-tion systems in this very slowly developing accident sequence. The HPCI and RCIC systems would be threatened by inadequate lube oil cooling if
~ they were used to pump pressure suppression pool water af ter two hours.
The steam supply valves to ' these systems would be' automatically shut at the 13-h point because of high sensed temperature [200*F (366 K)] in the
. respective turbine rooms. Finally, the steadily increasing primary-L containment pressure would reach 25 psig (0.274 MPa) at the 14-h point,
thereby 'providing a high exhaust pressure trip signal to the. RCIC turbine.
The containment vessel design pressure (56 psig (0.487 MPa)]- would be exceeded af ter 21.5 h, but actual failure is expected to occur at a
~ 'E ~ much higher pressure.2.2 After 23.5 h, the primary containment pressure reaches 65 psig (0.550 MPa), and the reactor vessel SRVs are no longer operable . in the remote-manual mode.t SRV discharge terminates and the reactor vessel begins to repressurize, reaching the 1105 psig (7.720
- .1 . MPa) setpoint for automatic' operation of the SRVs at 27.7 h. SRV dis-charge and the concomitant heatup of the pressure suppression pool re-sumes at this time.
- At 34.4 h, the drywell pressure reaches 132 psia (0.910 MPa) and the drywell is assumed to fail by the opening of a 2 f t2 (0.186 m2 ) hole located at the juncture of the cylindrical and spherical sections of the u ~drywell. The - pressure suppression pool temperature at this time is l 343*F'(446 K), so a.large amount of steam is generated by flashing of the water at the pool surface as the containment depressurizes. The l' steam escaping from the drywell creates harsh environmental conditions in the reactor building (see Appendix A), and this is assumed to cause failure of the control' rod drive hydraulic system (CRDHS). Thus loss of reactor. vessel injection occurs simultaneously with drywell failure.
L . [' *The lubricating oil for the high-pressure injection systems is j( cooled by the water'being punped. L [ . tin the remote-manual mode, service air pressure on one side of a l piston works against drywel1 ~ pressure on the other. A differential l pressure of at least 25 psi (0.172 MPa) is required for actuation.
10 2.3 Events Following Loss of injection , and Containment Failure As noted earlier, conditions in the reactor vessel, wetwell, and drywell following drywell failure and subsequent loss of coolant injec- . tion at 2064 min (34.4 h) are reported here as computed by the FnRCH code, assuming a 2-ft 2 (0.19-m2 ) ef fective break area. During the time following drywell failure, the events tabulated in Sect. 2.1 following the 34.4-h point occur. These include: Time min h Vessel bolloff begins 2064 34.'4 Core uncovery begins 2200 36.7 First cladding failure 2330 38.8 Core melting begins 2407 40.1 Core slump assumed at 75% of core melt 2492 41.5 Reactor vessel bottom head failure 2591 43.2 Core / concrete reaction to produce 2600 43.3 aerosols and sparge fission products Downstream SGTS HEPA filters tear 2917 48.6 Following bottom head failure, the core / structure rubble deposits on the concrete base of the drywell floor. Initially f rozen, the rubble bed heats up due to decay heat, separates into molten metal and oxide layers and is sparged by gases evolving from the degrading concrete below and
- surrounding the bed. Rubble bed behavior is discussed in Sect. 3.2.1 and Chap. 4. Table 2.2 provides an abbreviated event timetable for the entire sequence with emphasis placed on events significant to fission
' product transport. 2.4 Reactor Vessel Conditions 2.4.1 Predicted core heatup progression Core temperature distributions predicted by MARCH are listed in Tables 2.3 through 2.6. According to the highly idealized model in MARCH, each of the 100 core control volumes is presumed to " melt" when an average volume temperature of 4352*F (2400*C) is achieved. Table 2.3 shows this first occurs at time 2407 min in the 8th radial zone at the 8th axial position. Tables 2.4 to 2.6 show temperature maps at times 2428, 2461, and 2491 min, respectively, the last just prior to the projected core collapse onto the bottom head. - It is important to note f rom Table 2.6 that relatively cool zones in the core exist just prior to collapse. Temperatures in the bottom axial zone at this time range from 702 to 3470*F (372 to 1910*C),
,n 11 _g
~ '
reflecting - the continued presence of. water. near this region. Also, the outer radialL zones remain quite ~ cool relative to the melt temperature due : to both a =1ower. decay heat source in this region and high radial heat losses.- According - to the - MARCH formulation, the ' core is presumed to col-
*t lapse instantaneously when 75% (user-specified) of the core control
= volumes achieve the melting-point.
.The . temperature map shown in Table 2.6 indicates that it. is reason-able ; to assume that the . outermost radial zone . (10) remains in place-
.following core collapse. The low temperatures in the bottom axial zones indicate that the - collapsed core still contains about 10% (except for.
. decay losses)' of its original inventory of - even the more volatile fis- +
sion product groups. Since rubble bed temperatures in the. reactor ves-sel bottom head are also relatively- low, about 10% of the t original core inventory of- fission products passes to the drywell floor following re-actor vessel bottom head failure, according to this temperature history. 2.4.2 Predicted corium debris temperature in the reactor vessel Figure 2.1. illustrates the temperature history of the rubble . bed in - the ~ reactor' vessel as predicted by MARCEL. Note, only one mass-average temperature is available, and an almost instantaneous quench from 4352*F
-(2400*C) is assumed to occur. at the time of core collapse (2492 min).
- This -is because the mass-average temperature decreases rapidly when the hot corium material instantaneously combines . with . cooler- steel struc- -
^
.. tures below the core . and boils away the water ' remaining in the reactor vessel bottom head.
Bottom head failure is predicted to occur at 2590 min at which time the debris bed temperature has reheated to 3385'F (1863*C), principally due to . decay heat. During most of the ~100 min time ' interval between core sinap and reactor vessel bottom head failure, debris bed tempera-tures are relatively cool for fission product evolution; therefore, most of-the inventory of fission products in the collapsed core is retained. 2.4.3 Reactor vessel conditions
- Figures 2.2 through 2.11 illustrate some of the conditions -in the l '. reactor vessel up to the projected time of core collapse - at 2492 min.
Figure 2.2 shows , the effeet of 'the relief valve operation on' reactor
' vessel - pressure, which averages 1100 psia (7.58 MPa) until- the time- of. ~
bottom head failure. The - expanded water level (which includes the pro-jected bubble volume) in the primary vessel is shown in Fig. 2.3. Note L that the- top of active ~ fuel (TAF) is uncovered at ~2200 min and that I. some water is _ projected to exist in the lower axial zone of the core throughout - the period until core slump. Each relief valve opening pro-duces a temporary increase in the two phase water level, each time l3 l' quenching some of the previously uncovered portion of the. core.
- Figures 2.4 and 2.5 show the projected masses of water'and hydrogen in the reactor vessel up to time of core collapse. The hydrogen is pro-duced by metal-water reactions during periods when the core cladding,
- channel box, . or control blade material is at elevated temperature and
12 sufficient steam is available to feed the reaction. The average core , temperature is shown in Fig. 2.6. Figure 2.7 shows the energy input from metal / water reactions. Note, about 5.6 x 107 kJ are added to core materials due to metal water reactions, primarily from the cladding. Figure 2.8 shows that even so, only about 15% of the cladding is esti-mated to be oxidized by the time of core collapse. During the time in-terval extending from initiation of the Zr/H2O reaction at 2300 min until core collapse at 2492 min, about 30', of the thermal input is esti-mated to be derived from the metal / water reaction. The progression of core melting is shown in Fig. 2.9. Figure 2.10 illustrates the MARCH predicted variation of gas tem-peratures exiting from the core up until the time of core slump. The temperature generally increases with time, reaching a peak of 2642*F [~1450*C (1723 K)] just prior to core slump. The exit gas temperature from the core drops periodically to the saturation temperature corre-sponding to the reactor vest.el pressure during periods when a safety re-lief valve is open. Surface and gas temperatures in the steam separator are shown in Fig. 2.11. Following core uncovery at time 2200 min, gas temperatures are predicted to increase significantly above surface temperatures [1274*F vs 626*F (~690*C vs 330*C) at core slump]. Although not pre-sented here, other gas and surface temperatures in the upper reactor vessel are predicted to follow the trends shown in Fig. 2.11. 2.4.4 Containment vessel conditions Figures 2.12 to 2.15 show conditions in the wetwell and drywell from time 2000 min (prior to drywell failure) to time 3200 min, when the MARCH calculations were terminated. The drywell pressure history is shown in Fig. 2.12. Note that about 300 min are required to depres- , surize the . drywell to atmospheric pressure through the assumed 2 ft2 2 (0.186 m ) failure area. The drywell atmosphere temperature history is shown in Fig. 2.13. Drywell temperatures increase significantly af ter failure of the reactor vessel bottom head because of the oxidation of previously unreacted steel and zirconium in the corium on the drywell floor. The necessary steam to fuel this reaction is provided by the corium-concrete reac-tion. The suppression pool water and wetwell airspace temperatures shown in Figs. 2.14 and 2.15 follow the saturation temperatures at the corresponding pressure fairly closely. 2.5 Predicted Flowrates Between the Reactor Vessel, Wetwell, and Drywell Volumetric flowrates between the reactor vessel, wetwell, drywell and reactor building are illustrated in Figs. 2.16 through 2.18. Figure 2.16 shows the flowrate from the reactor vessel into the suppression - pool via the SRVs and the T-quenchers in the suppression pool. The ef-feet of the on/off action of the SRVs is shown; peak flow into the sup-pression pool while one SRV is open is about 1412 ft 3/ min (40 M 3/ min).
13
, Note the fraction of time that the SRVs are open diminishes
-after ~2300 min. The _large flow pulse at time 2492 min of ~106,000 f t 3/ min '(~3000 M 3 / min) is due to multi-SRV-opening at core slump into the water ' in the reactor vessel lower plenum, following which the reactor vessel is dry, effectively terminating flow to the suppression
*~ pool by this means.
Figure 2.17 illustrates the flow communication between the drywell and wetwell through the vent pipes. Note that af ter drywell failure, generally the flow direction is from the wetwell airspace to the drywell due to water boiloff from the suppression pool as the containment de-pressurizes, beginning at 2064 min, the time of drywell failure. Core slump at time 2492 min causes a large surge from the wetwell to the dry-well, whereas the predicted effect of reactor vessel bottom head failure is initially a large volume surge from the drywell to the wetwell fol-lowed by a flow pulse in the other direction. Figure 2.18 shows the predicted flow rates from the dtywell into the reactor building due to steam boilof f from the suppression pool and generation of concrete degradation gases af ter reactor vessel failure. Af ter reactor vessel bottom head failure, the solid line indicates the flowrates predicted by MARCH utilizing subroutine MACE for estimating steaming rates from the suppression pool and INTER for flows from the degrading concrete. The INTER results were not used in the present study. These were replaced by the CORCON-MODI flowrate estimates rep-resented by the dashed line in Fig. 2.18, where the steaming rate f rom MARCH has been combined with the CORCON results for times following vessel melt-through (2590 min).
- Prior to reactor vessel failure (Fig. 2.18), flow from the drywell to the reactor building is driven primarily by suppression pool boiling; hence the flow is primarily steam. These predicted steam flows are quite large, initially up to 71,000 f t / min 3 (2000 M / min),
3 a magnitude
. which ef fectively flushes the drywell every 2.3 min during the time interval f rom 2050 min to 2300 min.
The volumetric flows driven by the corium/ concrete reaction as pre-dicted by CORCON MOD /1 were used for the fission. product transport cal-culations and are represented by the dashed line in Fig. 2.18. Note that after the initial pulse at-time 2593 caused by reactor vessel fnil-ure, . the predicted flowrate of gases generated by concrete degradation combined with steam f rom the pressure suppression pool generally equals about 6200 ft 3/ min (175 M3 / min) from time 2640 to the termination of the calculation at 3100 min. 2.6 Reactor Building and Refueling Bay Response The release of steam and gases into the reactor building begins when the drywell fails at time 2064 min (34.4 h) as shown in Fig. 2.18. Drywell pressure is 132 psia (0.910 MPa) at the time of failure, and the drywell is assumed to depressurize by blowdown into the reactor l building through a 2 f t 2 (0.186 m2 ) hole. This causes a rapid increase
- l. in reactor building pressure.
e
14 At first, the flow escaping from the reactor building is limited to . l leakage by exfiltration through its concrete walls. However, there are l 300 ft2 (27.87 m2) of blowout panels between the reactor building and ' the refueling bay located above which open permanently when the differ-ential . pressure exceeds 6.93 inches of water (1.7 KPa). The subsequent flow into the refueling ba/ increases the pressure there until the 3200 I ft2.(297.28 m2 ) of blowout panels located on the stainless steel super-structure permanently open at a differential pressure of 9.63 inches of water (2.36 KPa) to permit direct flow to the surrounding atmosphere. All of this occurs within the first 20 seconds following drywell fail-ure. In this accident sequence, the Standby Gas Treatment System (SGTS) is'already operating at the time of drywell failure.* This system con-tinuously removes 12,500 cfm (5.90 m 3
/s) from the reactor building and an equal amount from the refueling bay as shown schematically in Fig.
2.19. (In actuality, the SGTS blowers take suction on the normal venti-lation system ducting within each compartment.) Af ter drywell failure, the pressure in the primary containment and the flow from the drywell into the reactor building both decrease ra-pidly. Furthe rmore , the reactor building fire protection system sprays are automatically initiated by the melting of sprinkler head fusible links in the areas of the reactor building into which hot gases are re-leased from the drywell. As a result of the reduced flow into the reac-tor building and the reduction of the gas temperature there, the action of the SGTS blowers is sufficient to maintain a sub-atmospheric pressure and to produce an inflow to the refueling bay and to the reactor building through the respective blowout panels, as shown schematically
- on Fig. 2.20. Once established, this pattern of flows is maintained throughout most of the remainder of the accident sequence.
A detailed description of the conditions within the reactor build-ing and the refueling bay af ter drywell failure is provided in Appendix - A. With a subatmospheric pressure in the reactor building during most of this period, the dominant wall-filtration flow is inward, i.e., from the outside atmosphere into the reactor building where it mixes with the much larger flow entering via the open blowout panels connecting to the refueling bay. .This combined inflow then mixes with the ef fluent from the failed drywell before passing into the ventilation ductwork leading to the suction of the SGTS blowers. Outflow (exfiltration) from the reactor building does take place during brief periods that occur af ter the onset of fuel melting at time 2407 min. One significant period of outflow begins at time 2492 min when an enormous quantity of steam enters the pressure suppression pool through flow distributers (termed T quenchers) upon the occasion of
" core slump" into the reactor vessel bottom head. A second significant outflow begins at time 2590 min when the reactor vessel bottom head fails. On both occasions, a large amount of steam enters the drywell
*The SGTS is initiated early in the accident when the drywell pres-sure exceeds 2.5 psig (0.119 MPa).
m- _
~
m 15
. andlis' immediately ~ released through the break into the reactor building,
. causing the reactor . building pressure to be greater than atmospheric in each ' case' for about ~ eight minutes.- (See Fig. A.6.)
The. hot gases- exiting from the drywell .are cooled by the action of the-- reactor building fire protection system sprays. The fusible link l 7 e- . sprinkler heads are initiated whenever the temperature -of the surround-ing atmosphere - exceeds 225*F '(380. K) and ~ continue to spray throughout the: remainder of the accident sequence. The average reactor building temperature consequently remains quite low during the accident sequence
~ [between 225'F - and 85*F (380 K and 303 K)]. The action of ~ the sprays also . causes the accumulation of a large amount of water in the. reactor building basement at the rate of about 500 GPM (0.032 m 3/s).-
~
Flow from the reactor building into the refueling bay only occurs during the period immediately following drywell failure, when there are no fission products in the drywell,-and during the brief periods of pos-
- itive ' pressure in ~ the reactor building after " core slump" and after failure of the reactor vessel bottom head. As a result, there will be .
almost no fission products in the refueling bay during this-accident se-quence, and the conditions in the refueling bay will be almost atmo-spheric.
- The predicted mixed-mean temperature at the inlet to the SGTS is shown in Fig. 2.21. The sharp drop immediately af ter- the initial rise
-is caused.-by actuation of the reactor building sprinkler systems. At no time during the accident sequence is the inlet temperature high enough
' to damage . the organic components of the SGTS or to preclude the dehumid-ifiers at the-inlet to the filter trains from performing their function.
L.. 2.7. Fission Product Transport Pathways 1* Figures 2.22 and L 2.23 illustrate - the . principal transport pathways prior to and ~ following failure of the reactor vessel, respectively. Dominant features are the early failure of the drywell (at 2064 min) and L actuationof the' fire protection sprays in the reactor building'(at 2080 min), events which occur-long before the first cladding failure at ~2330 min. n The main pathways prior to reactor vessel ' failure are shown in Fig. 2.22. Material ~ evolved from the overheated core enters the reactor !; ' vessel ' gas space from where it is convected upward by steam /H2 flow as
- 1. ,long as the boiling pool exists below the core. As it flows upward,
- - fission product vapors. may condense on or react with available solid i~ = surfaces,'e.g., aerosols or~ steel surfaces. Gases and aerosols in the l reactor. vessel upper gas space would leak through the shut 51Vs to the b ' condenser system or would pass into . the suppression pool via the T-
[ 4 quencher i devices at times when the SRV opens. due to excess pressure. Fission products leaking through the MSIVs are assumed to accumulate in
- _ the main condenser. - The volatile fission products Xe, Kr, and I also L collect in the pressure suppression pool and are assumed to equilibrate p' '
- with the ' wetwell gas space from where they may pass into the drywell with steam from the boiling suppression pool through the vacuum l ~
l: breakers. Fission products in the' drywell airspace, as vapors or as 1 .,
3 I 16 material condensed on aerosols, may either deposit on drywell surfaces , or, af ter drywell failure at 2064 min, pass into the reactor building i with the same steam flow from the suppression pool. I Flows in the reactor building are dominated by the SGTS blowers which operate throughout the sequence and maintain the building under negative pressure at all times, except momentarily at core slump * (2492 min) and drywell failuro (2064 min). At all other times, reactor building flows are inward from the refueling bay and drywell and outward into the SGTS, as shown. Aerosols containing fission products may settle in the reactor building and by the action of the fire protection system sprays, wash into the pool collecting in the basement. In addition, fission product vapors, mainly I2 and Org-1, are assumed to equilibrate with the reactor building water inventory. There is a potential water pathway to the liquid radwaste systems, but the sump pumps required to transfer water from the reactor building basement are assumed to experience motor fail-ure by flooding early in the sequence. Flow patterns change following reactor vessel failure, as shown in Fig. 2.23. Af ter reactor vessel failure at 2950 min and a brief period of pressure equilization, the principal flow path develops from gases evolving from degrading concrete into the reactor building through the failure in the drywell liner. Flows to or from the wetwell and to the condenser essentially stop after vessel failure. As reactor vessel surfaces heat up due to deposited fission products, gas expansion causes flow from the reactor vessel into the drywell, as shown in Fig. 2.23. 2.8 Standby Gas Treatment System (SCTS) Behavior Blowers drawing air into the SGTS filters from the reactor building and refueling bay are actuated early in the sequence and continue to
- operate throughout. Ar estimated 12,500 f t3/ min (5.90 m 3/s) are drawn from each control volume f rom various distribution locations using the normal air handling ducting system. Complete descriptions of the HEPA and charcoal filters within the SGTS are provided in earlier reports.2.3,2.4 The principal dif ference for the current case relative to the two earlier accident sequence studies relates to the use of CORCON-Hod 1 for estimation of rubble bed temperatures and concrete gas production rates and VANESA for estimation of the amount of concrete-based aerosols pro-duced in the drywell. The previous study estimates were based on the INTER subroutine in MARCH. By comparison, CORCON-Mod 1 yields signifi-cantly higher oxide-layer temperatures and concrete degradation rates.
llence, VANESA currently yields higher aerosol production rates than would be anticipated, based on earlier cases. The significance is that current esti, nates of drywell aerosol pro-duction for furtherratesdetail). would probably As discussed cause the in HEPA filters earlier to plug,2.4(seethe reports,2. Chap. 4 fol-lowing event sequence is believed to occur in the SGTS when it is ex-posed to an excass amount of aerosols: (a) the f ront HEPA bank plugs and tears immediately on deposit of 178.6 lbs (81 kg) of aerosols; t - - -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ------- - - - - - - - - - - - - - - - ------
17 (b) the rear filter bank plugs and tears on deposit of 178.6 lbs (81 kg) of aerosols; and (c) since air flow is therefore unimpeded, charcoal filter efficiencies for 12 and Org-I are assumed to be unaf fected by the tearing of the HEPA filters. As noted in an earlier report,2.4 SGTS behavior during any sequence o in which it is operational undoubtedly has a profound effect on the cal-calated accident consequence. The above-described fatture mode is to our best current judgement the manner in which it would fail when ex-posed to an excess amount of aerosols. We should note that other fail-ure modes are possible, even reasonable, that would lead to a higher calculated consequence. For example, if the HEPA filters plug but do not tear, the flows to the charcoal bed would greatly diminish resulting in heatup of the bed with possible evolution of the previously adsorbed iodine. Pressure in the secondary containment would be slightly above atmospheric and direct leakage to the atmosphere would occur. However, our current best judgement is that the filters would tear when plugged. The SGTS HEPA filters are designed to remove particulates 0.3 micron in size and larger with an ef ficiency of up to 99.97% and the Browns Ferry Technical Specifications require that the filters demon-strate 2,99% DOP removal during periodic testing. The charcoal filters are required to demonstrate 299% removal of halogenated hydrocarbons and 2,90% removal of methyl iodide. The following filtration and charcoal bed efficiencies are assumed for the Loss of DHR accident sequence: Prior to upstream REPA plugging (At times to 2768 min) Aerosol filtration efficiency 99.99% I2-removal efficiency (gas) 99.9% Org-I removal efficiency (gas) 95% Following upstream HEPA bank plugging and tearing (2768 to 2880) Aerosol filtration efficiency 99.0% 12-removat efficiency (gas) 99.9% Org-I removal efficiency (gas) 95% Following plugging and tearing of both REPA banks (time > 2880 min) Aerosol filtration efficiency 0% 12-removal efficiency (gas) 99.9% Org-I removal efftency (gas) 95% 9
18 References for Chapter 2 2.1 S. A. Rodge, et a1.,- Loss of DHR Sequences at Brotana Ferry Unit One
- Aacident Sequence Analysis, NUREG/CR-2973, ORNL/TM-8532 (May 1983) 2.2 L. C. Grelaann, et al., Reliability Analysis of Steel Containant Strength, NUREC/CR-2442~(June 1982) 2.3 R. F. Michner et al., Station Blackout of Brotana Ferry Unit One -
' Iodine and Noble Gas Distributicn and Release, NUREG/CR-2182, Vol.
2 (August 1982) 2.4 R. F. Michner et al., SBLOCA Outside Containment at BrotJne Ferry
' Unit One Volume 2. Iodine, Cesium, and Noble Gao Distribution and
.Releaes, NUREG/CR-2672, Vol. 2 (September 1983).
i h t t I I i L_ f
19 Table 2 1. Timetable and summary of events preceding o containment vessel f ailure and loss of injection Time g,,,g (h) 0 0 Inttisting reactor scram followed by Msly closure and f ailure of both pressure suppression pool cooling and reactor vessel shutdown cooling modes of the RHR system. 1 Drywell pressure reaches scram setpoint of 2 pets (0.115 MPa) causing automatte inttistion of diesel generators and the SGTS and actuation of Groups 2. 6. and 8 of the Primary Containment teolation System. Simultaneously, pressure suppression pool temperature exceeds 120*F (322 K) and operators begin = controlled depressurization of the reactor vessel. 2 LPECCS initiation signal on combination of high drywell pressure and low reactor vessel pressure. Pressure suppression pool temperature exceeds the mastnum recommended temperature for cooling of HPCI and RCIC system tube oil. 4 All necessary reactor vessel injection can be supplied by the CRDHS. S.6 Operators begin to throttle the CRD pump discharge to avoid overf tlling the reactor vessel. 13 HPCI and RCIC steam supply line ! solation. 14 RCIC turbine trip on htsh exhaust pressure. 21 5 Primary containment pressure exceeds its design value of 56 pets (0.49 HPa). 23 5 SRve become inoperative in remote-manial mode because of high dryvell pressure. 34.4 Drywell fails on high internal pressure of 117 pets (0.91 Mrs). Pressure suppression pool temperature is 141'F (446 K). e Table 2 2. Abbreviated sequence of events relevant to fission product transport , Time Min Huur Reactor trip. MSIV closure. P.HR f at te 0 0 SCTS on 60 I Suppression pool temperature = 318'F (~170*C) 2060 14 Containment falls and water ballof f begins 2n64 14.4 Reactor building sprays on -20M0 14.7 Core uncovery begins 2200 36.7 First eladding f atture 2110 18.8 Core melting begins 2407 40 1 Core slump assumed at 75% of fuel melt 2492 41.1 Reactor vessel bottom head fatture 2$90 41.2
. Core / concrete reaction, drywell aerosol ~2600 -41. )
production Downstream HEPA filter in SGTS tears 2911 48.6 Calculations terminated 1100 $1. 7 e
7 1 20 l i Table 2 3. Core temperature map ('C); time . 2407.3 min; % core melted = 1.0 l Top 10 1221 1327 1332 1354 1271 1327 1277 1354 1210 6 54
.9 1966 2093 2160 2160 2166 2160 2127 2316 1860 899 8 2099 2232 2243 2238 2271 2260 2227 Helt 1993 904 .
l 7 1716 1854 1871 1866 1866 1871 1871 1916 1593 799 6 1510 1677 1671 1693 1682 1677 1727 1777 1454 766 5 1543 1549 1482 1582 1560 1560 1582 1577 1410 727 4' 1471 1621 1488 1627 1638 1643 1482 1666 1416 677 1 3 1677 1877 1993 1882 1882 1882 1727 1977 1788 638 2 1099 1204 1216 1221 1221 1221 1204 1282 1038 560 , 1 292 293 293 293 293 293 293 293 292 292 Botton 1 2 3 4 5 6 7 8 9 10 Center Edge Table 2.4. Core temperature map (*C)1 time 2428.3 min; % core melted = 33.0
. 1 Top 10 1632 1699 1810 1777 1699 1710 1710 1721 1632 743 9 Melt Melt Melt Malt Helt Helt Helt Helt Melt 1060 8 Melt Melt Melt Melt Melt Melt Melt Melt Melt 1071 7 2260 Helt Helt Helt Helt Melt Helt Melt 1943 916 6 1866 2132 2082 2160 2027 2093 1849 ' Melt 1810 749 5 1854 1927 1810 1960 2027 1949 1938 2010 1649 821 4 1893 2071 1660 2143 2116 2093 1849 2266 1810 749 3 2321 . Melt Melt Helt Melt Holt 2338 Melt Helt 688 2 1532 1693 1710 1749 1604 1721 1688 1854 1721 610 1 291 291 291 291 291 291 291 291 291 290 Bottom 1 2 3 4 5 6 7 8 9 10 Center Edge 6.
21 a ' s Table 2.5. Core temperature map (*C); time 24613 min; % core melted = 54.0 Top 10 1721 1799 1893 1860 1799 1810 1810 1821 1716 816 9 Melt Helt Melt Melt Melt Melt Melt Helt Helt 1166 8 Melt Melt Melt Melt Melt Metl Melt Melt Melt 1193 7 Melt Melt Melt Melt Melt Melt Melt Melt 2316 1027 6 2177 Melt Melt Melt Helt Melt Melt Melt 1932 977 5 2154 Melt Melt Melt Melt Melt Melt Melt 1893 927 4 2210 Melt 1943 Melt Melt Melt 2154 Melt 2082 849 3 Helt Melt Melt Melt Melt Melt Helt Melt Melt 804 2 1866 2071 2049 2143 1938 2099 2021 2266 2243 732 1 426 434 437 437 436 434 434 441 421 343 Bottom 1 2 3 4 5 6 7 8 9 10 Center Edge Table 2.6. Core temperature map ('C) at time of postulated collapse; time 2491.3 min; % core malted = 74.0 o Top 10 Melt 2077 Melt 2249 2110 2066 2382 2093 2354 1032 9 Melt Melt Melt Melt Melt Helt Melt Melt Melt 140'e 8 Helt Melt Helt Melt Mnit delt Melt Melt Melt 1410 7 Melt Melt Melt Helt Helt Helt Melt Helt Melt 1227 6 Melt Melt Melt Melt Melt Melt Melt Malt Malt 1160 5 Malt Melt Melt Melt Melt Melt Melt Molt Melt 1093 4 Helt Melt Melt Melt Melt Melt Holt Melt Melt 949 3 Melt Melt Melt Helt Melt Helt Melt Melt Melt 916 2 Helt Melt Helt Melt Melt Melt Melt Melt Melt 843 1 743 971 1127 1116 1093 1060 58 2 1910 649 372 Bottom 1 2 3 4 5 6 7 8 9 10 Center Edge o l 0
s
,s d F
22 , t i ORNL DWO S4 40$1 EtD 2000 I I I I i 1900 - - BOTTOM HE AD 1800 - ASSUMED RAPID - F AILURE (2590) OUENCHFROM 2400"C $ 0 1700 - - I B ' i 5 1600 - - i 1500 - - I i e 1400 - - CORE COLLAP$C l2492) I I I I 2450 2500 25$0 2600 2660 2100 TIME F HOM $CR AM (mini Fig. 2.1. Core debris temperature in the reactor vessel as pre- , _dicted by MARCH. 1 1 [ e I t t i
. . - . 9 .
I ic k 2:*Q> :n(>.n(HMM A:QMnnD:QW bc< Tb _o o ne s, e
- *o
- - - - ~
, I
- Q n>n(HM% L:
Db ZC:>* r n
. DwMeD%M v,aM. Q oo ko @- r8o
- vo @S 2 - -
0 0 F 0 i g d 2 2 2 1
. 0' 0 b
[ l h P h r i m T I l i l i a M2 l i m 2 r 3 E$D i y m 0 ; m s - r l i y [ l s [ e t M I m FCI N2 IAL ATN I p r U$0 u L U DI A e T0 D RI L s E EN s S G u e r
.e 2 O 4'
0 R N 0 L O W SC G _ LO 8 UR 4 ME 5 2 P 7 0 5 9 0 E 0 T D m
24 ORNL DWG 84-5710 ETD E o
,3 _
N g m O m #l M M I N N '
'I , a h-W W3 g- .,
en V1 M CD W w > I
> 'l l W
- W
>8~
0
>o Oo lli mg CD E ~ 'l Y
v O- l M gg Wg +--TAF (364 in.) s .- gr 3 s,'l e a , N 8 2 y 4--- BAF (216 in.) e o h h* { $~ d O d N o D > tQ .; o o_ o 2000 2100 2I00 2300 2400 2500 TIME - MINUTES Fig. 2.3. Expanded liquid level in the reactor vessel. 0 l
e . . . . c 2<mm Ow $Pmcc Z 4e X<n>
- vwO rO'
_o o -
" on
- o_ o m_S GO i
i F e <mm Ow $Hmn : Z 4n 2<c> : va6 pb g oo O ",,, ** op 2 d. - _ _ 0 2 0
. 0 4
M a z s i s o' o o f w T a I t e Ma 2 r Eod 5 i o n - t h M e I r Na e Udo a To c E , t o S - r v z e i O R s o N s e
.l o
~ L D
W G 8 4 5 7 2 0 3 2 0 E 0 T D lll1 41li
, 1 '
i CC,2<C> L 4 ._ C (>nHm2 Z> azo 0NZ %<tenn n bO F g i o8 $3 g8 ~8 on 5R Ng g
. J ~ - - -
2 5 oo CC LM %<M> nn i>iHmX Z>QCO0WZ NRo d NS E<im n ho S3 o8 2 ~ - - - 0 B 0 0 u i l . d p u 2 o 1 0' f 0 h y d r T o I g e M2d 2 6 n Eoo w - i t h M i I n N2 t U$0 h e T0 E r S e a c O R t 2 N o / r 0 0 L. O w v G e 3 s 4 s 5 e 7 0 3 l 2
. E 5 T 0 D 0
[ 27 qg, o ORNL-DWG B4-5704 ETD o 6-N 8 v m o R_ 8_ 8 8 m n v M tr v in do
- 28. 28_
we E-w ;5 E-W W Cu ct: O O U n U o w*_ cza 8_ o= os tu ct: W cz]
>g >4 e o
.- g_ g_
- o 8_ o 2000 2100 2200 2300 2/00 2500 TIME - MINUTES Fig. 2.6. Average core temperature.
28 e a i o
' ~ * "4 's# * ' T "O O m e-
"O me o
k M b Cladding Channel Box 7 4.5 = 10 KJ 0.9 = 107 KJ b v Control Blade 0.2 = 107 KJ g Total 5.6 = 107 KJ Z e- 59e,- o*- ~ E o P
< U w -< % Q:
3: o e 5: n - y n-e b D: N O N % N
>** pNna -
O N- O Q: D: w N g W Z W W W
> 3 p ,
< P a <
D D E x u D o Oo o_ o ' ' ' 2000 2100 2200 2$o0 2400 2500 TIME - M1NUTES Fig. 2.7. Cumulative energy input.from metal-water reactions. e
29 ORNL-DWG 84-5706 (TD o 8-o A E-* 8 Ez] i U o-tzJ tt A
<e A o-U Z
o Y US r
)
<s: o -
CC N
- 8 -
o 8 o a I i i 2000 2100 2200 2300 2400 2500 TIME - MINUTES Fig. 2.8. Fraction of cladding reacted. l l
- l
'- _ _l.
30 o OW4L-DWG 84-5707 ETD
-a o
o O Ma b W
$~
W c: 8 Oo U Z-l2'; O ~ P U
- U:: g _
(X o Ex. k-a f e a o e a a 4 2000 2100 2200 2300 2400 2500 TIME - MINUTES Fig. 2.9. Fraction of core melted. 9
- r- .
m O * * . = . be UO%w tg,H s a 0<(n HMMi
~$er-S *Rm ~k0 1
_Nc?
",,_ s. n ~ _ -
F i g o bo Uo%m Mg. H O<V Ne,o o
) HWMN
~8oo S n$o 2
2 ~ _ - s_E o 0 1 0 0 0 T e m 2 p 1 e 0' 0 r a t u r T e I o MF 3 f EOI O 1 g - a 1 s e s M I . e N2 / xi U$0 l/ T0 l . t i E n g S t h e 2 4 i !, llIlj!i [ O R N c 0' 0 'i / L o D W r G
.e 8 4
5 7 2 0 8 5 E 0 T 0 O
6 32
. l l
l ORNL-DWG84-4304 ETD 7@ - g g 600 - - O
?
G t 9L 500 - A -
# 4
?
[ C @
= 5 3400 e -
s d t=f STE AM SEPAR ATOR SURF ACE CORE SLUMP m i I 200 2000 2200 2400 2600 TIME tmin) Fig. 2.11. Reactor vessel gas and surface temperatures at the steam separator. b I
e e , * .* . oZ>, JU EOE<A
* - %ZL(wDZC aD d v%<a rD ne s
A ne tr e- *F _O O JI
- - - - l
! i Cm>5wj. : . O.5<- J % % W )V D(D%M v.fln <m OO 8* *O ** C "3 * :$ -
2 - - - - 0 0 . 0 F i g D 2 R 2 Y 2 4 0 W E 1 L 2 L F T . A I I L D _a 3 r M24 C U R 3 y w E B m O R E - e 0 E (2 l - C 0 l 6 (IB O 1 M O L p r I Q-T T L A m ni e N2 . O P S
)
s M s UY2 . H E u T0 c E ( 2 r E A 4 e S .
. D 9
2
. F m
. A i I
.n -
L ) O 2 U R 9 6 - R E N L 0 - (
- 2 D 5 W
- 9 G
. 8 0 4 m 5 7
3 .ni
)
1 1 2 E 0 T 0 D l l l
34 ORNL-owo s4-sn2 tro *
@ o 8-r>
P 8 o 4 o_ 8 j J m" dly$j@ffh i !. l fi E a l p! y lt :
;s e
x ja= 1 6 4 1 k'fhp De s kx-h- $e os- n N W
- a. A M
-M /
f o gr N 3-W Q: 2: QS4~ Qow g W BOTTOM HEAD FAILURE (2590 min.) , S EJ -8. soo aa'e ano ano aso 3a00 TIME - MINUTES Fig. 2.13. Drywell ambient temperature. i-E r c . I L . i
35 3-ORNL-DWG 84-5713 ETO T o
$_ o
-, 18
, o 2-v Ok-v W W 0: Q:
D D Ee Ho
<p 0: <m_
n 2: N N N 2 X N N E- E~ A: o Q: m E-g- H"m-h h
- a. a.
2 %
. De De V2 g _ to g _
- o 8_ o.
2000 2240 2480 2Y20 2960 3200 TIME - MINUTES Fig. 2.14. Pressure suppression pool water temperature. e a
36 ORNL-C*G 84-5714 EtD , e e 4 N, 8 w n 1 1 e o 1 E_ m" M m" tz. v v.
. ;-) ,
H e- - H,
<t-n: "
<e ct: "
tz3 N C. Q M E tas taa P H w o
.O 9 J_i 2 ~ J %~
.3 .3 Iz] C23 -
k k . H -H Cz3 Cz3 ke >o , $- N-y Sl m _. o o 2000 -2240 2480 2Y20 2d60 3200 l TIME - MINUTES l Fig. 2.13. Wetwell airspace temperature. 4 I l e I I
-s , ,y, ::- . 4 J-y.
,. 1 ia ORNL-DWG 84-5715 ETD 8 CORE [ t SLUMP -
'(2492) -
. PEAK FLOW ~~ - ;
e 3 4 3000 M / MIN.' 14
~*
@ \ r 8-z r
s s.. m
- x AN
-_ y
- w:
e c x 2 O E% , G-e r e [ p_ "2 8 . gPEAKFLOW240M/ min
- u A A A h o A AA AVAm/mMAAAAAAAArvmAAAAAAA A A ^^ A A A A A e
2440.0 2040.0 2140.0 2240.0 '2340.0 TIME (MIN) r l Fig. 2.16. Volumetric flow from reactor vessel to suppression pool
- through safety relief valves.
? _____ _ _ _- - _ _ w _. , - _ _ , - - .-.
~;4 ORNL- DWG 84-5701 ETD 7.9
^
REACTOR VESSEL 80TTOM HEAD 4 n 3.7 o h I
.1 ,
-0.5 . ' U" 5
{' ..
/W f~
w l k -4.8 E S W 3 DRYWELL 00
* -9.0 F AILURE
-13.2 CORE SLUMP
-17.4 ,
2040 2140 2240' 2340 2440 2540 2640 2740 - 2840 2940 3040 3140-TIME (min) Fig. 2.17. Volumetric flow between the drywell and wetwell through
'the vent pipes. Negative flows are from the wetwell to the drywell.
I i i I s-
s 4 e tp g. .; - _o ORNL-DWG 84-6700 ETD 2800-CORE FAILURE OF INTER kESULTS DRYWELL ' SLUMP REACTOR VESSEL ------- CORCON RESULTS 2100 < FAILURE -l BOTTOM HEAD I I i ! 1400 - 5 :\ y I b
'd U/*', ,,ygf , e. .
,, ..vf s e..A_f ,, aj ,,,7,,,y,,,,,,, ,
0 . 2040 2140 2240 '2340 2440 2540 2640 2740 2840 2940 3040 3140 TIME imin) Fig. 2.18. Volumetric flow from the drywell to the reactor build-ing as calculated by MARCH . subroutine MACE. The depicted flows af ter reactor vessel bottom head failure include' both the effect of the high corium-concrete reaction gas generation rates predicted by MARCH subrou-tine INTER (solid line) and the CORCON-predicted gas generation rates (dashed line).
40 ORNL-OWG 84-6716 E TD
- FLOW TO
~
REFUELING FLOOR SGTS SYSTEM A A 5 C U ,
)
, G
_ ~ G
; J- LF b t Fig. 2.19. The SGTS filters the exhaust from the secondary con-tainment under accident conditions.
i ORNL-DWG 84-5717 ETD
)M (M
r- , REFUELING FLOOR [ FLOW F TO
" SGTS SYSTEM O
W
! I tu J [
Fig. 2.20. The SGTS blowers have suf ficient capacity to maintain a vacuum in the secondary containment after partial drywell blowdown; flow through the blowout panels is into the refueling bay and reactor build-ing. O
, y = .
~
e
- 4 ORNL-OWG 84-5718 ETD K 'F 408.16 - 275.0 394El- 250.0-N 380.38- 225.0 -
Z l3 P 4: 366.49- 200.0- DRYWELL C1 FALLS 352.61- fl5.0 - 1' n
~
E!
~ 338.72- 150.0 -
FAILURE OF CORE. REACTOR VESSEL 0 SLUMP BOTTOM HEAD cn tn 324.83- 125.0 - s 310.94 - 100.0 - 2r1.05 - 75.0 , , , , , , , , , , 2000.0 2100.0 2200.0 2300.0 2400.0 2500.0 2000.0 2700.0 2000.0 2000.0 3000.0 3100.0 TIME (min) Fig. 2.21. Temperature at inlet to the standby gas treatment sy=- tem.
ORNL-DWG 82-16498R2 A i MAIN
, > CONDENSER - 182.9 m SYSTEM (600 f t) 9 SrAcK .
AIR SGTS A DRYWELL - V REFUELING BAY SRV e REACTOR C D i
-O VESSEL ySiv V REA_CTOR BUILDING ;
. 5.' QC
; > y.
N
<f CORE .- > AIR O--O . . . s C > - -----------------
\ O >
s C > l __ - - = - - g_ l P-
\r
.--) l
+ +
LIQUID RADWASTE Fig. 2.22. Principal fission product pathways prior to time of reactor vessel bottom head failure (2590 min).
. - . , s .
e- Je .e -e , p '. -g :1 1 ORNL-DWG B2-16500R A MAIN N NSER 182.9 m SYSTEM (600 f t) O STACK
#R SCTS O A DRYWELL y REFUELING BAY REACTOR VESSEL C )
y 6 >
! REACTOR SUILDING ,,
h M A [ 5 -.d. C si - - 4.=--O AIR O ..
..: s -
c- . N O i > WETWELL l s V l
--)
coni & LIQUID RADWASTE Fig. 2.23. Principal fission product pathways af ter time of reac-tor vessel bottom head failure. (2590 min).
..=
44
- 3.~ FISSION PRODUCT -TRANSPORT ASSUMPTIONS ,
3.1 -Introduction Models and assumptions ^ used for ' estimation of (1) fission product - release . from fuel, - (2) chemical forms 'in the reactor vessel and in the
- primary. containment, . (3) interaction ~ of gaseous species with aerosols, (4) deposition of gaseous - forms onto solid surfaces, (5) volatility over.
water, (6) rates. of formation of organic ~ iodides, and (7) aerosol forma-
- tion ' rates in the reactor vessel and in the drywell have .been outlined
- in previous reports.3.1,3.2 There is- admittedly considerable room for
- improvement of rany of these models and we are continually reexamining c them as .new -information being developed by NRC research programs becomes
- available.
.. LIn 1this o chapter are described three modifications to the earlier
.models used in- previous studies. The new models are the result of re-l cently : developed information and are. described 'in Sects. - 3.2.1 through L3.2.3. In addition, Sect. 3.2.4 describes some new insights regarding the chemical behavior of cesium and: iodine which have not been incorpo-rated ;into this study but may at some future time alter transport model r
assumptions for' these ' two elements. 3.2- Changes in Iodine and Cesium Transport' Assumptions 3.2.1 Estimated release from drywell rubble
- Est' imat'es of the': rate. of - fission product release from the core / con-
, crete : debris '. bed on .the- drywell floor could be a critical factor in the *
~
'overall transport pathway. Materials evolving from this source location bypass 1the suppression pool, entering' directly into the drywell atmo-sphere.
Earlier estimates of release rates , from this inventory ' location were', based on use of the conventional- fuel release rate factor, mul-tiplied by a surface-to-volume ratio; factor to account intuitively ~ for
. theismaller surface i per unit volume rratio of the drywell ~ debris bed relative ' to : that of ' the fuel. element geometry 1on which the experimental
- values were based. For the drywell rubble bed, a surface area factor of '
25 f was . employed; i . e '. , release ' rate coefficients .obtained for the 1 calculated . bed temperature were divided by _25 to ' account for its smaller surface L to . volume . ratio. . Bed temperatures were obtained from the INTER
? subroutine'of.thi MARCH, program. .
The abovel rocedure was always considered- to be an interim method.
- _ ^ 1In partt 41ar,- it excludes- consideration of the sparging action caused -
+. byl gasere evolving ~ frot the -Econcrete which should significantly acceler-
" ate ther. fission product. evolution. rate.
For this study, we havel replaced this interim procedure by employ- *
-ing directly. results obtained using the VANESA code. 3
- 3 3 This com-iputational procedure
, developed at Sandia National Laboratory, can be a
q a _
45 y. characterized as follows:
- 1. VANESA _ employs the CORCON code to develop the thermal-hydraulics of the drywell rubble bed. Due to the differing densities, the ' rubble bed is predicted to divide into metal and oxide layers and O the CORCON results include predictions of metal and oxide layer temper-atures, CO2 and H2 evolution rates from the degrading concrete, and de-termination of the melt geometry.
- 2. An important feature of CORCON used by VANESA is the determina-tion of the melt layer locations relative to the gases evolving from overheated concrete. This feature is essential for determination of realistic sparging rates and estimation of the degree of availability of CO2 and H2 for reaction with the more reactive elements in the metallic layer; i.e. , there occurs various degrees of bypassing of gas flow.
- 3. The degree of reaction of H2O and CO2 with metals to produce H2 and CO is - then estimated. The principal reactive metal is Zr and when available, Zr is oxidized to Zr02 (which transfers to the oxide layer),
CO2 reduces to elemental C, and H2O reduces to H2
- 4. Fission product compounds are assumed to be distributed in so-lution between the oxide and metal phases. Cesiu:n is assumed to be principally Cs20, iodine is assumed to be CsI, and both are assumed to be dissolved in the oxide phase. Vapor pressures of dissolved materials are estimated to determine the driving force for transfer into the sparging gas flow. Raoult's Law is used to determine the vapor pressure of each fission product compound dissolved in either the oxide or liquid metal phase, a method strictly suitable only for ideal solutions.
- VANESA, in its current simple form, presumes complete liquifaction of all core / concrete phases, even when CORCON predicts otherwise.
- 5. VANESA then calculates the amount and composition of aerosol material which forms by subsequent condensation of compounds in the
, sparge gas. (These results are discussed in Chap. 4.) Results of the CORCON/VANESA calculation for tge loss-of-DHR acci-dent sequence are illustrated in Figs. 3.1 and 3.2. Figure 3.1 illus-trates the predicted oxide and metal layer temperatures. Note that most of the decay heat is assumed deposited in the oxide layer causing much higher temperatures relative to the metallic layer. Interface heat transfer resistance is evidently estimated to be extremely high, re-sulting in predicted temperature dif ferences in excess of ~750*C in the oxide layer relative to the metal layer. The level portion of the oxide layer temperature curve is evidently due to melting of. this layer during the time period, ~2620 min to ~2850 min; i.e., prior to 2620 min, the oxide layer is completely solid and later than 2850 it is completely liquid. For this case, the oxide layer is predicted to be more dense than the metal layer and thus is the lower layer. In addition to these two principal phases, oxide crusts form above and below the metallic layer. Further description of CORCON re-sults is provided in Chap. 4 in connection with aerosol transport esti-mates. o
*We gratefully acknowledge the aid of Sandia personnel; the VANESA calculations were performed by Dana Powers.
46 l _ The resulting ' evolution of cesium and iodine as computed by VANESA is illustrated 'in. Fig. 3.2 and compared with . estimates based on our
- earlier model. Note that VANESA predicts far more rapid evolution
-rates, as.would be anticipated by_the inclusion of the sparging effect.
The ' higher evolution rate of cesium relative to iodine results solely from the higher estimated vapor pressure for Cs20 relative to CsI. . Note, all of the cesium is predicted to evolve within 40 min after reac-tor ; vessel .-bottom head failure and during the period when the oxide phase -is predicted by CORCON to be still frozen. This is currently an unrealistic feature of the CORCON-MODl/VANESA results. We will Luse directly the predicted rate of cesium and iodine re-lease from the drywell rubble bed in our current transport rate esti-mate. ' Although no estimate of noble gas release is provided, it will be assumed that it is at least as rapid as the. predicted cesium evolution rates. This is formally accomplished in our computational procedure by assuming the following values for release rate coefficients from the drywell rubble: Release rate coefficients to approximate VANESA predictions Noble gases 0.046 min ~1 Cesium 0.046 min-1 Iodine 0.011 min ~l 3.2.2 Organic iodide production rate Organic iodides-[ principally methyl iodide (CH3I)], which form to.a small degree under containment vessel and reactor building conditions, possess far lower chemical reactivity than does molecular lodine (12) , which is the prin'cipal vapor form outside the ' reactor vessel, and have a far higher volatility than any metal iodide vapors that might exist. , Therefore, organic iodides will' deposit more weakly on all surfaces (structure and aerosol), are less soluble in water, and are less absor-bent in the SGTS charcoal beds than other iodine forms. The potential for release to the atmosphere is therefore significantly greater for or- , ganic lodides relative to other iodine vapor forms. In order to determine the amount of iodine present as organic io-dides, a rate of formation equation is needed to calculate the rate at which I2 is converted to organic iodides. Since the principal mech-anisms of formation. are not known, a well-founded rate equation cannot be ' developed. _ ' In available reviews on the subject of organic iodide formation,- the rate of formation was mostly ignored and only equilibrium levels were considered. Since the formation rate and not the equilib-rium level is needed, an estimate must be made from the data available. In earlier reports,3.1,3.2.the formation rate was estimated by using the equilibrium . equation presented by Postma and Zavodski.3. 5 to [ predict equilibrium levels of organic iodides, and by assuming that these levels are achieved in 4 h. These assumptions yielded a formation . -rate: conversion rate of I-vapor to organic iodine, %/h = 0.05 x c .26 o (3.1) __------r.- . . ~ . - . . . - , - - - , - - - ---------..-----.----e -.c- - - - , - - - -- .~, -
. . _- .-. _ _ _ . _ - _ ~_ - - _ _ . . _ - . . _ _ . _ . _ _ --
47 3 : where - C _ is the _ iodine concentration in the vapor in.mg/m3 . This . equa- l tion predicts a formation rate of 0.015%/h for a " typical" I2 concentra-tion of 100 mg/m 3. In more conventional units, the predicted equilib-
- rium level of organic iodide in the atmosphere of a control volume may be expressed as C* = 8.'11 x 10-6 Cgo .74 , (3.2) where C* = equilibrium level of organic iodide, (gmol/cm 3),
C0=initialconcentrationofI2intheatmosphere, geol /cm. 3 We now will assume an exponential formation rate approaching C* asymp-
~
totically, dC-
.de " *(N o) , (3.3) n
{ 3 where C, _= the concentration (gmol/cm ) of organic iodine at time t(s). The integrated form of Eq. (3.3) is:
~
- .- C,(t) = C* (1 - e " ) + C,(0)e"* . (3.4)
The factor a was determined by using the equilibrium results given by , _ , Postma 3 .5 and by assuming 4'h is required to reach equilibrium; i.e., if 99% of the equilibrium value is reached in 4 h and assuming no organic iodine was present initially , the constant a becomes equal to 3.2 x 10-4 s-l. Organic iodides will be removed from the vapor phase by absorption, decomposition, and by chemical reaction. Data from the Containment Sys-tems Experiment (CSE)3.6 were used to derive an equation for organic iodide removal as a function of time and temperature. Currently we will assume that the removal rate of organic lodide can be expressed as a first order-process: dCo de- "' '
- l-i = where I:
l ,- Y = Y(T) = exp[a-b/T} . (3.6) I I The CSE data indicate that the,value of Y is relatively independent ~ of Co. At temperatures of 358 K and 395 K, values of 3 x 10-6 and ; 4
- -r + ~ , , ------- r ,m, ,,,-,,,n-n--w-- - - , - - - , , - - - - --.w...-7, , - - - - . - - - , , - - - - - -- ,c,,.- sw - - - - - - - , , - - - - -
48 10-5'..e-1 resoectively are estimated, yielding values of a = 0.14, b = 4.6 x 103 for the constants in Eq. (3.6).
- The complete organic iodide formation / destruction equation may thus be written:
d = a(C* - C,) - YC, , (3.7) whoee solution is C,(t) = [1-e( ~
} ] + C,(0)e' E
. (3.8)
It.must be recognized that the principal mechanisms for the forma-tion and removal of organic iodine are not known, and the equations pre-sented above are estimates developed from the experimental data avail-able. In this regard, it should also be noted that factors other than iodine gas phase concentration might affect the formation' rate of or-ganic iodine. These factors include the type of paint used on the ves-sel walls and the amount of oxygen, moisture, and radiation present. While it seems fairly well established that moisture and radiation can increase the formation of organic iodides and the presence of oxygen de-creases its formation, there la still some doubt on the effect of dif- - ferent types of paint. For example, Bennett et al.3.7 suggests that the presence of organic paint on the walls increases the formation of or- . ganic iodides, but this would suggest a surface reaction while most re-searchers feel that the main mechanism is reaction with airborne or-ganics.
..3.2.3 Iodine volatility Iodine volatilities are usually expressed in terms of a partition coefficient defined by:
partition coefficient (PC) = c neen ra n ne in awe us solution , concentration of iodine in gas phase The aqueous-to-gas phase iodine partitioning is partly determined by the rate and degree of iodine hydrolysis and the effect of radiolysis on io-dine solutions. The classical representation of iodine hydrolysis is expressed in two successive reactions: I 2 + H2O ; HOI + I~ + 11+ (3.9)
~ ~
3 HOI ; 103 + 2I + 3H+ .
49 For purposes of evaluating volatility only, the molecular (un-ionized) o iodine forms need to be considered. Because of the polar nature of water and electrostatic effects, ions are not expected to vaporize d4-
~ ~
rectly. The final products of iodine hydrolysis are I , 103 , and H . Since these products are ionic, parameters such as high pH and high tem- . perature that favor their formation reduce iodine volatility. Radiolysis of iodine solutions results in the interaction of iodine species with radiolysis products from the solvent vater. Thus, the chemistry is concerned ,with the interaction of water radiolysis produc3s 08., e H+, H0, 22 02 , and H02 with iodine species such as I, 2 13 , I , an89,OI3~. The basic aspects of aqueous iodine chemistry and radiolysis have been studied for many decades and are still not well understood. How-ever, there are a number of experimental measurements of iodine parti-tion coef ficients which yield information that can be used to evaluate aqueous / gaseous iodine partitioning in nuclear reactor accidents. In a series of four tests at the Containment Systems Experiment,3.6 a parti-tion coefficient of 6.2 x 103 i 1.9 x 103 is given. These tests were conducted with a vapor temperature of either 358 K or 398 K and samp temperatures between 303 K and 317 K, and initial iodine vapor concen-trations from 1.17 to 165 mg/m 3. In addition, there have been a number of laboratory scale studies of iodine partitioning. In carrying out measurements of iodine parti-tion coefficients, it is far easier to accurately measure low values (high volatility) rather than high values. This is due to the nature of the tests whereby large values mean that very little iodine appears in e the gas phase. Also anomalous chemical behavior of low concentrations (<~10-6 8*mol/L) of radioiodine has been observed for many years. 3
- 8 The dif ficulty in measuring large partition coef ficients and the anoma-lous behavior of radioiodine at low concentrations indicate that many laboratory scale volatility measurements will be system dependent. That is, factors such as interaction with container materials and impurities could dominate the experimentally reported volatility. It is a paradox that such measurements may still be the best guideposts that can be ex-rected because nuclear reactor accidents do not occur in " hypothetically pure" systems.
Lin3 *9 reports that for iodine concentrations less than ~106 g mol/L, partition coefficients tend toward a value of ~9 x 103 and become relatively insensitive to concentration. Kelly 3.10 reports values somewhat lower than this; at 7.9 x 10-7 g mol/L and 303 K, a value of 8.5 x 102 is given. At this same temperature and a concentra-tion of 7.9 x 10-10 g*mol/L, Kelly lists a value of 2.1 x 10 . 3 Pel-letier and Hemphill3 *ll report results at concentrations as low as 7.9 x 10-14 g.mol/L. , The partition coefficients at this extremely low concen-tration are ~8 x 103 and thus substantiate the value given by Lin. At Oak Ridge National Laboratory, both calculations3 .12 and experi-mental measurements 3 *l3 have emphasized the importance of concentration, pH, temperature, and mixing time in determining iodine partition coef fi- , cients. Values in the range 102 to 7 x 104 have been measured under widely varying conditions. In general then, values of iodine partition coef ficients measured in the laboratory and in the large Containment Systems Experiment yield 0
50 values in the range 103 to 10 4. For this accident sequence, an iodine partition coefficient of 5x 103 will be used. This is close to the
- value found in the Containment Systems Experiment and is in line with values measured in the laboratory. It is a factor of 10 lower (factor of 10 higher volatility) than assumed for our earlier studies.
It is recognized that radiolysis of iodine solutions could alter
- the results obtained from tests involving iodine hydrolysis alone. The influence of pH on the radiolysis is quite complex. In acid solution, I can be cxidized by OH or H20 2 to give I. 2 ~ Conversely, in alkaline solutions I2 can be reduced by H 202 to form 1 . Until the complexities of radiolysis of iodine solutions are better understood, it appears necessary to use partition coefficients based on hydrolysis.
3.2.4 Cesium and iodine chemical behavior In earlier reports 3 .1,3. 2 it was noted that cesium compounds Cs0H and CsI may react with a number of available metal oxides, such as Cr20 3 , A12 03 , A12 03*SiO2, to form stable cesium compounds. The net ef-feet of these reactions would be to sequester cesium and liberate iodine f rom CsI. The degree to which these reactions proceed depends princi-pally on the temperature and oxygen potential. Most are predicted to proceed at relatively high oxygen potentials (H2/H2O << 1), but some are predicted to proceed at least partially under reducing conditions expected in the reactor vessel prior to failure, i .e. , H 2/H2O > 0.1. It has also been noted that the major portion of the iodine liber-ated from CsI by reaction with metal oxides would not long remain as , gaseous 12, but would likely chemisorb on available surfaces such as oxidic or metallic aerosols and metallic structural surfaces. We now note that in addition, control rod material, BgC, will likely hydrolyze in a steam environment to form a number of relatively volatile compounds such as HB02, H B3 036 B 23 0 , which are predicted to
- have a high thermodynamic driving force for formation of cesium metaborate (CsB0 2 ) or dicesium tetraborate (Cs2 Bg07).
We have performed equilibrium thermochemic calculations for the hy-drolysis reaction of BgC with steam and subsequent reaction with Cs0H and Cs1 for hydrogen to water ratics of 0.0, 0.1, 1.0, and 10, at tem-peratures of 1000, 1200, 1400, 1600, 1800, and at 2000 K and 70 atm to-tal pressure. At 1800 K and all hydrogen to water ratios, more than 90% of the cesium was calculated to be in the form of CsB02 and Cs2B407 Also at this temperature and for hydrogen to water ratios of 0.0, 0.1, or 1.0, more than 70% of the iodine was in the form of HI + 1. As noted earlier, HI or I released due to stripping of cesium from cesium iodide would likely not be permanently gaseous species. They are quite reac-tive and are likely to combine with other available metal forming an-other iodide or attach via chemisorption to virtually any available solid surface. A more complete description of the predicted degree of BqC hy-drolysis at various H2/H2 O atmospheres and the subsequent reactions to form cesium borates is presented by E.C. Beahm and R. P. Wichner3 '14
- p f A'
_)
. 4-
'-d 51 l
, References for Chapter 3 a
3.1 R. P. Wichner et al., Station Blackout at Brouns Ferry Unit One
'\ -
Iodine and Noble Gas Distr,'bution and Release, NUREGlCR-2182, y .Vol. 2 (August 1982),- Chapters 3 and 4. 3.2 R..P. Wichner et al. , SBLOCA Outside Containment at Browns Ferry
^ Unit One Volume 2. Iodine, Cesium, and Noble Gae Distribution and
. Release; NUREG/CR-2672, Vol. 2 (September 1983), Chaps. 3 and 4.
3.3 . D. - A. - Powers ' ar.d i J. E. Brockmann, Status of VANESA Validation
* (Sandia report, to be published).
3.4 D. A. Powers and J. E. Brockmann. , Release of Fission Products and Genention of Aerosols Outside the Primry System (Sandta report,
-to be published).
am 2-3.5 A. K. Postma and R. N. Zavodski, Revie9 of Organic Iodine Form-
- tion, WASH-1233 (1972) 3.6 R. - K. Hilliard and -L. F. Coleman, Natumt Tmnsport Effects on Fission Product Behavior in the Containment Systems Experiment, BNWL-1457, 197G. ,
~3.7 . R. L. Bennett et al., Peactions of I-vapor oith Paints, ORNL/TM-
,q - 2760, 1970.
~ , . j L 3'. 8 M. - Kahn and J. Kleenberg, Radiochemistry of Iodine, NAS-NS-3062, ati 1977. -+
$ d 359- C. C. Lin, " Volatility of Iodine in Dilute Aqueous Solutions," J.
y + s s. Inorg. Nucl . Chem. 4J, 3229 (1981) . M' 3.10 l ' J. L. Kelly, C. J. Bahad, and R. U. Mulder, Study of Iodine Parti-
..-' tion Coefficients, Report No. UVA/532256/NEEP 81/101,1981.
3.11 C. A. Pelletier and R. T. Hemphill, Nuclear Pooer Plant Related
~ ~
Iodine Partition Coefficients, EPRI NP-1271, 1979. s 3.12 J. - T. Bell, M. ' H. Lietzke, and D. A. Palmer, Predicted Rates of
*^
Formtion of Iodine Hydmlysis. . Species at pH Levels, Concentm-
, tions and Tempenture ' Anticipated in WR Accidents. NUREGlCR-2900 sL (ORNL-5876),- 1982. >
3.13 E. C. Beahm, private' communication, 1983.
-3.14- E. Ci Beahm and . R. P.'Wichner, Surveu of Chemical Reactions of
\ Cesium. Iodine. and Tellurium Which May Occur in WR ^ Accidents, u (to be published).
t k s s . ', j
. ;[ 4 ti
}
* % g g., 4 . + -
-*- e
\
g { i a e 1
52
- O R N L- DWG 84-4218 ETD 2600 l I e
2500 - - 2400 - - 2300 - OX1DE PH ASE 9
; 2200 - -
5-
- H REACTOR VCSSEL ;$ FAILURE e
3 2i00 - - Y e 2000 - - U METALPHASE 1900
- ~
1800 f.r-- l l l l l 1700 2500 2600 2700 2800 2900 3000 3100 TIME AFTER REACTOR SCRAM (min) Fig. 3.1. Drywell rubble bed temperatures predicted by CORCON. t e
, , _ , , - , y , _ . - _ - _ . . _ _ . . _ _ _
l l 53 0 o-ORNL-DWG 84-4219A ETD I I I I I VANESA PREDICTION CESlUM 1.0 7 IODINE g
/
O.8 -
/ -
'N
~$ f w
e j a j 0.6 - f -
~O E k O EARLIER MODEL
, E FOR CESIUM
" AND IODINE 0.4 - -
y 0.2 -
/ -
/
/
# I I o 1 I 2500 2600 2700 2800 2900 3000 3100 TIME AFTER REACTOR SCRAM (min)
Fig.-3.2. Comparison of the VANESA-predicted cesium and iodine re-leases from the drywell rubble bed with the releases predicted by a pre-vious model. e
54
- 4. AEROSOL PRODUCTION AND TRANSPORT .
4 .1 Introductior. e Af ter ; reactor ' vessel melt-through, molten fuel comes into contact
' with ; the : concrete basemat in the drywell, and .the heat from the molten
~ fuel e produces' concrete deccaposition, gas evolutfon, and evolution of
" core-concreto" aerosols. Evaluation of the transport and deposition of
.these aerosols -is an important factor for . determining fission product transport rates. . . This 'is so because volatile fission products can ad-sorb or condense on . the . aerosols and some fission product species may themselves form aerosols. In either case, aerosol transport behavior may'largely determine-fission product transport behavior.
The ' calculational procedures used in this study to determine core-concrete interaction phenomena, . core-coccrete aerosol production rates, and aerosol transport in.-the drywell' and reactor building are different
~
from those used in the previous two Browns Ferry accident sequence
- studies 4.1,4.2 The differences in calculational procedures include the
-following:
[1. - In . the previous studies, core-concrete rubble temperatures and gas evolution rates from - the decomposing concrete were estimated using the ; results from the MARCH subroutine INTER. In this study, the CORCON-MOD 1-code, developed at Sandia National Laboratories,4.3 is used for this purpose. . 2.: Core-concrete aerosol production rates were estimated in the previ-ous formedstudies ~ using*a at Sandia. 4 simple correlation based on experiments ' per-In this study, the VANESA code, developed by Dana Powers ~ at ' Sandia,4
- 5 is used to determine aerosol production ,
rates.
- 3. The HAARM-3 aerosol transport code was previously 'used to determine aerosol transport in ' the drywell, reactor building, and refueling bay. In this study, the QUICK aerosol transport code, developed at-Battelle-Columbus Laboratories,4*k is used. to calculate' aerosol
-transport.
A summary of the results. from the series of - calculations used to estimate aerosol transport in the LDHR accident sequence is presented in this chapter; the details of these calculations are' presented in Appen-dix B. -Also presented in ' Appendix B are results - from i group of sensi-
. tivity calculations performed using - the MARCH subroutine INTER, the CORCON-MODI code, and the VANESA code.~
R
.4.2 Drywell Debris Bed Behavior Calculation of the gas evolution rates and core material tempera-tures produced 1 due to core-concrete interactions in the drywell is I
1
w - 55 i important for the following reasons: ' '
- 1. The - rates L of release of volatile fission products and structural aerosols from the core material-concrete mixture are a strong func-0, tion -of the core-~ concrete temperatures and gas evolution rates.
- 2. . The rates of- gas evolution f rom . the core-concrete mixture largely determine - the thermal-hydraulic conditions existing in the drywell, reactor building, and refueling bay; these thermal-hydraulic condi-tions influence aerosol and fission product deposition in these com-partments.
In -analyses performed for the previous two-Browns Ferry accident sequences, results from the MARCH subroutine INTER were used ' to estimate the . time history of core material temperatures- and gas evolution rates due to . the interaction of the molten core with 4the concrete basemat.
'For the'present LDHR study,.the CORCON-MODI code .3 was made operational at ORNL and used to - estimate gas evolution rates and core material
- temperatures due to core-concrete interactions. In this section we present results from the CORCON and the INTER calculations to illustrate how they differ for'this accident sequence (more details on these calcu-lations are presented in Appendix B).
' Table 4.1 presents a summary of the core-debris initial conditions, as calculated by the MARCH code. Figure 2.8 shows that about 15% of the zircaloy fuel-rod- cladding is estimated to be oxidized prior to reactor vessel melt-through. The table also shows that the calculated core ma-
- terial temperature' prior to the start of core-concrete interaction is
'1737 K. Since the' estimated initial solidus temperature for the oxide portion of the melt .as calculated by CORCON, is roughly 2100 K, at least.' some portion of the time - the oxide -portion of the melt is pro-jected to be solid. Since the CORCON-MODI code was not designed to a
model a solidified melt, -results for early times with low oxide layer teisperatures may be unrealistic. The 'CORCON and INTER- calculations were performed to model a period of 8 h 'af ter the start of the time of core-concrete interaction for the
- LDHR ' accident sequence -(the start of . core-concrete interaction was cal-culated to occur- 2590 min af ter reactor scran was initiated). Table 4.2 presents a summary of the total amounts of CO, CO2, H2, and H2O evolved
= due to core-concrete ' interaction, based on the CORCON and INTER calcula-tions (the " base-case" CORCON calculation used in the accident seq'uence analysis is noted as. "CORCON1"). Figures 4.1 and 4.2 present the oxide .
and metal layer temperatures calculated using the CORCON ' and INTER codes. . The results of 'the comparison of ' the calculated CORCON and INTER results can be summarized as follows:
- 1. The ' amounts of gases evolved from the melt projected by INTER is
.significantly greater than predicted by CORCON. The largest differ-ence-occurs for C0; INTER predicts that about 10,000 times more C0 is evolved for the LDHR accident sequence than does CORCON. The low A +>
CO release in the CORCON calculation ' is, based on discussions with members of the Sandia staff,4 7 due to " coking," or the reduction of CO in the CORCON calculation to elemental carbon. As is illustrated
F k 56
-in calculations presented in Appendix B, coking is predicted to k
- occur when there is a large amount . of un-oxidized zircaloy in the core . melt (this is the situation . predicted for. the LDHR sequence).
For : the " base-case" ' CORCON calculation summarized in this section, 755 kg of elemental = carbon 'was calculated to have been produced due to coking. .
- 2. As billustrated in Figs. 4.1. and 4.2, CORCON predicts significantly higher' met'al^and oxide-layer: temperatures than does INTER. The cal-culated CORCON l metal;. layer . temperatures are roughly 300 K higher than;the-INTER temperatures. The calculated CORCON oxide layer. tem-
.peratures are roughly 400 to -500 K . higher than the INTER tempera-tures for, the first 5.5 th 'af ter the start of core-concrete interac-tion; -. however, Lat -8 h, the CORCON oxide temperature is roughly 1000
;K higher than.that calculated with INTER. CORCON predicts that the
" oxide portion'of the melt does not begin to melt until 3 h after the start of .the core-concrete interaction, and that the oxide remains heavier than the metal layer for the entire 8 hours of the calcula-
-tion. . In contrast, INTER predicts that the oxide layer becomes molten almost immediately af ter the start of the core-concrete in-teraction, - and E that= the metal layer becomes more dense than the
'> : oxide-layer roughly;l hour after the start of the calculation. Even when melting . of the Joxide . layer in bulk is predicted, a detailed
'look L at the output . shows that the interfaces between the oxide, metal, -and concrete layers are below tht- solidus temperature. Since the heat-transfer - correlations in ' CORCON-MOD 1 ' are not designed to model this situation,- we suspect - that the high oxide temperatures fpredicted near the end'of the 8-hour calculation are not realistic.
4.3 Aerosol' Production Rates in-the Drywell For the' previous Browns Ferry accident sequence studies, an 2mpir-
, , ical correlation developed based on experiments performed at Sandia4 4 was used to ' estimate . aerosol . production rates resulting from core-concrete interactions. .More recently a' mechanistic model, VANESA,-has
;been ' developed to predict aerosol generation and fission product release during ~ core-concrete interactions. .5 VANESA- results related to vol-
. atile fission product release rates are discussed in Section 3.2 of this report; iin this'section we will discuss the aerosol generation rate re-suits obtained using the VANESA model.
-The VANESA model was: developed at Sandia by Dana Powers who per-
~ formed the VANESA calculations summarized in this section' and in Appen-dix B. VANESA ~ includes models for aerosol formation ' due to evaporation y from the melt,' sparging of. the melt by gaseous products of. concrete de-composition, . reactive = vaporization of melt components, and mechanical formation when gas bubbles burst at the melt surface.
'.. Inputs required ( to run VANESA include the initial composition of the core . debris, the ' concrete composition, the melt temperatures (metal and oxide) during core-concrete interaction, the rate of gas evolution .
' f rom. the melt, , and ' the geometric top surface area of the molten pool.
This ' input ~ was ' supplied ; to Sandia from the LCORCONI and INTER calcula-
~
E,
'tions discussed in the. previous section.
y . . _ . . . ._ . __ ... - _ ._ ~ .. , ' t [y'p s c ,_ l .b b
.=
57
.# 1The aerosol ' generation - rates predicted by VANESA for the CORCONI thermal- data are summarized in Fig. 4.3 and Table 4.3. VANESA predicts thatt a total ,of 973 kg: of core-concrete aerosol would be produced. As illustrated: in Table 4.3, the majority (roughly 94%) of this aerosol is
._ lmade ? up of - SiO2,-- Ca0, K20, ~ Fe0, Al203, Na20, and Mn02 Details of the
* , ..VANESA a results obtained; based on inputs from the INTER data and from three. other CORCON runs 'are presented in Appendix B. It should.be men-tioned.that the calculated amount of aerosol produced based on INTER was roughly la- factor of 2 higher than that based on the CORCONI thermal
- data.-
4.4. Aerosol Transport in the Drywell Aerosols are produced in the drywell due to core-concrete interac-tion phenomena that start roughly 2590 minutes after reactor scram. Prior to the J start of core-concrete interaction the drywell is calcu-elated to have failed due to over pressurization (see Sect. 2.2). After
, drywell- failure, aerosols produced by core-concrete interaction can be transported from the drywell to the reactor building; the gases that would transport the aerosols out of the drywell are those produced by
. the core-concrete' interaction.
The ; QUICK aerosol transport code 4 .6 was used to calculate aerosol
- behavior; and transport in - the drywell.- Input requirements for QUICK include the time. dependence of the rate of gas leakage from the drywell 3' (obtained fron- the - CORCON1 gas evolution data), the rate of aerosol production : in .the drywell (obtained from= VANESA) and parameters related to .the sizes of - the aerosols produced due to. core-concrete inter-actions. . QUICK code 1 outputs include the aerosol concentration and size distribution in the drywell, the aerosol deposition on horizontal sur-
-faces b; 7 'ettling, and the aerosol - leakage f rom . the drywell; all - of s
these were determined as -a function of time. ~ Detailed . results are pre-sented:in Appendix B.
. A summary of the 'results for the QUICK drywell calculations for the .
LDHR sequence is. presented . in Table 4.4 and" Figs. 4.4-4.6. Table 4.4 lists the . amounts of aerosol predicted -to have settled . in the drywell, leaked.out of the drywell, and' to have remained airborne in the drywell at -3070 . min, or 8 h after the start of core-concrete ' interaction. Roughly 82% of the aerosol produced in the drywell is predicted to be
- transported7 to the reactor : building, and about 14% . of the aerosol pro-
- duced settles ' on , the drywell horizontal surfaces. .These results are
'somewhat ' different from those produced in ' the SBLOCA analysis; 4. 2 in
!that case, 93% of the - aerosol produced 'was transported to the ' reactor
- building while only. 0.9% of .the aerosol- mass settled to the floor. The differences- in the LDHR and SBLOCA' results are due to two major factors.
First, the 'LDRR : drywell ' leak rates are ' lower than those used in the SBLOCAisequence; . .because of' this, aerosols would reside in the drywell
, : longer in I the LDHR ' sequence. Secondly, the projected aerosol mass produced in . the LDHR sequence is roughly 5. times that' predicted in the SBLOCA sequence; because of ' this and the lower leak rates, aerosols
~
would be expected to grow to larger sizes and settle faster under LDHR accident conditions..
58 The time variation of the suspended aerosol concentration predicted , in the drywell is shown in Fig. 4.4. The concentration peaks at about 13.2 g/m3 at 2832 min, roughly 4 h af ter the start of the core-concrete interaction. The concentration airborne at 3070 min remains high and has a value of' 9.2 g/m 3. Figures 4.5 and 4.6 illustrate, respectively, the time variation of the leaked and settled mass between 2590 min and 3070 min. 4.5 Aerosol Transport in th'e Reactor Building As discussed in Sect. 4.4, roughly 82%, or 848 kg, of the core-con- . crete aerosol produced in the drywell is predicted to be released into 4 the reactor building. The aerosol transport pathways in the reactor building and refueling bay for the LDHR accident sequence are somewhat different than thosu produced in the SBLOCA accident sequence. For the LDHR sequence, the gas flow rates into the reactor building from the drywell are somewhat smaller than those produced in the SBLOCA sequence; f this is . largely due to the fact that CORCON-MODI was used instead of INTER for the present sequence. Because of the lower flow rates to the i building, the Standby Gas Treatment System (SGTS) is able to maintain a - slightly negative - pressure in the building continuously for the 8-h period after the start of the core-concrete interaction. This means that, except for a short period of time af ter 2590 min, all of the gases flowing into 'the reactor building go through the SGTS, and there is only a short period of time where there is flow to the refueling bay or di- . rect ; exfiltration to the environment from the reactor building. The reactor building thermal-hydraulic response is described in more detail in Sect. 2.6 and Appendix A. Because of the flow conditions in the reactor building, the main pathway for aerosol transport to the environment is through the -SGTS
. af ter the HEPA filters in the SGTS rupture. The SGTS has two banks of HEPA filters; based on analyses performed for the SBLOCA sequence , te . 2 each of the banks may be loaded with -61 kg of aerosols before failure by tearing. Thus according to this assessed failure mode, both HEPA filters would tear following transport of 162 kg to the SGTS af ter which there would be a direct aerosol transport path through the SGTS to the environment.
The QUICK code was used to calculate aerosol behavior and transport
- in the reactor building. The rates of aerosol input to the reactor building were determined from the drywell results. Gas leak rates from the . reactor building were determined from calculations performed with the secondary containment thermal-hydraulics model (see Appendix A).
QUICK code outputs were of the same type as those produced for the ucy-well _ calculations.- ' Aerosol leakage results from the reactor building determine the times when the SGTS filters would rupture. Details of the QUICK reactor building calculations are presented in Appendix B. A summary of the reactor building calculations is presented in
- Table 4.5 and Figures 4.7-4.9. Table 4.5 lists the amounts of aerosol Jt predicted to have settled in the reactor building, remained airborne in the reactor building, and leaked to the SGTS. Also listed are the
-- - _ . , _ , _ _ _ . - , _ - . , _ . _ . - . . , _ , , _ , , , . _ , _ , _ _ , - - , _ . . _ , v_,.__. _ _ . _ _ _ ,
i l l 59
, - amount of aerosol deposited on the SGTS filters, the predicted times for
- SGTS filter rupture, and the amount of aerosol that leaks through the
~
4
> SGTS after filter rupture. Table 4.5 illustrates that aerosol settling is the major -deposition mechanism; roughly 67% of the aerosols trans-ported to the. building deposit - there by settling. The table also
- illustrates ~.that the second filter bank is predicted to rupture at
> 2880.5 min. roughly 290.5 min after the start of the core-concret_e interaction. Prior to filter rupture, 162 kg, or 17.9% of the aerosol transported to the - reactor building from the drywell, was deposited on
. the SGTS filters. . The amount of aerosol transported to the SGTS af ter filter rupture, . 87.3 kg or 9.6% of the aerosol in the reactor building, is predicted to be released to the environment.
It should- be noted that the LDHR reactor building results differ
. from those for the SBLOCA sequence in that SGTS filter rupture was not predicted in the SBLOCA sequence. This difference is due to the fact
- that - the predicted aerosol source for the LDHR sequence is almost 5 times higher than the source for the SBLOCA sequence.
The time variation of the suspended aerosol concentration predicted in' the reactor. building is shown in Figure 4.7. The concentration peaks at 'about 2.1 g/m3 at 2836 min, roughly 246 min af ter the start of the core-concrete interaction. Figures 4.8 and 4.9 illustrate, respective-ly, the time variation of the leaked and settled mass between 2590 min and 3070 min. 4.6 Summary of Aerosol Transport Results The overall results from the QUICK calculations indicate that all aerosol release to the environment for the LDHR accident sequence occurs
, by transport through the SGTS after filter rupture. The predicted amount of aerosol release through the ruptured SGTS filters is 87.3 kg; this represents 9% of the total mass of aerosol predicted to be produced in the drywell (973.3 kg).
The main uncertainties in the aerosol transport calculations for the LDHR accident sequence are summarized below:
- 1. The_ temperature of the core material " melt" predict;ed by MARCH at the time of the start of core-concrete interaction was 1737 K; at this temperature none of the core material would actually be mol-ten. Uncertainties in this initial melt temperature would influence both the CORCON'and the'VANESA results.
-2. The CORCON-MODI code was used in this analysis to calculate gas evo-lution ' rates and core material temperatures resulting from core-concrete interactions. Although we believe that using CORCON for such calculations is an improvement over using the results obtained from MARCH-INTER calculations, the CORCON-MODI code is not designed to model conditions .where a portion of the core material is fro-
, zen. Because . of this, the predicted core material oxide tempera-tures based on the - CORCON calculations are believed ' to be too high. It should be noted that a new version of ~ CORCON, called CORCON-MOD 2, is being developed- and should be available in 1984;
60 this. version is designed to handle situations where portions of the , melt are frozen.
- 3. ~ The aerosol generation rates ' calculated by VANESA are dependent - on the CORCON results; therefore there is an uncertainty in the aerosol generation. rates , used for . the LDHR calculations. For the calcula-tions presented here, the aerosol . generation rates are important 1argely because they influence the time when SGTS filter rupture could occur, if at all.
- 4. It was determined that the reactor building fire protection sprays would be operating during the time when core-concrete aerosols would be produced. The - QUICK code does not include models to account for the influence of the sprays on aerosol behavior in the reactor building. The sprays would be expected to increase the rate of aerosol deposition -in the reactor building; however, the fire-protection sprays are not designed for aerosol washout. A more careful assessment of the effect of these sprays on aerosol trans-port is highly desirable.
b s
~--. . 61 References for Chapter 4 4.1 R. P. Michner et al., Station Blackout at Brotans Ferry Unit One - Iodine and NobleL Gaa Distribution' and Release, NUREG/CR-2182, ORNL/TM-455/V2 (August 1922). 4.2- R. P. Michner et a1., SBLOCA Outside Containment at Brotana Ferry Unit One, Voltwn 2. Iodine, Cesium, and Noble Gac Distribution and Release, NUREG/CR-2672, Vol. 2, ORNL/TM-8119/V2 (September 1983).
'4.3 J. F. Muir, . R. K. Cole, Jr. , M. L. Corradini, M. A. Ellis , CORCON-
- 1D1 : An Improved Model for Molten Core / Concrete Interactions, NUREG/CR-2142, SAND 80-2415 (February 1982).
.4.4 D. A. Powers, " Containment Safety Studies," Appendix SA, Empirical Description of the Rate of Aerosol Generation During Melt / Concrete Interactions, in Report of the Zion / Indian Point Study: Vol . 1, prepared by W. R. Murfin, Sandia Laboratories, NUREG/CR-1410, SAND 80-0617/1 (August 1980).
4.5 D. A. Powers and J. E. Brockman, " Status of VANESA Validation," in Report of the Status of Validation of the Computer Codes Used in the NRC Accident Source Term Reassessment Study (BMI-2104), pre-pared by T. S. Kress, ORNL/TM-8842 (to be published). 4.6 ' H. Jordan, P. M. Schumacher, and J. A. Gieseke, QUICK Users Manual, NUREG/CR-2015, BMI-2082 (April 1981). 4.7 Personal communication, R .' K. Cole, Jr., Sandia National Labora-tories, August 1983. O 9 i b i
.e,
62 Table 4.1. Summary of core debris, concrete conditions , prior to start of core-concrete interaction
~ 1. Initial Temperature of Core Materials
- Oxides and Metals 1737 K +
- 2. Core Material Metal Phase Composition:#
Cr : 18,820 kg
.Ni : 10,450 kg Fe : 120,100 kg for INTER
- 116,100 kg for CORCONb
.3. - Core Material Oxide Phase Composition:"
-J02 : 159,400 kg Feo : 0 for INTER, . 5150 kg for 00RCON
- 4. Concrete Composition:
Ca003 : 45.5% Ca(OH)2 : 7.0% Si 02 : 38.8% H2O (free) : 4.8% Rebar : 0.135 kg - Fe/kg - concrete
- 5. Initial cavity radius: 3.23 m for INTER, 3.09 m for CORCON
" Quantities of Zr and Z r02 in the melt were varied as shown in Table B.I.
b4 ARCH predicted no Fe0 in the core melt, about 5000 kg
~
was adpgd7 jor CORCON runs based on discussions with .Sandia staff. i. Table 4.2. Comparison of cumulative evolution of CO, CO2, H2, and H2O for "CORCON1" and INTER calculations, at 3070 min # . Total Total cas release, release 00RCONI INTER (kg) (kg) 00 0.9 13300 CO2 6456.0 20240 . H2 100.5 757 H2O 2098.0 6575
""CORCON1" was the
- base-case CORCON calcula-tion done for the LDHR sequence.
63 i [. Table 4.3. ' Core-concrete r aerosol. composition pre-L
, dicted by VANESA (using
. . .CORCONI input-data), at 3070 min Total Material. aerosol
(%) SiO2 26.6 l Ca0 20.4 .- K20 13.0 Feo 12.4 . s A1203- 7.32 Na20 7.22 Mn02 7.11-La203 1.87 4 Ce02 1.08 All other materials 3.0 o! Table 4.4. Summary of QUICK drywell results for.LDHR sequence,'at 3070 min Mass mass"a , (kg) (%)
- 1. Mass settled on floor 145.6- 14.1
, 2. Mass leaked to reactor building 848.1 81.9
- 3. Mass airborne in drywell 41.5 4.0 Total" 1035.2-aPercent of total based on total mass from QUICK y,- calculation = 1035.2'kg; this is 6.4% higher than the 973.3 kg input amount based on the VANESA data.
,.m.. ..
e-64 f-Table 4.5. Summary of QUICK reactor building results for LDHR sequence, at 3070 min 3,,, Precent of total mass ( 8) to building" 1.- Mass settled on floot 602.5 66.6 2., Total mass leaked to SGTS 249.3 27.6 3.: Mass deposited on SGTS filters 162.0 17.9
- 4. . Mass legked through SGTS af ter SGTS filter rupture 87.3 9.6
- 5. Mass airborne in drywell 53.1 5.8
- 6. .' Mass transported through building leaks and to refueling bay 0 0 Total:a 904.9 "The percent' of total mass to the- building was based on the total
- mass to the building from the QUICK calculation = 904.9 kg; this is 6.7%
higher than the 848.1 kg to the building based on QUICK drywell results. D Time for rupture of first SGTS filter bank (81 kg to SGTS) = 2768
-min; time for rupture of second SGTS bank (162 kg to filters) = 2880.5 min.
e e
65 ORNL-DWG 84-5719 ETO , 2400 O coRcost s X INTER 5 2100 -- k C 2 w
$ e n =_
g ~~-;;0:00s w
$1800- -
()( )( 1500 l l 2500 2700 2900 3100 TIME (min) Fig. 4.1. Core debris metal layer temperatures predicted by "CORCON1" and INTER. ORNL-DWG 84-5720 ETD 3000 e
' ' " O coRcoNi
*- X INTER E -
E 0,2400 -- w g 2100 -- _ ,mmmmm__m
-www=__
1800 -- e.,
.s . .. a. . s ,. s,,
<.,. v. c. . v. g ., , ., c, , ,
i5m : : 2500 2700 2900 3100 TIME (mm) Fig. 4.2. Core debris oxide layer temperatures predicted by "CORCON1" and INTER. e 1 l
66 ORNL-DWG 84-5721 ETO 60 50 . E a = = = =- w - w Q 4n . e z O < p e 30. . 5 0 . a o . y 20 5 g .. to . . 0 l 2500 2700 3t00 . TIME (rnin) Fig. 4.3. Core-concrete aerosol generation ra es predicted by VANESSA, based on 'CORCON1" data. 4
m" O T E 0. 8 '0 2 81 2 1
- 7
. 5 4 8 G W D
- 0 L 4 N $1 R
O 0 0
$1 .
e m i t 0 s i s v r n i o t a r o t
- 2) n I
rC e E c S n ( o c E l . oM o s aI dT o i r e a l l 0 e 4 w di y r D 4
- o. 4 o
di
- g i
F 0 6
$1 0
2 5
- 1 oS oej oS oe- o* od t o S* *NOv ri<X nn amoZwc,n(DE n
! 2I'!. a ! !
l eco D T 0f 6( E 81 3 1
- 2 7
5 4 8 G W O 0
- 4 L i81 N
A O 0 0 i81 e m 0 i 6 t i71 s v s s 0 a
- 2) m i7C 1
E d e S ( k a E e l . 0. 8I M l i6T l 1 e w y r D 0 . 4 5 i6 1 4 g i F . 0 i06 1 0
- 6 i5 1
0 2 5 1 Q% QR Qoo @ Q+ oie oo a mOvCinn<2 QC*g<E 41 bm.
. < i " !i;l;11< . 1 .< ! < ,4 ;
1!I ,11 1 . 1 i i1
- A h
69 O
- N E
i 3 0 9 - i - 8 O 9 B C u
-8 .
G en V) O h an bO e is3 e in ~ w u u
- M W OE
+
~g M,4 h n=4 c=4 3
h 9
.- 43 O
_g g, 4 e w O %
\
O O Odg gI Ofg Ob U " i .
,01. (c) inoTiva I
e
wo O 0' T E 80 81 5 1 *
- 2 7
5 4 8 G W D 0 L 4 N $ R 1 O e 0 m i 0 t 51 s v n o 0 i 6 t N a r t n e c n 0 o
- 2) c NC E l o
( S s o E r e . 0, M a 8I 5T g 1 n i d l i u . 0 b 4 d r i o t c a e R 0 0 . 6 1 7 4 g i 0 F 6 1 0 2 5
'1 wN oN e- m eo wd' Qo nn QWo*/N%r3cnn 33%Nev ic<M
D 0. f T 8C E 81 . 6 2 1
- 7 5
4 8 G W D 0 L 4 N d R i O 0 0 1 e m i t 0 s 6 v N s s a m d 0 e
- 2) k NC a e
E l S ( g n . E i oM aI d l i dTi u b r o . t c 0 a 4 e 1 R 8 4 0 d o g i i F 0 6 1 0 2 5 1 c8 oG I o$ $ o2 oo oo b ^$ mn<M e @ M<3AU
I - Y D 0' T E 80 81 7 1 * . 2 7 6 4 8 G W D 0 L 4 N 8 R 1 O / 0 0 e 8 m 1 i t s v 0 s 6 s 7 1 a m d e l t 0 t e 2 T )C 7 s E g S ( n i E d l . 0M i u 8I 6T 1 b r o t c a . 0 e 4 R i6 1 9 4 0 g 0 i 6 F
/
1 0 6 5 1 0 2 5 1 1 t
< e- 2 oC e+ 0 8
%p 8 H D O1.b*: u
- m. .. _ - .___ _ . .- _ - . _ . ,_ _ .
w 73
- 5. : FISSION PRODUCT TRANSPORT CAI4ULATIONS AND RESULTS 5.1 Initial Nuclide Inventories
*= 2 Estimates for the-initial' inventories of Kr, Xe, I, and Cs nuclides in . the core are identical ~ ~ with those employed in the two previous
; studies,5 1,5.2 and are described in detail in Sect. 5.2 of the station -
blackout study.5 1 The initial fuel- loading . (in - 1973), the refueling
; schedule Land the power history to the eleventh month of the fourth refueling . cycle were used 'with the ORIGEN2 code .3 to calculate nuclide
~
5
- inventories at shutdown.
In determining ~the particular.nuclides to be include'1 di n the study, both mass- and radioactivity contributions were considered. The total mass of each' element, which determines. transport and chemical activity, is comprised mostly of stable isotopes, although the calculations do ac-count for the small mass contributions of all radioactive nuclides. The calculations do not currently account for the specific behavior of Te or
. Br; both were treated as isotopes of I in transport and - chemical activ-ity.* With the exception of the tellurium - precursors of iodine, the calculations do include the behavior of precursors.
The nuclides currently included in the calculation are listed in Table 5.1, which also lists their half-lives and amounts present at the
~
time of shutdown. - As shown in the table, all stable isotopes of each element are considered as a single nuclide in the transport calcula-tion. The total masses and activities for the elements Kr, Xe, I, and
~
Cs are listed in Table 5.2 for selected times during the accident - se-
.quence. The table entries indicate that > significant quantities of Kr.
I, and Cs decay before the severe portion of the accident (after'2040 min) is reached.- '
' * '- It.-is important to note- that, as for the previous two stu-
' dies,5.1,5.2 included here are only those nuclides with half-lives greater than 30 min. For these three cases, shorter-lived nuclides would have no significant impact on either the transport characteristics within the reactor or on the ' amount of activity ultimately released to the atmosphere. The curves presenting fission product inventory vs time discussed in Sects. 5.5 through 5.7 are all normalized with respect to the initial total elemental activities listed in Table 5.2.
Decay chains accounted for in the fission product transport cal-culations used for this study are listed below:t t l *Te transport models are currently being developed for use in the
.. .next study.
II3I"Xe, 133m 135m Xe are . simply added to the masses of I , and
- 131 X
U3 Xe and I e re This appears to adequately ap- ' proxIdate the, actual [ec,ay cEetively. ns. i m -na ,.-m~ >w,-,- .r.e --..m -w,- w ,. - - - - --- -
.--.,~,+nw-p+w . - - , , . . . . -
. . . . . ~ - -- . . - -- .. ..- _
74 ; 130I. : 130xe 1311 ___,131x, 132 Te ---+ 1321 ~___,132xe e
'1331 ___ ,133Xe --+ 13 3Cs 134Te ---+ 1341 ----+ 134Xe 134Cs ---+
1351 ___,135xe __ ,135Cs
- 136C s --+
i 137Cs ---+ 138Cs ---+ 5.2 Control Volume Characteristics For fission product transport calculations, the entire nuclear
.' plant is divided into two basic regions, each of which is then sub-divided into control volumes corresponding to distinct, physically sep-arate zones.- The first region includes the reactor vessel and steam
, lines, i.e., all parts of the primary system that remain pressurized
- j. after the reactor scram. The second region contains various containment or building areas and basically any volume or. system not included in the ,
first region. Calculational procedures for different control volumes of each region are similar, although the actual models used may vary con-siderably among control volumes of a region. The individual control volumes are the. smallest subdivision for flow and chemical activity cal-culations, in that the distribution of nuclide species in each control volume is assumed to be uniform.. 5.2.1 Primary system volumes The primary . system is subdivided into eight control volumes, which represent specific portions of the reactor vessel internal structure and connecting steam lines. These volumes are listed in Table 5.3 along
- with their deposition surface areas 'and free volumes; their locations and interrelationships are illustrated in Fig. 5 1.
Although the core region is assumed to be a typical well-mixed con-trol volume in calculating inflow and outflow, it is subdivided into 10
+ radial and 10 axial zones for the : calculation of fission product re-leases from intact ' fuel. (The first radial zone. is the center of the core, the tenth is the outer portion; the first axial zone is at the bottom of the' core, the tenth is at the top.) These 100 core nodes have .
equal volumes, but do not have equal fissio's product inventories or tem- } peratures. As shown in Table 5.4, decay Tower is higher in the central regions of the core. Nevertheless, core loading patterns ~ provide that s
. , - - - e---r,m.7 ..,e - . , _ , , - - , - - - - , - , ,-,%-._,...,.-,---,.._,-,.,,y...._
t 75 the longer-lived isotopee are concentrated in the lower-power, long-burnup fuel assemblies that are located near the perimeter. (Details of the core loading pattern can be found in Ref. 5.1.) The nodal fuel tem-peratures also vary considerably in dif ferent areas of the core, with generally lower values at the extreme perimeter (9th and 10th radial
. zones) and in the lower regions. Because the fission product release rates depend upon both the temperature and fuel inventories, this nodal-ization produces a much more accurate descri,-tion than would core-averaged quantities.
5.2.2 Containment and building control volumes Because all portions of the plant outside the primary system are considered as parts of this general region, there will be many dif-ferences in the characteristics and modeling of fission product behavior in the different control volumes. The particular volumes used in this study are listed in Table 5.5 with their surface areas and free volumes, and include such diverse entries as the reactor building, standby gas treatment system, and wetwell. The physical locations and inter-rela-tionship of the dif ferent control volumes are shown in Figs. 2.19, 2.20, 2.22 and 2.23 of Chap. 2. The main steam lines (downstream from the MSIVs) and the SRV tailpipes are included here, since they are not at primary system pressure and therefore are not subject to the same fis-sion product behavior. As discussed in Sect. 2.2, the drywell fails prior to core uncov-
, ery; hence, there is extensive venting from the drywell to the reactor building during the later stages of the accident. The drywell-wetwell and drywell-reactor building flows are major fission product transport pathways and are illustrated in Figs. 2.17 and 2.18. Except for a few brief instances, the SGTS operation is successful in maintaining negative pressure in the reactor building and refueling bay during the accident. As a result, very little flow f rom reactor building to re-fueling bay occurs, and thus fission products have very little oppor-tunity to enter the refueling bay, or the atmosphere except through the filters of the SGTS.
5.3 Calculational Procedure The movement and activity of released nuclides in the various reac-tor control volumes are determined primarily by the flows between the volumes and the temperatures of structures and fluids within the vol-umes. Additional information is required in certain volumes (e.g., the mole ration H/0 and I/H2O are used in the reactor vessel to determine the distribution of iodine species.) Flow and temperature information are obtained from the output of four dif ferent computer codes, whoso contributions are as follows:
- 1. The MARCH code, as improved at ORNL for BWR analyses, furnishes all data for the core and pressure vessel, and most temperatures and flow rates for the wetwell and drywell.
m 76
- 2. The ORNL-developed secondary containment model (described in Appen-dix A of Ref. 5.2) provides necessary flow and temperature informa-
- tion for the reactor building, refueling bay, SGTS, and atmosphere.
- 3. The QUICK 5 .4 code is used to determine aerosol concentration, plat-ing, settling, and leakage in the drywell and reactor building.
- 4. Cas release rates from the corium-concrete reaction af ter reactor .
vessel failure are calculated by the CORCON-H001 5 5 code. All data from the above sources are input to a npecial input processing routine that interpolates and integral-averages all quantities, output-ting a single data set at standardized 2-min intervals. This data set then becomes the principal input to the main fission product transport computations. The time steps for the transport calculations are gener-ally smaller than the times between data input, requiring interpolation of the data. At each new time step, the transport calculation progresses through the following stages of computation:
- 1. DATA: Input of new temperature and flow data, if needed.
- 2. RADIOACTIVE DECAY: Isotope inventories are revised according to the exact solution of the coupled decay differential equations.
- 3. RELEASE FROM FUEL: Using current temperature data and isotope con-centrations, the release of nuclides from fuel to gas is calculated.
- 4. TRANSPORT: The movement of nuclidos between control volumes (based on flow rates between connected volumes) is done explicitly (i.e. ,
using only the concentrations f rom the previous time).
- 5. CllEMISTRY: Chemical activity (e.g., plating, dissolution, I species
- distribution) is calculated using updated concentrations in each control volume. Aerosol so.tling and plating in each volume is done after the calculation of deposition or condensation of nuclides onto aerosols. .
- 6. OUTPUT: Done every 2 min af ter the begir.ning of calculations (2040 min).
5.4 Aerosols in the Reactor Vessel The term aerosol is here restricted to materials originally incor-porated in the steel structures of the reactor vessel or the Zircaloy cladding and channel boxes of the core. These materials are of suf-ficiently low volatility that the aerosols can be considered to form by condensation in all of the reactor vessel control volumes outside of the core. The degree to which the fission product species Cs1 and Csoll add to the aerosol mass depends on the amount of condensation permitted by the local temperatures. , 5.4.1 _ Production and deposition rates Figure 5.2 shows the amount of structural aerosols produced in the reactor vessel as a function of time during the course of the severe ac-cident. The figure also shows the plated and suspended aerosol masses
77 o for a base case that ass med 80% aerosol deposition by the time of ves-sel failure (2590 min). Because of the uncertainty in certain deposi-tion rate parameters, two other cases were also investigated; these cases assumed a degree of aerosol deposition at vessel failure of 60% and 95% of the total produced. Suspended mass at vessel failure was 3.1, 1.9, and 0.4 kg for the cases of 60, 80, and 95% deposition, re-spectively. 5.4.2 Aerosol effects on fission product transport As described in the previous section, aerosol deposition was treated parametrically by consideration of cases with 60, 80, and 95% plating of all structural aerosols produced. The degree of aerosol trapping would have no effect on the Kr and Xe releases since no inter-actions between these elements and aerosols are postulated. Overall releases of I and Cs to the atmosphere are almost identical in each of the three cases, although the behavior of these elements in certain con-trol volumes was affected by the aerosol deposition fraction in the early stages of the accident. The major dif ferences in I behavior occurred in the reactor vessel before failure of the bottom head at 2590 min. As shown in Fig. 5.3, the suspended I activity (including I in the gas phase and plated on airborne aerosols) varies somewhat, but the values for the three cases are very close at vessel fatture. Figure 5.4 shows the ef fects of aero-sol deposition f raction on the total amount of I that deposits in the r
- reactor vessel (including I plateout from condensation directly onto surfaces and I plated onto deposited aerosols). Both figures indicate that amounts of plated or suspended I at the time of vessel failure are largely independent of the aerosol deposition fraction, although these 3 amounts vary at earlier times. The reason for this is that in the case of 60% deposition, more aerosols remain airborne and available for I deposition. Consequently, when these aerosols do finally deposit, they carry larger amounts of I with them. It is also important to note that both plated and suspended I activity depend not only on aerosol trapping but on I plateout directly to fixed surfaces and gas flows f rom the ves-sel during periods of SRV actuation.
The effect of aerosol trapping on the release of I to the environ-ment is illustrated in Fig. 5.5, which shows that the degree of aerosol deposition in the reactor vessel is significant only in the early stages of the accident (before vessel failure at 2590 min). The only release pathway available prior to this time is by leakage from the wetwell air-space to the reactor building; hence, this figure is also a good indica-tion of the variance in the amounts of I in the watwell atmosphere due to SRV operation. It is important to note that the releases f rom fuel rubble on the drywell floor in the later portion of the accident com-pletely overshadow the early releases, thus rendering inconsequential the variation during the early stages due to aerosol deposition fraction
. in the vessel.
The ef fect of aerosol deposition upon the behavior of Ce resembles that of I and will not be described in great detail. The activity of Ca suspended and plated in the reactor vessel is similar to that of I as a
78 depicted in Figs. 5.3 and 5.4. Also, similar to Fig. 5 5 for 1, the re-lease of Cs to the atmosphere shows essentially no variation between the
- three parametric cases. Due t a the lack of Cs evaporation and venting from the wetwell airspace, atmospheric release of Cs occurs only af ter the vessel fails at 2590 min.
Because of ths. cimilarities noted above for various acrosol depost- . tion f ractions, it is concluded that this parameter is not of great im-portance in assessing the overall consequence of this accident, i.e., the release to the atmosphere. Rather than overload the reader with an excessive number of tables and figures, only a base case with the aero-sol deposition of 80% will be used for reporting results in the follow-ing sections. 5.5 Noble Cas Transport Re..ults Calculated Kr and Xe activity innentories (PBq) in several key con-trol volumes during the course of the accident sequence are itsted in Tables 5.6 and 5.7. The only noticeable differences in the behavior of these two noble gases are due to the much higher ef fective decay rate for Kr and the somewhat higher solubility of Xe. Both gases are evolved from the fuel in large quantities and ef fectively flushed through the plant systems by steam in the reactor vessel and concrete degradation gases in the drywell. As seen from Table 5.6, less than 1% of the initial Kr inventory (given in Tables 51 and 5.2) is projected to be released to the atmo- , sphere. This low value is due to the rapid decay of Kr nuclides rela-tive to the holdup time involved in this accident sequence. On a mass basis, the released Kr at time 3100 min represents about 76% of the initial l'tventory. Physically, Xe behaves in a similar way. Ilowever, Xe nuclides de- ' cay more slowly, and significant amounts build-in due to procursor de-cay. Thus there is less decay loss in the 3100 min spanned by this study. As seen in Table 5.7, about 60% of the initial Xe activity is projected to be present in the atmosphere at sequence termination. Tables 5.6 and 5.7 show that the major repositories for noble gases within the reactor complex at time 3100 min are the in-place fuel, the reactor butiding atmosphere, ni.! the main condenser. Figures 5.6 through 5.10 show the time-variation of Kr activity, normalized to the activity at time 0, in the fuel material, reactor ves-sel, wetwell water and air, drywell, reactor building air and water, and the main condenser. Figure 5.11 shows the timing of the release to the surrounding environment, also normalized to the initial activity. As seen in figures 5.8 through 5.11, virtually all of the movement of Kr before vessel failure is governed by SRV actuations (cf. Fig. 2.16), which induce flow not only f rom the reactor vessel to the pressure sup-pression pool, but also from the wetwell airspace to the drywell and f rom the drywell to the reactor butiding. The major releases to the en-vironment result from these SRV actuations and from the burst of activity occurring at reactor vessel fatture. As noted above, less than 1% of the initial Kr activity is projected to be present in the
gi W ^
^ - ' '
; ~ ^ ~ " -~~ --
'~ ~ ~
t ~.
,s Y ,
e
*d [ ' 79 .
y w ,, a 'I N% =;- o
~g environmentJ at the 2 termination of the sequence, most of which release occurs af ter vessel fai!iure at 2590 min.
Q Similarly Figs. J 5.'12 through 5.17 show the time variation of the normalized Xe activity fn key control ~ volume locations, the last one ,
'(Fig. ~ 5.17) referring, tot the timing of the release to the atmosphere.
l - ' ' As noted above, . the 'm]ajor. dif ferences in the. results for Xe and Kr are qM
. attributable to the much slower. rate of Xe decay.
s L ,, 5.6 Iodine Transport Results Estimated iodine activities (in .PBq) are listed in Table 5.8. for times.up to 3100 min following event initiation for the following inven-I tery locations: (1) . in . fuel- within the original locations in the core, (2) in fuel ~ material slumped onto the reactor vessel- bottom head and j.4 later located on the ' drywell ficor, (3) plated on_ the reactor vessel surfaces in all forms (e.g.,- condensed CsI, deposited aerosols carrying some iodine form),- (4)- suspended in the reactor vessel (gaseous plus on suspended aerosols), (5) dissolved in the reactor vessel water, (6)
;g plated on drywell walls, (7) gaseous plus suspended in the drywell air, 1~ - (8) wetwell water, (9) wetwell air space,- (10) reactor building water,
.- '(11) reactor . building air, (12) in the main condenser, (13) on the HEPA-l- ' filters, (14) absorbed on charcoal - in the-SGTS, and (15) in the atmo-l ,. sphere. f- . .
L 1 l' Interp'retationgof the iodine inventory estimates given in Table 5.8 j(4., may' bei aided by ? viewing Figs. _5.18 through .5.27, which illustrata the
. Ltime-dependence' of' these quantities, and by reviewing the descriptions
!. : of : the leakage' pathways shown in Figs. 2.22 and 2.23.
- As seen in _ Table 5.8, the ~ largest todine repository at time
~j g 3100 min is the deactor vessel. Of -the total 6854 PBq available at this time, about-400 PBq is in the tenth (outer) radial region- of fuel, which 1 remains intact!throughout the accident, and about 6000 PBq is deposited j -
on interiorhydpsel surfaces. Concerning the latter, Fig. 5.20 indicates that 7 the? dominant mechanism is plating of iodine-laden aerosol parti-W cles',-- accounting . for 60f:of the vessel iodine. Other iodine remaining $ in 'the' reactordessel is. due to Cs1 condensation or HI deposition di-iG , -
- rectly onto /vessell structukes. - This . second mechanism is augmented at m 3 -the L time of core slump l(2492 min) by the rapid boiloff of water in the bottom-head,: leaving previously dissolved iodine deposited on lower ves-selb surf aces. lit: is interesting to note that in .a previous . study,5.2 -
'adst oQthe-iodine. remaining in the vessel resulted from direct plating, sinceMhigh gasJ temperatures inhibited plating onto aerosol - particles.
4 J , HoweM, . in th'e Dcurrent accident sequence, gas temperatures are much
, . . 11ower,fpermitting more extensive plating =onto aerosols.
> @j = / DAnoiber majorirepository -is the water in the reactor building base-mentipool, L which contains ~239 PBq of dissolved iodine at 3100 min. As ain'the SBLOCApadcident sequence,5.2 the reactor building fire protection
- c. , : system /sprayslar'e. assumed to continually flush surfaces in the vicinity.
of = tb/ drywell : rupture. Iodine in the reactor building atmosphere is 1 assumed to -l equilibrate with and - partly dissolve in water on floors or
-walls,'iand "-j be washed into the basement pool. The calculational 3gy :: ,
; g C. (
,f6' -Q'
- j n
;3 37;w ,
80 mechanism used is equilibration with a partition coefficient of 5000. This calculational mechanism is a great simplification. The assumption of gas / water equilibration is reasonable, but the partition coef f'.cient would in fact vary with location (local water temperature, pH, and io-dine concentration) and with time of contact in a way that might be unique for iodine. That is, the effective volatility diminishes with . time as the dissolved 12 reacts with the water to produce the nonvola-tile I- and 103- forms. Therefore, short-term contacts such as with building sprays can exhibit higher effective iodine volatility than the long-standing basement pool. The selection of a single value of 5000 for the iodine partition coef ficient applicable to the air-water contact in the reactor building may lead to over-estimation of the iodine dis-solution in the sprays and also to a compensating underestimation of the retention in the basement pool. The other location involving major iodine activity is the drywell, where 178 PBq remain plated at 3100 min. Most of this plating occurred shortly after reactor vessel failure, resulting from 12 deposition directly onto drywell surfaces. Although the wetwell is not a significant iodine repository, it is interesting to note the timing of I transport to this control volume. As can be seen in Fig. 5.21, the presence of I is governed by SRV actua-tion before the reactor vessel fails and by drywell venting af terward. Because the only release pathway before reactor vessel failure is through the wetwell, the early releases to the other containment volumes also reflect this dependence on SRV operation, as seen in Figs. 5.22-5.27. The great preponderance of iodine in the reactor building and the . primary containment volumes occurs as gaseous 12 However, the forma-tion of a small amount of organic iodide is calculated, according to the model described in Sect. 3.2.2. As shown in Figs. 5.24-5.27, the pre-sence of this species does not appear to be a significant factor in the , overall transport of I through the reactor or to the environment ac-cording to current modeling estimates. The total atmospheric inventory of radioactive iodine shown in Table 5.8 represents a cumulative total. The release of 0.05 PBq at 3100 min is a very small fraction of the initial inventory. Gaseous I is the dominant form, and results mostly from the fraction (1% of 12, 5% of organic iodide) that is not absorbed in the SGTS charcoal. Smaller contributions released directly from the reactor building occur during brief moments of positive building pressure at core slump (2492 min) and vessel failure (2590 min), as noted in Fig. 5.27. Also evident in the figure is the increased release of I2 on suspended aerosols af ter tear-ing of the upstream and downstream HEPA filter banks (occurring at 2768 and 2880 min). The iodine release to the surrounding atmosphere in this report is an estimate that depends strongly on the assumptions regarding the operation of the SGI'S and the effects of the fire protection system sprinklers in the reacter building. Without the SGTS, releases from the reactor building directly to the environment would be much higher. How- . ever, the timing of HEPA filter failure is not particularly important for I release, since most I occurs in the gas, not plated on suspended aerosols. The reactor building sprays are the mechanism for removing a
81 l large amount of iodine from the gas in the reactor building, although no
. * - effect on aerosol' particles is' assumed. Therefore, any degree of aero-
- sol ' washout that would be affected by the sprays ' would reduce. the
- release estimates provided in this report.
5.7 Cesium Transport
' The activity of cesium nuclides at various locations is listed in Table 5.9 for selected times during the accident sequence. The normal-
' ized (relative to activity at time 0) inventories for many locations are
-- plotted as functions _of time in Figs. 5.28 through 5.35. .
The- behavior of cesium resembles that of iodine in many respects. Table 5.9 indicates that most of _ the Cs at 3100 min is deposited in the reactor vessel. Figure 5.30 reveals that the strongest mechanism for ; fission ' product removal is again aerosol deposition. Direct condensa-tion of Cs0H onto fixed surfaces also plays an important role, augmented
' by the Cs remaining in the lower . vessel, (previously dissolved) after the boiloff of water at core slump (2492 min).
. As is the case with iodine, a significant inventory of cesium is plated _ on the drywell walls at 3100 min. The mechanism here is pre-dominately the direct condensation onto surfaces. Unlike the situation in ' the - reactor vessel,- condensation onto aerosol particles that subse-quently settle or deposit is not a significant factor here.-
The . third major . repository for Cs is that remaining in the intact g outer , ring of . the core. This inventory is about 10% of the total Cs activity at-3100 min.- _ . Differences in .Cs - and I behavior are evident from several of the , control volume : inventories shown in Tables 5.8 and 5.9. Very little Cs
- dissolves in the reactor building basement pool because no equilibration
- is assumed -between the- liquid and gas phases, and because most Cs in the -
(reactor building : is in Taerosol particles. The quantity of Cs in the reactor building water depends on the degree of aerosol deposition pre-dicted for . that control volume. . This -latter situation is also respon-sible for the ' greater trapping of Cs in the HEPA f11ters of the SGTS. Although the SGTS charcoal was arbitrarily assumed to retain 90% of the
~
' gaseous _ - C s , the very small gaseous inventory rendered this assumption
- almost - inconsequential. Perhaps -the most significant difference in I and:Cs behavior is the sparging release from fuel _ rubble in the drywell, illu'strated by the dashed lines in Figs. 5.18 and 5.28. Since Cs is as-sumed - to_ be released from fuel . rubble -by sparging at a much faster rate-than iodine,'it is not surprising. that the fraction of Cs released to
~the environment is about ten times _that of I at the end of computations.
. F rom - Table 5.9, the:Cs released to the atmosphere at 3100 min is 0.04 PBq, ' less than 0.01% of the total remaining inventory at that .-
ftime.= Figure 5.35 ' illustrates. the timing and form of this release. Since no Cs volatility _is assumed, all Cs entering the pressure suppres- _ sion pool during' the reactor vessel blowdown remains there; hence, there 1* -is no Lpathway for. early releases to the environment, as was the case with iodine ' (Fig. : 5.27). Thus,- the appearance of Cs in the atmosphere begins only af ter vessel failure at 2590 min. , largely in the form of O 1
. j.
82 aerosol . particles, with- negligible amounts released in the gas phase. The solid curve in Fig._5.35 illustrates the dramatic increase in Cs.re-1 ease at the failures of each HEPA filter bank (at 2768 and 2880 min), indicating that the timing and ar.ount of Cs released to the environment is- very dependent on assumptions regarding SCTS operation. 4
83 References for Chap. 5 5.1 R. P. . Michner et ' al., Station Blackout at Broona Ferry Unit One - Iodine and Noble Gas Distelbati.on and Release, NUREG/CR-2182, Vol. 2, ORNL/TM-455, Vol. 2 (August 1982). 5.2 R. P. Michner et al. , SBLOCA Outside Containment at Browns Ferry
. Unit One, NUREG/CR-2672, Vol. 2, ORNL/TM-8119/V2 (September 1983).
5.3 A. G. Croff, ORIGEN2 - A Revised and Updated Version of the Oak
' Ridge Isotope Genention and Depletion Code, ORNL-5621 (July 1980).
5.4 H. Jordan, P. M. Schumacher, and J. A. Gieseke, QUICK Users M2nual, NUREG/CR-2015, BMI-2082 (April 1981). 5.5 J. F. Muir, R. K. Cole, Jr., M. L. Corradini, M. A. Ellis, CORCON-MOD 1: An Improved Model for Molten Core / Concrete Intemations, NUREG/CR-2142, SAND 80-2415 (February 1982). -p . ,9-e
k l 84 iL
. Table 5.1. Nuclides of Kr, Xe, I, Cs, and Te represented in the
- l: transport calculatior.a h
r -
"E* "*" #I Nuclide Half-life (geol) .(PBq)D-Kr 83 + 84 + 86 = -277.9 0 85 a 4.4 h 0.0333 879.9
'85 10.7 year 20.1. 24.7 87 1.3 h 0.0182 1,666.0 88 2.8 h 0.0572 2,368.0 Xe 131 + 13'2 + 134 + 136 = 2,803.0 0 133 5.3 d 6.862 6,289.0 135 9.2 h 0.138 1,739.0-Cs 133 + 135 = 837.0 0 134 2.1 year 45.0 290.5 136 13.0 d 0.316 117.4 137 30.2 year 624.2 273.3 138 32.2 min 0.093 20,093.0 1 127 + 129 + Br81 = 147.3 0
- 130 12.4 h 0.0080 81.0
. 131 8.0 d 5.304 3,175.0 i - 132 2.3 h 0.0915 4,612.0
' 133 20.8 h 1.177 6,529.0 134 52.0 min 0.0545 7,293.0 135 6.7 h 0.350 6,058.0 Te 132 78.0 h 3.064 4,553.0 L 134 0.700 h 0.0378 6,262.1 a Includes nuclides with half-lives greater.than 30 min.
b Petabecquerei s.1015 sq = 27,027 curies. i Table 5.2.- Total asse (geol) and activity t (P8q) of fission product elements at selected timesa Time after shutdown Element (min) 0 2,040 2,500 2,800 3,100 j Kr , geot 298 297.9 297.9 297.9 297.9 PBq 4,939 29.3- ;26.0 25.2 24.9 Xe ~ L geol 2,810~ -2,811 2,811 2,810 2,810 8,028- 5,784 '~ P8q 6,786 6,280 6,012 ll gb l geol ~154.3' 152.5 152.3 152.1 152.0 PBq. 27,748 8,636 7,747 7,268 6,854 Cs l- geol 1,508 1,508 1,508 1,508 1,508 PBq 20,774 ' 672.7 670.7 669.4 668.2 ' a 1.Petabecquerel I 1015 sq = 27,027 curies. }
.b Includes Br 81.-
s.
7 85
.v; Table 5.3. Reactor vessel control volumes '
Surface area Volume Interconnecting Control volume- 2 (m ) (,3) regions 1 Core 1.28 x 104 45.1 (2) (3) 2 Lower plenum 1.00 x 103 90.7 (1) i 3 Upper: plenum 34.8 27.8 (1) (4)
'4 Steam' separator ~6.4 x 102 38.0 (3) (5) 5 Downcomer 1.26 x 103 183.0 (4) (6) 6 Steam' drier - 2 ^.9 5 : x' 103 -64.4 (5) (7)
-7 Upper-head 1.98 x 10 2 89.0 8 Steam lines (4) 3.34 x 102 55.1 (6)
(7) 04ST (8) f (TPf-amain steam lines, downstream from MSIVs. b Safety relief valve tailpipes. Table 5.4. . Decay power in-core control volumes, 2.4 h afte'r shutdownL(kW).
.o.
Center Radial-zone . Edge Axial' node 1 2 3 4 5 6 7 8 9 10 i Top 10 149: 136' 144 132 146 145- 114 112 96 80
-9 277 260- 269 253 276 267 241 232 190 141 8' 361- l341 351 E334 362 352 330 311 256- 185 J7 -407
- :387 397 381 412 400 384 358 297 213 26 .431 412 .422' 407 438 425 - 417- 387 322 230
- 5- 444 428 436 423 453 440 440 406 339 240 4 452 436 444 431- 462 448 4 56 418 345 -237.
3 452 436 444 433 465 .449 467-- 423 346 226 c 2 426 E411 417 410 440 423 448 -400 322 198' v: p Botton 1 245 231 241 .229 251 246 230 210 172 117 i [: - O '
,w-e - ,- , . - , ,y m -
,-c_- , ,_ . ,.
86 Table 5.5. Primary containment and reactor building control volumes Control volume $"* Connecting regions 8"#**$)**** 1 Steam lines" (4) 103 1.65 x 102 (STLhf(9) 2 Tailpipes (13) 2.0 (STLT (3) 3 Wetwell water 3.83 x 103 (4) (2) (5) 4 Wetwell air 3.53 x 103 3.66 x 103 (3) (5) (6) 5 Drywell- 5.20 x 103 4.50 x 103 (4) (6) 6 Reactor building 2.43 x 104 4.25 x 104 (7) (8) (10) (4) (5)
- 7. Refueling floor 1.56 x 104 7.48 x 104 (6) (8) (10) 8 SGTS (6) (7) (10) 9 Condenser 1.52 x 102 (1) ,
~10 Atmosphere. (6) (7) (8)
%ownstream from MS1Vs.
Steam lines between reactor vessel and MS1Vs. - ? t
, . ;y- sy 2,g' .e O' i Table 5.6. Krypton inventories and release to atmosphere (P8q)G Fuel _
Wetwell Reactor building Time E** E** - Condenser Atmosphere Total Core _ Rubble Gas Water Air Water 0 4939.00 4939.00 2332h 26.47 0 0.27 0~ 0 0 0- 0 . 0 0 26.75 24880 4.53 0 20.05 0.031 0.57 2 x 10-4 0.32 2 x 10-4 0.357 0.15 26.02 2590 d 2.'88 1,62 16.91 2x 10-10 1.25 0 0.75 5x 10-" 1.19 1.09 25.69 27688 2.85 9 x.10-4 0 4 x 10-4 1 x .10-6 5 x 10 4.79 0.005 1.19 16.46- 25.30 co i 2917f 2.85 .' 8 x 10-6 0 3 x 10-6 g x go-/ 5 'x 10 1.81 0.003 1.17 19.30 25.13 3100 2.84 8x 10-10 0 4 x 10-10 g x . go-4 5 x 10-6 0.27 5 x 10-4 1.15 20.67 24.92 al Petabecquerel E 10_5 1 Sq = 27,027 curies. DSoon after initial cladding failure. C Just prio- to core slump. kirectly following reactor vessel failure.. 8 First bank of HEPA filters f ail. [Second bank of HEPA filters fail. 4
7, , _ _ - - . '^: , ,
/
~
1
~
s b'
"ft i
~
.'a-1' ,
e a-6
. Table 5.7. ; Xenon inventories and release to atmosphere (PBq)#
Fuel- Wetwe11~ Reactor building.
; Time Core E E Condenser-' Atmosphere Total g -Rubble . Gas Water Air Water.
0 8028 8028 2332 b 7.s. ' 6443 'O 0 6450 2488 8 778.2 0 5375 0.031 78.5 0.02 27.6 0.02 65.7 12.8 '6292 2590d 356.7 405.7.- .4.320 2 x~10-10 L317.7- 0 128.0 0.12- 291.0 175.5 6195 o$ 2768 8 4x 10-4 3x 10-4
~
347 0.22I O 2 x 10-5 ' 1233.0'. 2.2 292.0 4164 - 6038. 2917I 341 0.002 0 3 x'10-6 3 x g g-S .- 2 x 10-6 -461.0 1.04 287.4 ' 4856. 5947 3100 331. 2x 10-# 0 3x 10-12 2 x 10-3 5x 10-6 68.1 0.22 279.3 5105 5784 al Petabecquetel 's 1015 Bq = 27,027 curies. b Soon after initial. cladding failure.
#Just prior' to core slump.
d Directly following reactor vessel failure.
~
First bank of HEPA filters fail. [Second bank of HEPA filters fail. e . ..- . . .
e 6' G 'o e O 4 Table 5.8. todine inventories and release to atmosphere (PBq)# Fuel Reactor vessel Wetwell Drywell Reactor butiding Tim Condensar g* " .
- Plated Core Rubble F1sted *'g g Water Water Cas Cao or e Water h oal HU4 Atmospherd Total
'0 3
27,748 27,748 2,332 8,047 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8,047 2,488 , 928 0 4,122 80.8 2,637 0.02 4 = 10-6 4 go-e 0 2 = 10-1e g , 10-5 0.083 7 gow o y go-10 7,768 4 2,590 436 '468 6,686 5.853 0 0.23 4 = 10-5 4m go-5 0 6 = 10*7 5 = 10* 0.092 3 = 10-6 o 3 ,3o-7 7,596 CD 2.768 , 420 65.I 6,439 0 0 3.3 6 = 10" 2.9 182.6 1.4 200 0.088 1.6 0.0011 0.016 7.316 @ 2,917 7 410 18.6 6,293 0 0 3.5 7 = 10" 1. 5 198.7 1.3 221 0.086 2.8 0.0014 0.029 '7.151 3,100 393 1.7 6,032 0 0 3.4 7 = 10-' 3.68 178.1 1.1 239 0.083 4.9 0.0014 0.050 6,854 1 Petabecqueret a 1015 Sq = 27,027 curtes. Soon af ter intttal cladding fatture. Just prior to core stump. kirectly following reactor veaset fatture. First bank of HEPA filters fatt. t
'Second bank of HEP 4 filters fall.
-7 . . . . _ . - - - - - .--
c r.; , -
' 8 b
/
, .ii -
s
- 3. , ,.
i- -J i: a : -
.r.
1 4 l' L i- , Table 5.9. Cesium inventories and release to atmosphere (PBq)a 6
. _pu,g Rosetor vessel Wetwell- I - ****' II I"8 SCTS .
0** ** #I**d -' O** ~ *" '"*** Charcoal- letIp- Atmosphere. Total.
) Core Rubble Plated ,'ovepended ' Water . Water Ces euspended' - suspended Water 0- ^20,774 '20,774 6' 671.5 0 'O 0 0 'O L O O O- 0 'O- 0' 671.5-2.332 O 0. 0' 8 325.4 , 26.2 0.11 0- 0.054' O: 670.7' 2.468 .105.7 0 213.2 0 0 0' O O 0.
2,590d 66.3- 39.0 .561.2 .3.4' O: 0.46 -0 0 0 0 0 0.059 0. 0. .0: 670.3 $ 8 0.023? 'O' 0.46 , 0.12 10*' 5 = 10 - 2.764 66.3 560.5 0 0. 0.004 - 42.1' ~ 0.020 0.059 O.052 669.5'-
' 2,917f' 66.2
-2 = 10** 560.1 0 0' O.46 - O. ~ 0.002 42.0 0.019 0.17 0.058 10-' O.068 2 z ' 10* 669.1 3.100 66.2 ' 12 = 10-e 559.3 '0 'O 0.46 ' .0 0.004 41.9 0.022 ..' O.2 5 - 0.058 2 = 10-' O.068O.04 668.2 ,
81 Petabecquerel E 1015 sq =.27.027 curies. DSoon af ter 'inittal etad' d tng feiture.
#J ust prior to core eluep.
dDirectly 'following reactor vese'el feiture. First bank of HEP 4 filters fail. [Second bank of HEPA filters f ail. I e . ~a . . .
V r 91 ORNL-DWG 84-10866 e
.:*/ SHADED AREA REPRESENTS
"' DOWN-COMER CONTROL AREA @
- @ UPPER
. aEAo
~ ~
{ljlOf (
@ DRYERS a -
l 1 .. . . . . . . . . STEAM vc.... . . .. .r re.:. c. ..~................c ygts
). b
"'Aalaea-AND STAN0 PIPES ..
's ffff 999fff f 5.%
b s b
=
,- ,,,,@ UPPER etEsuu 5
fi, l
,1 i
, e en 1 g-, ' '
-umis.
,its e
e l l
@ PLENUM LOWER
=
ld v DRYWEU. WALL l CONTRCL ROD GUIDC TUBES l
'O:c. g;;:
usiv I I WWWWWWWWWWEWWW1 Fig. 5.1. Reactor vessel control volumes. e
. , . . . - . . , _ - . - , . . - _ _ , . , _. ;. ,.7, _,
i . - 4 9 40 O~
.c ' ORNL-DWG 84-5728 ETD
~ REACTOR.
VESSEL-
, ' CORE SLUMP FAILURE
. . (2492) - '(2590)
U - ' TOTAL AEROSOLS PRODUCED 't AEROSOLS DEPOSITED
?g ,. .____.....__. ......__.._______.. ...............
M a w ,e 4 .L (n -
'/
rn e
. co e -
i g j_ .r.
.e
.J e
. O s -e 1
<n ,
o ; N (E e y a g . ,o . o
' e s
c
, s I '
f ,*d i I 1
- i. .. -
.. 7 .
, r t
AIRBORNE. MASS
-:/
O
- g .._.
2100 2200 2300 2400 2500 2600 2700 2000' 2900 3000 3100 TIME (MIN) g l Fig. 5.2. Aerosols produced in reactor. vessel; base case, 80% i ! deposition. ! L I t 1 s . '? i 1 w g
4
' ' _io , ,e .cs
~
a L. ORNL-DWG 84-5729 ETD REACTOR VESSEL o
*i CORE SLUMP FAILURE 5
. (2492) (2590)
"o( = 6
, . . n 5 ?V .,
-1 t '.
E srb su _
>I H 31 Degree of Aerosol Plateout e
([ "
$'O, in Reactor Vessel c
*o
- ' BASE CASE, 805 DEPOSITED
- -- ------- 60 5 DEPOSITED I
o"
-i - - -
- - 95s DEPOSITED e-o,
=
i v-g 2100 2200 2300 2t00 2500 2600 2700 2000 2900 3000 3100 TIME (NIN) Fig. 5.3. Predicted airborne -(gas plus suspended) -iodine activity in the reactor vessel as a function of assumed degree of aerosol deposi-l tion.
'r s
ORNL' DWG 84-.5730 ETD b -
~1
~
b]:
~
', e b~
.-]
~o
~1:
a 31 : j-H S1
- m.
- s'a1-O C j Degree of Aerosol Plateout in Reactor Vessel
~
REACTOR
' .o.] BASE CASE, 805 DEPOSITED VESSEL CORE L P. FA RE f - --------- 605 DEPOSITED o.]: ---< - - - - 955 DEPOSITED
., : u.
'O
~ , , . . . . i e i I
2100. 22$X) 2300 2100 . 2500 2600 2700 2600 '2900 3000 3100 TIME-IMIN) Fig. 5.4. ' Predicted iodine activity plated in the reactor vessel, as a function of the degree of aerosol plateout. g g 4 *
'- * . ,a ..
*: .e'- y
-o "
'_ ' e ; .
e.
' ORNL-DWG 84--5731 ETD
.o
]
5 REAe7R. VESSEL
' .a'o,- CORE SLUMP FAILURE
'-' s -(2492) (2590) 1,
'o.]: o
- *o
-1
.i
03 n :
- =
3 ,
% - ",,s o .
$*o, ,
,- u e e
- -l... Degree of Aerosol'Plateout i; o P in Reactor Vessel
= ?f"p -
f ,e
,i / ./ BASE CASE. 8JS DEPOSITED o : /' . . . . -- -- -- 605 DEPOSITE D o]
. :( .. ... .-..-95s DEPOS ngo
- .f
'o., a i a e i i i a a e' 2100 2200 2300 2400 2500 2600 2700 2000 2900 3000 3100 TIME (MIN)
Fig. 5.5. Predicted iodinc' release to environment as a function of assumed degree'of aerosol:platcout in.the reactor vessel.
. .. 4 _
~g ORNL-DWG 84-5732 ETD REACT 0g COMPLETE VESSFL HEPA FILTER n CORE SLUMP FAILupE FAILURE
' O- INVENTORY (2492) (2590) (2917) IN CORE ir it
- e. OUTER RING OF CORE '
/ REMAINS IN PLACE w -
- / ___
M - O . W-z O e
.N. g: -
cn _-) - 8 INVENTORY IN
. i, g FUEL DEBRIS
.{
e , , o z - l
. \.
b_ -
- - 'g
.. i
- : s,
- i l \
_ : i
- i
- i
- b. . . . . . . . .
2900 3000 3100 2100 220C 2300 2t00 2500 2600 2700 2000 TINC (HIN) Fig. .5.6. Normalized krypton activity in fuel material. The ac-tivity is normalized to the initial activity level for nuclides with half-lives greater than 30 min. This normalization is also used for the results presented in Figs. 5.7 through 5.35.
1 p f
,p, D .0
,0 T
E 3
,- 3
- 3 7
5 4 8 0 G~ ,0 0 W" 3 D- . e L-N~ c R- a p O s 0 0 i92 s a g l e s E C 0 s ' 0 e A P 0 2 v S S r A o c t n c I a n 0 e o :0 7 r i 2
.a r Y i n
- - s I K -
N y I t R E - . M i OLR) .
, TEU0 CSL9 0(
0 i v 4 ASI5 o 6 EEA2 _ 2E t RVF( M c I a T P n M o U) . t s L2 . 0 p S9 0
.e 4 , 5' y E2 2 r
~ r R( O'- - k C~ d e z 0 i
/
0 l 4 2 a a t r o N 0 . 0 7 3 2 5 g i F 0
' 0 2
2 5' 0 0I o..__ 2'
'.'o .i ro ._ Y o-4 - .
Mxoezu>z ouN a_xrzoz
;.e_
A
~
~ ~
1 s n p 5 . _( 3 'l , g t
~ ORNL-DWG 84-5734 ETD REACTOR VESSEL-
. ' CORE' SLUMP. FAILURE
- ~
(2492) -(2590) n
.O ]'
~
~
, - .1r o T ~ o WETWELL GAS SPACE .
~=
.: J
.g.- .
o . SUPPRESSION-
~l # POOL 4
? - i 8
>4 1 i
' ~
xE
- o : m l' gs .
l 6-% yz 'O ,
\
6 ,e . , i i. 1
.a . . g ,e l
.> E Z ; 1 l'
, N,,. i t
\
.. s
*g
_ s ., I i 4
, co -
'(Jc iE e. s .I s s .: l'-
l-j
\,
)oa' c
E =
- i. ;-
is .
.l
's g
' E : ,,' i s g
. OS , l'i-
, i i
- z'o -l l 's '
i l \ 3
- l l
, \, ,Is' \,
; :': E 'f= l l 4
o ,_ l ; \; , ir\'i 2 . i l : 'A.
.lL '
'O
~ '
'l *
~
WATER N' ' , l l 's l
'\
i
' l c.g _ ;
e- .-
~
2100 22'00 2I00 2I00 ~ 2d00 '2d00 27'00 2d00 2d00 3dOO 3l'00 TIME-(MIN)
. Fig. 5.8. Normalized krypton' activity in the wetwell water and gas
! space.-
T-i ~ 99~ n,.;-
,,h
g, O-.~
= _g.
.a n e
.. , .s.
. 3-
.e
_8o i n
.z
- 8 =; -
y - e
> w e=4 e
.u E
o m-
-E e-
~ .e a
.e,A h
8 s u N ~ I> e . M u
.-z-g
.. . r-
.Rdgg e
r., -w - u n.
. -..J e
=ges 8'u-
.w
~
O a
' x
.r.- w
. g /
- 23. o
-8 :[*
w
.: 4~- a :q
.gw u
c- ' 2 4 _ _8 .. N ' g.
,D
- m. ...
en
~ .w
_8e
-i-
_8-
.4...
g 8. it y ....,,.,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, g c.01 ,_01 ; s01 y 0I: 01 01 1. ,,.0 l =
,O l ' ".01 180fN3ANl.01:032bWW3N m-
- W -
~ -
- _ = _ ,,
4
t , i g
< . s , ,, s E
w
- e. j j .i REACTOR.
- M- Em.
VESSEL-CORE SLUMP FAILURE' (2492) - '(2590) M
- r. ,
'o
-l ir , u ..
3 REACTOR BUILDING Aih0 SPHERE
,n .
*o al i . . . . . .. .: , . . . . . . . . . . . . . . . . . . . . .
T * '
... ' MAIN CONDENSER SYSTEM
-o}:
j- .. o I
- y. -
>* 31:
E .1 , o :: l - .,-
-g-st u _, ,.
> : 5 Z ' '
~
.i REACTOR BUILDING BASEMENT POOL
- m. . . ,-- - .. .!' . . .
. o.
c3 cw 31=
. I' s : ;- , .--,
]o' l l'
.(E r -? : ...
x : : co' - .- z'o -i
-r l
e- ,
'o
-i ~
r - . o ., ,-.,
=
a
'o
~
, , , , , , i 2100 2$00 2I00 2/00 2500 2600 2700 2000 2900' 3000 3100 TIME [NIN)
Fig. 5.10. Normalized krypton activity in the reactor. building at-mosphere'and basement pool, and in the main condenser system in the tur-bine building.
. + . . . .
7
;e - :..'
- r
- e. - ;; -
v
-l .
l ' 3 j ORNL-DWG 84-5737. ETD REACTOR. VESSEL CORE SLUMP . FAILURE (2492) (2590) _?o)
,, [ u o.
o
-1 .
o,
=
T -
> O.i e .: 0-o : -
z w i
> 5 z - :
n - c'O-i w : S N : 3-
~ci0 r :
.m. :
So~ -1 I
?o
-1 :
9- C i
=
2 -
'o~
2100 2200' 1b 2 2I00 . 25'00 2N00 2700 i 2000 m i s 3000 M' TIME (MIN) Fig. 5.11. Normalized' krypton activity released to the outside' at-mosphere.
^
^'S'^ s w..
~
'I ' n * ?
.y - -
<,.e <
!.ORNL-DWG84--5708 ETD -
. ' INVENTORY-1 O-. IN CORE -
-1 ' REACTOR'
~- .
VESSEL
'. CORE SLUMP,: FAILURE
- ,(2492) (2590) o p E
O ' H - Z. u Z
.O -
.L.
U - N -
' -o m N J -
-T -
E OUTER RING OF CORE-
- "'"'-~~~-*
'@' , REMAINS IN PLACE Z _ .
-i i 1 i L l
i
. INVENTORY IN.
FUEL DEBRIS t 6
.5 8 e i g s.
7 8 5, - O '
~ ' ' ' '
2100 2NDO 2I00 ' 2400 2I00 . 26'00 2700- 2600 2dOO 3000 3100 TIME (filN) Fig. 5.12. Normalized xenon activity in fuel material.
O C O Q R O t ORNL-DWG 84-5739 ETO REACTOR VESSEL CORE SLUMP FAILURE (2492) (2590)
'b _ _
u o w ce 'O _ O-XENON IN GAS SPACE z o _ Z O _ S w
.5 -
_2 cc E? 0 0.. z~-
?
C3 '
~ ' ' i i i i 3 2100 2200 2300 2i00 25'00 2600 2700 2000 2900 3000 3100 TIME (MINI Fig. 5.13. Normalized xenon activity in reactor vessel gas space.
y s i ORNL-DWG 84-5740 ETD REACTOR-VESSEL CORE SLUMP' . FAILURE'.
- (2492)- . (2590) 7
.o o- v
-1
- i L ,WETWELL AIRSPACE
.?o :
k
-1 l
/
~1:
l
? - J SUPPRESSION
-1 -
4 i POOL 5
. -i T - i
>o x -m : n n
4-s s o.: l\ l- , 5- 0 pll!.'N ', & ,
- W' 1 l' ,
z 5
,e -
n
'q:
\, it oi u
s : O l l e l i l- l _ro ? (C i s r, r : l i [,, e , e : . < , og - i, . l l z ~l o li .
. i i
't 4,
l l l \
- l. . ; . 't, e:
i
! \l ';
l l C3 i 8 * * \
~l -
l
%w e : . l \
l
'o l i i
~1 -
l l l
. \.
~
l l l \
?o-- . . . '. '. . i i i .
2100 2200 2300- 2400' 2500 2600 2700 2000 2900 3000 3100 TIME (MIN) Fig. 5.14. Normalized xenon activity in the wetwell-water and air-space.
g.
, , 105; .: o:
i 8-- i to '
.. . .i g 1
.e a
.g- -g
-a.
'Z t e ..
> . .O
-k. .
N
' u(
- _@ m o ,.
.c
.a
. g.
.re
_8 x. m - u
> -i
' z u g
-adss m E-M a d S; -__ g
. ;o =ge
~
y ~$ e N
.k-AU
,g---
??5 m ,- s g-- .@
O s ; ;'.: - c_ 7 -
.8g in in
.o,eo re m.
. _8 N-r 8'
.....,,....,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,r--s
'+ : y -pl' ,.0I .j0i OI . '01' -
0i 0I 01 ,,.OI-f.01. AB0 fN3ANI032b8W8d.0I N
+
,,.q. --
,a-A
- ORNL-DWG 84-5742. ETD REACTOR' . , -
VESSEL CORE SLUMP- FAILURE (2492) (2590)
*b., n n.
;j REACTOR BUILDING ATMOSPHERE,
'o ,
=
7 o, K-j
. MAiihoiMi3Eisir3Eti'"
= -.
p - f.
> Si oc 5 <---
o s - za
- j .
W' if ' REACTOR BUTLDING BASEMENT. POOL .. , z : l - 3 ,.- -- - .,. - g o w Si N : ~* K-Ci _i *
,a
, r : : M : : l C1 - : . z'o i - 5
,/
y - . o ,= .
- e-o, i
=
o - y , o . .
> i . . . . . .
2100 2200 2300 2t00 2500 2600 2700 2000 2900 3000 3100 TIME (MIN) Fig. 5.16. Normalized xenon activity in the reactor building at-mosphere and basement pool, and - in the main condenser systeam in the tur-
- bine building.
i e e e e o e v _ _
g, _c,,
. . . ..- ., n ..
; 7,n,y - - .- n- = .y - . g, -
- - ; mss ~.,- 7,n" < ., := .; y
- n. ._ .
A j. ; . f V,z n. % q-~s.)p i .fA~ . ' ' " '
~,..,e q , '. , ;--
YNN x , e ky
\'--
e
, 4. m. o ..WN 'W:-
;; , QC ~
Q, ' J {' , ' ' - J, .,
,X;h,;f:lm .f . A.: ,5'
^
s g
. l'
.; y 9 - /, -. ,5_s^, $5N A l -
4, .
~
- % h .L :. l * ;7 '>4 '
'b m f-: W'Q%
~ ,
h.' ,.ywy3 y 7.;.y k..) ' [,.f' - Rl y. ,&n : ,L .. ,' C
~ &~ ; , . Q .,, .l:. ,
-_ .~ 4 n.,-
, nx:
n, . q-i yy _ w,.y c __
- .c7y. .
p; ,;a ;.. . <? y, - .., ,
.~ n 3 s
~ .
+ . . .x a
- f. . ,,_y.s,.
a<g. = a f,.,...
.4. ,
8
, ),<
,.-~.u- t i, , ,-_. r .,, y,o e
. ,m , a
' *4't#:
..g.7.-'
'a i
*4
# . ' " ' ~ ...
--5,^M"'. 'W
; .n'v-'j ,
.cl e
,.4+#+.v
'A P [, T .
;. s . T d' q ., '.r.-s ,< ;ar *%i ~
'.,.p.',s
",f,~',9,./p,,,',... :.. * ,-'
c m.-.,,..'... I,)' ; l j~ w') f;
,o
- ~ -f. --t = -
s t ,
.;;}
. Ay . ,S". ' ,cr o^~; a.,.,.. . %~w.mp. . .o ,
-: ;-., , n
.,,..,_ 2,-.m ~ ,;' ~2 m . _ -
< ~ -,e z / ?* . .
m, y. p*- u
. a,. ;g: - ; !,: . .s . f
-m u ry ,. _: ,., ~ ,, n. cm,, eq G, .: .' -, ,:
,s,y s, og,-
- -. .6, a mm7 yycy ~.vm .-m> .r-c R.
; Nc., ,..s-m3;, -
% Ge ., f .i. . . .f.~ . ~-~ s' ,:.
,. .w , n, . . ; .:.,,
g . ,- , ., , . .g.
.,. cn -
, g.%.. -:9 .6 % g-.. .s. .--.
w . g-- w
.,,3 y .g-J...MM'. a ; ..g:*7..
. e_ . ."*: 3
~,, C ORNLaDWG 84-5743 L
' ETO y p :. .g%. .,O
.~
-c. _ . . J-y . ' ": ,,i
':?,."
W,.' iREACTOR~. ' y- > o. m' 3VESSELL T.>-U .*
. I~ - ; .1 : CORE SLUMP ' ' FAILURE- 4tf .#,. :d
?. , , - (2492)< ~ (2590) :
~
y S %- ^C~
.o
+ %. . o !._m w , n
,a
; ;q
,-e ?
U
^
.o a p ;
~,7
- : u- . <
'o
. ,. x
.w
, Wn -p-
. . _ . .. .,. .p -_ m w ;< r ny
-=_ ,
"u ,
.~ . y- -
~
, t.
' a >
-.1 e.,
1 6 ., y. r
.t j-
. o , , .
-Qo g
p _ ,- o . -- H.~ , _i.azoo . . , -
~ ,-
m g
> mi : .
w 7-o o-oA u ta .: N : 3a~ xA r : m : cx - - z'o.! :
?o?- :
f - O A _=
- 44
'o , i , i i e a 2300 2NO - 25'00 2600 . 2700 2000 , 2900 3000- 3100
'~'2100 L2200 LTIME'(MIN)
Fig. 5.17. Normalized -xenon activity . released to Ithe outside at-
. mosphere.
- s. ~
A
'i dRNL-DWG 84-5744 ETO 1 b_ -4 :
INVENTORY IN CORE
. REACTOR ~
. . VESSEL'
* ~ CORE SLUMP FAILURE
.o_
(2492) (2590)
~ " e o p . , OUTER RING OF CORE
.g O - REMAINS IN PLACE -
uz
> .: *o.: .
.z. .- ;- ,
a - u ! N - ,
's, l
30 .l
,' FUEL DEBRIS .$
gi-! m - :
, ,/
-O - ,,
-z - ;-
b_ ! '
~5 ' ,,
e : o, 2:00 2200 2300 2t00 2500- 2600 2700 2000 2900 3000 3100 l TIME (MIN) Fig. 5.18. Normalized iodine activity in fuel-material ~.
e e- e , .' g v' ORNL-DWG 84-5745 ETD
'o_
~5 REACTOR VESSEL
~
' C0hE' SLUM P FAILURE Y (2492) (2590) o_:
~
'-! .t. ',
'4 y . .
M - + { o .
,' 2C'
'% 7.:.
= l :.
> : i: -,
Z - j .
*~*
2 >1 o - ' .k i w
.r 9 - ,.
,e' @-
No= g: - m , p-o - Z - '
- TOTAL GASEQUS IODINE
~
\s ..........CsI o_
\' * -
- AIRBORNE AEROSOLS
~
T I : o
~
2I00 2d00 2600 2/00 28'00 2N00 3dOO 3100 2100 2IOC 2IDO TIME (MIN) Fig. 5.19. Normalized . iodine activity suspended in the reactor vessel. Iodine produced near the time of core melt is rapidly adsorbed on aerosols.
, p r
Ij 1 REACTOR
' VESSEL-b., CORE SLUMP : FAILURE
*E -(2492) .(2590) i- i, r ,.......-................:............................................
.-e 5
- l.
.. .f
$ l A : ,: .
E O O.-* : g H a + gz . ..- ,.d z-7 - j., e O S-
.(J -
?
N e-o
.J -
o C q E - e 1 E 4 t O z .-e : t
- j
- g IODINE CONDENSED ON SURFACES
- 'FROM THE GAS PHASE t
- l. - --------- AEROSOLS PLATED ON SURFACES '
%-?
,i "*""
- IODINE'IN WATER
- l:
Li t l t o,_. I r
. . . . . . i . . .
2100 2200 2300 2t00 2500 2600 2700 2000 2900 3000 3100 TINC (NIN) Fig. 5.20. Normalized iodine activity in the reactor vessel water and plated on interior reactor vessel surfaces.
. .y ,
e o e. , - s ' >- ~e' :
. . - ~
d s ORNL-DWG 84-5747 ETD VESSEL'. fo1 CORE SLUNP FAILURE j (28i92) (2590) r-O] IN PRESSURE SUPPRESSION POOL f - SRV. , l o] ACTUATION l
?- 1r
>< 31
.E 5 I
'O : 'l h .
Z' W i a
> 5 z : ,
, " , - - . . - - - - . IN. WETWELL
. , - - - AIRSPACE a i
- w 5 N -: -
. .l y- _ ;---------.'.,_,..; -
cc r :i e -: ' o zge. - ,-----. e l,,.l
-1 ; ;;
- , ie
=: *
'a
-1 i
ll p: l' ;; oi *
. 'l
= .
* ,i c ': .
,l :,.:
a-4 I i 1 5 I I E E 3 3 3 2100 2200 2300 2400 2500 2600 2700 2000 2900 3000 3100 TIME (NIN) Fig. . 5.21. Normalized iodine activity in the wetwell water and airspace.
r a w 4 f ;t n REACTOR ORNL-DWG 84-5748' ETD - O CbRE SLUMP . A E ' DEPOSITED
-l: (2492) (2590) *,,...
" " - " ' " "~
~
.y oI u o
. /,, -
o'
-i: GAS ip -
'~
>* 31=
x O -
~
z '. w i
> E i
Z : ' i w O, E " N : . ~
's'
)O1 C
r = 'I x , 9 :- 2 AIRBORNE AEROSOL 31 1 ! S, o Y ' l { COMPLETE O
-1
- j. g HEPA FILTER r
j ,
' FAILURE l l (2917)
- ?: l O
-j ./
( i
'. -i g o
- o ; l
~ '
2100 22'00 - 2M 2400 2M 2M 27'00 2M 7Ao 3do 3 [00 TIME-(MIN) i Fig. 5.22. Normalized iodine activity in the drywell. 1 I e e o e o e
xi .
,s ..
-g- e- ,', , .,
'I
%e
~ ~
REACTOR VESSEL
? CORE SLUMP , FAILURE DISSOLVED IN WATER 3] (2492) (2590) j ,,,,..--- ---- ----- --------------------
,' , s s
- ?o -
~1 l
.E l.
r-31 : .
! TOTAL GASEOUS 2 .'
f - l
. >. a, .
I
.M:
'O :
a c-,' , Z : 7-o . . O y1 -
~;
. 'i.f. . AIRBORNE ON AEROSOLS t4 5 I I!!! a
^
.! t A,4 :
,--- i '! e ":1 s C
'J go-,- ,
*3 bu, .;i4<
- ;iw<!.ljh;o
- W! .
m agj!iji!,:!.A. i o?.;:g.,g gjj p
= : .. v :s O
-l e 2
E s
#. ' .. COMPLETE
'O
**3
[ HEPA FILTER E FAILURE l (291T) r - , e o! E : p: ; i o.- 2100 2200 2300 2900 - 2500 2600 2700 2000 2900 3000 3100 TIME (NIN) Fig. 5.23. Nortaalized iodine activity in the reactor building: dissolved in water, on airborne aerosols, and total gaseous, i
f.r : . .4
,'m - !
m
$ _ t r
~ * .
/ i '_ ' + 2p : ' ' ' . , y;_ -- ; {,s <.,.Y , _
~
_:s: >L i _ ~
?
' ' t
...:Q ,
"' k- _
_ ,. p;
' REACTOR.
ORNL-DWG 84--5750 ' ETD M VESSEL.. . J ' o '
' Co#E SLUMP , FAILURE'
! TOTAL GASEOUS 1001ME- '
-(2492) -(2590)
.T - ,
-a
... o 3
, o
,. O . .
-l -
i/
.-.ess 'E.- 1
)
u1 r : ~
. c3 :
-> l ..................................................
z .: - GASEOUS ORGANIC 10DIDF o i u g. 5 '
; ~
e_ . Jo 9,a*
' C -i r -= l m
cn . ro -1 :
';- ,..,~~~~j COMPLETE C l i [9 HEPA FILTES E rAILuaE
, , . . . ,! (2917) :-
o d' *
- l
- e. P r: o p..
~
m N00 . N00 2t'00 2dx} 2600 2/00' . N00 2900 d 3 00 TIME (MIN) Fig. 5.24. . Normalized . gaseous iodine activity ..(total and portion as organic lodide) in the reactor building. e
- g. 4 ,' ,6 t *
' '~ e .:j
! - e -
q( , ,
., g; .: g.;
.' o '- -l e ,
~
t ', ,
' e .I ,f "
. !.f e':-
Y ORNL-DWG 8415751 ETD ' - REACTOR' -COMPLETE VESSEL' '
-HEPA FILTER:
. CORE SLUMP FAILURE. F AILURE -
Y -(2492). (2590) ( 2917) O_ ' ;
. := .
o o
+'*
'o_ =
TOTAL GASEOUS IODINE - , O-
> . =
- m. :
- o. -
sn .. .-
. j ..~...., *
*- GASEOUS ORGANIC IODIDE:
o_
,z .=
~....,~~..,,.'
.og .
.~..~.....,'"- ~
~
tw ..... V.- w o? _.s ac i': ~ N .
@'o_- l ' ,, AIRBORNE ON AEROSOLS
- .g ,.
.-,,,. . ~~ .~.~ ,,
,~~ .~~. -
v .
.-5 o_ ~ e
.- +
.a
~.~. .,
'o . . .I . !. . . . . .
2100 2200 2300 2400 2500 2600 2700 2600 2900 3000 3100 TIME (NIN) Fig. 5.25. Normalizedsiodine activity in the ' refueling bay. +
y _ r A o ,
. ORNL-DWG 84-5752 ETD
.O '
~l
- REACTOR'
. : VESSEL TOTAL IODINE IN SGTS CHARCOAL C CORE . SLUMP FAILURE
~i ~
-(2492) -(2590)
Y -
,j <r
- j. . . . . _ . ,
IODINE IN MAIN CONDENSERS
. /,........
QO i /* E 5 / O -: :
'Y-t 31 ,
TOTAL' IODINE IN SGTS HEPA FILTERS Ei ) ,.-- -
'T-- / ,,,__,....------
O O ,_,i ; ~ , . . . - h! 710: /
/
ORGANIC IODIDE IN
/. ,,*',, **~ SCTS CHARCOAL y
#$ : h ,* , '
- E : i l <
OS . . ,# Z'o : ,
~1 l l
- / e
~'
?5 O
'e l
COMPLETE
~i . l HEPA FILTER 5
l FAILURE i (2917) r - O *
~1: o y: ,
o i i i i i . . , 1 2100 2200 2300 2900 2500 2600 2700 800 2900 3000 3I00 TIME (MIN) f Fig. 5.26. Normalized iodine activity in components of the SGTS trains and in the' main condenser system in the turbine building. I 9 I e . . . . .
. -Q N
D T E 3 iM 5 . E
- 7 . D .
5 E N S. L O. I D 4 I , O . 8 D S O ,,, I O R G I E , C W D S U A N I . N . A im -
- O .
., G E O R R t L
N S A E O. EE a N / TTE) H G R . ELR7 e O O LIU1 PFL9 lf d B -
, R ** M I 2 i I
A , :!
. OAA(
CPF E iE2 s t
!. . H u
~- o
- e
. 0 h
t
- ,00
*.- 2 o t
. d e
. s a
e
, ,M l 2 e
- r
,i ) y
.-.
- N t
. I i
- R OLR)
- E a, ..
M v J TEU0 CSL9 a < , , , l':i i,::
.I.. ,6 0I 0
i t ASI5 r
,,!' c EEA2 i
!' 2E
- RVF( -
M a
- I T e
- P n M i U) d L2 0 o
S9 0 4 i i ,5 i E2 2 R(
- O d
, C' e z
i l
,o o a t m 2 r o
N 7 0 2
,30 2 5 ge i r F e 0 hp
,20 s 2 o m
0 0 1 l ] - y i:- ~ iE : . i=: . = 1;: lE . 1:: 2 O *- O.- 3 y O. e.eaEN.yryCZo d- -c* o ?. - JS
~ c'o * = o ~ yO *~
S.EO
> d ZI.>Z.
e
i REACTOR ORNL-DWG 84-5754 ETD Y VESSEL o? CORE SLUMP FAILURE (2492) (2590)
~
'o_
~: .
. OUTER RING OF FUEL REMAINS IN PLACE e
o-o? H e e z -
- i i i.,
u - l z i-
\. -
7 o_ _ l \ a i i U"5 ' INVENTORY IN FUEL DEBRIS N - i n
~
l '. O d- l \ m b~ o o_ l i
'N z ~E - l \
i ~ i g l \ l b_ l \
-E l \
i i l i
\.
i.
*o. . i .
l i i e 2100 2200 2300 2100 2500 2600 2700 i 2000 i i i 2900 3000 3100 TIME (MIN) Fig. 5.28. Normalized cesium activity in fuel material.
, g .. -
-s
. .,. w- ,_ , ,
- e. .e O
4 REACTot ORNL- PWG 84-5755 ETD
' VESSEL' CORE SLUMP FAILURE
'o (2492) (2590)
#i 1r 1,
, _ f'
'o, _
- e
. ::...h., < -
. ="o_ : ,
o_. i z : i iu % ll ;. u !;.;;. :::: e .:
' :: ....s., e .
z 7 a_ o
- : la "
c: . u -3 . ., N _ J e. l.. .
.e. .,* . .l ,
C I: '.. . Er . ;; '.' f. S: l !;
- oo_ .
, z ~5 ,:<::
- v ::
4 _
$.,,. . L, :
1 TOTAL CASEOUS INVENT 0nY l "5 ;.
}'
1
'o, *yi. I., ---------- Atason=E on AEnosols
.' .' 'g 4 -
l l. .... 4 e : :. i o w-e . .
. . . . . I . .
J 2300 2200 2300 2400 2500 2600 2700 2000 2900 3000 3!00 TIME (NIN) 1 Fig. 5.29. Normalized cesitun activity suspended in the reactor i vessel. I 1 C _ _.
~
- DEACTOR WESSEL COSE SLUMP FAILURE (2492) (2590) 7 o
-l: o- "
AE9050LS PLATED ON SURFACES
? ,,,,,,,............................................................
~,! l/- CONDENSED ON SURFACES FROM THE CAS PHA5E y - '
o, r .
~=
- z <
r. T - E~. i
>+ o l :
$ -15 l .; WATER Do~
u -1
; s ?
7 - g' . oi O e'!. ! w g = O N 5 .!: ! C ro ;! : r -1: .i u: : .: : o,' -
= : -
, . .i o,, ll t
2
- *: i
!! i
?o: 2
-1 .
t.
?E !.
o.aus 5 5 5 5 5 B B g 5 5 210D 2200 2300 2900 2500 2600 2700 2000 2900 3000 3100 TITE (MIN) Fig. 5.30. Normalized cesium activity in the reactor vessel water and plated on interior reactor vessel surfaces. t 9 . . . 4 . 9
, ,. 'E e .e?,
.,: >~ <
ORNL-DWG 84-5757 ETO
- REACTOR VESSEL
coat SLUMP- FAILUaE ^ (2492)' (2590) T
~]: ~SRV o o a ACTUATION
'o,' - 4 f >
p _ o,'
? - y
>* 31 5i D-z z :
a)
*T - 5 o
u 3 Si - i w-N : -
.-e .
Jo E x
~!
cm . zo 1 - 7 . o
-1 e_E
'o?
C3
" 2100 -d d 2 2 30 2M10 2700 a
2000 2900 M' N' TIME (MIN)
~ Fig. 5.31. . Normalized cesium activity in the pressure suppression pool.
4 -
.I REACTOR L-N M-SM Em o CORE StullP E-
"l (2492) (2590),
"5 "E "
i .
'g, o 5 : COMPLETE
- : HEPA FILTER 7
31-N$- 3'
;h3] '
cisEous
$3l
$: ,f
_\s,
$64 #Y& Q C dSi '.. "s k[ '
@o- - ',
~f ',, AIBBORNE AEROSOL 3' Id hf ((
- 5 'l, - ! l' ' l1.*.
o, l l l l'- l l\;;
~s D l l;;;
?c E . ll
.. . i.:.l.l; ;. l f' .I!.!.
2100 2200 2300 2100 2500 2600 2700 2000 2900 3000 .3100 TIME (MIN)
-Fig. 5.32. Normalized cesium activity in the drywell.
t e # . 9
4_Q, - - .jg w - 7gNyi7 -'----e --uw -- - [
. V .,'
., :/
b. ORNL-DWG 84-5759 ETD T REACTOR O] m
- CORE SLUMP ~ FAILURE
' VESSEL
,y (2492) (2590) DISSOLVED IN WATER-O1 5
4, ir -
'T #
o ,a '*- s.,,,,,,... .ee *'" ' ' ' * * " " * . - * ,.i
~E <--' , AIR 80RNE ON AEROSOLS
>O i
-0" :
.e a
~ O : *
~
W1 o .
. z : .-,
7 -
-W O1 o i
i
- N : o
*-G l J'o_ .
COMPLETE
~
N i k 2'
*] HEPA FILTER R-FAILURE
[g- (2917) e
-i : .
y .
~
o . .-e l
."c TOTAL GASEOUS '
3-4
-l i m !!: 2 *I !
i s ,,
- e }: nr:: {!. . .:!! ! .': :}-
!.::II}7 E
- : t' :
i f: o
\;-
is .: Ef .:i . ;i!d $i .. !' Ili t N
~
2100- 2M 2M . 2 5'(10 28'00 2700 2000. 2900 3000 3100 TITE (MIN) t Fig. -5.33. Normalized cesium activity in the reactor building: j dissolved in water .and on airborne aerosols. i i i E i - E r , - , .- y---, -.c ~ r ~ v _
?
ORNL-DWG 84-5760 ETO REACTog COMPLETE vg33tL HEPA FILTER CORE Stunt FAILUBE FAILURE (2492) (2590) -(2917) g . C O
-l: y 1. 9 y . ,
O
- al :
/
/
t - ' O al ./ T #
>. 0, l
' E 5 '
O ; W :- z wo, .
> = e z : l MAIN CONDENSER SYSTEM 1- .
~
oO ,, ..........SGTS HEPA FILTERS
- u E /
Q
.f g
,2 l ,"-~- ~ CESIUM COLLECTED ON l SGTS CHARCOAL c
n -1 : , e ,. oc :. e Z'O !..
-l :
?:O ,, .
s O l. ..,..
~
c:
.e O f-2100 2200- 2300 2100 2500 2600 2700 2000 2900 3000 3100 TIME (MIN)
Fig. 5.34. Normalized cesium activity in . components of the SGTS train and in the main coadenser system in the turbine building. Q . Q O
.j sN*
'e D 0 50 T 1 E - 3 1 - 6 - 7 S . 5 L .
- O 4 S .
8 O . R . 0 G E 0 W A M 50 3 - D N U. t
- O I L E R S E ._ a N N EE C _
R R O TTE) ELR7 S -
- e O B LIU1 U - d R PFL9 o O - 0 i I
A M OAA( I2 E S - B9 0 s CPF A - 2 t E C - u
.H o
._ e
_ h
, t
_ 0 0
. B0 o 2 t
- d e
s
. a
_ 0 e
, 0 l I7 e 2 r
) y
- N t I i R E
, M v OLR) ' 0( i TEU0 50 t CSL9 6 ASI5 2E c e EEA2 RVF( - M a I
T m P u M i U) s L2 S9 50 0 e 4 5 c E2 2 R(
*. O d C e z
i l 0 a 509 m r 2 o N e 5 0 3 0 53 5 2 ge ir F e h 0 p 0 s 52 2 o m 0 0 I 1_ ~ ' 1 ]: . . i5 : ~ i5 ; . 1:: 1:I 1- 2 1.: : . y: _ - o- o.2- ,'o -
+
cO o e'o 3'o yow
- d
>&o Zd>Z4 QEN.2.c[c'Z t c r
1
u-d'-. se- a/. i c% , 126 ( 6.~ .
SUMMARY
AND CONCLUSIONS h. f-6.1- Summary of the Work Performed _1. : S'equence . parameters (temperatures, - pressures, flow rates, etc.) were recalculated, again using the ORNL version of . the MARCH code, but
-employing 'a smaller drywell. failure size based on the assumption that
; overpressure ; failureo of 7 the drywell would create s. 2 ft2 (0.186 m2) failure . sone instead _ of -10 ft2 (o,929 m2 ) as had been assumed in the L '
companion study. -As explained in Sect. 2.1, this is believed to have an insignificant'effect on the fission product transport'results. I' 2. ' . Conditions in the reactor building were estimated for the pro- '
'jectedfdrywellileakage rates using_ the ORNL-developed Browns Ferry secondary containment-..model. Flows, temperatures and - pressures ~1n the
~
l, reactor, building are given in Chapt. 2 and Appendix'A.
'3. Some -' fission product transport 'model assumptions were altered
' based on.new information developed from NRC-sponsored research programs:
.(a) ' A L revised. estimate of iodine volatility over water was adopted
- , i based .on recent measurements conducted in a . radiation field. The new volatility 'is ' ten times ~ higher (partition ~ coef ficient one-tenth) than originally - assumed. ; (b) The manner in ' which drywell rubble tempera-tures , and ? fission product evolution - from ' the drywell rubble are esti-mated'was changed.-.The' newer method employs CORCON/MDD l'for estimating rubble temperatures = and VANESA for _ estimating evolution from the rubble due to sparging. (c) Similarly, CORCON/HDD and VANESA are used for .
l ' estimating the rate and amount : of aerosol produced from the core / con-crete. interaction on . the 'drywell -floor replacing an older procedure based on- the INTER subroutine of. MARCH and an unpublished correlation.
; (d). A . modified - form of the organic fodide production rate model has
- been developed. *
-4. Some newly-recognized features of iodine and cesium chemistry under reactor vessel conditions have been- noted which in the future may r
impact transport . assumptions' for these materials in the reactor vessel.
'5. . The transport characteristics in the drywell and reactor build-ing of aerosols produced by the corium-concrete reaction on the drywell floor were estimated using the QUICK code. - These results are presented in. Chap.:4 and Appendix B.
6.-;The transport of krypton, xenon, . iodine. and cesium throughout
- the reactor system and to the atmosphere was estimated based on (1) a
- fission ' product transport model for these elements, (2) postulated SGTS
' equipment and reactor vessel failure modes, and (3) flow and temperature conditions predicted in the reactor vessel,' wetwell, drywell, and reac-
, tor building. These results are presented in Chap. 5.
6.2 Suammary of Results and Conclusions
- 1. Krypton transport - characteristics are presented in Table 5.6 and Figs. : 5.6 through 5.11. As expected' from previous studies, there is i
u- -:.
= - 127 3 little barrier for krypton movement through the reactor system. Steam evolution from the reactor vessel and suppression pool and concrete deg-radation gases evolved in the drywell effectively flush krypton from the primary and secondary containments. However, significant decay of
-krypton radioactivity occurs due to the long delay time in this se-
- quence. Of the 4939 P3q present at scram, only ~20 PBq of krypton are projected to enter the atmosphere, principally due to radioactive decay. This amount of krypton activity at time 3100 min represents 82%
of the total krypton activity at that time.
- 2. Xenon transport results, presented in Table 5.7 and Figs. 5.12 through 5.17, parallel those for krypton except for a smaller degree of radioactive decay. In addition, an appreciable degree of Xe radioactiv-ity builds in from decay of iodine precursors. About 5100 PBq of Xe are projected to be released to air by time 3100 min, which represents ~88%
of the Xe activity at that time. The remainder resides principally in the fuel elements remaining in place, the reactor building air space and dissolved in water leaked into the main condenser through the shut MSIVs.
- 3. The computed distributions of iodine at various locations are presented in Table 3.8 in terms of total radioactivity (PBq) and in Figs. 5.18 through 5.27 in terms of relative activity normalized to the initial level (i.e., at reactor scram). We note that a far smaller de-gree of iodine is predicted to enter the atmosphere than noble gas be-cause of the chemical reactivity of iodine. As shown in Table 5.8, about 0.05 PBq of iodine are predicted to be in the atmosphere at time 3100 min, which represents 7.3 x 10~4% of the iodine activity existing at that time and 1.8 x 10~4% of the amount existing at reactor scram.
- 4. The principal repositories of iodine at time 3100 min are pre-dicted to be plateout on reactor vessel surfaces (88%), retention in un-failed fuel (5.7%), as dissolved material in the reactor building water
. pool (3.5%), and plateout on drywell surfaces (2.6%).
- 5. The principal iodine pathway to the atmosphere employing the current modeling assumptions is the following: (a) retention in the fuel during core degradation and passage with fuel rubble onto the dry-well floor, (b) sparging release fran the core debris on the drywell floor, (c) convective transport through the drywell and reactor build-ing, both as gaseous iodine and as chemisorbed iodine on aerosols, due to the convection and aerosols produced by the core / concrete reaction, and (d) passage through the SGTS filter / absorber system.
- 6. The pressure suppression pool does not capture a large share of the iodine evolved from fuel because the reactor vessel surfaces are predicted to capture most of the iodine via plateout (of various types) during the time ~ span when the suppression pool is in the direct path, i.e., times prior to reactor vessel failure at time 2590 min. Following time 2590 min, the suppression pool is bypassed.
- 7. According to current models, the principal form of iodine cap-tured by reactor vessel surfaces is Cs1 condensed on aerosol particles that subsequently deposit, accounting for about 60% of the total dep-osition.
- 8. Even though about 60% of the iodine deposited in the reactor vessel is due to plateout of aerosols, Fig. 5.5 shows that the assumed
r j 128
. degree of aerosol trapping in the reactor vessel does not greatly affect '
the ' calculated iodine release. This is because sparging release from
. core rubble on the drywell . floor (at . time > 2590 min) overwhelms any release range created by varying assumptions on aerosol trapping in the
- reactor vessel.
!' 9.: Iodine transport behavior in the reactor vessel and pressure .
. suppression pool could be significantly altered by possible revisions in
-assumed chemical forms currently being considered (see Sect. 3.2.4). It is possible that reactions with boron compounds evolving from overheated control- rods (and other materials) could destabilize CsI. The net
. result would likely be a smaller degree of iodine deposition en aero-sols, a smaller ' degree of capture in the reactor vessel and a corre-l spondingly larger degree of capture in the suppression pool prior to l vessel failure.
- i. 10. Only a minute fraction of the' iodine raleased to air occurs as l organic iodide (0.001%) according to current net production . rate esti-mates. This is admittedly highly preliminary and is not in accord with estimates presented by German analysts for an approximately equivalent event sequence.
- j. 11. Estimated cesium inventories (in PBq) at various times and l several key locations are presented in Table 5.9. Figures 5.28 through l 5.35 illustrate the time variation of the cesium inventory, normalized l to the initial activity level, for a number of locations. As shown in
- Table'. 5.9, an estimated 0.04 PBq of cesium activity is projected to enter the atmosphere by the end of the sequence; this represents about 1.9 x 10'% of the initial core inventory. Virtually all of this re-lease is predicted to occur af ter time 2917 min, the estimated time for HEPA filter failure in the SGTS.
- 12. The principal cesium pathway differs somewhat from that of.
iodine due to its much lower volatility in the drywell and reactor building., Cesium transport in the1e areas occurs almost exclusively as , f material condensed on aerosols whereas iodine transport is projected to
- j. occur both as a gas and associated with aerosols. Hence, the critical L- . factor in cesium release to air is the time for HEPA filter failure; virtually all of the cesium release is predicted to_ occur following this event. Summarising the principal cesium pathways (a) retention in the L fuel'and passage onto the drywell floor with the core rubble, (b) spara-i ina release from the core debris on the drywell floor, (c) transport through the drywell and reactor building associated with aerosol parti-cles generated by the core / concrete reaction, and (d) passage through the SGTS ' to the atmosphere following HEPA filter failure.
- 13. The principal cesium repositories at the end of the sequence
- are 'in order of importance
- (a) plated material on the reactor vessel upper surfaces, both as condensed Cs0H and cesium associated with depos-ited aerosols (2.7%); (b) in the portion of the original fuel elements that is ' predicted to remain intact (0.32%);' and (c) plated on drywell walls (0.20%). Very , little cesium appears in the ' suppression pool l (2.2 x '10-3%) ' due to the ef fective trapping on reactor vessel surfaces i of condensible asterials. About half this amount (1.2 x 10~3
% ) is pre- .
l dicted to reside in the reactor building water pool collecting in the basement at the end of the sequence due to aerosol settling in the building and washdown by the fire-protection system spray water. (The l. l-
m , 129 l y above percentages are activity levels referred to the core activity A ,
-level at screa.)
14 . -Expressing cesium releases as percentages normalized to the
' activity of ~ nuclides with half-lives greater than 30 min at scraa shows high degrees of = radioactive decay mainly because most of the original
,j . cesium activity (%.7%) is due to Cs138, .which has a 32.2 min half-life. -(Nuclides with half-lives less than - 30 min are assumed to - be completelysdecayed prior to cladding failure.)
~6.3 Principal Uncertainties-y hugh no formal sensitivity analysis was performed, the results indicate . fairly clearly the areas of greatest uncertainty, given below
.in approximate order of importance:
(1) h SGTS failure model, including the HEPA . filter loading causing failure, the type of HEPA filter failure and the functioning of. the charcoal beds under the projected environment at the bed inlet. h SGTS failure model assumes prime importance because the SGTS is the last
~. barrier to the atmosphere in this accident sequence. .
(2) Iodine volatility over water. N degree of iodine capture in the thoroughly-wetted reactor building depends - directly on the effective solubility Lof iodine-in water under those conditions. The reactor , building water' pool is the second-froa-last barrier to the atmosphere j for iodine..
'(3) Capture in the reactor vessel by aerosol' plateout and ' con-
_=*' densation. The current .modeling assumptions project the largest reposi- ! tory for cesium and iodine at the end of the event sequence to be Lplated - j- asterial on the reactor vessel walls,- principally . associated with plated - !- ' aerosols (60%) and as ' condensed Cs0H and CsI. All factors 'affecting [- e this calculated result bear heavily . on the computed distributions' of l ' lodine and cesium throughout the reactor systems and consequently on the [ , . : predicted release to air. The major factors ' that- af feet this computed . (- result are (a) the - presumed . chemical form of ' cesium and iodine . gas in l' the . reactor vessel on which the driving force for condensation on aero-I ' sols and walls directly depends; '(b) the ' surface ' temperatures, ' particu-larly late -in the sequence when revaporized deposits . can pass directly . into the drywell space following reactor vessel failure, and-(c) the na-p ture and rate of interaction between' fission product vapor and aerosols. (4) Rate of "sparging" release.- Because cesium and iodine evolved l ' from the core rubble - on the drywell floor bypass the suppression pool, j this transport step assumes a high significance. The rate of sparging ' release dependa on both the projected rubble ' temperatures,' currently
~ predicted by 'CORCON MOD 1, and the manner ~ of equilibration of fission -
product species with 'the sparge gas ' flow as predicted by VANESA. At this time it seems that currently-estimated rubble temperatures (of the oxide phase) calculated by CORCON MDD 1 are much ' too high', causing a -
-higher than realistic. release rate from the rubble.
p . * - (5) The supporting thermal-hydraulic calculations. A major-diffi-
- culty lies in the use of the MARCH code to develop the thermal-hydraulic p conditions in the primary system and the drywell ' during the accident l-Le' l- ,
i
- s . w.._ . .m. . . _ _ - . _ , - _,...,__.,_,_..__._,_,_.-_._,._,_,___.-..___..m ..m_. . . . . . , _ . - . .
130 sequence. For example, the core meltdown and slump into the reactor vessel bottom head as modeled in MARCH does not reflect the actual BWR
- under-core -structure of control rod guide tubes and is therefore not realistic for such application. Of equal concern is the inadequate
.modeling of the heatup of the BWR steam separators and driers, whose temperature .during the accident sequence is a major factor in fission . prodact transport. For these reasons, the degraded core and containment response analysis must be considered to be no more than a reasonable approximation to the events and event timing that would occur after core uncovery in this accident sequence. Since the fission product transport analysis presented in this volume is derived from the accident sequence analysis, this also can be no more than a reasonable approximation. 6.4 Significance of the Study The elements' iodine and cesium have a complex chemistry, and their passage through the containment structures is correspondingly compli-cated.. They evolve from fuel as a gas, partially condense upon aerosol particles or reactor structures, only to be re-evolved if the prevailing
- temperatures and chemical environments so favor. They can react chem-
.ically to form bonds with other elements and then, if conditions permit, break those bonds to form compounds with quite different transport prop-erties. These effects must be considered in each control volume of the calculation. Thus the determination of the transport of cesium and iodine is not straightforward, and the results presented in this report , for the transport of these materials cannot be considered to be more than reasonable estimates. The authors of this report do not believe that the principal value of' this work lies in the calculated numbers for fission product trans-port to the containment boundaries. Rather, the report is intended to a identify the weaknesses in the state-of-the-art methodology and to pro-vide user feedback to the code developers and to- show where future ex-perimental work is needed. At ORNL, the Chemical Technology Division SASA team members who participate in these fission product transport calculations are all part time on this effort; the remainder of their time is spent on applied research related to advancement of the under-standing of fission product transport. Thus their association with the SASA program calculations provides an avenue for transmitting the most recently developed research information to the methodologf- for esti-mating LWR accident consequences. D
y 131' g Appendix A REACTOR BUILDING ' AND REFUELING BAY CALCULATIONS
- lThe purpose of this appendix is to describe the results of calcula-tions ' performed with the ORNL-developed Browns Ferry secondary contain-ment .model for the Loss of Decay: Heat Removal accident sequence. The computer code employed for these calculations is essentially the same as that described in Appendix A of Ref. A.I. Modifications were made to improve the .model for condensation and evaporation and to provide for calculation of the . temperature of the pool that forms in the reactor building basement as a result of the actuation of the fire protection sys?:em sprays.
A.1 Introduction Calculations have been performed for the Loss of Decay Heat Removal accident sequence, both for the case with reactor building sprays, and for the case without sprays. In an actual accident sequence, the ele-vated temperatures in the reactor building af ter drywell failure would 7 .cause actuation of the sprays and this case is described in Sect. A.2. Combustible gases are released during the accident, both f rom the - metal-water reactions in-vessel and ex-vessel, and by the corium-con-e -- crete interaction on ~ the drywell . floor. Combustion is not of concern-within the primary containment'because.it has an inerted atmosphere. On the other hand, the concentration of oxygen-in'the secondary containment
~ during the : course; of the ' accident varies as a function of the relative amounts of infiltration and exfiltration. The resulting combustible concentrations in the reactor building and refueling floor are discussed
'in Sect. A.3.
The accident mitigating effect of the reactor building fire protec-tion system sprays can best be demonstrated by. contrast with the case without sprays; the latter is described.in Sect. A.4.
.A.2- Results for the Loss of Decay Heat Removal Accident Sequence The steam and gas flows included in the secondary containment model during the period af ter the refueling bay blowout panels have lif ted are shown in Fig. A.I. Flow through the refueling bay blowout panels is calculated af ter they have lifted. The reactor building basement pool shown in Fig. A.1 is formed by the accumulation of the water from the fire protecting system sprays. The design and operation of the reactor
. . . building. fire protection system is described in Appendix B of Ref.
> A.I. There' are no sprays on the refueling floor.
l The inleakage to the reactor building f rom . the 2 ft2 (0.186 m )2 hole in the drywell wall is provided as input to the reactor building l model. . The~ rate of mass inleakage of steam is taken directly from the l .- 1 p N s
132 MARCH code output and is shown in Fig. A.2. Most of the steam entering the reactor building is evolved from flashing of the water in the pres-sure suppression pool during the depressurization of the primary con-tainment. Additional steam sources are created in the drywell when a large quantity of water is vaporized in the reactor vessel lower plenum upon " core slump", and later upon blowdown of the reactor vessel into , the drywell when the vessel bottom head fails. After this, the corium-concrete reaction begins, and concrete decomposition provides a steady source of steam to the drywell which is contiauously released to the reactor building. The rate of mass leakage of hot dry gas from the drywell into the reactor building is shown in Fig. A.3. The large release at the time of drywell failure is nitrogen plus the small amount of oxygen in the inerted primary containment; these gases are flushed f rom the drywell by the large steam flow during the initial phase of the blowdown. There is a continuous release of hydrogen from the drywell after the metal-water reactions begin within the reactor vessel at a, bout time 2250 min. This leakage is too small to be discernible, but the sudden flushing of all hydrogen from the reactor vessel at the time of " core slump" results in a pulse of gas flow into the reactor building at about time 2492 min. The corium falls onto the dry concrete floor of the drywell upon failure of the reactor vessel bottom head at time 2590 min, and the corium-concrete reaction begins shortly thereaf ter. Steam released from the concrete permits oxidation of the previously unreacted zirconium in the corium, releasing hydrogen. Carbon monoxide and dioxide are re-leased by the direct concrete heating. The result is a continuous re-lease of hot dry gases from the drywell into the reactor building during - the later phase of the accident sequence as shown on Fig. A.3.* The total inleakage to the reactor building is shown in Fig. A.4 and the mixed-mean MARCH-computed temperature at the (drywell) source is shown in Fig. A.S. The temperature increases significantly af ter reac- , tor vessel bottom head failure due to oxidation of the previously un-reacted zirconium in the corium mass on the drywell floor. The reactor building and refueling bay response during the period from 2000 to 3100 min af ter inception of the Loss of Decay Heat Removal accident sequence has been calculated by the secondary containment model developed at ORE for use with the inleakage flows and temperatures de-rived from the output of the MARCH code. Core uncovery, cladding fall-ure, the onset of fuel melting, " core slump," reactor vessel bottom head failure, and the subsequent continuous corium-concrete reaction all occur during this period. The ORE model, although still crude, permits a much more refined analysis than would be possible if the reactor building and refueling bay were simulated by use of the existing con-tainment models in the MARCH code.
*It should be noted that CORCON MOD 1 predicted gas generation ,
rates for H2, CO, and CO2 were used in lieu of MARCH subroutine INTER results for this analysis. The rates predicted by CORCON are believed to be more accurate and are much lower.
4 133 The calculated response of the reactor building is shown in Figs. A.6 . through ' A.12. 4 As shown on Fig. A.6, the pressure variations are small. The standby gas treatment system (SBCTS) maintains a slight negative pressure irs the building during most of the period af ter dry-
- well failure, but positive pressure. spikes occur at the time of " core slump" and at-failure of'the reactor vessel bottom head.
. The reactor . building ambient temperature is shown in Fig. A.7. The
. temperature increases rapidly at the time of drywell failure because of
- the entry of steam and hot gases from the drywell at a temperature of Labout 425'F - (490 ' K) . The reactor building sprinkler system is auto-matica11y actuated when the building temperature reaches 225*F (380 K);
this causes a sudden decrease in building temperature just af ter drywell f ailure ' as shown in Fig. A.7 . The reactor building tempt.rature con-tinues to. decrease as ' the leakage flow from the drywell falls off (see Fig. A.4), and the sprinklers continue to operate. Increases in build-ing temperature occur in conjunction with the temporary increases in inleakage from the drywell occasioned by " core slump" and reactor vessel
. bottom head failure.
-The reactor building fire protection system directly cools the building atmosphere and simultaneously increases its heat capacity by raising the water - vapor content to saturation and by introducing sus-pended water droplets into the atmosphere. The sprays also promote turbulent mixing so - that all surfaces in the vicinity would be com-pletely wetted. The . water falling onto the intermediate floors of the building would flow through drains and down walls and stairways to col-1ect in a pool in the reactor building basement.
- . The asss flow to the SBGTS is shown in Fig. A.8. As modeled, the volumetric flow from the reactor building to the SBGTS is constart at 12,500 f 3t / min (5.90 3m /s)* so the variations shown in Fig. A.8 are caused by density changes in the reactor building atmosphere. The tem-perature at the-inlet to the S8GTS filter trains is shown in Fig. 2.21.
, ' The volumetric inf11tration flow to the reactor building is shown t
' in Fig. ' A.9. This is_ the flow of air through leakage pathways in the
. building walls and through the building vacuum breakers. Infiltration
)' . occurs - during most .of the period af ter drywell failure because of the negative pressure maintained in the reactor building by the SBCTS. As
. shown in Fig. A.6, the reactor building pressure becomes greater than atmospheric for brief periods just af ter core slump and just af ter fail-ure 'of the reactor.. vessel bottom head; infiltration ceases during these periods.
Whenever the pressure in the reactor building exceeds atmo-spheric, there .is an exfiltration flow through the leakage pathways in the building walls directly to the outside environment. The volumetric rate of exfiltration is shown in Fig. A.10. No exfiltration occurs , af ter time .2600 min because the reactor building pressure is subatmo-spheric throughout this period. The relative humidity of the reactor building atmosphere is shown in Fig.~A.11. The relative humidity decreases rapidly immediately after
*An equal amount flows to the SBGTS from the refueling floor.
l i l
134 drywell failure as the reactor building atmosphere is heated. Shortly thereaf ter, actuation of the reactor building sprays lowers the ambient
- temperature and increases the relative humidity to 100%. Since the sprays continue to function throughout the remainder of the accident sequence, the building relative humidity remains at 100%. (It should be noted that the calculational procedure permits relative humidities .
slightly in excess of 100% as seen in Fig. A.11. This slight departure from realism 'is accepted to avoid the use of extremely small timesteps in the calculations.) As, discussed in Sect. 2.6, the blowout panels between the Unit I reactor building and refueling bay are predicted to relieve almost immediately af ter drywell failure. This provides a large flow area be-tween these two control volumes; the calculated inter-volume flows are shown in Fig. A.12. Except for the period immediately af ter drywell failure, and for brief periods immediately af ter core slump, and af ter failure of the reactor vessel bottom head, the flow is from the refuel-ing bay to the drywell. Since fission product release from the fuel does not begin until well af ter drywell failure (about time 2330 min), few fission products will be transported to the refueling bay during this accident sequence. The calculated response of the refueling bay to the flows to or from the reactor building through the blowout panels is shown on Figs. A.13 through A.18. As indicated on Fig. A.13, the refueling bay pres-aure is very close to atmospheric during the accident sequence. Except for brief periods at drywell failure, core slump, and reactor vessel lower head failure, the refueling bay pressure remains slightly below atmospheric.
- The refueling bay atmosphere temperature is shown in Fig. A.14.
Since the refueling bay responds to the inflow from the reactor build-ing, it is reasonable that the refueling bay temperature trends follow those of the reactor building shown in Fig. A.7. However, the refueling , bay free volume is about 1.76 times that of the Unit I reactor building and its temperature remains significantly lower throughout the accident sequence. The mass flow from the refueling bay to the SBGTS is shown in Fig. A.15. As in the case of the reactor building, the volumetric flow from the refueling bay to the SBGTS is constant at 12,500 f t / 2min (5.90 m3 /s). The infiltration of air to the refueling bay is shown in Fig. A.16. Infiltration occurs whenever the pressure in the refueling bay is below atmospheric so that outside air enters through leakage pathways in the superstructure, through the refueling bay vacuum breakers, or through the blowout panels (if previously opened). Exfiltration, which occurs during the brief periods when the refueling bay pressure is above atmo-spheric, is shown in Fig. A.17. A large amount of water vapor is carried into the relatively cool refueling bay with the initial flow f rom the reactor building at the time of drywell failure. Consequently, the relative humidity in the refueling bay is predicted to be 100% for about four hours following . drywell failure as shown in Fig. A.18. Subsequently, the relative hu-midity is reduced by the continued infiltration of cool, dry air from the outside.
L V " =- ' - [_ [
=^
E E 135 , L A.3 Combustible fias Concentrations in the l-f - Secondary Containment Atmosphere g w E - Combustible gases would be present in the drywell atmosphere under b
- h. , _ the conditions of a Severe Accident. These include hydrogen gas formed _
i', by metal-water reactions botn inside the reactor vessel and subsequently EE E within the corium mass on the drywell floor and carbon m'ohoxide, formed F by the corium-concrete reaction. The combustible gases within the dry-f , well would be released into the reactor building af ter drywell failure, I and it is necessary to consider whether or not the conditions required E y for deflagration would be reached within the secondary containment at - m any time during the accident sequence.* F f :/ Th'e mole fractions of steam, carbon dioxide, carbon monoxide, hy- k { drogen, nitrogen, and oxygen in the reactor building and refueling bay = are computed throughout the accident sequence by the secondary contain-
- - p ment model. The model assumes perfect mixing of the gases which is a -
f reasonable assumption given the turbulence created by the drywell blow- r down and the action of the fire protection system sprays. - Based on information supplied by F. E. Haskin of Sandia National - E
; Laboratories ,(Ref. A.2), deflagration would be expected to occur in the -
secondary containment atmosphere if the following conditions are met: [ g 1. The sum ~of the hydrogen mole fraction an(' 60% of the carbon monoxide -_ mole fraction exceeds 0.090.
- 2. The oxygen mole fraction is greater than 0.05.
s s 3. The sum of the mole fractions of the inerting gases steam and carbon - dioxide is not greater than 0.55. _
=
The results of the secondary containment model calculations indicate - E that these conditions are never met. The closest approach occurs in the reactor building during the 5 min period from 2502 to 2507 min when the - E combustible gas mole fraction (H2 + 0.60 CO) has reached 0.088, the j q , oxygen mole fraction is 0.16, and the inerting gas mole fraction is also - p 0.16. After this time, the combustible gas mole frt' tion falls off L 5 rapidly, being only 0.003 at the end of the calculation. Even if de- _ s flagration were to occur in the reactor building, its effects would be
. minimized by action of the cont.1'nuously cperating sprays. -
, As discussed in Sect. A.2, flow from the reactor building into the l E refueling bay occurs only during a few very brief periods in the Loss of - E f. Decay Heat Removal accident sequence. The predicted combustible gas l _
-mole fractions in the refueling bay never exceed 0.004; accordingly, -
[ deflagration would not occur there. - w E - a A.4 Secondary Containment Response Without Sprays
- g ,
'
- w F As discussed in Sect. A.2, the automatic actuation of the reactor -
L > building ' fire protection sprinkler system plays a major role in deter- - [ mining the reactor building response to the inleakage after drywell failure in the Loss of Decay Heat Remov&' accident sequence. ; [ - For this L i k - .\
.r t
5 i s
*It is known that this would not occur in the primary containment, g .
j i >which'is inerted. - g.- 5 g
l.
~ 136 study, it . has been assumed that the sprinkler system is actuated when the* average building temperature reaches 225*F (380 K), producing a con- -
.tinuous spray of 500 gpa (0.032 m /s) of water at 100*F (311 K) into the 3
reactor building atmosphere. This occurs when the reactor building tem-perature increases immediately after drywell failure. The reactor building response without the sprinkler system sprays , will be briefly. discussed in this section. The reactor building temp-erature is shown in Fig. A.19. As can ,be seen by comparison with Fig. A.7, the temperatures within the reactor building are significantly higher for the case without sprays. With the sprays, the temperature decreases *mmediately and remains well below 200*F (366 K) for the rest of the accident sequence. Without the sprays, the temperature reaches a
. maximum of 235'F (386 K), and then remains in the vicinity of 200*F (366 K) for about five hours.
The reactor building pressure response without sprays is shown in Fig. A.20. Comparison with Fig. A.6 shows that the reactor building pressure is higher for the case without sprays, as would be expected. The higher reactor building pressure results in more exfiltration through the building walls and more flow into the refueling bay through the blowout panels. Thus, more fission products would be expected in the refueling bay during the latter part of the accident sequence for the case without sprays. The combustible gas concentrations in the secondary containment were calculated for the case without sprays to determine the potential for deflagration. Deflagration is not predicted to occur in either the reactor building, or the refueling bay.- The closest approach occurs during a two-minute period (time 2500 to 2502 min) in the reactor build- . ing when the combustible gas mole fraction (H2 + 0.60 CO) is 0.086, the oxygen mole fraction is 0.10, and the inert gas (H2O + CO2) mole frac-tion is 0.41. The combustible gas mole fraction subsequently decreases rapidly. The highest combustible gas mole fraction reached in the re-
- fueling bay is 0.012, at-time 2515 min.
l[" W
?-
137. References for Appendix A o A.I. R. P. Michner, et al., SBLOCA Outside Containment at Browns Ferry
' Unit One, Volume 2. Iodine, Cesium, and Noble Gas Distnbution
. and Ralsass, NUREG/CR-2672, Vol. 2, ORNL/TM-8119/V2 (September
* ~1983).
A.2. Private communication, F. E. Haskin letter of September 14, 1982 to S. A. Hodge. O V
- O.
l u __ _ _
om-o o u-ono ma . (s00 FT) room twoou" O 5"
,"'"'r'$Es 4 -
REFUELING BAY Flow DetouGM LEAKACC m
" ,, flCACTom BJtLDi$
astiLinAfloN ' Iq" Blowout PAsELs cXrit nation 0" sNiisttas REACTOR T m on 9 9 , @fI,
- - . .=. .
BUILDING sc s sysitu R 1 }
/\
K "' Ic=1 ,, 58 7 ,'& b LEARACC ist LTIIAft(pe s 1 on e % .. .. tXFILinAftom __m W sam m ui Poot W e J J d C U [ sais Boeseis Fig. A.I. Flows included in the reactor building model af ter rup-ture of the refueling bay blowout panels.
w* 8 0 0 1 ( 0 2 0 1 6 1 3 5 9 4 B G 0 W 0 D 0 L 0 N 3 R O r 0 o
" .00 t c
9 a 2 e r
~
~
0 0 e
.0 h
~
0 t
~ 2
~
o t
' 0 n 0 i
,0
~
7 L - 2 l E ' l F SSAD e O E E 0) w E VH y R ,00 n r RM U OO 6 i d 2 L I TT ( m ACT e F AO h E B P 0E 0 M t R E
,0 R M O Ugi 5I m L
C S 2 T o r e f 0 0
,0 m 4 a
_ 2 e t s 0 0 f
,0 9
3 o 2 e g 0 0 a
,0 k 2 a 2 e L
0 L 0 .
,0 L S 1 2 E L 2 WI ' A
' YA 0 0
R F 0 .. D- 0 gg 2 i n
/n0 si 0
-0 0 0 0 0 0 0 0 0 0 Fi d
8 6 4 2 0 8 6 4 2 0 bm2 l 1 1 1 1 1 l i u b ga, - - - - - - - - - -
/0 8 0
6 9 4 2 7 0 6 8 G 4 2 0 0 g1 2 3 1 0 4 2 6 0 5 0 5 0 5 0 5 0 k5 1 3 1 21 0 1 9 7 6 4 3 1 g NOCN<N4d wQNQ g
JE ORNL-QWG84-8763 EYD lbs kg/s min / 37.80 - 5000.0 DRYWELL FAILS 34.02- 4500.0-3024- 4000.0-26.40- 3500.0-- m
< 22.68 - . 3000.0 -
x i
'M a 18.90 - 2500 0--
Z
~
s-
.. o 15.12 - 2000.0-Q c
1134- 1500.0-F AILURE OF RE ACTOR VESSEL 7.56- 1000 0.. BOTTOM HEAD CORE SLUMP 3.78 - 500.0-- 0.00- 00 . . . . . .- . . . . 2000.0 2100.0 2200.0 2300 0 24000 2500.0 2600 0 2700.0 2000 0 2900.0 3000 0 3100 0 TIME (min) Fig. A.3. Leakage of hot dry gas from the drywell into the reactor building.
. . . - - +
liI , j)l iiI ' Il!) i! !ll l! O D f 0 E 0 0 1 4 6 3
~
l
- b 4
8 G '
~
0 W 0 D 0 M
~'3 0 d
H O ~ l i
^
~
0 0 u - 0 b
'9 2
r
^
o t 0 c 0 0 a 0 e 2 r e
~
^
0 h
^
t 0 0 7 2 o L t F SSA0 E n L i O E E 0) E VH 0 0 fn l d U RM G l mwe 1 OO 2 L TT ~ ( I y ACT F AO E B 1 0E d r P R E R M 0M 0 5l e O C L Uf 2T h S t _ e _. m 0 0 o 0 4 r . f _ 2 e g 0 a e 0 k b a 3 2 e l L 0 l L S 0 a E L b t WI 2 o YA 2 T
~. R F c D 0 .
0 4 0 1 2 A 0 . 0 0 g 0 i 2 F
/n0 0 0 0 0 0 0 0 0 0 2
0 0 g si 0 0 6 4 2 0 8 6 4 bm2 l 1 1 1 1 1 n i gp - - - - 2 0 8 6 4 2 0
,/ 0 0 0
6 9 4 8 7 6 4 3 2 1 0 g2 1 6 0 5 0 5 0 5 0 5 1 O k5 1 3 2 0 9 7 6 4 3 1 1 1 e 4 N04*.Adz - n O + _QN$ r e l't l' l1
, osuu owc a4. siss ETo ..
810 9 - 1000.0 6998- 800 0-p FAILURE OF E-* REACTOR VESSEL h r.g 588.7 - 600.0-BOTTOM HE AD 0 J CORE l l d j
, 477 6 - 400.0-
[ ! h
< c 366.5 - 200.0 -
t 255.4-8 0.0 , , , , , , , , , , P0000 2100.0 2NOO 2300.0 24000 2500 0 2600 0 2700 0 2800 0 2900.0 3000 0 3100 0 TIME (min)
- l Fig. A.S. Mixed-mean temperature of drywell atmosphere.
t 9 3 # # D
1l\ l l 11 j1ll Il, !1 1llli 2w 0 m 0 t 0 1 s 3 e
=
i s
, 4 n
0 _ o m _ ,0 0 _ ( 0 t 3 w s o 0 0 q ,9 0 2 7 z __ - ; 0 _ 0 .
,0 _
- 0 -
x 2 - x; _ . 0 0
,0 e 7
2 r L _ u - F SE D s s O S E E A
- 0) e E VH 0 n r R ,0 p _
U RM i L OO TT p 6 2 mg - I ACT ( n F AO i E B p 0E d R E ,0 0 M l Ry r 5I i Ogt ' u C S 2T b
, r 0
0 o
,0 t 4 c 2 a e
R 0 0 . e ,0 3 6 2 A 0 . 0 g 0 i
,2 F 2
0 0 L ,0 1 L S 2 E L WI o 0 Y A 0 R F 0 D 0 02 - - - - 2 1 1 0 0 0 0 0
' 0 5 0 5 0 1
' 1 -
e g - - - - - a0 5 0 5 0 0 4 9
- g9 4 4
2 0 2 4 2 1 0 1 2 a NWJnbW D 0 ( a Mo
- m <NW.
a a i l )ll llf i
. +
;k oan owca4 sier eto 408.16- 275 0 s
39427- 250 0- DRYWELL F AILS C4 E-, 380.38 - 225 0- -
- SPR AYS INITI ATE D Z
W CL. M 36G.49 - 200.0-N E-O Z
~
352.61- T/50 FAILURE OF RE ACTOR VESSEL
- h. BOTTOM HEAD $-
330.72 - - 150.0 - S UMP Z , O Y E-Q 324.83- 125 0-- W ' i 31094 - 100.0 - 297.05 750- , , , , , , , , , , 2000.0 2100 0 2200.0 2300 0 2400 0 2500.0 2600.0 2700.0 2000 0 2900 0 3000.0 3100 0 TIME (min) Fig. A.7. Reactor building temperature.
& I & ., D R
1)1)\1 ll\ y* D 0 T E 0 0 1 8 6 7 3 i . 4 8 G 0 W - 0 0 D ,0 L 3 N s H O a g 0 0 - ,0 9 y b 2 d n a 0 t 0 s _ ,0 _ 0 2 e h t 0 o 0 t
,0 _
7 g - 2 n i d
- 0) l
_ ,00 n iu i 6 2 ( mb r 0E o _ 0 t
,0 M c _
5I a _ 2T e - r 0 e
,0 0 h _
t . 4 2 . m . o _ 0 r _ _ f 0 - . ,0 3 w - 2 o l f 0 0 s
,0 2
s 2 a M . m 0 e 0 .t
,0 8 s 1
2 y As
- 0 0 .t 0 gn 0 i e 2 F m t
/n0 si 0 .
0 0 0 0 0 0 0 a 0 e bm020 l 0 5 1 0 0 2 0 5 r t s - 4 s 8 0
/ 21 3 5 7 0 g5 1 7 3 0 k1 1
. n eE0 @ ot- 5o 0f,5dDcn mob <a.m r
!lI!ll!I ll ll l 1ll 11\\l\ l l'
O 0 f 0 3a 0 1 7t 3 5 4 8 G 0 . W o D b 0 N L 3 t F G o 0 0 9 2 0 0 . b g a n c i d l o i 0 u o b
'7 L 2 r E D F SS A o t
O E E o) c E VH R 0 0 n a U RM 'ci e OO 2 r L I T T m ACT ( e F AO h E B E P oE o t R R M O U bM5I o L t C S 2 T r i i 0 a . 0 0 f
'4 2 o n
o o b 0 0 3 i t a l 2 r . l
.l t
l i l o f
'l 0
0 n
'2 I
, 2
' t I
f 9 o i o A L , i - b l L E S i ; a . g i L ,l WI , _ 0 i Y A 0 F R F 0 D 0 2
/n0 i 0 0
0 0 0 0 0 0 0 0 0 o tm50 f 3 0 o 3 0 5 2 0 o 2 5 1 0 0 1 0 5 s - - - - ~ - - -
/2 5 6 1
o B 4 4 8 7 2 6 3 o 7' m6 1 4 1 1 1 9 4 2 o{QdDAmoHQE%
- @]x V6P<mDgd $4 .
l ;lll1 '1ll' we" O 0 T 0 E 01 0 1 3 1 5 4 8 0 G 0 c 0 ( 0 L 3 N m s 0 0 0 9 2 0 o . b g 0 2 n i d l 0 i 0 0 u 7 b L 2 r F SS A E D o t O E E 0) c E VH 0 0 n a H U RM
'G i e L OO 2
m r I T T ( ACT e F AO h E B 0E
) t P
R E R M d' XM m I O U 2T o L C S r O f 0 0 n 0 o 4 i 2 t a r 0 t 0 l e 0 i 3 f 2 x E 0 0 . 0 0 2 1 _ 1 2 1 . A _ 0 . 0 g L L h 0 1 i E S 2 F L Wl ' f 0 YA 0 R F 0 D 0 2
/ l 0 0 0 0 0 0 0 0 l
i 0 0 0 0 0 0 0 tl 0 0 0 0 0 0 0 0 5 0 5 0 5 f I I 53 3 2 2 1 1 j - - S - - - - 2 G 0 0 [5n6 2 6 1 4 1 0 8 1 1 4 4 9 0 7 7 4 3 2 1 I a
. OA Dm dObC W ^O4N gop<MCt 3 E C
e f l l lI I l J1lllIl lll
OHNL- WG80'III' 100.0 hth 60 *[ '..'w" a " i 90.0-b m A 80.0-2. D Z 70 0-W P-E. 60 0 - f W IU 50 0-O : SPR AYS INITI ATE D
*/,
Y-a" 400- # d ORYWELL D F AILS IQ 30.0-(4 O h p 20 0-W 10.0-00 ' , 2000.0 2tc' 00 23'00.0 2400 0 0
- W' 2Y00.0 2b.0 d0 M0 W TIME (min)
Fig. A,11. Relative humidity of the reactor building atmosphere.
mNb@ c5Ag47o n N4h 0 $kUd5o k @Z
- 2 2 4 7 9 1 4
1 6 1 0 2 1 2 1 n 3 3 7 0 4 1 1 1 5 8 2 3 3 G 0 s 2 0 4 0 6 2 0 4 0 /.
- ~ - - J - - - - -
- S f - 1 1 2 2 3 3 4 4 s 5 5 0 3 0 5 0 5 0 5 0 u 0 0 0 0 0 0 0 0 0 0 0 mf e 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 i
- t l
i 0
- 0 0
0 0 0 0 0 0 0
- 0 n/
~
nF 2 g i 0 D g 0 F R 0 f. l 0 ~ AY LI W oA
- o. 2 1,
S LE r1 0 L
. 0 2 0 V 2 2,
. E o 0 0
l u 0 _ m e t 2 r 3, _ i 0
- c 0 ::
0 _: f l i o 2 4, _ w 0 0 f 0 r . o S C m T2 t I 5 m oO t M 0,0 N uR e E R h E0 BE e OAF ( TCA r e m26, z TTI L OO U a i MR R c n0 0 H VE t o
)0 C E E A SF S O r C:
DE 2 L - b u i 7, 0 0 - l 0 d i n 2 T g 8, 0 0 t 0 - o _ T a t 2 h 9 e 0 0 r 0 e O H
- N 3 L 0, D 0
0 W . G 0 8 4 . 5 3 77 1 7 0 E 0 T 0 D T
- 4 1 1ll l iI!lljl1J 11 11)l .
go 0 0 7 0 E 0 1 5 8 7 3 6 4 . a G 0 W 0 D ,0
- 0 L 3 N
H O 0 0
,0 9
2 0 0
,0 8
2 0 0 .
,0 7
2 e L r F SSA E D u s O EE 0. ) s E VH 0 n e R RM ,0 r U OO
- 6 i p 2
L I TT ( m y ACT a F AO b EB P 0. E R E
,00 M g
R M OU
- 5I n L
C S 2T i l e . 0 u 0 f
,0 e 4
2 R 0 3 0 1
,0 .
3 2 A g 0 i 0 F
,0 2
2 L L S E L 0 WI 0 YA ,0 1 R F - 2 D r 0
~F 0 0
0 03 - 2 H0 0 0 0 0 5 0 0
" 1 5 0 1
e g - - - - a0 9 5 0 5 0 4 0 4 9 g4 2 0 2 4 2 1 0 1 2
=
h gAOMgC1 Mo3A eZ C( aEWg z
=
I illl1lll.l!lll(lil e
%NNDN3ZO aOo% HN^ON%<hD%N 2 3 3 3 3 3 3 3 4 9 1 2 3 5 6 8 9 7 0 4 0 6 0 0 9
2 4 8 0 8 7 E 4 3 2 1 5 4 3 2 1 9 8 7 6 7 1 1 1 1 2 2 2 2 0 2 5 7 0 2 5 7 5 0 5 0 5 0 5 0 5 . 0 0 0 0 0 0 0 0 0 2 0 D 0 F R 0 0 ~ AY I L W E 2 S 1 L 0, L 0 0 2 F 2, i 0 g 0
. 0
.A 2 1 3, 4 0
_ 0 s 0 R e 2 4, _ f 0 _ u 0 e 0 - _ l i . n S C g T25, L I , UO b a M00 M P R E R E0 y ( N BE OAF TCA t m26 TTI m p e e i n
) 0 0,
0 e - OO LU MR R HVE E E O r AS a D E S F t u 7, 2 L r 0 0
.e 0 2
0 0, 0 0 2 9, 0 0 0 o n 3 s 0, - 0 o 0 w 0 c a 4 s 3 1 0 m 0 e T 0 o 6 u~
~ .
I)1l l1Il1IlllI l jIf j I1\
1 ljjj I Gn 4
. 0 0
0 1
. 3
. 0 0
- .0 0
3 s a
- 0 g
0 y
.0 b 9
2 d n a t 0 s 0
.0 0
2 e h t 0 o 0 t
.0 7
2 y a b
- 0. ) g
,0 0 in n 6 i 2
( ml e u 0E f
,0 0 M e 5I r
2T e h . t 0 0
,0 m 4
2 o r f 0 w . 3 o
,0 3
2 l f s 0 s 0 a
,0 2
M 2 . m 0 5 .et 1
.0 0 .
s y 1 2 A s 0 .t 0 gn 0 ie 0 2 F m
- - - t
/nc 0 0 0 0 a si 0 0 0 0 0 e r
bm020 l 0 5 1 0 0 1 0 5 t s - - e 8 0
/ 2 1 4
3 s. 7 0 g5 1 7 a 0 k1 1 1omW 2 o" wSw goa a3dgmx .
\llllli !llil!ll l 1l1I!1Ill lIlll!l1illlll1l1<
' ll
'l l!ll l ll e
g<~c : *</*g g
- CpO, h g a p,0 (o %
- 1 1 2 2 2 2 3 3 3 g
3 4 7 7 0 9 4 u. 8 1 4 1 6 5 8 8 1 2 3 6 5 9 8 3 0 6 3 0 5 4 m 0 3 2 8 4 6 2 8 4 0 6 2 8 1 0/ s 2 2 3 3 4 4 5 5 6 6 7 7 mg 1 1 0 5 0 0 5 0 5 0 5 0 5 0 5 0 0 0 0 0 2 ,- F D R Y 0 0 0 0 0 0 - 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 i3 0 i l' j 0 I W - 0 L S E 0 L . 0 L - F 2 , i 1, C g 0 . 0 .A 1 2 6 2, 0 0 l 0 I n f i 2 3, l 0
- t r
a 0 0 - t i 2 o 4, 0 n 0 0 o 4L e f S a T2 I 5 U i r M 0,0 M R P B E t E0 OAF o ( TCAI TT t h m26, i OO LU MR R e n00 HVE
)0 E E r
e A S O S F f D E L u 2 7, e 0 l 0 i 0 n g _ y b a 2 8, 0 0 m
. 0 2
9, 0 0 0 m 3 0, W 0
- 0
- 0 *
- a 1
3 ' 0 0 U 0
- e "g
I Il , 1
1 9
# NMc[Z OZ kmh DN3'ZO , OOM 1 1 1 1 1 2 2 3 5 7 9 1 3 5 6 8 0 2 0
1 8 7 6 5 4 3 2 1 9 8 7 6 m 0 0 8 8 7 6 6 4 4 5 2 4 0 2 0 1 6 0 4 9 2 8 0 6 8 5 6 eO
* /*
s 0 4 8 1 2 1 6 2 0 2 4 2 8 3 2 3 6 4 0 4 4 4 8 mf i. t 0 0 0 0 0 0 0 0 0 0 0 0 0 n/ 2 0 0 0 0 D . 2 F R F 1, 0 AY i 0 I W g 0 L E
. S L L .
.A 2 7
1 2 0 0 Y 0 '^ E x 2 3 . 4 f 0, i 0 l 0 t r a 2 t 4, i 0 o 0 4 n 0 ' f S r C o T2 I 5, L UO R m M0 0 M E P BE R t E0 OAF h ( TCA e TTI r m20,6 i OO LU MR R e n0 HVE f )0 E E u O e ASS F , l D E i 2 L n g 7, 0 0 0 b a y .
. 2 8
0, 0 0 2 9, 0 0 0 O H N 3 L 0, D 0 0 W G 0 B 8 d 5 , 7 3 1 l 6' lI 0 E , 0 T 0 D 9
~
j]1 l iljlI ! {ll i l: 1
.vc
. ii 0 0 1
6 0
. . 7 0 1
7 7 3 5 4 8 G 0 W 0 D ,0
'L 0
N 3 R O 0 e 0 r
,0 9 e 2 h p
s o - 0 m _ 0 t
,0 8 a 2
y a b L F SE D O SA EVH R EE W ,0 0 0 7 2 0) 0 n i l f g n e u e U RM L I ACT OO TT F AO
'sq'A ,0 6
2 i ( m h r e EB P t R E 0. E
. R M ,00 M f OU 5I o c SL 2T y t
0 i 0 d
,0 i 4 m 2 u h
0 e 0 i v
,0 3 t 2 a l
e 0 R - 0
,0 2 .
2 8 1 - 0 A - 0
,0 .
1 2 g
.f mS L I
A 0 i F 0 mF 0 0 _ ~ - 2 -
,. 0 0 0 0 g 0 0 0 0 0 -
,i 0 0 0 0 g 0 0 0 9 0 g 0 0 0 1
7 '4 3 2 1
>bQ hpg mB g i c oO A o+g3ND w% -
C* 0
" 0 0
' 1
" 3 S
* ~ 0
" 0
,0
"" 0 3 0 0
,0 9
2 s y a
- r
- 0 p
,0 0 s 8 t
^ 2 u
o h 0 t 0 i
,0 7 w L 2 E e F S D s n
O SA EE 0) o EVH 0 n p R ,0 s U RM i 6 mree l L OO 2 I TT ACT ( FAO E B 0E r R E P 0 M u R M ,0 t OU 5I a L C S 2 T r e p . 0 m 0 e
,0 t 4
2 g n i 0 d l 0 i .
,0 3 u 2 b r
o x
% ,0 0
0 2 t c a 2 e R 0 . L 0 9 L ,0 1 E S 1 L 2 WI o l
!l i ;I A
Y A f 0 R F 0 g D 0 0 i 2 F F
- 0 -
0 0 0 0 0 0 0 0 5 0 5 0 5 0 5 0 5 7 5 2 0 7 5 2 0 7 2 2 2 2 1 1 1 1
- - - - - - - - J K6 7 8 9 1 6 2 3 4 5 1
8 2 3 4 7 8 9 0 2 0 4 0 6 8 4 01 7 9 8 6 5 3 2 9 4 3 3 3 3 3 3 3 2 i S9EMP 7ZH O{ 9=L= UH .
- s. X
e 's *. ,. e
- OhNL- DWG B4 6779 E TO kPa gage "H 2O 2.490 - 10 0 DRYWELL F AILS FAILURE OF RE ACTOR VESSEL BOTTOM HE AD 2
% 1245" 50- . CORE SLUMP
% -t 'k C I O
.E i i l' i
-)
e , . pl 0000- 00- & %%N
'&trku = - -
= - - = = -'
D " x 9 m U 4 Z -1245" -5.0 - X
-2.4 90 J -10 0 . . . . . . > i . .
2000 0 2100 0 2200 0 2300 0 24000 2500 0 2600 0 2700 0 2800 0 2900.0 3000.0 3100 0 TIME (min) Fig..A.20. Reactor building pressure response without sprays. l
-gi 159 Appendix B:
DETAILS OF CALCULATIONS FOR CORE-CONCRETE DEBRIS BED BEHAVIOR, CORE-CONCRETE AEROSOL PRODUCTION RATES, AND AEROSOL TRANSPORT IN THE DRYWELL AND REACTOR BUILDING B.1 Introduction The purpose of this Appendix is to describe the methods used to calculate the results presented in Chap. 4 of this report. This in-cludes core-concrete debris bed behavior calculations performed using the CORCON-MODI code (Ref. B.1) and the INTER subroutine from the MARCH code, (Ref. B.2) aerosol production rate calculations performed using the VANESA code, (Ref. B.3) and acrosol transport calculations performed using the QUICK code (Ref. B.4 ) . In addition to the results presented .. in Sect. 4, three additional CORCON calculations were performed so that ' j the sensitivity of CORCON calculations to input parameters could be j investiated in a limited way; these results are presented in Sect. B.2 ; of this Appendix. The VANESA code was used to predict aerosol L l'
. generation rates vs tim for all of the COPCON runs and for the INTER -
calculation; these results are presented in Sect. B.3. B.2 Core-Concrete Debris Bed Behavior Calculations As discussed in Sect. 4.2, the aerosol transport results predicted for the LDHR sequence are based on core-concrete gas evolution rates and , rubble temperatures calculated with the CORCON-MODI code; for past ; , Browns Ferry accident sequences the results calculated with the INTER i subroutine of the MARCH code were used. In this section we present a - comparison of the " base-case" CORCON results presented in Chap. 4 with ! calculated INTER results and also with results from three additional > CORCON calcul.ations. . Table 4.1 in Chap. 4 presents the major input parameters used for ' the base-case CORCON run ("CORCON1") and for the MARCH INTER calcula-
^
tion. Table B.1 presents a summary of the major parameters varied in - The MARCH code calculated that the four CORCON runs and the INTER run. a roughly 13% of the zircaloy cladding was oxidized before melt-through of : - the reactor vessel; this amount of oxidation was assumed in the CORCONI and INTER calculations. The CORCON2 and CORCON3 calculations were per- i. ;
~~
formed to determine the influence of the initial amount of iircaloy oxi- ; dation on the calculated gas evolution rates and core debris tempera- : i tures. The initial amounts of zircaloy oxidation assumed in the CORCON2 and CORCON3 calculations were 50% and 90%, respectively. The CORCON4 - calculation was performed to determine how an increase in the initial i core-melt temperature would influence the calculated core-concrete in-
. teraction phenomena.
Tables B.2 through B.6 summarize the results from the CORCON and _ INTER calculations. Presented in the tatles are metal and oxide layer = i temperature, cumulative releases of CO, CO2, H2, and H20, and cumulative j amounts of SiO2 mixed into the melt due to interaction of the core-melt
4 l' 160 material with thel concrete.- A . number of observations can be made rel-5 ative'to'the information presented _in these tabIes: -
- 1. As' discussed; in. Chap. 4, CORCON tends to predict lo~wer gas releases "
and higher metal and oxide temperatures than INTER does. p 2. The -largest - difference in ' gas evolution between the CORCONI and INTER results was for C0; the CO release predicted -in the CORCONI , results is roughly four _ orders of magnitude less than that predicted by INTER. It was stated 'in Chap. 4 that the low CO release in the
- CORCONI run is due to " coking," or the reduction of CO to elemental
. carbon. 1Results from the . CORCON2 and CORCON3 runs saan to indicate that the initial' amount-of oxidized zircaloy in the. core melt has an influence on the calculated total amount of CO release. The total ~
CO release predicted.in'the CORCON2 calculation (50% of the zircaloy initially oxidized) was roughly 9 times that predicted in the CORCONI run, and the total CO release in the CORCON3 run (90% of the i zircaloy. initially oxidized) was more than 3500 times that predicted , in.CORCONI. . It is interesting to note in Table B.4 that there is 'a rapid increase in CO . release af ter 2830_ min; this time corresponds exactly to the time when there is no more metallic zircaloy in the
- core melt.
- : 3. The oxide layer temperatures predicted in the CORCON2 and CORCON3 runs tend to be lower, particularly near the end of the calculation,
.than the oxide - temperatures predicted for the CORCONI and CORCON4
, .. runs. This may be due to the fact that a " layer flip" is predicted
"~
to -occur in the CORCON2 and CORCON3 runs and not in the CORCONI and
'CORCON4 runs. 'In CORCON calculations, the core-melt is segregated a 7 _
into oxide 'and metal layers; _ for the conditions in the LDHR CORCON y runs the oxide laver is initially heavier and so is below the metal ~ layer. CORCON - assumes that all reactions in the metal layer af ter
- the start of core-concrete interaction produce oxides- that form a separate oxide layer ' above ' the metal. As the lower oxide layer *
. heats and changes . composition, it may eventually become less~ dense p .and " flip" with the metal layer; 'if this happens, then all of the
-1 oxides mix together. A layer flip was predicted in the CORCON2 run at 2890 min, and was predicted in the CORCON3 run at 2760 min. This layer flip was most likely due to the fact that there initially is more ' oxidized - zircaloy in the melt ' layers for the CORCON2 and CORCON3 runs. - The -mixing of the two oxide layers. due to .the layer
~
i flip . probably results -in the lower oxide temperatures predicted in CORCON2 and CORCON3. 4.- The increase -in the -initial melt temperature in the CORCON4 run leads to increases -' in the. release - of CO, CO2. H2, and H2O when compared to the CORCONI run.
.-5. cThe total amount of SiO2 predicted to be formed by INTER is a factor i-
- of 2 - to 5 times that predicted from the CORCON runs. - Within the
-CORCON runs - themselves, the ' total ~ SiO2 produced varied by roughly a factor of 2.5. The amount.of SiO2 produced _is an important output,
-because _ SiO2; may make up a major ' portion of the core-concrete aero- ,
sol. jx v a
- - - m- -,
, ,,,, ,,,wwwwam -
- -~
N N 161 B.3 Core-Concrete - Aerosol Production Rate Calculations
~
tec
- Core-concrete aeros'o1 production rate calculations were performed
;for .this 1 accident sequence using . the VANESA code; these calculations were -performed' at- Sandia National; Laboratories by Dana Powers, . who de-
+- .veloped VANESA.- The required input for VANESA calculations include con-crete f specifications,' the specification of fission product inventories and other materials ~in . the core melt at the time of start of core-iconcrete~ interaction,Jand core-concrete interaction. gas evolution rates,
- layer 1 temperatures,' and SiO2 production rates. The data . presented in
~
LTables 1B.2 p through B.6; provided la large portion of the' input for the VANESA1 calculations. . It - should be noted that the concrete specifica-tions L used ~' for the VANESA . calculations were for " degassed" limestone concrete; - these; specifications are presented in Table B.7. As will be seen:in discussing the.VANESA results, some concrete components that are present 'in smail: amounts contribute significantly to the predicted aerosol-source. Tables B.8 . through B.13 and Fig. B.'1 present . a summary of results from the.VANESA calculations._ Some of the important results from these
-calculations include the following:
- 1. ' Tables' B.8 throughl B.12 -and Fig. B.1 present aerosol release rates
.and . cumulative aerosol production. predicted by VANESA. The total amounts -of core-concrete aerosol predicted by' VANESA to be produced during the h following . the - start of core-concrete interaction varied by roughly a factor of -4. However, - the assumption used in' -
4 the CORCON3 calculation that 90% of the zircaloy is oxidized at re-actor - vessel melt-through. is probably unrealistic. for this accident sequence; ;if the VANESA-CORCON3 results are ignored, the total pre-dicted aerosol production for e the other four calculations varies by
- . only slightly greater than a factor of~2.
- 2. It is -interesting to find E that .the . VANESA run performed using. the
. INTER data as input gave the highest total aerosol release of 2107.
kg over the 8-h period. This was somewhat surprising since the pre-dicted INTER temperatures are quite,a. lot lower than those predicted. in the . CORCON runs.' At : least a partial explanation for the high re-lease-predicted in the VANESA-INTER calculation is that the. gas evo-
-lution rates from INTER calculations are significantly larger: than .
? those predicted in the CORCON runs.
L3. The VANESA-CORCON4 predicted -aerosol release ~was ' higher than that m . predicted for the VANESA-CORCONI run. This was expected, since the
- a. initial melt temperature assumed ~ for CORCON4 was more than 200 - K higher than that assumed for CORCONI.
-4. - Table :B.13 illustrates the predicted make-up of the core-concrete aerosol at - 3070 min. ; ' The predicted make-up of the core-concrete aerosol varied significantly ' for the summarized calculations.- It should.be noted that although K 02 and Na20 made up only 1.75%. of the assumed concrete composition (Table B.7), because they are quite ~% . volatile they constituted a significant fraction of the aerosol com-position in the VANESA calculations. In particular, for the VANESA-m INTER ' calculation K20 made . up 93.8% of the predicted final aero. sol mass.
.-1 162' B.4.?hUICKDrywellAerosolTransportCalculations The ..results from - the QUICK code aerosol transport . calculations , per-formed . for; drywelll conditions in - the LDHR sequence were summarized in Sect; 4.4. The purpose of ~ this section is simply to.' summarize the input-
. data ? used h fori th'ese, - calculations,. and~ to present any 'results -not .
~ summarized in--Sect. 4.4.
_ . _ The: input : data fore the QUICK drywell . calculations fall into the
~
following ) categories: (1) " fixed" -QUICK input parameters, (2) aerosol
- production-rate data,o (3) drywell . leak-rate data, _(4) drywell . pressure
~
and temperature data, .and (5) drywell geometric; data. _ Input data used for each of these categories is described in this section.
"B.421s Fixed QUICK parameters ETable B.14 presents QUICK input - parameters that do not vary - during the.drywell and reactor building' calculations.
B.4.2 Aerosol production-rate data
-The VANESA-CORCONI calculation was used.to develop the aerosol pro-
. duction-rate l data for the LDHR _ sequence. . Aerosol production rates for QUICK _ calculations are input in the form of cumulative' aerosol mass con--
centrations as a function of time. This' data is obtained from the data ipresented in ~ Table B.8 by dividing the column labeled " total ' aerosol . , produced" by the drywell volume which -is 4,500 m3. The drywell aerosol production-rate input data is presented in Table B.15. for- --An this aerosol sequenceparticle fis the size sameisasinput that to QUICK. Theparticlesizeuseg used in the SBLOCA s andisbasedondata.fromexperimentsperformedat-Sandia.B.pquence,B.
~
The aer- . osol ' source -' parameters used, assuming that the core-concrete ' aerosol ' density is;2.8 g/cm 3,; are:
. Geometric: particle radius: gr y 0.113 um-J Standard deviation:* a = 1.9-8
!A . ' B.4.3 Drywell leak-rate data
. The MARCH code computes gas leak ~ rates from the drywell; however, at present MARCR'uses INTER gas evolution rctes to calculate the drywell fleak rates. For this accident sequence, as is described in Appendix A,.
the :, MARCH-INTER drywell leak rates were modified by the ratio of the CORCON1-to-INTER gas: evolution rates; the modified drywell leak rates
. were . used .in f the QUICK calculations. The leak-rate . data used for the
*84% size /50% size.
4
163 QUICK drywell calculations are presented in Table B.16; the fractional gas leakage rate is obtained by dividing the volumetric leakage rate by the drywell volume. B.4.4 Drywell pressure and temperature data Based on MARCH drywell calculations, a temperature of 530 K and a pressure of 1 atmosphere were used in the QUICK calculation. B.4.5 Drywell geometric data Table B.17 presents drywell geometric data used in the QUICK calcu-lation. B.4.6 QUICK drywell results Table 4.4 and Figs. 4.4 through 4.6 present some of the results from the QUICK drywell calculations. Figures B.2 and B.3 present addi-tional results for the aerodynamic aerosol radius and the aerosol standard deviation as a function of time. B.5 QUICK Reactor Building Aerosol Transport Calculations
. The input parameters used in the QUICK reactor building aerosol transport calculations are presented in this section.
B.5.1 Reactor building aerosol production-rate data Aerosols that are leaked out of the drywell become the aerosol sou-rce to the reactor building. Table B.18 summarizes data related to the aerosol source to the reactor building. The aerosol source data used by QUICK are again the cumulative aerosol concentrations in the building; these data are obtained by dividing the total amount of aerosol released to the building by the building volume of 42,470 m . 3 In principle, the time variation of the agglomerate geometric ra-dius and standard deviation, listed in Table B.18, can be input to QUICK for the building calculations. However, including this form of input makes the code calculations very long-running and expensive. Because of this, a time-average of the size data presented in Table B.18 was used as input for the reactor building calculations: Average geometric radius: rg = 0.2154 pm Average standard deviation: o = 1.796 8
y . gQ l *. -
- v. 6 :. :> f., ' ' - I - : ' +. n .-- .> ' ' . . ' :n ,, -
. . * . :. . . r..;,n. : gs y : ... . ; .i ;p ;(. =..
7 .~,
- .3 _
164 .i
.e e- j
,, .:9 ,
' R. 5. 2 Reactor building leak-rate data . .
.. t >
, >[ Reactor building thermal-hydraulic calculations described in Ap-g
' pendix A indicate that for the first 4 min after the start of core- g
./ . concrete interaction there is gas exfiltration to the refueling bay and i. '.
to the environment in addition to the gas transport to the Stand-By Gas ' 41
-[ Treatment System (SGTS). However, af ter 2594 min all gas leakage in the
.~ .,
; reactor building is directly through the SGTS at a rate of 12,500
~ - :i SCFM. Table B.19 presents the leak-rate data used in the QUICK building [1:
L b ,' calculation; the fractional leakage rate is obtained by dividing the . volumetric ~ rate by the reactor building volume.
~ .- - Np ll@
e, - JQ : :.:, sf ;; y : B . S.3 Reactor building pressure and temperature data }; 9 j}. -,;J. -
.- Based on secondary containment model calculations, a temperature of G
% Agft,[" ,. ,... 283 K and a pressure of 1 atmosphere was used in the QUICK calculation. p. .
, p ':
. B.S.4 Reactor building geometric data . :
( "- . 5
..,.s', '
Table B.20 presents reactor building geometric data used for the f.. S '1 - QUICK calculation. ,,.
- ,, a ,
. .. c.
g,. J,. B . 5. 5 QUICK reactor building results 48 ..' "l.. og ' s
- 4. .:
.,
- Table 4.5 and Fign. 4.7 to 4.9 present some of the results from the ,
y, QUICK reactor building calculations. Figures B.4 and B.5 present :ly.
.MJ11 . additional results for the aerodynamic radius and the aerosol standard ,f
'+
- 'A deviation as a function.of time. It should be noted that the predicted ,,
1 ' * *,' aerodynamic radius in ~ the building is somewhat smaller than that pre- .* dicted in the dryvell; this is due to the fact that the aerosol source . 'y.y w , I.U.2-..f$ to the reactor building was assumed to have a time-independent averaged size. h_j
' ,g
*, .?.e. . .,- Q 4
.t 6 * . ,e.g - m;,
.k . , , .a 4 ?.
l .
'~
'y.* .g, *h' - i. m u
- ?, N .f f h-jf?. :' i*
') * .[ g
- t. . -
$r
' ,M N
.s*'-'. < g, ,. .
/v f
.. v' .,
A.:&$ $ g s,;#q N '.<
?-
p
- . _ 3
' :),9 ? if. l.', j f l.Y q
- j % 'l Q . o.'O l '.'* R.. .; 4= vyQQ: \ f b,. ~_o (- [:-f..;~.y. p ' 4 g,[. y . ;. - _y , 3 } ;q
165-References for Appendix B B.1 .J. F. Muir , R. ' K. Cole, Jr., M. . L. Corradini, and M. A. Ellis, CORCON-MOD 1: An Improved Model for Molten Core / Concrete Interac-
.-- tions, NUREG/CR-2142, SAND 80-2415 (February 1982).
B.2 R. ; 0. Wooten and H. I. Avci, MARCH Code Description and User's
- Manual, Battelle Columbus Laboratories, USNRC Report NUREG/CR-1711 (October 1980).'
B.3 D. A. Powers and J. E. Brockman, " Status of VANESA Validation," in Report of the- Status of Validation of the Computer Codes Used in the NRC Accident Source Term Reassessment Study (BMi-2104) , pre-pared by~T. S. Kress, ORNL/TM-8842 (to be published). B.4 H. Jordan,--P. Schumacher, and J. A. Gieseke, (UICK User's Manual, NUREG/CR-2015, BMI-2082 ( April 1981). B.5 Personal Communication, D. A. Powers, Sandia National. Laboratories, January'1984.
-B.6 R.' P. Wichner et al., 'SBLOCA Outside Containment at Browns' Ferry Unit One, Volume 2. Iodine, Cesium, and Noble Gas Distribution and Release, NUREG/CR-2672, Vol. 2, ORNL/TM-8119/V2 (September 1983).
.B.7 - Letter - f rom R. A. Lorenz to A. L. Wright, "Recent Comunications from Sandia' Laboratories Regarding Core Melt-Concrete Interaction Aerosol Sizes," July 12, 1982.
o. e
+
9 10 1
166-Table B.1.- Summary of major parameters varied for the CORCON and INTER calculations Initial Initial Initial Initial Initial Case temperature Fe0 Fe Zr02 Zr
.(K) (kg) (kg) (kg) (kg)
CORCONI 1737 5,150 116,100 11,530 56,960 CORCON2 1737 5,150 116,100 44,240 32,750 CORCON3 1737 5,150 116,100 79,620 6,550 CORCON4' 1950 5,150 116,100 11,530 56,960 , INTER 1737 0 120,100 11,530 56,960 O
lllli)Illlllll)l 1ll l ll ll ;llI 1 ll ' t _ N' gw _.'e t a e e ' l' 2)
~
O7966067439426468601 92306 a Og 81 28891 47491 1 098775321 09 io i t k 1888902345691 0371 59371 58 T S( 1 235678901 344455566777 1 1 1 1 1 1 1 1 1 1 1 1 1 1 O788187842O00O00OOOOOOOO0 l . . . . . . . . . ~ ................ ' a O )g t 2 501 855571 6574245544321 98 201 1 2345781 5906284062839 To H (k 1 234567891 2355667889900 1 1 1 1 1 1 1 1 1 1 1 1 22
~
l' ) O61 55952G2555727384O74O85 t a2g 40865421 8529445567/8990
> o Hk( 1 1 23456678899999999990 s T 1 t
l. u s e O91 89O00O0O00O00OOOOOOOO0 r l . . . . . . . . . ................ . a2) t0g 801 8O8701 41 6021 9381 23556 I 71 571 4838336921 986531 975 To 0 (k N 3693693604826891 3579024 O C 1 1 1 22333444455555666. R O
.. C l a )
g 001 3 6' 81 2234677788888889 470424267902346780
'2. t@k o (
0000001 1 1
~
B T e
- l.
b - a T
- e 7606289001 571 996256640590 d) i 390466677766770'62727271 501 xK o( 780000000000001 223344556 T
1 1 22222222222222222222222 l a . 757891 3468572425939529764 te( )K 345556666725801 0998887777 s 7777777777888999888888888 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 T
)
en 0000000000000000000000000 mi 91 35791 35791 35791 35791 357 ' i m 5666667777708G88999990000 T( 2222222222222222222223333 4 i e I' f
5.;
- a.
~
+
Table ,8.3 00RCON2 results -; r
'u Total Total Total- Total' Total' Time ~Teetal. -Toxide -
2 H_ 2 H2 0. SiO2 (min)- (K) (K) (kg) -(kg) (kg) (kg): (kg)
.00 .O .O .O O 2590 1737 .1737-- 4-2610 1745 1863 . 00. .64.5: .' 4 - 21.0 '145-2630- 1757 1964 .01 .223.9' 2.1 ' 72.8- 549 2650 1757 2025 .02 -528.8 6,2 -171.9 1357 2670 1759- 2056 .04 873.4 12.1 283.9' 2338-2690 1760 2070 .06 '1218.O 18.9 395.9 3376 2710 1762 2076 .08 1580.O 26.3 513.5 3404 2730 1763 2078 .1 1. 1953.0 33.9 -634.'7 5602 2750 176':> 2079 .14 2326.0 41.6 755.9 6737 .
2078 .' 16 '2'704.O 49.3 878, f3 7883 $- 1767 2770 2790 1807 2079 .20 3086.O 57.1 1003.O 9035 2
' 810 1854 2084 .28 3519.0 64.9 1144.0 10297 2830' 1892 2081 .44 4013.0 73.3 1304.0 11700 2850 1919 2069 .70 4549.O 81.3 1478.O 13170 2870 1935 2063 .97 5097.0- 88.5 1657.0 14618-2890 1944 2017 1.32' 5637.O 96.4 1832.O. 16090 ,
2910 -1940 2028 2.21 6133.0 123.6 1993.0 18510 2930 1935 '2 0 3 8 - 3.01 6639.O 150.2 2158.O 20910 2950 1933 2046 3.76 7172.O 176.4 2331.O 23340-2970 1932 2049 4.48 7761.O 202.5 2522.O 25880 2990 1934 2048 5.21 8416.0 228.7 2735.0 28550 3010 1932 2044 5.92 9155.0 254.1 2975.0 31340 3030 1931 2039 6.60 9913.0 279.3 3222.0 34160 3050 1933 2033 7.30 10680.O 304.8' 3471.O. 37020 3070 1934 2027 8.02 11458.0 330.8 3724.0 39920 l l l l t
- 6 .o
y_ ,- w. -. i
;, , ; :e; ., },' -
+ , p . [o; , ' ~ ,
- i.!
S > J Table B.4, .00RCON3 " results s :1
. Total- Total: Total -Total Total ~. -
Time ;Tmetal Toxide CO H. HO SiO2 -
~
-CO2 2 2 (min) -(K) (K) -(kg) (kg) :(kg)' (kg)' '(kg)-
2590 1 7 3'7 1737. .00 .O .O- .O O 2610 1746
~
1839 .00 56.1 .3 18.' 2 - -124 2630 1757 .1928 .00 171.2 1.3 55.6 400 2650 1757 1994 .12 403.3 3.7 131.1 982 2670 1758 2035 .03 725.3 7.8 235.7 1822 ~l 2690 1760 2058 .04 1065.O 13.O 346.O 2759 j 2710 1761 2071 .06 1430.O 19.O 464.6 3787 ! 2'730 1767 '077 .08 1813.O 25.4' 589.3 487S 2750 1'765 2079 .11 2211.O 32.1 710.6 600' 2770 1766 2054 .13 2592.0 41.3 842/5 7?34 5 e
)
2790 1767 2049 .16 2965.0 52.2 963.5 8543 2010 -17 W) 2045 .lo T357.0 63.6 1091.0 9915 2830 1827 2036 .25 3828.0 78.2 1244.0 11610 7850 Ivo7 2002' 574.10 4393.O 93.7 1428.O 13530 'l 2870 1768 1 r?74 - 1147.00 4886.O 105.4 1588.O '15110 2890 1760 1955 1656.00 5336.O 115.8 1734.O 16540 2710 1767 1955 2168.00 5711.O 126.3 1856.O 17830 ; 2930 1766 1979 2396.00 5912.O 136.7 1922.O 18770 j 2950 1~766 1991 2579.00 6115.O 147.1 1988.O 19730. j 2970 1766- 2003. 2761.00 6322.O 15'7.4 2056.O 20600 l 2990 1765 2015 2929.00 6528.O 167.O 2123.O' 21590 5010 t765 2027 3089.00 6734.O 176.1 2191.O 22480 3030 1765 203e 3250.00 6946.O 185.2 2260.O 23380 WM 1754 2049 3410'00 7164.0' 194.3 2331.0 24290 I 3070 1764 2059 3570.00 7387.O 203.5 2404.O 25210
=1 v ,
1 Table B.5;.00RCON4 results4 Total- ' Total Total ' Total- Total-Time- Teetal Toxide. 0 CO2 H-2 302 MO2 (min)- -(K) (K) (kg)- (kg) (kg): (kg) (kg): 1950-
.00 .O- .O' .O- 10- i 2590- .1950
~2610: ' 1941 .2044. .27 465.6- 5.5- 151.3 1199' 1630- 1951 ~ 2086' .70 1016.O 14.O 330.3 2722
. 1966 2104 i t . 30 -; 1625.O' 24.2 528.2 4444 -
2650
~%70 . 1981- 2112 2.09 2277.O 35.0- 740.O '6285-3.05- 2967.O 45.9 964.2 8208-2690' --
.1993 2116
.10155 2710 2000 2111 4.11 3675.0 56.6' 1195.0 2001 2100 -5.09 4359.0 66.1, 1417.O' 11986 2730 2097 6.00 5032.0 74.9 1635.O 13763 _
2750 1799 _ 15472- g 2770. 1996 2099 6.85 5677.O 83.4 1845.-O
-2107 7.68 6271.O 92.O' 2038.O 17082-27'70 1995
~18214 2810 1985 2137 8.15 6718.O 97.O 2183.O.
2830 - 1969 2195 8.19 6980.O 97.5 2269.0 18750 195S 2250 8.23 7235s0- 98.O 2351.O 19272 2850 2303 8 26 7481.0 98.6 2431.0 19780 _2870- 1944 20280 2890 1933 2354 8.28- 7723.0 99.2 2510.0 1925 2403 8.31 7958.O 99.8 2586.O 20770 2910 21'250 2930 1917 2451 8.33 8190.0 100.4 2662.0-1910 2497 8.35 8419.0 101.1 2736.0 21740 2950 22210 2970 1904 2541 8.37 8645.O 101.8 2810.O 1898 2584 8.39 8870.O 102.5 2803.O 22680 2990 1893 2626 B.41 9092.0 103.2 2955.0 23150 3010 3030 1889 2666 8.43 9315.0 104.0 3027.0 23630
- 1886 2724 8.45 9546.0 104.9 3102.0 24120 3050 1883 2791 8.47 9785.O 105.9 3180.O 24640 3070
i.j
. ~
,r _v"
*[ U 5
s l l aOg 2) O300000000e00000000000000 t ik 57669051 1 1 41 961 1 00807724 oS( 25743421 09764320751 83037 T ' .9372703691 4703691 4792479 1 1 223333444555566667777 l' O226771 1 1 3230202905345795 a 0 )g . 83587097539631 9776666667 t 2 41 4592357(1 0246791 35791 35 To H (k 22223333344444455555666 1 l . a 2g ) O091 42358261 6047888527257 t 1 1 6381 471 481 581 4'7 03681 35 o Hk( 1 1 2223334445556666777 T
's t
l u s e l p a 2) O573799000000000000000000 R tog 555554495001 495322346804 E o C (k 55591 84951 61 61 73951 73962 T TL . 467799001 2233445567788'? 0 N 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 I k 6 l O9331 2350770761 61 00000000
'B aOg
)
t 3821 61 76557899625762701 0 e o Ck( 1 2031 728406284061 61 60593 l 2223 b T 1 2334456677899001 1 1 1 1 1 1 1 1 1 f a T e '74047281 321 21 83940629631 8
'd}
i 3453043444444332221 1 00009 x 7777766666666666666666665 T o( 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l a 709301 t 888889777665554433 3 t
'e 3944432222222222222222227 D
n 7666666666666666666666666
' 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 e"
}
0000000000000000000000000 mI 91 35791 35791 35791 35791 357 i" 566666777778088899999000O T( 222222222222':22222272333T [ 7 e e;
~
2. 172 Table B.7 "De-gassed" concrete specifications used for VANESA + calculations-Total
.*- e n rete dal mass
(%) cao 43.0 A1203 4.9 Na20 0.1
- a . ,_
-K20 1.65 SiO2- 48.4 Fe0 1.95 e
+.-
W 173 m. o; f
'l Table B.8. VANESA aerosol generation rate results,- based on CORCONI 4
.L \
Aerosol Total-
. Time . generation aerosol-(min) rate produced (g/s) (kg) 2590' 6.70 .O'
'2610 22.69 17.6 2630 34.93 52.2 2650 4O.92 97.7 2670 42.16 147.6 2690' 43.32 198.9
-2710 43.93 251.2 2730 44.19 304.1 l 2750 44.53 357.3 2770 ~44.21 410.6
-* z2790 45.37 464.3
'2810 49.24 521.1-2830' 50.66 581.0
. 2850 44.22 637.9 2870 29.38 682.1 2890' 20.91 712.3
'2910 22.09' 738.1 2930 22.13 764.6 2950' 22.53 791.4 2970 22.93 818.7 2990 23,51 846.5 3010 25.13 875.7 3030 26.28 906.6-3050 27.74 439.O
. 3070 29.50 973.3 4
. - i 0
,.s I
O'
'i = .
.J h................ ..i.,..___....m. _ . . . . . .
4 179
. Table B.9. -VANESA aerosol generation -
rate results, based on CORCON2
-Aerosol Total
' Time generation aerosol
' (min)- rate produced (g/s) (kg) 2590 5.13 .O
. 2610 13.65 11.3
= 2630 35.44 40.7 2650 35.73 83.4 2670 39.69 128.7 2690 41.70 177.5
- 2710 43.57 228.7 2730 44.32 281.4 2750 44.63 334.8 2770 45.02 388.6 2790 47.64- 444.2 ,
2810 53,28 504.7 2830 58.00 571.5 2850 59.44 641.9
, 2870 59.22 713.1 2890 66.71 788.7 .
2910 75.54 874.1 2930 76.62 ' 965.3 2950 79.94 1059.3 2970 85.25 1158.4 2990 91.28 1264.3 3010 94.91 1376.0 3030 95.69 1490.4 3050 96.45 1605.7 3070 96.69 1721.6
175
- o.
-= .; Table B.10. VANESA aerosol generation rate results, based on CORCON3 Aerosol' Total Time generation aerosol (min) rate produced (g/s) (kg) 2590 4.22 .O 2610 8.72 7.8 2630 21.95 26.2 2650~ 24.34 53.9 2670 34.48 89.2 2690 37.79 132.6 2710 40.41 179.5 2730 42.16 229.1 2750 43.41 280.4 2770 43.93 332.8
, 2790 39.83 383.1 2810 4.15 409.4 ;
2830 .6.39 415.8 l 2850 7.07 423,8
'2870 5.75 431.5 'e '2890 4.88 437.9 ,
2910 3.61 443.O I 2930 2.66 446.8 2950 2.64 450.0 2970 2.65 453.1 2990 2.63 456.3 3010 2.68 459.5 3030 2.78 462.8 3050 2.89 466.2 3070 2.98 469.7
-e
'-e ,.
u a.
176 Table B.11. VANESA aerosol generation a rate results, based on CORCON4 Aerosol Total Time generation aerosol (min) rate produced (g/s) (kg) 2590 40.26 .O 2610 57.46 58.6 2630 69.56 134.8 2650 77.48 223.1 2670 82.66 319.2 2690 85.63 420.1 2710 83.62 521.7 2730 79.33 619.4 2750 76.22 712.8 2770 71.96 801.7 2790 60.06 880.9
- 2810 39.56 940.7 2830 28.68 981,5 2850 29.44 1016.5 2870 29.29 1051.7 2890 29.11 1086.8 ,
2910 29.25 1121.8 2930 30.65 1157.7 2950 31.59 1195.1 2970 32.84 1233,7 2990 34.27 1274.O 3010 36.30 1316.3 3030 39.71 1361.9 3050 44.57 1412.5 3070 53.68 1471.5 O
p . s . 177 4
' - l
,1 Table B.12. VANESA aerosol generation rate results, based on INTER Aerosol Total Time -generation aerosol (min) rate produced (g/s) (kg)
'2590 279.83 .O 2610 245.10 315.0 2630 152.41 553.5 2650 158.39 739.9 2670 167.91 935.7 2690 122.23 1109.8 2710 88.42 1236.2 2730 85.03 1340.3-2750 84.73 1442.1 2770 84.01 1543.4 2790 81.53 1642.7 2810 79.80 1739.5 e 2830 '78.48 1834.5 2850 74.97 1926.5
-2870 66.02 2011.1 2890 28.17 2067.6 2910 2.70 2086.2
-* 2930 2.57 2089.3 2950 2.42 2092.3 2970 2.24 2095.1 2990 2.11 2097.7 3010 1.97 2100.2 3030 1.83 2102.4
,3050, 1.70 2104.6 3070 1.59 2106.5
'O'.
.. 9
178 Table B.13. Core-concrete aerosol components predicted by VANESA based on CORCON and INTER data gg g,1 Total aerosol mass at 3070 min, (%) CORCON1 CORCON2 CORCON3 CORCON4 INTER SiO2 26.6 23.5 1.15 19.8 0.11 Ca0 20.4 21.9 6.12 18.3 2.% K0 2 13.0 20.2 69.0 9.57 93.8 Feo l'.4 21.5 1.1 9.31 0.05 A1293 7.32 9(10~4) 0.01 8.33 0.05
- Nas o 7.22 12.1 3.69 5.37 1.57 Mn 7.11 0.53 12.3 8.97 0.77 13203 1.87 3(10-5) 6(10~4) 5.10 8( 10~ ) .
Ce02 1.08 1(10~3) 8( 10~) 1.90 1(10 3) Te 0.33 0.13 2.76 0.42 0.96 Sn 0.66 0.03 2.65 1.11 0.1 . All'other 2.0 0.1 1.2 11.8* 0.5 materials
*For. the VANESA-CORCON4 run, Nb 02 5 made up 9.45% of the aero-sol mass at 3070 min.
e e i..... __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _
i b 179 i Table B.14. " Fixed" QUICK input data Parameter Value CASE COMPL LSTEP 20 NDIM 1500 DELR 0.001 RMAK 100.0 N 30 DELT 0.001 EPS 106 MF 22
' BMINI 10 15 BMIN2 10~4 IRST 0 RG 0.0 SIG 0.0
.o NO 0.0 T 155400.0 TMAX 184200.0 TMASS 0.0 ENERGY 0.0 TNERG 0.0 PRESS 1.0 RHOP 2.8 CRATIO 0.0006 DELD 0.001 CEFF -1.0 TS 184200.0 CHI 1.0 GAMMA 1.0 VF 0.0 l-1 f
L-s h t L
j . ( t 180
\
Table B.15. Drywell aerosol data Cumulative Time- aerosol
. (min) ' concentration (s/m3 )
2590 .000 2610 3.919 2630 11.601 2650 21.715 2690 44.189 2710 55.823 2750 79.401 2770 91.233 2790 103.'177 2810 115.792 2830 129.112 2850 141.763
.2870 ~151.576 2890 158.281 ,
2910 164.015 2950 175.865 2990 188.118 5 3010- 194.604 3030 201.458
-3070- 216.293 g4 4
-p
p. 1 181 ?' s: w Table.B.16. Drywell leak-rate data Volumetric Fractional leak rate leak rate ("} (m3 /s) (1/s) 2590 54.361 .0120802 2594 27.069 .0060153 2597 19.818 .0044041 2606 14.585 .0032412 2618 11.027 .0024504 2630 9.051 .0020114 2653 3.712 .0008248 2670 5.038 .0011196 2685 3.338 .0007419 2709 3.518 .0007817 2769 3.039 .0006752 2829 3.381 .0007512 2889 2.715 .0006034 g 2949 2.834 .0006298 3009 2.554 .0005675 3069 2.523 .0005606 3070 2.523 .0005606
-o Table B.17. Drywell geometric data Floor area / volume = 2.89(10 3) ca~1 W11 area / volume = 9.49(10'3) cm 1 Drywell volume = 4,500 m 3
-e t_
Table B.18. Reactor building aerosol data Cumulative Aerosol- # '# Time . concentration geometric standard
* "E (min) (kg)
-to building radius deviation (gf,3) ()
2590.O .000 .000 ,1130 1.900 2600.5 4.311 .102 .1468 1.810 2617.2 20.502 .483 .1756 ~1.739 2633.8 46.186' 1.087 .1945 1.713 2667.2 105.110 2.475 .2406- 1.740 2700.5 178.230 4.197 .2499 1.835 2733.8 255.810 6.023 .2284 1.880 5' 2767.2 332.540 7.830 .2281 1.899 2800.5 411.240 9.683 .2260 1.912 2833.8 497.570 11.716' .2278 1.914 2850.5 539.710 12.708 .2308 1.914 2867.2 576.690 13.579 .2312 1.912 2883.8 606.670 14.285 .2288 1.905 2901.3 632.080 14.883 .2248 1.902 2933.8 673.750 15.864 .2218 1.875 2967.2 715.050 16.837 .2226 1.849 3000.5 755.670 17.793 .2238 1.846 3033.8 798.090 18.792 .2252 1.856 3070.0 848.120 19.970 .2402 1.847
' ~ ~
. . < _ , . . 'e e. 4 Table B.19. lleactor building leak-rate data Time Total volume Total volume Fractional e leak rate leak rate volume leak rate (min) (cg,) g,37,) (17,) 2590 99139.8 46.788 .0011017 - 2591 50605.7 23.883 .0005623 $ 2592 31281.0 14.763 .0003476 2593 13866.2 6.544 .0001541 2594 12500.0 5.899 .0001389 3070 12500.0 5.899 .0001389 M h
184
- Table B.20. Reactor building geometric data
' ~
Floor area / volume = 1.387 m~1 W11 area / volume = 1.334 m1 Reactor' building volume = 42,470 m3 w a b a f 4
185 I a ORNL-DWG 84-5780 ETD 2400 [ I Q CORCON1 2mo - Q CORCON2 ' l K CORCON3 I
-{ CORCON4 .
~ _
g A INTER 5 3 42w - - a o 5 4 am - - m 1--=== M em - =_^ * "=[y,- , - - J# w I
',. l 27m 2900 31 "
TIME (min) Fig. B.l. Total aerosol mass vs. time from VANESSA calculations. F
\\:
e 0' 80 81
. 1
- 0 4
$1 0 e 0 m
$ i 1
t s v 0 s . 6 u N i d a r - i c 0 m
- 2) a NC n y
E d ( S o r e E a . 0. 8I M l
$T o 1 s o
r e a l . 0 l
$1 4 e w
y r D 0 2 0
$1 B g
i F 0 6 1 0 2 5 1
. ,e. n- o <. .o .. ,
g 55- 22< . l1 l1 ll lll jll 1l l llll1
i ll lll l ) ) g* O T 0. ' e E 80 81 2 8 1
- 7 5
4 8 G W D
- 0 L
_ N 4 I $1 W _ 0 0 e
$ m
- 1 i t
. s _ v n _ 0 o _ 6 i N t a _ i v e . d 0 d
- 2) r NC a
_ d E n S ( a
, t s
E oM aI l o s dTI o r e a l l 0 e 4 w 1 y r D 3 0 B _ ! 0 1 . g i F
\ 0 6
1 0 2 5 1
~ - _ - - -
b ! ,_ a- eC b E
- =
< 2 en
, ~
J. Eoo D 0' T E 80 81 3 1 * . 8 7 6 4 8 G W D 0 L 4 i8 N R O
, 1 e
m i t 0 s 0 v i81 s u i d a r 0 i71 6 c i m a y n d o 0 r
- 2) e i7 C 1 a E l S
( o s E o . r
- 0. M e 8I a i6 1 T g n
i d l . i 0 u 4 b i61 r o t c a e 0 R 0 i61 . 4 B g 0 i 6 F i51 0 2 5 1 _ - _ ~ _ ~ 8- 8o No E Eo 8o Nd OZo%O.2v:a22< . l
,)
l iI Eo - s D 0. f T E 8C 4 81 - 8 1
- 7 4
4 8 G W D - 0 L 4 N $ R 1 . O e m i t 0 s 0 v 1 n o _ i t a e, i 0 v N 6 e I d d r a d n 0 a t N 2 I )C s E l S ( o s E o r 0M e a aI
- dTi i
d g n l
- i 0 u 4
b d r i o _ t c a e R 0 0 d . 1 5 B g 0 i 6 F
$1 0
2 5 1 b Ra b Ea $ 3 $.. a <MO rn s
191 Appendix C T
-ACRONYMS AND SYMBOLS ANS American Nuclear Society
,; ; ANSI American National Standards Institute BAF. Bottom of Active Fuel
.BCL Battelle Columbus Laboratories BFNP Browns Ferry Nuclear Plant
.Br Bromine-BWR . Boiling Water Reactor.
CBP . Condensate Booster Pump Co . Cobalt
, CP- Condensate _ Pump Cr Chromium CRD~ Control Rod Drive CRDHS' Control Rod Drive Hydraulic System Cs Cesium CSEL Containment Systems Experiment Cs0H Cesium hydroxide
'DF Decontamination Factor DW- DrywellL ECCS- Emergency Core Cooling System
- E01 Emergency Operating Instruction EPA Electrical Penetration Assembly EPRI- Electric Power Research. Institute
~0 Fe: !
Iron
'FSAR Final ~ Safety Analysis Report
- GE General Electric Company HAARM Heterogeneous Aerosol Agglomeration Revised Model
- c. HEPA High Efficiency Particulate Filter HOI' Hypoiodous Acid HPCI High Pressure Coolant Injection I Iodine I2(d) Dissolved molecular iodine ID Internal Diameter Kr- Krypton
'L Liter
.LDHR Loss of Decay Heat Removal
. LPECCS Low Pressure Emergency Core Cooling System LPCI- Low Pressure Coolant . Injection mode of the RHR system -
--LOCA Loss of Coolant Accident LWR Light Water Reactor MARCH _ Meltdown Accident Response Characteristics Mn Manganese MPa .Megapascal MS'V Main Steam Isolation Valve Ni Nickel s NRC Nuclear Regulatory Commission O 0xygen ORNL Oak Ridge National Laboratory 1
192 Pa Pascal PBq Petabecquerel , PC Partition Coefficient PSP Pressure Suppression Pool PV Pressure Vessel PWR Pressurised Water Reactor RCIC Reactor Core Isolation Cooling
- RES Office of Nuclear Regulatory Research RHR Residual Heat Removal RPV Reactor Pressure Vessel RV Reactor Vessel SASA Severe Accident Sequence Analysis S5GTS Standby Gas Treatment System SBLOCA Small-Break Loss of Coolant Accident SGTS Standby Gas Treatment System Si Silicon SI Systeme International Sn Tin ,
SNL Sandia National Laboratories SRV Safety Relief Valve TAF Top of Active Fuel Te Tellurium TVA Tennessee Valley Authority U Uranium W Wetwell Xe Xenon Zr Zirconium , 3 4 f
-. ~ .. . . .
...w
a: o l 193 NUREG/CR-3617 e ORNL/TM-9028 Dist. Category RX, 1S Internal Distribution
-t
- 1. E. C. Beahm 17 . F. R. Mynatt
- 2. S. D. Clinton 18. L. J. Ott 3.. T. E. Cole 19. R. D. Spence
- 4. G.' F. Flanagan 20. H. E. Trammell
- 5. S. R. Greene 21. D. B. Trauger
- 6. D. Griffith 22. C. F. Weber
- 7. W. O. Harms 23-24. R. P. Wichner
. 8. R. M. Harrington 25. J. H. Wilson
- 9. J. R. Hightower 26. A. L. Wright 10-11. S. A. Hodge 27. R. G. Wymer
- 12. C. R. Hyman 28. Patent Office
! 13. J. E. Jones 29. Central Research Library t; 14. T. S. Kress 30. Document Reference Section
- 15. R. A.,Lorenz 31-32. Laboratory Records Department
- 16. A.' P. Malinauskas 33. Laboratory Records (RC)
External Distribution
- 34. Director, Division of , Accident Evaluation, Nuclear Regulatory Commission, Washington, DC 20555 35-36. Chief, Containment Systems Research Branch, Nuclear Regulatory Commission, Washington, DC 20555 37 . Chief, Fuel Behavior Branch, Nuclear Regulatory Commission,
- c. Washington, DC 20555
- 38. M. J. Jankowski, Fuel Behavior Branch, Nuclear Regulatory Commission, Washington, DC 20555
- 39. L. . Chen, . Fuci Behavior Branch, Nuclear Regulatory Commission, Washington, DC 20555
- 40. .W. Pasedag, Accident Evaluation Branch, Division of Systems Integration, NRR, Washington, DC 20555
- 41. Director, Reactor Safety .Research Coordination Office, DOE, Washington, DC 20555
- 42. Office of Assistant Manager for Energy Research and Development, DOE, ORO, Oak Ridge, TN 37830
- 43. L. O. Proctor, Tennessee Valley Authority, W10D199 C-K, 400 West Summit Hill, Knoxville, TN 37902
- 44. Wang Lau, Tennessee Valley Authority, W10C126 C-K, 400 West 7 . Sunuait Hill, Knoxville, TN 37902 45.- J. A. Raulston, Tennessee Valley Authority, W10C126 C-K, 400 West Summit Hill, Knoxville, TN 37902 46-47.
' R.-; F. ? Christie , Tennessee Valley Authority, - W10C125 C-K, 400 s Suiamit Hill,. Knoxville, TN 37902
- 48. H. L. -Jones,' Teanessee Valley Authority, W10A17 C-K, 400 West Ssamit x Hill, Knoxville, TN 37902 f: i sT ,
a u .
i 194
- 49. J.. D. . Woolcott, Tennessee Valley Authority, 1530 Chestnut
. Street,- Tower II, Chattanooga, TN 37902 ,
L 50. Z. R. Rosstoczy,' Research and Standards Coordination Branch. Office of ' Nuclear Reactor Regulation, U.S. Nuclear Regulatory Commission, Washington, DC 20555 51-52. Technical Information Center, DOE, Oak Ridge, TN 37830 53-577. Given distribution as shown under categories RX, 1S (NTIS-10) 7 4 i J f .2 .
r<sC , v., ao .< ~, gegoaWm u s. =vetin anzutaroav eo issen i anco 1 ~wv ta a . N'i SE' Bl;LIO2RAPHIC DATA SHEET - SEENstauCT6045 0g twg aeveast ORNL/nt-9028 d VITLE A%Q sL91tTLg 3 (E AVE ti A%4 Noble Gas, Iodine, and Cesium Transport in a Postulated Loss of Decay Heat Removal Accident at Browns Ferry
. o A n a .oa r p.u n o vo~r, veAa
* ** '"C " ;5' R. P. Wie ner, C. F. Weber, A. L. Wright July f 1984 t .S. A. Hodg E. C. Beahm *'"Y"""*
WQN r se VEAa August 1984 7 e tMEQ81Mi4Q Qa3ANIZArlON MAUT ANQ MAIL G AODat$5 fsarewee le Ceser a Paostc1.T Asa wca[vNir Nuvet. Oak Ridge Natio 1 Laboratory 1 -
,,,,,,c,,,,,, ,,,,
P. O. Box X y Oak Ridge, Tenness 37831 B0453 10 5+0%$3aiNG OaGA%il ATION Navt AND M AILING ADDats$ ff de lf Cape, 1 S te tv*E O{ at#0af Division of Accident Ev uation /bpical Office of Nuclear Regulat y.Research f I' U.S. Nuclear Regulatory Co ission = n +oo coviaao oa~.~ => Washington, D. C. 20555 p [N/A 2 su,.avi~r Aav =ons g
/
5 n sesvucr ,zx. -.: , j This report presents an analysis of the m ement of noble gas, iodine, and cesium fission products within the Mark-I containment B reactor system represented by Browns '
# Ferry Unit 1 during a postulated accident sequence initiated by a loss of decay heat removal. capability following a scram. This acciden oEild be brought under control by various means, but the sequence with no operator act ultimately leads to failure fol-lowed by loss of. water from the reactor vessel, core A radation due to overheating, and reactor vessel failure with attendant. movement of co debris onto the drywell floor.
_c The fission product transport analysis is base and provides an estimate of fission product inventof iesonasth' no-operator-action function of time, within 14 sequence controlinvolumes volume the transport outside sequence. the core, We find with small the atmosphIred as co,nsidera,8 the final control gas ejection to air, these gases being ef fectively purged from the dr ,b%rrier for no 11 ding by steam and ell and reactor concrete degradation gases. In contrast, large degrees of holdup iodine and cesium are projected due to the chemical reactivity o -these elements. On1 bout 2 x 10'% of the initial iodine and cesium activity are iredicted to be released o the atmosphere. Principal barriers for release are depositio on reactor vessel and cont nment walls. j is occuva%v A%A6vs s -, nev*oaos oasca.*roas '
/
d"' 's a v Anaaitit y
- srAnwa%T
'BWR Severe Accident Analyses "
s
'6 SEcu sTV cL AS5'secAT+0N Fission Product Transport ,r.,,,,,,
.6 . ,ce s .ie ,e m e%een navs Unclassified UhhTa5sified o w ... o.. Ages
. . ,aia
' ' ' ' 'l '
._-____-__.__--___m___-.-_---__.______._m_ - - - - . . _ _ _ _ - - _ _ - _ _ - - -- -
\
A r 120555073377 tJ 5 1 1ANIRX115
'i?. C AD'4-DIV GF TIOC POLICY f. P U B .-50i .% T 3 R -P D P P.UREG WASHINGTCN OC 20555 e
( 1 s}}