ML20106C403
| ML20106C403 | |
| Person / Time | |
|---|---|
| Site: | Peach Bottom  |
| Issue date: | 07/31/1984 |
| From: | Cybulskis P, Denning R, Gieseke J Battelle Memorial Institute, COLUMBUS LABORATORIES |
| To: | NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| References | |
| RTR-NUREG-1150-2-V2-B.42 BMI-2104-DRFT, BMI-2104-V02-DRFT, BMI-2104-V2-DRFT, NUDOCS 8410240152 | |
| Download: ML20106C403 (225) | |
Text
,..
DRAFT l
BMI-2104 Volume ll l
l l
l Radionuclide Release Under Specific LWR Accident Conditions Volume 11 BWR, MARK I Design Prepared by J. A. Gieseke, P. Cybulskis, R. 3. Denning, M. R. Kuhlman, K. W. Lee, H. Chen Battelle Columbus Laboratories Columbus, Ohio 43201 July 1984 8410240152 840732 DR TOPRP EXIBMCL PDR
l i
DRAFT .
BMI-2104 Volume 11 l
Radionuclide Release Under -
Specific LWR Accident Conditions Volume 11 !
BWR, MARK I Design '
Prepared by J. A. Gieseke, P. Cybulskis, R. S. Denning, M. R. Kuhlman, K. W. Lee, H. Chen Battelle Columbus Laboratories Columbus, Ohio 43201 l
July 1984 l
l Prepared for
- ' Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission l Washington, D.C. 20555
'-a w ACKNOWLEDGMENTS Battelle's Columbus Laboratories wishes to acknowledge and express appreciation for the computer codes made available for this program by Sandia National Laboratory, Battelle's Pacific Northwest Laboratories, and the
- Kernforschungszentrum Karlsruhe, and for the computations and consultation I provided by Sandia and the Oak Ridge National Laboratories. Further, members of the Peer Review Group have contributed significantly to this effort by providing comments, suggestions, and information on various reactor systems desigh.
The support of the U.S. Nuclear Regulatory Commission is gratefully acknowledged, and is the untiring leadership of the NRC staff, particularly Mel Silberberg and Mike Jankowski.
The diligent efforts of many Battelle. staff members contributed to the preparation of this report. The following list identifies those staff making major contributions:
RJ Avers GT Brooks EP Bryant R Freeman-Kelly CS Jarrett H Jordan RG Jung DJ Lehmicke MB Neher DR Rhodes PM Schumacher R0 Wooton.
n- ,e TABLE OF- CONTENTS -
P.aS!!.
1
. 1. . EXECUTIVE
SUMMARY
....................... 1 -1 l A p p ro a c h . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2 The Peach Bottom Plant . . . . . . . . . . . . . . . . . . .
1-3 Accident Sequences Chosen for Study. . . . . . . . . . . .' . 1-3
' Computer Codes Used in the Study .............. 1-4 S umma ry o f Res u l ts . . . . . . . . . . . . . . . . . . . . . . . 1-5 2.- INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2.1 References ........................ 2-3
- 3. GENERAL APPROACH . . . . . . . . . . . . . . ... . . . . . . . . 3-1 3.1 Pl ant Sel ecti on . . . . . . . . . . . . . . . . . . . . . . 3-]'
3.2 Selection of Accident Sequences . . . . . . . . . . . . . . 3-2 3.3 Computer Codes Us. i in the Study ............. 3-2
~
3.3.1 Assumptions .................... 3-3 3.3.2 Uncertainty Considerations . . . . . . . . . . . . . 3-5 3.4 References .........-............... 3-7
- 4. PLANT SELECTION AND ACCIDENT SEQUENCES . . . . . . . . . . . . . 4-1 4.1 General Plant Description . . . . . . . . . . . . . . . . .. 4-1 i 4.2 Selection Basis and General Description of Accident
- Sequences . . . . . . . . . . . . . . . . . . . . . . . . . 4-1
- 4. 2.1 Sequence AE (Loss of Coolant Accident, Failure of ECC System) .................... 4-5
'4.2.2 Sequence TC (Transient, Failure to Scram) ..... 4-6 l 4.2.3 Sequence TW (Transient. Loss of Decay Heat Removal ...................... 4-9 4.3 Containment Failure Mode and Reactor Building Response .. 4-9 i 4.3.1 Containment Failure -- Mode and Pressure Level . . . 4-11 l 4.3.2 Analysis of Reactor Building Response ....... 4-13 l 4.4 References ........................ 4-16 L
- 5. ANALYTICAL METHODS . . . . . . . . . . . . . . . . . . . . . . . 5-1 5.1 Thermal Hydraulic Behavior ................ 5-1 5.1.1 Overall System Thermal Hydraulics: MARCH 2 .... 5-1 5.1. 2 Primary System Thermal Hydraulics: MERGE ..... 5-5 i
ll- ,
4 TABLE-0F CONTENTS y (Continued) .
Page.-
i
[ 5.2 Radionuclide Release from Fuel . . . . . . . . . . . . . .- 5 5.2.1 Source-Within Pressure Vessel: CORS0R , . . . . . . '5-6
,. 5.2.2. Source from Melt-Concrete Interactions: VANESA . . 5-10 5.3 Radionuclide Transport and Depositio . . . . . . . . . . . 5-11 5.3.1 Transport in' Reactor. Coolant System: TRAP-MELT . . 5-11 5.3.2 Transport in Containment: SPARC . . . . . . . . . . . 5-16 5.3.3 Transport in Containment: NAUA 4 .........
- 5-17 5.4 References ..-....................... 5-21
- 6. BASES FOR TRANSPORT CALCUL ATIONS . . . . . . . .- . . . . . . . . 6-1 6.1 Plant Geometry and Thermal Hydraulic Conditions . . . . . . 6-1 6.1.1 Seq uen c e AE - . . . . . . . . . . . . . . . . . . . . 6-1 ~
6.1.2 Sequence TC .................... . .
6-11 6.1. 3 Sequence TW . . . . . . . . . . . . . . . . . . . . 6-27 i
6.2 Radionuclide Sources ................... 6-27 6.2.1 Source Within Pressure Vessel . .. . . . . . . . . .- 6-27 6.2.2 Source Within the Containment ............ 6 6.3 References ........................ 6-40 4
J
- r 7.1 Introduction ....................... 7-1 7.2 Transport and Deposition in Reactor Coolant System
! (RCS) . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1
- 7.2.1 RCS Transport and Deposition for Sequence AE . . . . 7-3 7.2.2 RCS Transport and Deposition for Sequence TC . . . . 7-11 7.2.3 RCS Transport and Deposition for Sequence TW . . . . 7-20 7.3 Transport of Fission Products Through Containment . . . . . 7-27 7.3.1 AEy' Sequence ................... 7-27 7.3.2 TCy' Sequence ................... 7-33 '
- 7.3.3 TCy Sequence . . . . . . . . . . . . . . . . . . . . 7-36 1
7.3.4 TWy' Sequence ................... 7-43 i
7.3.5 Results for Release of Reactor Safety Study G ro u p s . . . . . . . . . . . . . . . . . . . . . . . 7-49 7.4 Discussion ...........'............. 7-52 APPENDIX STAND 8Y GAS TREATMENT SYSTEM OPERATION AND EFFECTIVENESS UNDER SEVERE
. ACCIDENT CONDITIONS: PEACH BOTTOM AND GRAND GULF NUCLEAR STATIONS . A-1
- 11
- , , . . m, ,, , , - - . - = _ . . - , , . - - - . , . _ . - - _ _ . . - - . - - - , , , , . - - - -
m' ,~ Ap -
m
- m. .1. EXECUTIVE SUMARY
/
-This is Volume.2, dealing with the Peach Bottom nuclear power- plant, of aiseven-volume report of work done at Battelle's. Columbus Laboratories to
~
estimate thetamount of radioactive material that could'be: released from light'
~
~ water reactor (LWR) power' plants under specific, hypothetical accident condi-L t' ions. To make these estimates, five power plants were. selected that represent
~
the major. categories of LWRs: .three pressuriz'ed water reactors ~(PWRs)'and two boiling water reactors (BWRs). Specifications and data from these plants,-
~
along with data from laboratory experiments, were input to computer codes design _ed_to describe various conditio'ns prevailing inside-an operating reactor, Ultimately,'these computer codes provide an estimate of how much radioactive-
~
t
- material would be able to escape to the environment ~if a specific series of events-(an " accident sequence") took place.
Volume 2 of this reporteals d with the Peach Bottom 2 Power-station, a General Electric-BWR, Mark I containment design. The specific accident sequences investigated for the Peach Bottom plant were selected to represent
- cases of high risk, severe consequences, and most importantly, a wide range of physical conditions. The computer codes used to analyze the accident sequences-were the best available, including the new MARCH 2 code. Other power plants i included in the study'are Surry PWR (Volumes 1 and 5);' Grand Gulf-SWR (Volume 3);
, Sequoyah PWR (Volume 4); and Zion PWR (Volume 6). The seventh volume will
! address technical questions raised during the peer review of this effort.
- j. The possibility of radioactive material being released to the environ-ment has long been the focus of considerable public concern about the-safety of nuclear power' plants. Since 1962, several major reports have addressed l that concern by using computer codes to estimate the release of fission products
- (radioactive material produced during reactor operations) to the reactor i I
containment building, and thence to the environment, during a hypothetical-l_
severe accident. Although these analyses have-improved over the years, in
- terms of how realistically they describe what happens during a hypothetical j accident, it has not previously been possible to apply the various codes
[ consistently to follow the transport of fission products along their flow path i-i 1
1-1
!I
m trom the core to the environment. This limitation resulted in piecemeal, parametric estimates-of release.
Tne research results reported here are intended to provide a system-atic, sequential application of the codes as well as to present analyses performed with improved computational procedures. It is to be recognized that this report-describes an analytical approach for estimating radionuclide trans-port en'd deposition which incorporates individual physical and chemical processes or mechanisms. This approach is being evaluated for use in predict-ing the amount fission product release (the " source term") to the environment for specific reactors and accident sequences. When verified, these prediction techniques are expected to be more specific and perhaps to supersede generic tables of release fractions provided by previous analyses.
The purpose of this report is then to:
(1) Develop updated release-from-plant fission product source terms for four types'o' nuclear power plants and for accident sequences giving a range of conditions. The estimated source terms are to be based on consistent step-by-step analyses using improved computational tools for predicting radionuclide release from the fuel and radionuclide transport and deposition.
(2) Determine the effects on fission product releases associated with major differences in plant design and accident sequences.
(3) Provide in-plant time- and location-dependent distributions of fission product mass for use in equipment qualification.
Approach This study was conducted by selecting specific plants and accident sequences and then using consistent and improved analyses of fission product release from fuel and radionuclide transport and deposition to predict fission product release to the environment for these specific cases. The approach comprises a sequence of steps; in the combined analysis, the results are speci-fic to a particular set of accident conditions, and each step is based on results from analyses of the previous step.
l 1-2
=
The Peach Bottom Plant Peach Bottom Unit 2 was selected to characterize BWR Mark I designs.
The contain ant of Peach Bottom 2 is typical of BWR Mark I designs: its drywell is s,haped like a light bulb (with the " bulb" pointing down) and is connected to a toroidal wetwell that encircles the " bulb". As in most BWR Mark I designs, the Peach Bottom 2 drywell and wetwell are made of steel, with an internal design pressure of 56 psig (0.38 MPa).
Accident Sequences Chosen for Study The following accident sequences were selected because they repra-sent high-risk situations with potentially severe consequences and because they involve a considerable range in physical conditions:
AE Sequence:
e Large break Loss-of-Coolant Accident (LOCA), Failure of Emer-gency Core Cooling System.
e Break occurs in a recirculation line.
e Suppression pool remains subcooled throughout the accident.
e Containment is assumed to fail by overpressurization from non-condensible gases produced by steam-cladding reactions and core-concrete interactions. .
TC Sequence:
e Transient, failure of control rod insertion (failure to scram).
e Emergency Core Cooling Systems operate.
e Containment failure results from the imbalance in heat genera'-
tion and heat removal due to the continued high power level of the reactor.
TW Sequence:
e Transient, loss of decay heat removal l-3
i T
, ~e Emergency Core _ Cooling Systems operate._ i e Contahmeent failure by overpressurization precedes' core melting o It is assumed that operators will depressurize the primary coolant system before core melting occurs.
Computer Code #Used in the Study-The present efforts built on previous computer modeling work performed at Battelle-Columbus, at Sandia, and in the Federal Republic of Germany, and on experimental and model evaluation studies performed at Oak Ridge, EG&G Idaho, Sandia, and Pacific Northwest Laboratories. In addition to the calculations performed at Battelle-Columbus, calculations of thermal-hydraulic behavior and fission product release related to molten core-concrete interactions were performed by Sandia. Research efforts specifically, directed toward increasing our understanding of fission product release and transport under severe acci-dent conditions are under way at the laboratories listed above, as well as at other research installations around the world. Over the next few years, it is expected that considerable progress will be made in this area. Therefore, this report must be considered as an expression of current knowledge, with the expectation of future validation or modification of the calculated fission product releases.
The first step in analyzing accident sequences was to collect plant design data and perform thermal-hydraulic calculations. Thermal-hydraulic ,
conditions in the reactor over time were estimated with the MARCH 2 code, and detailed thermal-hydraulic conditions for the reactor's primary coolant system were estimated with the MERGE code, developed specifically for this program.
. The time-dependent core temperatures from the MARCH 2, code were used as input to another code developed for this program, CORSOR, which predicts time- and temperature-dependent releases of radionuclides from the fuel inside the reactor pressure vessel. Releases of radionuclides from the interaction of the melted reactor core with the concrete outside the reactor vessel were estimated by Sandia National Laboratories using their computer code, VANESA.
- Using the MARCH / MERGE-predicted thermal-hydraulic conditions and the l
CORSOR-predicted radionuclide release rates as input, a newly developed 1-4
L version of the TRAP-MELT code was used to predict vapor and particulate trans-port in the primary coolant circuit. Transport and deposition of radionuclides i in the containment were calculated using the NAUA-4 code. For flow through'a 1
- pressure suppression pool, scrubbing of radionuclides in particulate form was calculated using the SPARC code.
The calculations performed in this study were of a "best estimate" type.- Whenever possible, input was derived from experimental measurements.
Data employed in these analyses include vapor deposition velocities, aerosol deposition rates, aerosol agglomeration rates, fission product release rates from fuel, particle sizes formed from vaporizing / condensing fuel materials, engineering correlations for heat and mass transfer, and physical properties of various fuel, fission product, and structural materials.
Summary of Results The results for the release of fission products from the fuel during the period of fuel heatup and melting in-vessel and during ex-vessel attack of the concrete in the present study are generally consistent with the results of analyses previously reported in the literature. Essentially all of the ulatile fission products are predicted to,be released from the fuel during initiai core melting in all cases. The release of tellurium was found to have some dependence on the details of the accident scenario, and its predicted total releases are somewhat lower than previous estimates.
The retention 'of fission product aerosols during their transport through the reactor coolant system in the present study was found to be on the order of 10-20 percent of that released from the fuel. Primary system reten-tion of tellurium was found to be quite high,-but since only a fraction of the total tellurium release occurred during the in-vessel phase of the accidents, the effect on the overall tellurium release from the plant was not great. In general, the predicted retention for Peach Bottom is not as high as for some of the PWR sequences considered in the other volumes of this study; the differ-ences in the predicted primary system retention are associated with differences in accident thermal-hydraulics for the two types of designs. The prediction of primary system retention of the released fission products was found to be 1-5
f sensitive to the details of accident thermal-hydraulics although a wide variety of modeling assumptions were not considered for this plant alone. Further, j the ultimate fate of fission products deposited on primary system surfaces is quite uncertain.--Follow-on analyses which include the heating of these surfaces by the decaying fission products may show significant fission product reevolu-tion.
l Fission product aerotol removal by the BWR pressure suppression pool j has been found to be sensitive to the aerosol particle size distribution, the
- nature of the flows through the pool, and the thermodynamic state of the pool..
, For accident sequences such as AE, in which the suppression pool remains i subcooled, quite high pool decontamination factors are predicted. Due to the timing and location of the predicted containment failure in this sequence, however, not all of the released fissin products pass through the pool, thus limiting overall decontamination and resulting in subst'antial releases to the
, environment. In sequences such as TC and TW, all of the melt releases are discharged to the suppression pool, but here the pool is boiling or saturated j at the time of core melting, greatly reducing its effectiveness for fission product retention. Overall approximately one-half of the aerosol releases were predicted to have been retained in the suppression pool for the accident sequences considered here.
, In the accident sequences evaluated here for the Peach Bottom reactor, i the failure of the primary containment was assumed to lead to the failure or
- j. bypass of the secondary containment (reactor building). The potential effec-
} tiveness of the secondary containment and the Standby Gas Treatment System i (SGTS) were explicitly considered for one of the accident sequences. The f potential for retention in the reactor building and removal by the SGTS was
! found to be limited by outleakage from the reactor building, as the gas and I
vapor inputs carrying the fission products exceeded the flow capacity of the
, SGTS. The inclusion of the secondary containment was found to result in only
- l. a fractional reduction in the predicted fission product source term to the
{ environment. The overall fission product releases to the environment in the i present study were generally found to be somewhat lower than those reported in j WASH-1400 although in some cases they were predicted to be somewhat higher.
The primary system retention and removal by saturated pools included in this 1-6
study tend to reduce the predicted releases to the environment. The contain-ment failure pressure and the location of the assumed failure used in the present study tend in some cases to significantly increase the predicted releases compared to the WASH-1400 assumptions.
In viewing the results presented in this report, it should be recog-nzied that uncertainties could be quite large. The prediction of fission '
product release to the environment has been shown to be sensitive to accident thermal-hydraulics, the mechanisms of fission product release and transport, as well as the structural behavior of the containment. It should also be recognized that prediction of the course and consequences of the low probability hypothetical situations considered here is inherently uncertain; at best, the small number of potential accident scenarios considered here can only be representative of a wide spectrum of possible outcomes in the event such acci-dents do take place.
ll 1
1-7 l
1
- 2. INTRODUCTION The possibility of radioactive material being released to the envi-ronment frcm LWRs has long been the impetus for considerable concern and research. Most reactors in the United States were designed, and their sites
-were chosen, on the basis of research report TID-14844.(2 1) Published in 1962, TID-14844 makes certain assumptions about the release of fission pro-ducts (radioactive material produced during reactor operations) to the reactor containment area during a hypothetical severe accident. Although these assump-tions are representative of the state of knowledge at the time, the behavior of fission products has become better understood in the intervening years.
Accordingly, the Nuclear Regulatory Comission conducted the Reactor Safety Study to reassess the accident risks in U.S. commercial nuclear power plants.
The report of that study, known as WASH-1400,(2.2) was published in 1975 are provided a more comprehensive and physically accurate description of fission product behavior. The amount of fission product release (the " source term") .
estimated in WASH-1400 has since been used extensively in planning and evalu-ating reactor operations.
The WASH-1400 source term to the environment for accident sequences has had broad implications for operating LWRs--in licensing, emergency plan-i ning, safety goals, and indemnification policy. However, additional research continued to provide even better methods for estimating fission product release
. and transport. In 1981, the Nuclear Regulatory Comission issued the report
" Technical Bases for Estimating Fission Product Behavior During LWR Accidents",(2.3) a review of the state of knowledge at the time. As part of the Technical Bases report, the assumptions, analytical procedures, and avail-able data were evaluated, and new estimates were made. One advantage of the new estimates was that they took into account the fact that some radioactive material would be deposited inside the reactor primary system during an accident and would therefore not be available to escape to the containment and from there to the environment. On the other hand, because of the limitations of the computer codes available at that t;me, the new estimates could not follow the transport of fission products along their flow path from the core to the environment by applying the various codes consistently. This resulted in piecemeal, parametric estimates of release.
. 2-1 rn- - .--
d T
The research results reported here are intended to provide this
. systematic, sequential application of the codes as well as to present analyses performed with computational procedures improved since the " Technical Bases"
-report. It is to be recognized that in this study, x analytical approach was '
developed for estimating radionuclide transport and deposition which incorpor-ates individual physical and chemical processes or mechanisms. This approach is being evaluated for use in predicting fission product source terms for release to the environment for specific reactors and accident sequences. When verified, predictions made with the approach used here are expected to replace the generic tabular release fractions such as tho;e in Table 6, Appendix V of
- j. WASH-1400, where release fractions are given for broad classes of accidents. - '
The purpose of this report is then to: ,
(1) Develop analytical procedures and use them to predict updated i-release-from-plant fission product source terms for four types of nuclear. power plants and for accident sequences giving a i range of conditions. The estimated source terms are to be based l on consistent step-by-step analyses using improved computational j tools for predicting radionuclide release from the fuel and l radionuclide transport and deposition.
} (2) Determine the effects on fission product releases associated '
, with major differences in plant design and accident sequences.
- (3) Provide in-plant time- and location-dependent distributions of fission product mass for use in equipment qualification.
i It is not necessarily the intent of this work to produce an all-
! encompassing definition of source terms, but rather to make best estimates of j source terms for a range of typical plants and several risk-significant sequences covering a wide range of conditions. These analyses are to be made with the best available techniques, in a consistent manner, following along j release pathways for fission products, and at a level of detail consistent j with current knowledge of pertinent physical processes. Based on state-of-the-art techniques, these best-estimate analyses should provide an indication of the conservatisms inherent in current source term assumptions and guidance for the development of new source terms. The analytical methods and corre-sponding predictions presented here are based on currently available informa-tion and are subject to revision and improvement as better analytical l procedures are developed and as a more extensive experimental base evolves.
2-2
T
-The preparation of this report, therefore, is an evolutionary process which will be carried out over a period of time, with verification and possibly l revision:of the procedures continuing over several years.
-As a part of this evolutionary process, it should be noted that this report is to be revised using improved analytical. procedures and incorporating
< concents and suggestions from participants at a series of NRC sponsored " Peer Review Meetings".
2.1 References (2.1) DiNunno, J. J., et al, " Calculation of Distance Factors for Power and Test Reactors Sites", TID-14844 (March 23, 1962).
'(2.2) " Reactor Safety Study--An' Assessment of Accident Risks in U.S. Com-mercial Nuclear Power Plants", WASH-1400, NUREG-75/014 (October, 1975).
(2.3) " Technical Bases for Estimating Fission Product Behavior During LWR Accidents", NUREG-0772 (June, 1981).
e 2-3 i
F
~ ~ '
- 3. GENERAL APPROACH The general philosophy behind this' study is that mechanistic predic-tions of:radionuclide release'and transport are possible if proper modeling'is performed to represent the physical and chemical processes occurring during LWR accidents. The study, then, represents an attempt to. describe in a reason-ably complete but tractable fashion the processes influencing radionuclide l release to the environment for selected plants and accident conditions.
The objectives of this study originally called for_a consistent analysis of radionuclide. behavior by following fission product transport along flow paths, starting with_ release _into the core region and ending with final release to the environment. To meet these objectives, numerous decisions-and assumptions were required for the analyses: selection of plants and sequences for consideration; choice of analytical tools to be used or upgraded; evalua-tion and incorporation of experimental data; and determination of major physi-cal effects which would be considered on a parametric variation basis to determine the sensitivity of calculations to such variations. Some of the major considerations will be reviewed and discussed in this section.
The general approad in this study was to select specific plants and accident sequences for consideration and then to use consistent and improved analyses of fission product release from fuel, transport, and deposition to predict fission product release to the environment for these specific cases.
The approach consists of a series of steps performed in sequence such that in the combined analysis, the results are specific to an individual set of acci-dent conditions, and each step is based on results from analyses of the pre-vious step.
3.1 plant Selection The first major step in the process was the selection of types of nuclear power plant designs to be considered and a specific plant to represent each type. The types to be considered were: large, dry PWRs; Mark I BWRs; Mark III BWRs; and ice-condenser containment PWR designs. The' specific plants chosen to represerit each type, respectively, are the Surry and Zion, Peach Bottom, Grand Gulf, and Sequoyah plants. These selections were made on a 3-1
combined basis of typicality of design and availability of design details needed for analysis.
3.2 Selection of Accident Sequences A
Accident sequences were chosen for each plant such that significant contributions to risk and a wide range of physical conditions were represented in the analyses. The selected plants and accident sequences are listed below:
PWR: Large Dry Containment PWR: Large Dry BWR: Mark I (Surry-Volumes 1 Containment (Peach Bottom- '
and 5) (Zion-Volume 6) Volume 2)
AB TMLB' TC SD2 SD 2 AE V TW TMLB' PWR: Ice Condenser BWR: Mark III Containment (Grand Gulf- (Sequoyah-
- Volume 3) Volume 4)
! TPI S2HF TQUV T&B' TC TML SE2 The accident sequences for each plant are described in detail in Section 4.2 of the volume of the report dealing with that plant.
3.3 Computer Codes Used in the Study Following the selection of plants and sequences, the required plant design data were collected and thermal-hydraulic analyses performed for each accident sequence. Overall thermal-hydraulic conditions on a time-dependent basis were estimated with the MARCH code,(3.1) and detailed thermal-hydraulic conditions for the primary system were estimated with the MERGE (3.2) code
! developed specifically for this program.
I The time-dependent core temperatures were used as input to ancther code developed for this program, C0RSOR(3.3), which predicts time- and temperature-dependent mass releases of radionuclides from the fuel within the 3-2
i I
i
. pressure vessel. Releases during core-concrete interactions of radionuclides j
= remaining with the melt were'provided by Sandia National Laboratories using their. newly developed model,'VANESA(3.4),.
Using the'MARr.H/ MERGE-predicted thermal-hydraulic conditions and the CORSOR-predicted radionuclide release rates as. input, a newly developed.ver-sion. of- the TRAP-MELT code _ (TRAP-ELT 2)(3.5) was used to predict' vapor and
-particulate transport in the primary coolant circuit.
Retention of aerosols in suppression pools was calculated using the SPARC(3'0) code and transport and-deposition of radionuclides in the containment were calculated using the NAVA-4(3.7) code.
The basic stepwise procedure described above is illustrated in:
Figure 3.1, which shows the relationships _among'the computational models. The calculations were of a "best estimate". type using input derived from experi .
mental measurements whenever possible. Types of data employed in the analyses include vapor deposition velocities, aerosol deposition rates, aerosol agglo- ;
meration rates, fission product release rates from fuel, particle sizes formed from vaporizing / condensing fuel materials, engineering correlations for heat and mass transfer, and. physical properties of various fuel, fission product and structural materials.
3.3.1 Assumptions In preparation for performing calculations of thermal-hydraulic con-ditions and radionuclide transport and deposition, it was necessary to make a t number of assumptions or to select conditions from.among several options.
Major assumptions used in this study of the Peach Bottom plant are listed below
~
in the categories of geometry, thermal hydraulics, and mechanisms.
Geometr_y
! (1) Surfaces within the containment building available for radio-t nuclide deposition include only the major geometrical features of the building.
l (2) There is no attenuation of radionuclides as they pass through l
leak paths-in the containment shell.
- - t i
3-3 i
, . ~ . ,
.-.___.,...i._.--._.-....-...-._.._.--.---._._-.~
_x
.5
..g -.
LSELECTION 0F TYPES.
.OF PLANTS
^
\
. . ,r l
. SELECTION OF
.5PECIFIC PLANTS l'
1 SELECTION OF ACCIDENT SEQUENCES
,e >
SPECIFICATION 0F PLANT INVENTORY GEOMETRY AND ACCIDENT --------
SEQUENCE PHENOMENA ORIGEN o
OVERALL THERMAL HYDRALLICS
- MARCH y u 't PRIMARY SYSTEM RELEASE FROM FUEL THERmt NYORAuuCs ------------
COR$0R MERGE 1f CORE-CONCRETE PRIMARY $YSTEM TRANSPORT [EEWT{m
---__-_-____-_ ; co,c, .
TRAP-MELT E 3r CORE CONCRETE RELEASE
~~~~
CONTAINMENT TRANSPORT AND
- VANESA POOL SCRUSSING
~~~~
7~
AyA 4 AND $PARC :
I 4
4 v
RELEASE TO ENVIRONMENT l
[
FIGURE 3.1. INFORMATION FLOW FOR RELEASE, TRANSPORT AND DEPOSITION CALCULATION 3-4 l-
c" d Thermal Hydraulics.
(3) Before pressure vessel failure, flow from the primary coolant system is restricted to direct leak paths. ]
(4) The. upper plenum geometry is modeled in terms of surface areas, J steel thickness, and compartment heights rather than with exact
. geometries. n (5) Decay heating of surfaces by deposited fission products is neglected in the calculations.
Mechanisms (7) Neither deposition nor resuspension of radionuclides occurs during rector coolant system depressurization at the time of pressure vessel melt-through.
(8) In the long; term (after pressure vessel. failure), deposited radionuclides remain in the primary system indefinitely.
(9) No change-in fission product physical or chemical properties results from radioactive decay.
Some of the above assumptions have been relaxed or changed to accommodate best estimates of conditions and occurrences in specific cases. These are dis--
cussed in greater detail for each plant in Section 6.1 of the volume of the report dealing with that plant.
3.3.2 Uncertainty Considerations The computation of radionuclide release and transport using mechanis-tic models is subject to many uncertainties of various magnitudes and importance.
Quantitative estimates of uncertainties in individual parameters, and hence the overall importance of such uncertainties, has been outside the scope of this study. .Where practical, however, qualitative (and in some cases quantita-tive) estimates of uncertainties have been noted.
Some of the uncertainties in the analyses and procedures can be identified that are currently considered significant. The following is a list of some uncertainties that are believed significant and warrant further evalu-ation through more detailed analyses:
3-5
4 4
l I
(1) The simplified fuel melting model in MARCH (i.e., a single l melting temperature) could bias the predicted release of )
material from overheated fuel, particularly regarding the source of inert and low volatility fission product aerosols.
.(2) The rate coefficients for the release of fission products from overheated fuel are empirical, rather than mechanistically based, and rely largely on scaled, simulant experiments.
(3) The model for the release of fission products and inert materials during the attack of concrete has a very limited experimental basis.
(4) The flow patterns in the reactor coolant system are uncertain.
The adequacy of the simple thermal-hydraulic models used in this study will require experimental verification.
(5) Primary system transport models used in these analyses have r.ot been validated against integral experiments. '
(6) The mode and timing of containment failure in severe' accident sequences can have a major influence on fission product behavior but are subject to large uncertainty.
(7) The calculation methods for water condensation in the contain-ment are based on limited, small-scale experiments and require verification at larger scales.
(8) Deposition velocities for vapor species used in the TRAP-MELT
" calculations were taken as a mid-points in order-of-magnitude ranges o; experimental data. More accurate data would reduce the uncertainty in these parameters and in the resulting rates for deposition by sorption.
3-6
3.4 References (3.1) Wooton, R. O. and Avei, H. I., " MARCH (Meltdown Accident Response Characteristics) Code Description and User's Maniial", NUREG/CR-1711, BRI-2064 (October, 1980).
.(3.2) Freeman-Kelly, R. G. and Jung, R. G., "A User's Guide for MERGE" (February 10,1984).
(3.3) CORSOR Manual.
(3.4) VANESA Manual.
(3.5) TRAP-MELT 2.1 User's Manual.
(3.6) Owczarski, P. C., Postma, A. K., and Schreck, R. I., " Technical Bases and User's Manual for SPARC -- Suppression Pool Aerosol Removal Code",
report to the U.S. NRC, NUREG/CR-3317 (May, 1983).
(3.7) Bunz, H., Koyro, M., and Schock, W., "A Code for Calculating Aerosol Behavior in LWR Core Melt Accidents Code Description and User's Manual".
3-7
- 4. PLANT SELECTION AND ACCIDENT SEQUENCES 4.1 General Plant Description The Peach Bottom 2 reactor was selected here, and in the Reactor -
Safety Study, to characterize the boiling water reactor Mark I plant designs.
The basic BWR Mark I design, referred to as an inverted light bulb and torus, is illustrated in Figure 4.1. In the Peach Bottom unit, the drywell and torus are made of steel, with an internal design pressure of 56 psig (0.38 MPa).
This construction is typical of Mark I designs, although two reactors of this class are reinforced concrete with a steel liner. The reactor coolant system design is referred to as BWR/4. The thermal power is 3293 MW(t) and the net electrical output is 1065 MW(e). The free volume in the Mark I BWR containment varies substantially from the large dry PWR containment. The gas volumes in the drywell and wetwell are 159,000 ft3 (4503 m3) and 119,000 ft3(3370m3),
respectively, in comparison with a volume of 1.8 x 106 ft3 (50,970 m3) in the i
large dry containment (Surry) previously investigated.
The primary containment structure in the Peach Bottom 2 design is enclosed by a reactor building. Leakage from the primary containment may be filtered by the Standby Gas Treatment System (SGTS) in the reactor ouilding before being released to the environment. The response of the reactor building and Standby Gas Treatment System to severe accident conditions, particularly following failure of the primary containment, could significantly affect the consequences of accidents.
4.2 Selection Basis and General Description of Accident Sequences Three accident sequences were selected for analyzing the Mark I design: AE, TW, and TC. (Table 4.1 relates the letters used to identify the accident sequence with the systems that failed during the accident.) The TW and TC sequences were found to dominate the predicted risk in the Reactor Safety Study.(4*l) They have also been identified as important sequences in the AccidentSequenceEvaluationProgram(4.2). Both sequences have similar 4-1 i
j-tWD) y
~
l % T' n
- E
- 1 p E I samma
- 1 i 5 hm 7
/ PREDICTED LOCATI0t,
[
f 7 ,OF DRYWELL FAILURE l,
(j 7 g/ '
REACTOR -
SUILDING p p i
ORY WELL f
V g l l L 'f!$ .--
I
- T
" r, WETWELL i
F suPPRassioN W W L L 900L J d k _J L i
FIGURE 4.1 BWR MARK 1 CONTAINMENT DESIGN
\
42
' e
TABLE 4.1. KEY TO BWR ACCIDENT SEQUENCE SYMBOLS I
A - Rupture of reactor coolant boundary with an equivalent diameter of greater than six inches.
8 - Failure of electric power to engineered safety features.
C - Failure of the reactor protection system.
D - Failure of vapor suppression.
E - Failure of emergency core cooling injection.
F - Failure of emergency core cooling functionability.
G - Failure of containment isolation to limit leakage to less than 100 volume percent per day.
H - Failure of core spray recirculation system.
I - Failure of low pressure recirculation system.
J - Failure of high pressure service water system.
M - Failure of safety / relief valves to open.
P - Failure of safety / relief valves to reclose after opening.
Q - Failure of normal feedwater system to provide core makeup water.
Si - Small pipe break with an equivalent diameter of about 2" - 6" S2 - Smell pipe break with an equivalent diameter of about 1/2"-2".
T - Transient event.
U - Failure of high pressure coolant injection or reactor core isolation cooling system to provida core makeup water.
V - Failure of low pressure emergency core cooling system to provide core makeup water.
W - Failure to remove residual core heat. :
Containment Failure Modes, a = steam explosion in reactor vessel.
8 = steam explosion in containment, y = containment failure due to overpressure; release through reactor building.
Y'a Containment failure due to overpressurei release direct to atmosphere.
4 = containment isolation failure in drywell, c = containment isolation failure in wetwell.
C = containment leakage greater than 2400 volume percent per day.
n = reactor building isolation failure. ,
e = Standby Gas Treatment System failure.
4-3
- i l
behavior in that containment failure precedes core damage, and the pathway for i
release of fission products during core heatup is through safety / relief valves to the suppression pool. To examine a broader range of accident conditions, the AE sequence, which involves a release pathway from the drywell to the sup-pression pool, was also selected for analysis.
In the Reactor Safety Study, analyses indicated that the most likely location of containment failure due to overpressurization would be in the torus regioncontainingthesuppressionpool(seeFigure4.1). Thus, for accident sequences such as TW and TC, in which the containment is predicted to fail before the core begins to heat up, there is more question as to whether water would actually be in the pool when fission products were being released. The torus forming the suppression pool could fail, with resulting spillage of the water, and the pool water would flash after containment failure. Furthermore, part of the ceiling of the room containing the torus is near the periphery of the plant, and if that part of the ceiling failed because of loads induced by rupture of the torus, a direct release pathway would be opened to the environ-ment. The large release fractions of volatile fission products predicted for some accident sequences in the Reactor Safety Study all involved this combina-tion of failures: failure of the torus preceding fuel melting, bypass of the suppression pool, and failure of the reactor building to make a direct pathway to the environment.
More recent analyses of the Mark I containment indicate that the location of failure under high overpressure conditions may be the drywell rather than the wetwell.(4.3) Failure was predicted at approximately 117 psig (0.8 MFa) at the knuckle where the spherical lower region of the drywell joins the upper section (see Figure 4.1). It should be noted, however, that the criteria used to define failure in the later study were not the same as that for the WASH-1400 analyses. The potential for fission product retention in the reactor building after release from the primary containment, as well as the potential effectiveness of the SGTS, is very sensitive to the location and mode of primary containment failure. At one extreme, the reactor building could fail, rendering the SGTS ineffective; on the other extreme, a relatively small opening in the primary containment may limit gas flows into the reactor l
building to rates that can be completely accomodated by the SGTS.
l 44
4.2.1 Sequence AE (Loss of Coolant Accident.
Failure of ECC S_vstem)
In this accident sequence, a break occurs in a recirculation line.
The reactor blows down into the drywell, which relieves through vertical vents into the suppression pool. TheECCsystems(spraysandflooding)areassumed not to operate. Following blowdown, residual water remains in the lower plenum of the reactor, but the core is dry. During heatup and melting, gases flow from the core, through the steam separators, down the outer annulus, out the intake of the broken loop, and out the break. This flow is driven by gas expansion and by generation of steam from thermal radiation'in the core. Gas 1 flows from the drywell through vents into the pool. The pool remains subcooled throughout the accident.
The buildup of noncondensible gases from the reaction of the cladding with steam and from molten core-concrete interaction is predicted to fail the containment by overpressurization. The predicted consequences could be senst-tive to both the timing and the location of the primary contsinment failure.
In the Reactor Safety Study, the failure was assumed to be in the wetwell, and failure was predicted to take place after reactor vessel molt-through. Thus the AE sequence in the Reactor Safety Study analyses included significant scrub-bing of the melt release. However, current analyses, based on more recent but lower containment failure pressure, indicate that containment failure could take place before reactor vessel melt-through. Also, if the containment fails in the drywell, much of the melt release would not pass through the suppres-sion pool in this sequence.
After the core melts through the lower
- head of the vessel and begins to attack the concrete, fission products are released to the drywell as the result of sparging of the melt by gases released from the concrete. In addf-tion, fission products could continue to be released to the drywell from fuel remaining in the reactor vessel or from reevolution of fission products from surfaces inside the vessel. If the primary containment falls in the drywell prior to pressure vessel failure, the pathway for the release of fission '
products would not include the suppression pool during this phase of the acci-dent.
45
Figures 4.2 and 4.3 illustrate the flow paths for release to the environment during the different phases of the accident with containment failure preceding (Figure 4.2)orfollowing(Figure 4.3)pressurevesselmelt-through.
In each case, the failure in primary containment was assumed to occur in the drywell. If failure were to occur in the wetwell, the flow pathway would look like Phase 1 and Phase 2 in Figure 4.3 but would continue from the wetwell with branches to the environment and reactor building following failure of the primary containment. The cases analyzed in this report involved the pathways indicated in Figure 4.2.
M.2 Sequence TC (Transient. Failure to Scram)
In this transient, the control rods fail to insert and shut duwn the reactor, but the emergency core cooling systems function. The reactor power equilibrates at a level that is balanced by the coolant makeup rate. Heat is removed from the primary system by releasing steam to the suppression pool.
Since the power level of the reactor (approximately 30 percent) exceeds the heat removal capability for cooling the pool, the temperature of the pool rises, and the pressure in the containment increases to the failure level. After the containment fails, the system depressurizes, the suppression pool boils, and the makeup pumps stop delivering coolant to the vessel. As the core heats and melts, fission products flow with the gases through the steam separators. The flow splits, with a major fraction (85 percent) going through the steam dryers and the balance (15 percent) bypassing the dryers through the outer annulus.
The split flows merge at the steam line and pass through relief lines to the suppression pool. With an assumed failure in the drywell, some suppression pool water would flash at the time of primary enntainment failure. Unless the forces generated by depressurization resulted in structural failure in the torus, some water would be in the suppression pool during fission product releasel however, this water would either be boiling nr Saturated. From the top of the suppression pool, gases and entrained aerosols would be transported through vacuum breakers back into the drywell before leaking thruugh the break in containment to the reactor building or directly to the environment.
46
Phase 1. Up to Containment Failure Helt Release Steam Lower Suppression Wetwell
'" Separators Drywell Annulus Pool Vapor'
_ Phase 2. Containment Failure to Vessel Failure Wetwell Vapor Melt Release Environment
- Steam Lower Drywell Reactor Core Building 4 Separators Annulus Environment Environment Phase 3. After Vessel Failure Environment Vaporization Release Cavity Drywell toF -
Build {ng Evolution from RCS -
SGTS Environment RCS FIGURE 4.2. FLOW PATH FOR FISSION PRODUCT RELEASE IN SEQUENCE AE MIEN CONTAI E NT FAILURE PRECEDES VESSEL FAILURE e
Phase 1. Up to Vassal Pznetration Melt Release a Steam Lower Suppression Wetwell Core Drywel1 pg,j y, pop.
j Separators Annulus
/
- Phase 2. Vessel Penetration to Containment Failure Vaporiz'at on Release
. . Cavity Drywell Suppression Pool Wetwell Vapor i i --
Evolution from RCS RCS
?
=
Phase 3. After Containment Failure Environment Wetwell Vapor Drywell g Reactor Building -
n ronment Vaporization Release L. -- Environment i Cavity Evolution from RCS RCS FIGURE 4.3. FLOW PATH FOR FISSION PRODUCT RELEASE IN SEQUENCE AE WHEN CONTAlff4ENT FAILURE FOLLOWS VESSEL PEMETRATION
1 I
Following reactor _ vessel melt-through, the airborne fission products in the vessel flow directly to the containment. Fission products released from RCS surfaces and from attack of the concrete by the molten core also contribute to the source term to the containment and, through the break in containment, to the environment.
The flow paths for fission product transport during'different phases of the accident in the TC sequence are illustrated in Figure 4.4. If the containment were to fail in the wetwell rather than the drywell, the pathway.
through the drywell would be bypassed during Phase 1 (up to vessel penetration) and the flow would also pass ints the wetwell regions from the drywell during Phase 2 (after vessel penetration), Under these conditions, however, the path-way might not pass through a water pool.
4.2.3 Sequence TW (Transient, Loss of Decay Heat Removal)
In this sequence the reactor shuts down and the emergency core cooling systems operate, but the suppression pool heat removal system fails, so that the suppression pool overheats and the containment pressure ri;?s. As in the TC sequence, the containment fails prior to core melting, but since the core is at decay heat power level, the time to failure is substantially longer.
The flow paths to the environment are the same as described for the TC sequence.
In the analysis of sequence TW, it was assumed that the operators would depres-surize the reactor coolant system prior to core meltdown.
4.3 Containment Failure Mode and Reactor Building Response In the Mark I containment design, the steel walls of the drywell and wetwell provide the.high pressure leak-tight boundary that assures the contain-ment of the pressure surge and confinement of fission products released in the design basis accidents. The reactor building that surrounds the primary containment also has some potential for the attenuation of the fission product source term to the environment in an accident. This building is not designed to withstand high internal pressures. It does, however, have a filtered exhaust 4-9
Phase 1. Up to Vessel Penetration (Containment failed prior to core melt).
Melt Release Environment I
Steam Steam Relief Suppression - Wetwell Reactor ~
Core - ~~ -" -
Vapor Dr m 11 T Building Separators Dryers Line Pool -
SGTS Environment I __
L -- -Environment *
. Up r ,
Annulus ,
Phase 2. After Vessel penetration.
Vaporization Release and RCS Evolution y, Environment O
Cavity Drywell -r I B 11 ding i
I SGTS Environment L _ _ _ Environment.
- Global failure of RB.
FIGURE 4.4. FLOW PATH FOR FISSION PRODUCT RELEASE IN SEQUENCES TC AND TW 4
system called the Standby Gas Treatment System which would be effective in controlling limited leakages from the primary containment. Following contain-ment failure, the survival of the reactor building and the subsequent effec-tiveness of the reactor building in attenuating the source term is uncertain.
If the primary containment depressurizes from high pressure (e.g., 132 psia), l the resulting pressure in the lower portion of the reactor building would exceed the level at which blowout panels would fail internally and relieve to the refueling floor. If the reactor building were to remain intact following the blowdown, the Standby Gas Treatment System would be available in limiting direct outleakage through the walls of the building, leading to plateout in the building, retention in the charcoal and HEPA filters, and an elevated release. In addition, if the fire protection sprinkler system in the lower portion of the reactor building were activated by the hot steam, washout of fission products could also occur. Thus, a number of mechanisms could be effective in the reactor building, if the building is able to withstand the loads induced by primary containment failure. The expected performance of the SGTS under accident conditions is described in Appendix A.
4.3.1 Conta'nment Failure --
Mode and Pressure Level In the Reactor Safety Study, nonlinear elastic analyses were performed with the MONSAS code for the Peach Bottom primary containment design. The analyses indicated several potential failure locations; these included the inner diameter of the toroidal suppression chamber, the knuckle region between the cylindrical and spherical portions of the drywell, the expansion joints on the vents between the drywell and the suppression pool, and the drywell supports.
Reaction forces from surrounding structures or attached equipment were not l included in the analysis. Some of the design details were not available for l the analyses and were estimated. Burst pressures of about 250 psig were indicated based on the development of the ultimiate strength of the material.
The actual failure pressure that was adopted was 175 + 25 psia, with the most likely location of failure being the inner diameter of the toroidal suppression 4-11 a'
h ,
~-
chamb'er.s:The: uncertainties.in-the pressure at which aailure was predicted and the location'of failure were therefore-large.
- More recently, analyses;.were-performed at Ames Laboratory for a number' of steel containment designs.(4.3)! In this study, the' predicted failure pres-sure for the Browns Ferry Mark I. design was 117 psig with failure predicted to occur at the intersection of the. cylindrical and hemispherical segments of the
-drywell.~ The computer code ANSYS was used in the analysis. Failure was assumed
-to occur at two times the yield strain.- If consideration is given to strength
~
beyond yield and plastic strain, it is likely that the actual failure would be at higher pressure and might occur at a-different_ location. The importance of the assumed. failure criterion should be recognized. The' failure pressure 4
developed in the above study and that developed in the Reactor Safety Study.-
4 are based on different failure criteria.
The analyses.descrited above examined the structural failure of the wall material under.high internal pressure. Other failure mechanisms must also be considered such as failure of penetration. seals subjected to.high temperature, failure of-penetrations under pressure, blowout of hatch seals, F
and failure of valves. These mechanisms have only recently received attention' '
- by the NRC for conditions exceeding design levels. A number of related research programs are currently being performed at Sandia National Laboratories.(4 4)
The mode and location of containment failure can have major effects on the predicted subsequent release of fission products to the environment.
In the Reactor Safety. Study, it was assumed that failure of the containment in the suppression p'ool region would result in displacement of the pool water'and that no subsequent scrubbing of fission products by the pool would occur.-
l- After release into the suppression pool room, it was assumed that fission products would either be transported through subcompartments of the reactor building, through the blowout panels and to the environment, or would be released directly to the environment, if the walls were to fail in a quadrant of the room with direct access to the outside. In WASH-1400, the failure mode of the first type was referred to as and the second type was referred to as y'. 'The most severe releases predicted were all of the latter type (e.g., TC-y' and TW-y'). In these sequences, containment failure preceded core meltdown,
-no scrubbing of fission. products by.the remaining saturated water was assumed
. 4-12 1
._..,,a -m-.. , , , . , - . ,.- . . ~ . - , - - - , . , - . . , , , - , . - , c , ,
[-C k-to occur,' and the fission products were released directly to the environment from the suppression pool room.
4.3.2 Analysis of Reactor Buil' ding Response l In order to evaluate the likelihood ~of different levels of performance of the reactor building following containment failure, a number of analyses were performed with the MARCH 2 code. Specifically, analyses were made to determine under what set of conditions the reactor building would:
(1) Remain intact -- where the SGTS is effective;in keeping the reactor building pressure below the point at which the refueling floor blowout ~panals would fail.
(2) Retain long-term integrity -- in which blowout panels relieve to the refueling floor and from the refueling floor to the' atmosphere, but continued operation of the SGTS results in leakage to both the refueling floor and the lower portion of the' reactor building, mixing and deposition of fission products, some filtering of fission products, and an elevated release.
(3) Undergo ma.ior structural failure -- in which it would be questionable that any retention mechanisms should be considered effective.
The principal variable changed in the analyses was the hole size at ,
the time of containment failure. For the TW and TC sequences, in which the potential for retention in the reactor building is particularly important, a i
minimum leak rate can be identified for the primary containment following fail-ure which corresponds to the steam production rate. If the leak rate were smaller than this value, the pressure would rise until the laakage area increased to match this rate. For TW and TC, these areas are approximately 1/4 ft 2 and 5 ft2, respectively. In either of these cases, the steam leakage i
rate into the reactor building is greater than the capacity of SGTS and, at a minimium, both sets of blowout panels would be expected to fail prior to core meltdown.
There is very little basis for the prediction of the hole size in the primary containment structure following failure. In a steel containment i
4-13 l
structure, however, if a region of the shell is loaded to near the ultimate strength or undergoes large_ deformations before failure, a very large leakage area can be expected following failure. If a localized weak element exists such as failure of a hatch seal at lower-pressure, the subsequent leak area would be limited.
For the purposes of examining the response of the reactor building to_a large leak area following failure, the AE sequence was analyzed with a 10 ft2 hole. In the analysis, a leakage in the lower portion of the reactor building associated with the operation of two trains of the SGTS was employed up to a pressure of 15.047 psia (50 lb/ft2 differential) at which point the opening was increased to 300 ft2 to correspond to failure of the blowout panels.
A peak pressure of 16.0 psia (1.3 psig) was obtained at approximately 20 seconds following failure.
In the TC sequence, a peak pressure of 16.4 psia (1.7 psig) lasting a longer time was seen, since the primary containment pressure did not decrease as rapidly due to flashing of the suppression pool water. The above calculated maximum pressure is based on the use of the entire volume of the lower portion of the reactor building and does not account for pressure differentials that would exist between subcompartments. The design pressure of the lower portion of the containment is 3 psig to withstand tornado loads. Thus, if leakage from the annulus is confined to the lower portion of the reactor building and the flow areas between subcompartments are large enough, the lower portion of
- the containment would relieve through blowout panels to the refueling floor area, but the structure of the lower part of the building would remain intact.
Within the uncertainties associated with the mode of failure of the primary containment and the subsequent blowdown in the reactor building, we have concluded that a significant potential exists for a direct leak path into the refueling floor area of the building (e.g., by lifting the plug above the primary containment) or for exceeding the pressure capability of a subcom-partment of the lower portion of the reactor building (e.g., the torus room that is in direct communication with the annulus). Either of these modes of failure could lead to a direct pathway to the environment with very limited potential for additional fission product retention. Resolution of this issue 4-14
i is'beyond the. scope of the current effort. Both scenarios will be examined to determine the effect on accident source terms.
In sequences such as TC and TW, failure of the primary containment is followed by the release of large amounts of steam. The steam release would continue.following initial containment depressurization due to the flashing of the suppression pool and continued boiloff from the primary system. If the reactor building were to survive primary containment failure, the SGTS would be subjected to this.centinued steam input; the ability of the SGTS to handle a large steam throughput without losing its effectiveness is not clear.
A separate consideration involved the possibility of a major hydrogen deflagration event in the reactor-building following core meltdown. In the AE accident sequence, flammable mixtures of hydrogen and air are predicted to exist in the building shortly after containment failure which would produce pressures in the neighborhood of 50 psig if ignited. In the TC and TW sequences, air would be depleted in the reactor building following blowdown of the steam in the primary containment. Whether the fire protection spray system operates or not could have a major influence on the amount of air available to mix with
~
hydrogen and the potential for steam inerting preventing deflagration.
To summarize the above discussion of reactor building response, it is to be concluded that the likelihood of severe structural damage in the reactor building as the result of blowdown loads appears to be significant, and if the building survives blowdown of the primary containment, subsequent failure as the result of hydrogen deflagration is also possible. Because of the expected-importance of building integrity on source term attenuation and within the uncertainties in our current understanding of the involved phenomena, sequences in which there would be little attenuation of the source term within the reactor building appear to be credible. The quantification of this effect
, follows from calculations discussed in the remainder of this volume.
I \
l 4-15 l
References l
(4.1) " Reactor Safety Study: An Assessment of Accident Risks in U.S. '
Commercial Nuclear Power Plants", WASH-1400 (1975).
(4.2) Kolaczkowski, A. M., et al, " Interim Report on Accident Sequence Likelihood Reassessment (Accident Sequence Evaluation Program), Draft (February,1983).
(4.3) Greimann, L. G., et al, " Reliability Analysis of Steel Containment .
Strength", NUREG/CR-2442 (June, 1982).
(4.4) Larkins, J. T. and Cunningham, M. A., " Nuclear Power Plant Severe Accident Research Plan", NUREG-0900 (January, 1983).
i i
4-16
- 5. ANALYTICAL METHODS This section describes the analytical methods used in assessing the source term to the environment for the Peach Bottom plant, a BWR Mark I design.
The methods employed here differ significantly from those used to analyze the Surry plant, as described in Volume 1 of this report (5.1) The Surry plant has been reevaluated-using these revised methods; the results of the Surry reevaluation are reported in Volume 5(5.2) ,
The first major difference between the methods used here and those described in Volume 1 is that the MARCH 2 code is used here for the overall thermal-hydraulic calculations, replacing the MARCH 1.1 code used for Volume
- 1. MARCH 2 incorporates a number of improvements in the treatment of accident thermal-hydraulics over the earlier version of the coae. The second major change is that in the present analysis, an accounting is made for the effect of Zircaloy cladding oxidation on the release rate for tellurium. Finally, a new code called SDARC is used here to predict the retention of aerosols in suppression pools.
Some other, less significant changes have been made as well. These are discussed in the text.
5.1 Thermal Hydraulic Behavior This section describes the computer code MARCH 2, which, along with the MERGE code, was used to analyze the thermal-hydraulic response of the reactor core, the primary coolant system, and the containment system for the I
selected accident sequences.
l 5.1.1 Overall System Thermal
! Hydraulics: MARCH 2 The MARCH 2(5.3) (Meltdown Accident R_esponse giaracteristics) computer
! code describes the physical processes involved in severe fuel-damage accidents l in light water reactors. Version 2 of the code replaces Version 1.1.(5.4) The differences between the two versions include changes in models, code struc-ture, and programing language. The new models in MARCH 2 were developed at a 5-1
Y
~ ~
number of institutions, including Battelle, Sandia National Laboratories, Oak
. Ridge National Laboratories, Brookhaven National Laboratory, and the_ Tennessee Valley Authority. In many cases, these model:. are provided as options to existing models. The changes'in MARCH were largely undertaken to address recog-nized deficiencies in the early version related to modeling approximations, time-step control, and transportability of the code to other installations.
The ERCH 2 code was developed primarily for use in probabilistic
, risk' assessment. The uncertainties in many of the MRCH 2 models are large, and in many cases the extent-to which the models have been validated against experiments is limited. More mechanistic codes are being developed by the NRC, such as SCDAP and HECTR, but they were not available for use in this program.
The MARCH 2 code examines the behavior of a large variety of accident processes including depressurization or leakage from the reactor coolant system, core uncovery, core heatup, oxidation of Zircaloy cladding, fuel melting, fuel slumping, fuel-coolant interaction in the lower vessel head, vessel head fail-ure, fuel-coolant interaction in the reactor cavity, debris bed coolability, core-concrete interactions, production of combustible gases, gas combustion in the containment, containment heat transfer, intercompartment flows, and the effect of engineered safety features on containment thermal hydraulic behavior.
- Some of the principal modeling improvements in Version 2 of the MARCH code are described below.
5.1.1.1 Containment Response. The containment response modeling in MARCH 2 includes the following principal changes: provision for expanded blow-down input via subroutine INITIAL, the ability to accept two input terms from the primary system, completely revised treatment of burning of combustibles, addition of a heat sink for radiation heat transfer from the debris in the reactor cavity, and removal of a number of restrictions in the earlier code.
The expanded blowdown table input capability is intended to facili-tate the interfacing of the MARCH code with more detailed thermal-hydraulic codes that may be used to describe the initial portion of the accident sequence.
The containment response subroutine, MACE, has been changed to accom- ~
modate simultaneous break and relief / safety valve flows from the primary system.
The two inputs can be directed to different compartments if desired, e.g., J l 5-2
]
es.
- ,..n, .v.,-. n . - . . . , , , , - . , , - . - - n , - . . - - , , n ._ l
I.
breas f low to . thel drywelliand relief / safety valve flow to the suppression pool of a~BWR.'
5.1.1.2 Primary System Response. The MARCH 2 treatment of the primary system includes both improvements in the treatment of initial (early)
!- primary system response and the addition of several phenomenological models to treat the processes following core collapse into the bottom head. Included are changes in the steam generator model to remove some of the restrictions and limitations of the earlier version, improved break flow models, changes in the flashing model in response to primary system pressure changes, provisions 4 for simultaneous break and relief / safety valve flow, changes in the treatment of heat transfer to structures, and consideration of the transport of fission products within the primary system.
5.1.1.3 Water and Steam Properties. The representation of the properties of water and steam has been improved in MARCH 2. This has included
. expansion of the property tables and correlations incorporated in the code as well as inclusion of additional properties required by the new phenomenological I models. The input parameters are based on the ASME steam tables.
L 5.1.1.4 Decay Heat. MARCH 2 incorporates the current American I National Standard (5.5) for evaluating fission product decay heating as a func-tion of time after shutdown and time at power, including the contributions I
from heavy element decay. This replaces the earlier, simplified version incor-porated.in MARCH 1.1. Alternatively, decay heat as a function of time may be.
input in tabular form; this approach would be particularly appropriate for transients with. failure to scram, where the power history would be provided by more detailed system codes.
5.1.1.5 Core Heat Transfer. MARCH 2 retains the basic model of the l core as developed for the earlier version, but incorporates additional models for a more detailed treatment of heat transfer processes. Heat transfer between the fuel rods and the steam-hydrogen gas mixture is now calculated using either the full Dittus-Boelter correlation (5.6) for turbulent flow or a laminar flow correlation.. A subroutine has also been added to approximate axial conduction 5-3
. heat transfer.in the fuel rods using the Fourier law of heat conduction and the BOIL-calculated node temperatures. The effect of axial and radial thermal
' radiation heat transfer'within the core, as well as between the core and sur-rounding structures and water surfaces, can now be calculated. The heatup of-the core support barrel by thermal radiation is included. Additional changes include corrections in the heat transfer analysis of partially covered core nodes and improvements in the metal-water reaction model.
5.1.1.6 Core Debris. A number of phenomenological models have been added for the treatment of the core debris in the reactor vessel bottom head.
These include a flat plate critical heat flux model, a fragmented debris-to-water heat transfer correlation, and several options that consider formation of debris beds within the vessel head while water is still in the vessel. The bottom head heatup model utilizes a calculated heat transfer coefficient between the molten debris and the vessel head.
A major area of concern and controversy in the analysis of core melt-down accidents has been the behavior of core and structural debris upon contact with water in the reactor cavity. The highly simplified .<,dels of MARCH 1.1 have been supplemented with a flat plate critical heat flux model, a particulate heat transfer model with more mechanistic heat transfer coefficients, and several debris bed heat transfer correlations. 'If desired, the switchover from one model to another can be based on calculated conditions, e.g., debris temperature. The production of hydrogen from steel-water reactions has been incorporated into these models in addition to the zirconium-water reaction previously available. Also included are the heating of the evolved gases by the debris beds and the effect of hydrogen flow on bed floodings.
A heat sink has been provided for the thermal radiation from the top of the core debris as calculated by the INTER subroutine. The decomposition of concrete due to radiated heat flux is treated by an ablation-type model with the resulting gases added to the containment atmosphere. Also, the geometry of the corium-concrete mixture is fixed following solidification of the melt.
5.1.1.7 Burning in Containment. The treatment of combustible gases now includes consideration of the burning of hydrogen and carbon monoxide if 5-4 i
- a Jtheiriconcentrations exceed flammability-limits'. Included are explicit
- . considerations:of inerting'due to high: steam concentrationsLand oxygen 'deple--
L ition, direction-dependent compositions for flame propagation.between compart-
<ments, and burn velocities'astfunctions of composition. .Various options are-available to explore the effects of assumptions-about the' burning-of hydrogen and' carbon monoxide.-
5.1.2- Primary System Thermal L Hydraulics: - ERGE-l When the MERGE (5.7) code was written, the existing computer codes
[ describing the thermal-hydraulic behavior of a core meltidown accident were not i capable of analyzing the flow and temperatures in the individual volumes of the reactor coolant system downstream of the core in the pathway for release-l to the containment. The report " Technical Bases for Estimating Fission Product 2' '
BehaviorDuringLWRAccidents"(5.8),. published by the NRC in 1981, indicated ;
that in at least~some accident sequences, the retention of fission products in the reactor coolant system (RCS) could be significant. To support more realis-l tic analyses of fission product retention.with the TRAP-ELT code discussed in
- j. ;Section 5.3.1, an effort was undertaken to write a simple stand-alone code, MERGE, to predict gas temperature surface temperature, and flow within the i reactor coolant system.
{' MERGE calculations are based on the output of MARCH,- and the output
{ of MERGE is input.to the TRAP-MELT. code. The MARCH results used by MERGE are:
i; the primary system pressure, the flow rate of hydrogen leaving the core, the
[ flow rate of. steam leaving the core, and the average temperature of gases
! leaving the core. The MERGE analysis accounts for conservation of energy and ,
! conservation of mass by. species. It is assumed that the gases within a volume-are well mixed and have the same temperature, and that the pressure differen-f tial between volumes is negligible.-
In MERGE, the equations are solved with an explicit time difference l
! scheme. At a particular time step, conditions within the first volume down-I . stream of the core are calculated first, and the solution proceeds from each I
';olume to the next downstream volume. Knowing the initial and inlet-flow conditions for each volume, MERGE. solves for the value of.the outlet flow from 5-5 I'
h
the'. volume that yields the known pressure. Heat transfer from flowing gas.to l structures is accounted for. Forced laminar, forced turbulent, and natural
-convection heat transfer coefficients are utilized as appropriate,' with a radiative term added to the coefficient. In addition, the E RGE-calculated radiation heating of the first structure is calculated based on a MARCH-
. calculated radiative flux.
The MERGE code involves certain approximations and limitations. In the ERGE analysis, the flow of gases in the upper plenum is assumed to be one-dimensional; in reality, circulation patterns could probably be' established in this region due to the strong temperature gradients. Whether a more detailed analysis is required for this region must be determined by the results of sensitivity studies with the TRAP-ELT code. The need for validation experi-ments must also be evaluated.
l 5.2 Radionuclide Release from Fuel 5.2.1 Source Within Pressure Vessel: COR50R CORSOR(5.9) is a simple correlative code which estimates aerosol and fission product release rates from the core during the period of core melting in a light water reactor. Quantifying the aerosol and fission product release from the core region is an important first step in determining the radionuclide source term to the containment during a hypothetical severe core-damage acci-dent. The timing of the release of various materials influences their reten- l tion in the teactor coolant system because it determines which species emanat-
] ing from the core will be able to interact. The timing also determines the residence time of the released materials and the temperatures in the reactor coolant system, since these are both dynamic parameters. Simplistic source ter:ns, such as constant or linearly increasing release rates with concurrent releases for all radionuclides, may therefore lead to unrealistic estimates of-radionuclide transport behavior.
For the present analysis, the core has been divided into 240 nodes, l 10 radial and 24 axial, which have distinct temperatures as predicted by MARCH.
l The core inventory, determined from the program ORIGEN(5.10) , has been divided 1
5-6 l
s l
l
n l
- - equally'among the' nodes.- In~an actual reactor, the distribution would vary.
' both axially and radially-and.would change with time. Typically... fuel.is
- shifted between three radial zones during its. irradiation history...To flatten ithe power distribution across the core,_the freshest fuel is placed in the
?outside zone of the core and the most highly burned-up fuel is placed in'the-
' central region.' Thus, .an abrupt ' change.'in the' spatial . distribution of radio- ,
nuclides occurs.at the time lof refueling but then continues to shift during
[ Ithe cycle as the fissile inventory is preferentially depleted in the. regions of higher flux.
Alternative distributions.of~ fission-products can be use'd in the CORSOR program, and the effect on fission product release rates of the " flat-flux"' assumption can b'e quantitatively assessed by examining.the results of-parametric studies-such as -those described in Appendix B of Volume 1(5.1) ,.
Uncertainties in the release-rate coefficients are expected to have a more-I significant effect on reicase rates than will the assumptions regarding fission product distribution among core regions.. .
n
-Temperatures at each of the nodes are obtained from the MARCH code l
for each of a number of time steps, beginning at the start of the accident and continuing to a user-specified time. An average temperature is-computed over each time span during core heatup and melting, and if the temperature is less than 900 C for any node, no release will occur from that node. The average temperature for failure of the cladding of a fuel rod is taken to be 900 C.(5.8) 1 The sensitivity of C0RSOR release estimates to the temperature set for cladding failure was also discussed in Appendix B of Volume 1.(5.1) When any. axial l
position in a fuel bundle achieves a temperature of 900 C, CORSOR calculates a l
i- gap release of certain volatile fission products for all fuel rods in that
! radial zone. This is intended to simulate the gap release accompanying the bursting of individual fuel rods. This release occurs because certain fission l
products accumulate in the fuel-cladding gap because of migration within the I
( fuel. The amount of the gap release is taken to be 5 percent of the initial I
amount present for cesium, 1.7 percent for iodine, 3 percent for the noble fission gases, 0.01 percent for tellurium and antimony, and 0.0001 percent for
! barium and strontium. Since this emission is very small in comparison with l
the melt release, and is concurrent with the melt release, it is not treated separately in any of the transport analyses. Clearly, the gap release would I 5-7 :
1 i
require more careful analysis-if less severe hypothetical accident conditions l
were considered.-
Subsequent mass release as~the nodes progress toward melting is cal-
'culated on a nodal basis as the product of the amount of each species remaining, the release rate coefficient, and.the time interval of integration. The mass released is then summed over all the nodes in the core for each species to give the total mass released.during the time step. It should be noted that the MARCH code predictions for core temperatures do not take into account the heat of vaporization of materials released from the cere.
The computation of the fractional release rate coefficients for fis-sion products is based on empirical correlations derived from experiments performed by Lorenz,-Parker, Albrecht, and others.(5.ll-5.17) The data from these experiments were graphed and curves developed for the releases. -A frac-
, tional release rate coefficient, K(T), is derived for species by fitting an equation of the form K(T) = AeBT to each of these curves. The resulting values of A and B for three different temperature regions of the graph are Dasically the same as those defined in Appendix B of the " Technical Bases Report"I * )'but have, in many cases, been adjusted to account for updated evaluations.(5.18) It should be noted that the fractional release rate is a function of temperature and elemental species only, and any effects of pressure and specific surface area of the melt on the release rate are not considered. Additionally, details of complex phase inter-actions of various components within the melting core are, for the most part, not known quantitatively; hence the release rates are valid only to the extent that the experiments upon which the release rates are based adequately modeled a core meltdown situation.
The release rate coefficients used in C0RSOR are the same as those used in Volume I of this cort, with the following exception. Tellurium release from the fuel elements appears to be strongly dependent on the extent of oxidation of the zirconium cladding. At this time the effect of zirconium oxidation on the tellurium release rate is not well quantified, nor is it known with any certainty whether ttlis phenomenon is exhibited in the release rates of other metals' releases. In an attempt to factor into these the inhibition 5-8
r of tellurium' release caused by the presence of un_ oxidized zirconium, the j following. sets of. release rate coefficients (5.19) were employed for calculation f
ofithe fractio'nal release rste for tellurium according to the usual equation, K(T) =~AeBT.
Zirconium 0xidation: <90% _ 290%
A B A B ;
T < < 1600 'C 1.65 E-11 0.01061 6.50 E-10 0.01061 T > 1600 *C 9.04 E-8 0.00522 3.62 E-6 0.00522
' o e rn a ut e ea r ten co e ma s Several uncertainties associated with the CORSOR predictions must be-2 mentioned. These uncertainties most strongly impact the predicted release-7 rates-for aerosols, rather than for the more volatile materials. One difficulty in predicting aerosol release'is that as core melting progresses,-the tempera-tures increase throughout the core until, eventually, a loss of-geometry would be expected to occur. In the BWR analyses, core slumping occurs in such a way ,
4 as to remove radial regions from the core in an incremental fashion. Thus, the emission of fission products from the fuel rods which have fallen to the lower structures is included in the calculations of the source to the primary system. This represents a change from the PWR analysis reported in Volume 1, i in which emission into the primary system was halted at the time of core slump-ing. A further difference.between the two sets of analyses is that for the BWR sequences, significant periods of time elapse between vessel dryout and bo'ttom head failure. Thus, during a portion of the melt period, the core is
- emitting fission products into an essentially stagnant volume. !
The behavior of the control rods during core melting is also a source of uncertainty with respect to aerosol generation. In the sequences modeled
{ here, the rods are fully inserted into the core, and it is assumed that these rods are at the same temperatures as the core nodes in which they reside. Thus the release of control rod materials is simulated in CORSOR by the addition of the tin and steel to the inventory of materials available for release. The
, burnable poison rods are not considered as a source of aerosol material though 4 it is understood that the boron in them may play a role in aerosol formation.
5-9 4
, -,-,-n -,,.,-,-------,,..--en, .-,----..---,.-,...--.-~,...-,-,.-,.r--------,n~.-,,,.--,.,>-n-.-+ . . . - .
. ,n---... . u, - , - , .
. -5.2.2 Source from Melt-Concre'te--
Interactions: VANESA The release'of' fission products'and nonradi.oactive aerosols.during the interaction of molten cote materials with concrete plays an important role in determining the risk of severe reactor accidents and is modeled with the
- VANESA code. Aerosol.' production and. fission product release from core debris
. outside the reactor. vessel can persist for many hours. The aerosols produced in this way do not usually have to traverse a convoluted pathway before they enter.the reactor containment as do aerosols produced in the reactor vessel. i The: increased inventory of aerosols in the reactor containment brought on by ,
ex-vessel core-debris interactions ~could lead to rapid agglomeration and settling of the condensed fission products released during the in-vessel phases of an accident. If containment failure is delayed, the primary source of
+
radioactivity released to the environment would come from ex-vessel sources.
Release of fission products from core-concrete interactions can com-j pensate for any inhibition in the release of volatile species during the in-vessel phase of an accident because gases from the thermal decomposition of
- concrete sparge through the melt and drive the release processes. Ex-vessel i processes can also lead to the release of fission product elements that are ordinarily quite refractory. This, again, is because of the strong driving
- force produced by gas sparging and the unusual melt chemistry that arises during l ex-vessel interactions of core debris with concrete.
i Also of importance is the generation of aerosols from nonradioactive materials, such as concrete and steel, during ex-vessel interactions. The additional concentrations of suspended particulates in the containment brought i on by these aerosols naturally mitigate the inventory of radioactivity released
- from the fuel that would then be available for release to the environment.
This additional material, on the other hand, poses yet another threat to equip-ment in the containment whose performance is degraded by the presence of aerosols.
'VANESA is a mechanistic model of fission product release and aerosol generation during core-concrete interactions. This model was based on observa-tion from experiments involving high-temperature melts on concrete and informa-tion from analogous industrial processes. Two broad mechanisms of aerosol 5-10 b
l
--wir . pie $ war t- zw w y+9-ip t--7i$I -% r ar-r + 91 --ga +r t'N+ r--'F~e-tt - g m (----*T' -
'TTt-e b - - - - - T W*
^
formation are considered in the model: vaporization o'f me1t species accentuated
.by gasisparging,.and' mechanical formation of aerosols-by' violent agitation of
- the molten debris sparged with decomposition gases. Vaporization processes produce-the-most intense; aerosol generation during ex-vessel' core debris-inter-
' actions,'while mechanical. processes provide.a mechanism for. aerosol formation that persists even when debris temperatures.are so low that.little. vaporization'
-of species in,the' debris'can' occur.
. Input to this model. includes melt temperature,. concrete erosion rate - A
.and gas generation rate-predicted by the CORCON model.of melt-concrete inter 1 l
' actions. .It computes _the thermochemical limits of vaporization from the melt,.
- - and then compares the extent of vaporization recognizing kinetic barriers,. ,
4 such as mass transport, to.the approach to the thermochemical limits for-vapori-
, zation. Mechanical aerosol generation'is estimated by. analogy'to experimental ,
data'with simulant systems.
More complete descriptions of the model are provided in the users' manual (5.20) and its uncertainties are discussed further in Appen' dix C of
~
p Volume'1.(5.1) l 5.3 Radionuclide Transport and Deposition 4
5.3.1' Transport in Reactor i Coolant System: TRAP-MELT i
The TRAP-MELT code that was used for the primary system radionuclide l-transport analyses of this study was developed from the published TRAP-MELT code (5.21) used for the " Technical Bases" report (5.8). Major changes were.
- made in the treatment of aerosol particle transport and behavior and in radio -
l nuclide condensation on and evaporation from particles. In addition, the
! internal data base of the code was increased to. include physical property data l for tellurium and cesium hydroxide. An outline of the code, highlighting these
. changes, is given below.- A more-detailed description is given in the TRAP-MELT. Users' Manual (5.22) f The TRAP-MELT model is designed to treat radionuclide transport in.
! 'an-arbitrary flow system whose-thermal-hydraulic conditions are given as func-f tions of time.- For this study, the data needed by TRAP-MELT to define the 5-11 l
_ _ _ _ _ .~, _- __ . _ _ _ . . _ _ _ _ _ _ _ _ _ _ - - _ _ -_ _ - _ .._.. ,
thermal-hydraulic conditions of the primary system were generated by MERGE.
In addition, TRAP-MELT requires the definition of source terms for each radio-nuclide; these terms were developed by CORSOR.
Once the' flow system is defined, it is subdivided into a series of control volumes that can, in principle, be arbitrary in number and flow connec-tions and that are chosen on the basis of characteristic geometry, thermal-hydraulic conditions, and suspected significant radionuclide behavior such as change of phase, agglomeration, or deposition. Radionuclides in each control volume are assigned, with uniform distribution, to one of two carriers: the wall surfaces and the gas phase. Each radionuclide is allowed to reside on these carriers in either particulate (liquid or solid) or vapor form so that by combining carrier with form in the concept of " state", the condition of a radionuclide in a given control volume is completely determined by its state.
TRAP-MELT thus considers five states:
e Radionuclide vapor carried by gas a Radionuclide particle carried by gas e Radionuclide vapor carried on wall surface e Radionuclide particle carried on wall surface
. e Radionuclide vapor reacted with wall surface.
This list of states is not exhaustive (for instance, in two-phase flow, the carrier water must be* considered) and the logic of the code has been chosen to accept an arbitrary number of states readily .
Radionuclide transport can occur among the five states of an indi-vidual control volume or between certain states of different control volumes connected by fluid flow, lhe former types of transport are modeled or corre-lated in the code itself. The latter are assumed to occur in phase with the fluid flow (as developed by codes such as MERGE) and are imposed on the system.
Sources of radionuclides to the system may occur in any volume and any state, and they must be input to the code as mass rate functions of time.
At present, the intravolume transport mechanisms contained in TRAP-MELT are:
l e Competitive condensation on, or evaporation from, wall surfaces and particles of cesium iodide, cesium hydroxide, and tellurium 5-12
e Irreversible reaction of molecular iodine, cesium hydroxide, and tellurium with stainless steel surfaces e Particle deposition on surfaces due to
- Settling
- Diffusion from laminar and turbulent flow
- Inertial impaction from turbulent flow
- Thermophoresis.
Particle transport (and evaporation or condensation from or on particles) depends on particle size. TRAP-MELT takes this into account by considering a discretized particle size distribution that is subject to change, in each volume, by the deposition processes themselves, by possible particle sources, by flow of particles from other volumes, by flow of particles out of the volume in question, and by agglomeration. The last can be due to many mechanisms.
TRAP-MELT considers the following agglomeration mechanisms:
e Brownian e Gravitational e Turbulent (shear and inertial).
Considerations of stiffness and linearity split the system of first-order differential equations resulting from the above-listed transport mechanisms into three classes. Most of the deposition mechanisms (transfer from gas to wall surface) are taken as first order in the concentration of radionuclide species on the carrier (gas, particle, or wall) from which the transfer occurs. T'.cy constitute the first class, whose transport scheme can be written in the form:
5 = S + MC, (5.1 )
dt d
where C is the concentration vector of the species in question for each state and volume, S is the source rate vector for each state and volume, and M is the transport matrix between all states and volumes. Because the deposition terms are taken as first order, M is independent of C and depends, with S, on time only. It is thus possible to solve Equation (5.1) as a set of first-order differential equations with constant coefficients by standard techniques.
This is done in TRAP-MELT for the class of linear mechanisms. Condensation 5-13 P
--' ___ -___.__._____.______-__-___m_- -
L
..and evaporation, which have a much shorter time constant than the linear prc-l ' cesses, constitute the.second class and are treated outside this framework but
- parallel to it,~ as is' particle agglomeration, which constitutes the third class
, of mechanisms in the TRAP-E LT code.
f The approach to this. parallel treatment is as follows: Equation t: - (5.1).is taken'as the master time-translation operation of the radionuclide system. Time steps are adjusted so'that S and M change little over a time step and so_that the time step does not exceed one'-third of the smallest flow
~
L residence time for any control volume. . The latter assures that the system
- does not translate excessively between couplings to the other two classes of
- i. mechanisms. In addition, the characteristic cosgulatifon time for the aerosol
!^ in each volume.is evaluated and compared to the master time step. If the former is'short compared to the latter, the master time step is appropriately reduced.
i At the beginning of each time step, phase transitions of radionuclides are modeled by examin'ing each control volume in turn and solving the molecular.
!. mass transport equations for vapor transport among the gas phase,' particles,
{ and wall surfaces. Because of the low heats of vaporization of the radionu- t clides in question, this transport is assumed to be isothermal. Transfer to l the walls assumes the Dittus-Boelter correlation (5.6) for pipe flow and trans-fer to the particles occurs by diffusion based on the size distribution at the i beginning of the time step. Redistribution of the vapor phase occurs in a time that is small compared to the master time step; therefore, this redistri- -
bution is essentially decoupled from the other processes considered which
< justifies the use of a time parallel solution treatment.
Once redistribution of the vapor phase has been effected, its effect j i on'the existing particle size distribution (in the volume in question) is I calculated by assuming that each size class gains (or loses) mass in proportion to the rate of vapor transfer to (or from) that size class. Conservation of
- number for each size class then dictates redistribution between, in general, j- two new contiguous size classes, the number in each size class being determined
- by mass conservation. >
l- At the end of a time step, the particle size distribution in each volume is reevaluated over that time step to account for possible particle
- agglomeration, sources, and flow terms. 'The agglomeration algorithm has been t
4 5-14
, -e vw- y v, w,- wrv w , y--,m -+w,----,w--yw -,-,w w, vorm w- --v-. -- y w- v v y- y-eT-,, --- - - -vr- +-v-
- - .. - - ._ - =..- - - - - -- - - .- - - .
l excerptedfromtheQUICKaerosolbehaviorcode(5.2N,whichisbasedonasize l
{_ discretization scheme.
The approximations inherent in this parallel treatment are minimized
- .by_ relegating mass redistribution and conservation to the master Equation (5.1),
except for redistribution due to_radionuclide phase change. Agglomeration and particle evaporation / condensation serve only to modify the particle size distri-4 bution and therefore affect particle deposition indirectly through mass -
distribution-averaged deposition velocities. Thus the aerosol aspect is solved
- (over a master time step) completely in parallel to Equation (5.1), using all sources, flow terms, and particle removal terms evaluated for each size class considered. The resultant distribution is used'to evaluate average particle deposition terms for use in the master equation only. Similarly, reevaluation
. of tne particle size distribution due to radionuclide phase change affects
.these average deposition terms only.
In eddition to the time-dependent thermal-hydraulic conditions and
- l mass input rates by species, the TRAP-MELT code requires input information on
! the initial particle size distribution of the source, the control' volume ;
f geometry, and the physical properties of species (including deposition veloci-ties on surface materials). The code provides output in terms of time- and
) location-dependent mass by species and state, as well as size distribution of '
{ suspended particulate material. -
l There are a number of uncertainties which affect the TRAP-MELT code f predictions of primary system retention of materials. Any errors or impreci-i sions in the input to the code will clearly affect the quality of the results,
, both for the primary system thermal-hydraulics provided by MERGE and for the ,
l core release rates determined by CORSOR. The extent of interaction among the
, materials released from the melting core is determined largely by the timing l of their releases, and this represents a less straightforward, but no less I important, potential effect on the code's results due to input inaccuracies. ,
i The experimentally determined vapor deposition velocities for Te, '
Cs0H and 12 on hot surfaces may not represent an accurate description of the process as it occurs in the reactor coolant system (RCS) because of the impre-l cision in the available data and because the experimental systems may differ f from the actual RCS conditions. Nevertheless, what data are available have l
, been incorporated, since these analyses are intended to reflect the state of '
l 4
5-15
,g - m - , , , -w , , . . --w --- , -w,-,n +,e,m m- -c. -----------,-nen,- - , . . , - , - - - - ,
the art. Additional uncertainties affecting vapor and aerosol deposition arise from possibly inadequate specifications of primary system geometry and-flow ,
patterns.
The disposition of materials suspended in the coolant system at the time of core slumping or at depressurization of the pressure vessel can have significant impact on retention calculated for some of the sequences analyzed.
This is because some fission products and aerosols emitted from the core have not escaped the RCS at the time of core slumping and are still available for injection into the containment. The large burst of steam which accompanies core slumping or depressurization when the pressure vessel fails will rapidly sweep out the coolant system, and the very short transit time to the contain-ment is expected to lead to minimal retention of these materials. Thus, in the analyses in this document, the material suspended in the RCS at the time of core slump or pressure vessel failure is assumed to be injected into the i containment as a " puff" release, with no further retention in the primary system.
The analyses in the main body of this document are subject to some uncertainties which may overpredict retention in the primary system. One mechanism not included in the current analysis is the structure heatup due to decay heat from the deposited fission products. Heatup of surfaces where species of intermediate volatility (e.g., Csl and Cs0H) are deposited would lead to reevolution and transport of the previously deposited materials through the reactor coolant system to regions of lower surface temperature or to the containment. Thus, the deposition of these species may be self-limiting to some extent.
1 5.3.2 Transport in Containment: SPARC Many BWR accident sequences involve a fission-product flow path which
- passes through a suppression pool. Although the importance of the suppression j pool in removing fission products has long been recognized, comprehensive analytical models that consider all pertinent parameters (such as particle size, bubble size, pool dimensions, and pressure and temperature conditions) have not been available. The SPARC code was written specifically for the 5-16 I
L__ , , ,
analysis reported here and was used to calculate removal of particulate matte: ,
by a suppression pool.
The SPARC code was developed by Owczarski, et al(5.24) . The model includes particulate removal due to steam condensation, gravitational settling, inertial impaction both inside the gas bubbles and in the gas injection regime, diffusion deposition, and mechanical entrainment of pool liquid at the pool surface. In addition, a mechanism which retards particle deposition due to evaporation of steam is considered. Details of the model are provided in the SPARC Users' Manual (5.24) . The bubble shape in the SPARC code is assumed as an oblate spheroid. For all calculations of pool scrubbing in this study, the ratio of minor to major axes (aspect ratio) for the bubble was taken as 1:3.
The bubble diameter based on a spherical shape was taken as 0.75 cm in all cases.
Recognizing that the role of the suppression pool in removing parti-culates was neglected in the past, or a fixed removal efficiency or decontami-nation factor was arbitrarily assumed, use of the SPARC code in this program represents an improvement over previous source term analyses by accounting mechanistically for the effects of the suppression pool.
5.3.3 Transport in Containment: NAUA 4 The NAUA code was developed at the Kernforschungszentrum Karlsruhe, West Germany, for calculating aerosol behavior in LWR core melt accidents.(5.25)
It is based on mechanistic modeling of aerosol agglomeration and deposition within a containment vessel where a condensing steam atmosphere may exist.
The model for steam condensation on particles was validated by small-scale experimental measurements (5.26), and larger-scale validation is being planned.
The NAUA code calculates physical processes, excluding chemical changes and radioactive decay. The removal processes considered include gravitational settling and diffusional plateout. Interactive processes include Brownian and gravitational agglomeration and steam condensation. Aerosol sources and leakage are also included. Compositional changes resulting from time-dependent composi-tions for the input aerosol are tracked by the code.
The particle size distribution is defined by a number of monodisperse fractions. In this approach, the governing integro-differential equation is 5-17 h
- transformed into a system of coupled first-order differential equations. In
'effect,' the particle size fractions interact and deposit according to.the included mechanisms, generating a time-dependent. distribution of mass among the various size fractions. Steam condensation is handled in a separate inte-gration. Output from the code includes mass concentrations of condensed water and dry aerosol materials (airborne and on surfaces), as well as particle size distributions at various times throughout the calculation.
Since the original version of the NAUA code has no provision for engineered safeguards, calculations were made to account for removal of aerosol particles by sprays, as follows:
2 h=-cwRN(V,-v)n, g (5.2) where n-is the aerosol particle concentration, e is the collision efficiency.
Vg and va are settling velocities of the spray drops and aerosol pArticlei, respective'y, R is the radius of the spray drop, and N is the water drop concentration.
Due to hydrodynamic interaction between a falling water drop and airborne particles, only a small fraction of the particles within the cross-sectional area of the water drop is removed by spraying. To account for this hydrodynainic effect, the collision mechanisms due to inertial impaction, interception, and Brownian diffusion of aerosol particles were used by defining c in Equation (5.2) as:
c = eg + cR * *D, (5.3) where et, cg and c0 are the collision efficiencies due to intertial impaction, interception, and Brownian diffusion, respectively. The following collision efficiency models were utilized for the three mechanisms:
2 Stk
- ! " (5tk + 0.35)g (5 0 5-18
. n
e = '*5(t/R)
(5.5)
(1+r/R)l/3 d
c = 3.5 Pe-2/3 .
0 (5.6) where Stk is the Stokes number for aerosol particles based on a characteristic length of water drop with radius R; r is the particle radius; and Pe is the Peclet. number. The Stokes number and'the Peclet number are defined as 2
2rpVC Stk = gj 9 and (5.7) 2VgR Pe = _ (5.8)
D where D is the diffusion coefficient of aerosol particle vg is the settling velocity of water drop.
C is the Cunningham slip correction factor op is the particle density p is the gas viscosity.
In general, for relatively large particles, the inertial effects on the overall collision efficiency are larger than the interception term because the water drops are much larger than the aerosol particles. As particle size becomes smaller, the Brownian diffusion term will become increasingly important. It should also be mentioned that Equation (5.4) is given by Hetsroni(5.27)- and Equations (5.5) and (5.6) are based on the work of Lee and Gieseke(5.28, 5.29) ,
Another particle deposition mechanism, diffusiophoresis, was added to the NAUA code. Diffusiophoresis results from steam condensation onto >
{
containment walls and involves twc mechanisms: a net flow of gas toward the \
wall surface (known as Stefan flow), and a molecular weight gradient caused by the steam concentration gradient. In general, the effects of Stefan flow are
[ -
much larger than those of the molecular weight gradient and result in deposition ' '
of particles on the wall surface. The condensation rate toward wall surfaces ' '
calculated by the MARCH code has been used to calculate deposition due to
- diffusiophoresis.
., ~,
5-19 y N
/
4 - , . ,
. 4. t 3-
a v
, T.-
In-utilizing the NAUA computer code for calculating aerosol behavior during various accident sequences, it was noted that in certain cases the code requires a long computing time to calculate the rate of condensation of water vaporontoparticles.}'Thistypeofproblemtakesplacewhenalargeamount-of condensible' water vapcr was used as art input. It was noted that a saturation ratio of much greater t,han 1.0 was frequently encountered even after the
~
condensation calculatidn was completed.
Some literature suggests that-pure water vapor at 20 C will spontane-ously form water droplets in the absence of condensation nuclei when the satura-tion ratio exceeds 3.5, and.at 0 C a saturation ratio of 4.3 is required for homogeneous nucleation.. This, mechanism has been implemented in NAUA.in addition-w to the existing condensation calculation. As the critical supersaturation _for the homo eneous nucleation, the following correlation equation given by Green
, and Lane 5.30) was used:
s s
, y
( :7 S = exp(0.55'7-(o/T)3/2,g),
s
<where '
- +
S is the critical supersaturation 1
l., s ,
., o is surface tension 3 I T is the temperature in *C' M is the molecular weight 6f water.
Nonucleationorself-condeksationratesarecalculatedinthecode.
Rather, if critical supersaturation is realized at a given time, the excess 1 ~
,ater w vapor is assumed to form water, particles of a uniform size spontaneously.
Of course, these small pure wate,r, droplets are subsequently subject to NAVA's usual condensation and' coagulation processes both among themselves and with T other particles containin'g' solids. Although the effects of this mechanism on
' N theoverallaerosolconcentrationchgngeareinsignificant,thecomputational time is,, reduced considerably by thit o imp'lementation.
- ? e .. ,\
A s g, ,
y z
b'f 4
5 1s A 7
- i , '
i \
_ , 5-20
~,' a W
'? '\
3 as
s .
k r;
' References
- (5.1);
Gieseke, JJ Al. et al, "Radionuclidel Release Under Specific LWR!Acci-
. 'dentiConditions, Volume 1", BMI-2104_(July, 1983).
-(5;2) Gieseke,;J.~~A.,~et al, "Radionuclide Release Under: Specific LWR Acci-fdent Conditions,, Volume 5", BMI-2104 (July, 1984).-
-(5.3)_ , Wooton, R. 0.,-et al, " MARCH 2 Code Description and User's' Manual",
' Draft:(December,.1982).
(5.4) _ , Wooton,.R. O. and Avci, H.:I , " MARCH _(Meltdown Accident. Response
. Characteristics) Code Description and Users' Manual", NUREG/CR-1711, BMI-2064 (October, 1980).
(5.5) "American NationalLStandard for' Decay Heat Power.in Light Water Reactors", ANSI /ANS-5.1-1979.
.(5.6) '
Geankoplis, C. J., " Mass Transport Phenomena", Holt, Rinehart, and Winston (1972). <
(5'7)
. Freeman-Kelly, R.,-"A Users' Guide for MERGE", Batte11e's Columbus'-
Laboratories, October, 1982.
(5.8) " Technical. Basis for Estimating Fission Product Behavior During LWR 'l Accidents", NUREG-0772 (June, 1981).
-(5.9). CORSOR Manual.-
(5.10) Bell, M. J., "0RIGEN, The ORNL. Isotope Generation and Depletion Code",
ORNL-4628 (1973).
(5.11) Lorenz, R. A., et al, " Fission Product Release from Highly Irradiated '
LWR Fuel", NUREG/CR-0722 (1980).
(5.12)' .Lorenz, R. A., Collins, J. L., and Malinauskas, A. P., " Fission Product Source Teims for the LWR Loss-of-Coolant Accident",
NUREG/CR-1288 (1980).
(5.13) Lorenz, R. A., et al, " Fission Product Release from Highly Irradi-ated LWR Fuel Heated to 1300-1600 C in Steam", NUREG/CR-1386'(1980).
(5.14). Lorenz,-R.-A., " Fission Product Release from BWR Fuel Under LOCA-Conditions", NUREG/CR-1773 (1981).
~(5.15) Parker, G. W.,-Martin, W. J., and-Creek, G. E., "Effect of Time and
' Gas Velocity of Distribution of Fission Products frca U02 Melted in a Tungsten Crucible.in Helium", Nuclear Safety Program Semi-Annual Report for period ~ending June 30, 1963, ORNL-3483, 19-20 (1967).
^
5-21.
b
. s-(5.16). - Albrecht, H., 'Matschoss, V. , .and . Wild, H., " Experimental :Investiga-tion of Fission and Activation Product Release from LWR Fuel Rods at Temperatures Ranging from 1500-2800 C", proceedings of the Special-
~
~ists' Meeting on the Behavior of Defected Zirconium Alloy Clad Ceramic Fuel in Water _ Cooled Reactors, 141'-146 (September, 1979).
~(5.17)! Albrecht, H., Matschoss,-V., and Wild, H.~, " Release of' Fission and Activation Products:During Light Water Reactor Core Meltdown", Nuclear.
Technology, 46,559-565_(1979).
(5.18) Niemczyk, S. J. and McDowell-Boyer; L. M., "TechnicalLConsiderations-Related to Interim Source.-Term Assumptions for Emergency Planning and Equipment Qualification", ORNL/TM-8275 (1982).
(5.19) .Lorenz, R.' A., Beahm,~E. C., and Wichner, R. P.,." Review of Tellurium Release Rates from LWR Fuel Elements'and Aerosol' Formation from Silver Control Rod Materials", ORNL, letter report, February 28,-1983.
(5.20) VANESA Manual..
(5.21) Jordan, H., Gieseke,-J. A., and Baybutt, P., " TRAP-MELT Users' Manual", NUREG/CR-0632, BMI-2017 (February, 1979).
(5.22) TRAP-ELT 2.1 User's Manual.
(5.23) Jordan, H., Schumacher, P. M., and Gieseke, J. A., " QUICK ' Users' Manual", NUREG/CR-2105, BMI-2082 (April, 1981).
(5.24) Owczarski, P. C., Postma, A. K., and Schreck, R. I., " Technical Bases and Users' Manual for SPARC -- Suppression Pool Aerosol ^ Removal Code",
report to the U.S.'NRC, NUREG/CR-3317-(May', 1983).
(5.25) Bunz, H., Koyro, M., and Schock, W., "A' Code for Calculating Aerosol Behavior in LWR Core Melt Accidents Code Description'and Users' Manual". .
(5.26) Schock, W., Bunz, H., and Koyro, M., "Messungen der Wasserdampfkonden-sation an. Aerosolen unter LWR-un-falitypischen Bedingungen", KfK 3153 (August, 1981).
(5.27) Hetsroni, G., " Handbook of Multiphase Systems", McGraw Hill Book Company-'and Hemisphere Pub. Co. (1982).
(5.28)- Lee, K. W. and Gieseke, J. A., J. Aerosol Science, 11, 335 (1980).
-(5.29) Lee, K. W. and Gieseke, J. A., Environ. Sci.'& Technol., 13, 446 (1979).
(5.30) Green, H. L. and Lane, W. R., Particulate Clouds: Dust. Smokes and Mists, D. Van Nostrand-Co., Princeton, New Jersey (1957).
5-22
- n
- 6. BASES FOR TRANSPORT CALCULATIONS 6.1 Plant Geometry and Thermal Hydraulic Conditions The MARCH 2 code
- was operated for each of the three accident sequences analyzed. The results of the MARCH analyses are used as inputs for the fission product release and transport calculations. - A summary of important reactor characteristics, containment parameters, and MAP.CH options is presented in Table 6.1.** The values of parameters used to describe the reactor coolant system such as the masses, surface areas, volumes, etc., were primarily obtained through information provided by the General Electric' Company and are
-not reprnduced here due to their proprietary nature.
In the'following sections of the report, the results obtained with the MARCH and MERCE codes are described for each of the acc.ident sequences.
6.1.1 Sequence AE A large break in the piping of the recirculation loop with failure
-of the standby core cooling system would result in loss of reactor coolant and uncovery of the core. Since core uncovery occurs early in the accident,.the decay heat level is high and a rapid core meltdown would be expected.
Table 6.2 indicates the times of key events as predicted by the MARCH 2 code.
Table 6.3 presents details of the core and primary system response at key times during the sequence. Core uncovery, heatup, and melting occur at low primary system pressure. The temperatures of selected fuel regions are illustrated as a function of time in Figure 6.1.
In the MARCH analyses for the Peach Bottom accident sequences, fuel slumping (movement of molten fuel from the core region to the support struc-tures) was assumed to take place when the lowest mode in a core region was fully molten. As can be seen from the event times in Table 6.2, initial slumping of the highest power (hottest) regions in the core was predicted A description of this version of MARCH can be found ,in 5.1.1.
i ** All tables in this section of the report have been placed at the end of the section.
6-1
4500 4000 TR0(1,6.5) 2000 (5. 6.5) 3500 3000 -
1 F,00
- 2500 - '
4 4 5 5 3 TR0(1.1) 3 2000 ,_
TR0(9,11.5) -
1000 1500 TR0(R. L) where R = Radial zone number L
- Oistance above core 1000 - bottom, ft.
500 500 -
0 8 8 ' ' ' I ' '
O 400 800 1200 1600 2000 2400 2800 3200 3600 FIGURE 6.1 TEMPERATURES OF SELECTE0 FUEL REGIONS AS A FUNCTION OF TIME -- PEACH BOTTOM AE SEQUENCE 6-2 4
l relatively early'in the sequence. Collapse of the entire core into the' bottom head was assumed when 75 percent of'the core was molten or when the structures g supporting the core reached their' melting temperature. The support structures-
- would be heated by the core debris that have fallen out of the core region.
Since the outer portions'of the core operate at lower power levels than the-central region, they experience slower heatup and thus require a longer time to reach melting. The molten fuel that falls into the vessel head will evaporate
'the residual water there and will be partially quenched.- In the AE sequence,
- the primary system and containment are at the same pressure,.thus there is no stress on the bottom head other than the weight of the core debris; vessel I
- ' failure in this case takes place by melt-through of the bottom head. .Under the conditions assumed, melt-through of the vessel head is predicted about an hour after the collapse of the core into the bottom of the vessel.
The prediction of the onset and progression of core slumping with the model described above can be sensitive to core power distribution, core nodalization, and assumptions regarding metal-water reactions after the onset of melting. The core power distribution used in the Peach Bottom analyses was more peaked than would be typical of a core with high burnup.
Prior to the accident, the piping and structures in the reactor cool-l ant system would be in the temperature range.of 290 to 300 C. Heatup of the
- fuel and the release of fission products occur at approximately 210 kPa (30
- l. psia). Thus, at the toime of fission product release the structures are expected to be heated above steam saturation temperature. The flow path for fission product release within the reactor coolant system is illustrated in Figure e ME E a i Pred te g a s u u e e o a vo me are illustrated in Figures 6.4 and 6.5, respectively. It should be noted that
! time zero ip Figures 6.4 and 6.5 is the start of core melting or 11.5 minutes into the accident. The decrease in temperatures occurring at the later stages ,
of.the accident results from cooling by the large amount of steam generated as
] the core is predicted to slump'into the lower head of the vessel.
7 The flow path from the vessel to the suppression pool is illustrated in Figure 6.6 for the time period-prL 'o containment failure. Figures 6.7 and 6.8 describe the temperature and pressure in the drywell and wetwell vapor space throughout the accident. Key event times for each of the sequences are 6-3 p
a
9 I
/
d k l l Steam Line
- Steam Dryers r ,
l ,
l Steam l
- Separators ,
I :
- Core -
f3 ' '
l Jet Pump Recirculation ( _
l FIGURE 6.2 FLOWPATH FOR FISSION PRODUCT TRANSPORT IN THE RCS -
SEQUENCE AE 6-4
l PIPES /
SEPARATORS 4
4 LOWER OUTER ANNULUS SHROUD HEAD 5
3 TOP GUIDE l 2 CONTAINMENT 4
CORE 1
FIGURE 6.3 SCHEMATIC 0F CONTROL VOLUMES FOR THE PEACH BOTTOM AE SEQUENCE i
6-5 l
l l
Core Exit
- Plate
--- Head 3500 - - - - Separators Lower Annulus 3000 - i 1
s 2500 O /
J 2000 -
Ii 5
= ; ,,l <
j 1500 -
l 3 1000 -
/j
/,/
kc! g
, ~ ~~~ - },i r \
i i 500
-~ ~ C
- Jl l l I I l l l
.J 0 20C 400 600 800 1000 1200 1400 1600 1800 Time, Seconds FIGURE 6.4 GAS TEMPERATURES IN RCS VOLUMES - SEQUENCE AE 6-6
l l
- - Plate
~~~
3500 -
_ _ _ _ Sepa LowerraAnnulus to rs 3000 -
2500 - T-\
/ \
p 2000 - / \
\
s ! / ~~ N\
B 2 1500 -
/
l l '% '
K j
$ I /
j -
1000 -
/.
,/ , ,
500 -
~~~~~~~~~" ~ ~ ~ j ,8 -
I I I I I I I I 0 200 400 600 800 1000 1200 1400 1600 1800 l
Time, Seconds FIGURE 6.5 STRUCTURE TEMPERATURES IN RCS VOLUMES - SEQUENCE AE I -
6-7
__-_____-___-_____--_L.
[ ..
l umas Blowout Bkmout
- Peneis Refueling Boy Ms su m z.. 7:== = t ..,~~ '
'< , p e A[$.1
- if' i 1
- d. - - - - - l1
/l I
v.
l 5 ..,.
- o! 't !
i
- % i e $ 9 l bl~ ~ ij h 9 j Reactor ? - -- - ^ i I 1
9 g y l l
l vs es ; :
.., ymm :
j' l k5 ll l k. . _ . . _ _
i l- .-
[
- 1 T ~s!m.mmanm:
_ . . . . _M "
.: g l i .
f 6=
h p i l i ' .;
. .f.. r ( .
l
{ k; !$ $
f.e . .P. .,
l {,
$93nM19&'*$g s.< .
- s. . m n.?dag.:-a,, -
33 a :tbiers% F5
, m. %
l
. c =_-
.. 9.
s.e_._- p.1.
.i Q#k if** i0
,rFKisM g i,0i&5dMM*
~
1 FIGURE 6.6 FLOWPATH FOR FISSION PRODUCT TRANSPORT BEFORE CONTAINMENT FAILURE - SEQUENCE AE ,
6-8 i
i
t 2500.0 DRYWELL
........... WETWELL i
l 2000.0-N, N
N 1500.0-D E<-*
i M l N
m (14 e' 1000.0-9 _ _
i l 500.0-
! L t .-
4 0.0 , i i i i i i
- 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 000.0
, TIME - (MINUTE) l FIGURE 6.7 TEMPERATURES IN CONTAINMENT VOLUMES - SEQUENCE AE i
l
.--. -_ _ . . - - __ . _ ~ . . - - _ _ - - - _ . _ _ _ -
i 140.0 DRYWELL
........... WETWELL 120.0 -
l 100.0 -
l l A J M i M i A. 80.0-i N D
rn en in 60.0-
.L M o f4 A
l 40.0-
! t 1
i
] 20.0-i
! 0.0 i i i '
i i O.0 100.0 200.0 300.0 400.0 500.0 650.0 750.0 800.0 l
TIME - (MINUTE) 4 FIGURE 6.8 PRESSURES IN CONTAINMENT VOLUMES - SEQUENCE AE
il l t
provided in Table 6.2. Table 6.4_gives details of the containment conditions
-at key l times during the sequence. Prior to containment failure the flow passes from the drywell into the suppression pool after leaving the vessel. Tempera-
.tures in'the drywell are nominal and in the wetwell are cool because of the cold suppression pool.
After the start of core slumping,-the pressure in the containment 4
volumes rises rapidly due to the production of hydrogen and the transport of
-the noncondensibles into the wetwell. At 132 psia, a leak area of.7 ft 2 is assumed to open in the drywell. The prediction of the. occurrence of containment h failure at this point in the accident sequence is sensitive to the failure pressure as well as the core slumping scenario assumed. Figure 6.9 illustrates
~
.the subsequent flow paths for the fission products. -Fission products airborne in the vapor space of the wetwell would flow back through'the vacuum breaker into the drywell and then out the break into the annulus surrounding the steel shall of the containment. ~ As discussed in Section 4.3, the fission products i may then follow a number of possible pathways depending on the response of the reactor building to failure of the primary containment shell. Fission products released to the drywell after containment failure would flow out the break in 2 the primary containment shell without passing through the suppression pool.
The predicted consequences of this accident requence will therefore be very ;
3 j- sensitive to the estimated time of containment failure. After vessel failure, .
l the vaporization release of fission products will be to the drywell. Again, the flow path does not include the suppression pool. The temperatures in the drywell atmosphere are predicted to become very hot in this phase. The
{
question of.the overheating and failure of penetrations is not a significant
[ issue, however, because the integrity of the primary containment has already. t been breached.
Table 6.5 summarizes the leak rates derived from the MARCH analyses f that were used in the analysis of fission product release to the environment.
6.1.2 Sequence TC i
l In the transient sequence TC, the reactor power level equilibrates with the primary coolant makeup provided by the emergency core cooling system.
l l A nominal power level of 30 percent of full power was assumed, but the actual i i
i 6-11
--____.________..m
}
muma-
)
sama Blowout Blowout
""" M Refueling Bay Ms szza w= - ._ ~ .
i ff e
Q5
,- b. ;
i
- j') P h i d
,e hg_v,.
i.* h , ,
,.- , l jo . . m . . um . o o _--
! i . 5l l5 h! !
q
- ,;6
... 6. .
,c i
---~^--9; f;, ll Reector D
---3 l h g y ; ; veuei l ; .g y_w w ten .
- h. p< ,?;
'; ; l5 y @ t h s < ? ,-
s ..- +
?
--dd",*.RT;.,'0' ;
l
,i , ,_ .
! Vt ll c
'6
'.e v*
7.:
v
,.d P
.n . ,
bb . % [IAMAI'O c
+ .
- d b _? s s ee ;.
. . G..,
k.3 . !Ct. E n N E M DC
=.._- -;r = = ==-
@pg o - .
y c . -
l 5 f* w 1 $
vA ggggg.gg g,gg.gg.gygi ... . _.
E a w g FIGURE 6.9 FLOWPATHS FOR FISSION PRODUCT TRANSP0RT AFTER CONTAINMENT FAILURE - SEQUENCE AE 6-12
power level in the~ analyses would vary with the water level in the reactor vessel. Heat was removed from the primary system via steam flow through the safety / relief valves into the suppression pool. Since the core power level is considerably above the capacity of the residual heat removal system, the temperature of the pool rises and the pressure in the containment increases to Lthe failure point. At the time of predicted containment failure, the emergency core cooling system is drawing water from the condensate storage tank. For the purpose of this analysis, the operation of the emergency core cooling system is not affected by the failure of the containment. As the inventory of the condensate storage tank is depleted, the emergency core cooling system is switched to take suction from the suppression pool. The water in the suppres-sion pool has been heated to a high temperature and would be boiling or at saturation following containment failure. Since the emergency core cooling system pumps are not designed to operate under these conditions, they ue assumed to fail due to cavitation. Core uncovery and melting follow.
Table 6.2 indicates times of key events as predicted by the MARCH 2 code.
Table 6.3 gives core and primary system conditions at key times in the sequence.
The temperature transient of selected fuel regions is illustrated in Figure 6.10. Core uncovery, heatup, and melting occur with the primary system at approximately 7.93 MPa (1150 psia). While the core power level is substan-tial during the time to contair. ment failure, it drops rapidly as the core begins to uncover. At the start of core melting as well as thereafter the power is down to decay heat levels.
The flow path for fission product transport within the reactor vessel is illustrated in Figure 6.11. After passing through the steam separators, the flow splits with the greater fraction (85 percent) passing through the steam dryers and a smaller fraction (15 percent) bypassing the steam dryers.
The flow then exits the vessel through the steam lines and safety / relief lines to the suppression pool. A schematic of the control volumes used in the MERGE analysis is provided in Figure 6.12. Gas and structure temperatures are shown in Figures 6.13 and 6.14, respectively. Time zero in Figures 6.13 and 6.14 corresponds to 97 minutes (5820 seconds) into the accident. As in the AE sequence, temperatures near the end of the core melting phase drop following l slumping of the fuel. The peak gas temperature in the relief line to the suppression pool is approximat?Iy 1900 F (1040 C). The predicted temperatures 6-13 I
110,000
-TRO (R.L) where R = Radial zone number L = Distance above core bottom, ft 5000 9,000 - -
8,000 -
_ 4000 7,000 -
6,000 -
- u. e TR0(5,6.5) - 3000j
$ TR0(1,6,5) q
$ 5,000 -
}
b 3 8 *
~ ) a 4,000 TR0(1,6.5)
(
TR0(1,1 )
2000 3,000 R
g vi ,
e' I 2,000 -
TR0(9 1000 4
1,000
\
~ - >
0 l I I I I I I I O 1200 2400 3600 4800- 6000 7200 8400 9600 10800 Time, seconds FIGURE 6.10 TEMPERATURES OF SELECTED FUEL REGIONS AS A FUNCTION OF TIME - SEQUENCE TC 6-14 l
l 1
/ \
/
Steam Line Dryers m /
e
=
-w-
- Seperators
\,
- i he A -
s A E FIGURE 6.11 FLOWPATHS FOR FISSION PRODUCT TRANSPORT IN l
6-15
UPPER STEAM DRYERS OUTER ANNULUS 5 6
)
LOWER FIPES/ SEPARATORS OUTER STEAM LINE ANNULUS 8
4 7 RELIEF LINES 9
SHROUD HEAD
. 3 TOP GUIDE 2
POOL i
CORE I
l l
FIGURE 6.12 SCHEMATIC 0F CONTROL VOLUMES FOR THE PEACH 00TT0f1 TC AND TW ACCIDENT SECUENCES 6-16 i
- - _ _ _ r,_. - . , ~ . . . , , - _ _ _ , . _ - . . _ _ _ _ . . _ . . , , _ _ _ . _ _ _ _ . . _ , _ _ . _ _ _ - . _ ,- - . ~. -_ , _ . . . _ _
l l
1 Core Exit i Plate l 4500 --- Head
Separators i
--- - Dryers i
. - Upper Annulus
.4000 -
_-- Lower Annulus
- ..- Steam Line
- Relief Line 3500 -
3000 -
o' 2500 -
J E
~ I E /. 7 % s _ . . ._ _ Z-1500 -
/
// e ,
s
~
\ w >..JS,
/
y
/ ll '....., \
/,,,,,,,,,, lr
______-m%
S 1000 : -- ' ' / * '
, , _ srum.=. -v- ,7 500 -
1 I l l l l l 1 0 400 800 1200 1600 2000 2400 2800 3200 3600 i
Time, Seconds l
FIGURE 6.13 GAS TEMPERATURES IN RCS VOLUMES - SEQUENCE TC 6-17
,. . , - . w , -..-.-.-,-.y - , . . - - , - . - - - _ _ . _ ,,
I
! - Plate 4500 - - - Head
Separators
Dryers 4000 - -- Upper Annulus
Lower Annul us
-..- Steam Line 3500 - -~.- Relief Line 3000 -
- /\
o- t
. 2500 -
p
/
E y j '{
15
- l ,
o 5 2000 - /
o.
8 >
/
- N .--.. --- .. _ ...
-- . . J.
+ / , q.- - - , , -
1500 -
/" " ~ "#^Mm
/
4 o I
/ :
,s'/ /
' ~
__ _ . - - c. -
},/,'f'Y..d's
.? .
l I l l I I I I O 400 800 1200 1600 2000 2400 2800 3200 3600 Time, Seconds FIGURE 6.14 STRUCTURE TEMPERATURES IN RCS VOLUMES - SEQUENCE TC 6-18
u k
x are high enough to pre' vent significant retention.of cesium hydroxide and cesium iodide on the reactor. coolant--system surfaces.
, The; flow path from the vessel to the suppression pool is illustrated
~
.in Figure 6.15. The flow path is then from the vapor space of the wetwell
-thrcughivacuum breakersJinto the drywell.and then into the reactor building
- sthrough the' break. -F igures 6.16 and 6.17 show the temperatures and pressures ;
in the containment volumes in the accident. Table 6.4 gives the containment conditions at key times during the sequence.- Containment failure precedes core melt and the subsequent thermodynamic condition of the suppression pool
[
is. boiling or saturated.
Figure 6.18. illustrates the flow paths from the vessel and reactor cavity after reactor vessel failure. .The vaporization. release of fission i
products does not pass through the, suppression pool prior to exiting the. dry-r i well. In the TCyf sequence, the fission products are assumed to be released
! from the drywell to'the environment without further-attenuation in the reactor building. In the TCy sequence, the reactor builoing is assumed to remain intact t' following failure of the primary containment. '
I Table 6.5 sumarizes the leak rate data derived from the MARCH analy-ses and used for the evaluation of the fission product release to the environ-ment.
1 Reactor Building Response l
Following blowdown of the primary containment, the blowout panels between the lower portion of the reactor building and the refueling floor region 1
- w
- :re assumed to fail, connecting the two volumes. The pressure in the reactor i building subsequently was calculated to rise to 0.25 psig, resulting in failure l of the blowout panels in the upper region of the reactor building with a
!' corresponding leak area of 3000 ft2 to the environment. MARCH calculations (Figure 6.19) indicate a peak reactor building pressure of about 1.0 psig.
F Because of the continuing large steam input to the suppression pool with the core at a power level of about 30 percent, the drywell depressurization l (Figure 6.20) requires about 20 minutes. At about 75 minutes, the containment
{ pressure begins to fall more rapidly. At this time, the ECC makeup switches
- l. from the condensate storage tank to the boiling suppression pool and the pumps .
6-19 .
i i
I
. . . _ . . ~ . _ _ _ . _ _ . . . . - _ _ _ _ _ _ . . . - . _ . - . . _ _ _ . . . _ , _ , _ - _ _ _ _ _ _ _ _ _ _ _ _ . _
i l
l Clowout Blowout Panels Refueling Bay Panels W
l l
~
- .
- .-1., v .= : ^M'c ?. " ^ '
i j
i n-
.. ?,{
J h ..- p s i
_ _ _ _ - _ _ _,%.. 0 i Lt ._
@ I3-l i 4
. .m .- u. m _ 3
! 6 . 1l l l d! I !
, 1.r i
y -
,f I 9
g s L, r : ,
ye,,e, :; n "rw.m.wm
,.m... _-. . _ _ -
, 6 5 i
.o.. -
'1
- - --gy*.i 4 :*.
%l s
s ;
,4
- M'o'.Fwo'a$"etT.:
Steam Line 9*.i l<\
bI h
Q'"
Jf f i o c m ,_z 6
J eg
,.s >
y I
p
,A 4 '
el e :
[*N.ENA*O
. -. 4 4 :.. .?'. h y ,, ?Q.'* o ~ : . $
- f. c, 1..
pl M 5 - .l
~
~^= =- M 'O<<
t --
ni o . :- --- h
.5 *y-
- U5
. eh'gogia;;' g g.gf N 'A *
- W' '
~0itgf.pM* g .'Qf l
bj 0 c
\
FIGURE 6.15 FLOWPATH FOR FISSION PRODUCT TRANSPORT BEFORE VESSEL Fall.URE - SEQUENCES TC AND TW
. 6-20 l
l
- - - - . .--- ----.-,,-,-- -.,-- - - - - , . . . - - - - - . . . - . - - ,,--.---._,,.--.---.--,-___.n.
. - - - ._ . _ . . ._ .- . - _ _ _ _ _ . _ - ~ - _ - . . _ . . _ __ ._ . __
1800.0 -
DRYWELL
.......... WETWELL s
l 1400.0-l I
1 1200.0-I N
N 1000.0-M .
i O
E-4 800.0-m.
~
w M 800.0-i
, N
! 8 r 400.0-1 200.0- /
i -
0.0 , , , , , , , ,
) 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 I TIME - (MINUTE) l l
FIGURE 6.16 GAS TEt1PERATURES IN CONTAINMENT VOLUMES - SEQUENCE TC I i
i l
i i
I l
l 140.0 W
i .... ..
wrrMEL t
120.0 -
i i
100.0 -
i ;
4 m
tn 01 80.0-i M
! M TD M
800-i M
- 40.0-1
{
j 20.0-1 1
1 EO , , , , , , , ,
i 0.0 100.0 200.0 300.0 400.0 500.0 000.0 700.0 800.0 900.0 l TIME - (MINUTE)
I i
FIGURE 6.17 PRESSURES IN CONTAINMENT VOLUMES - SEQUENCE TC i
1
l
}
mama suma Slowout Blowout musa Panels RefWig by Penels numa N /**"""'****
- _ = ,
i d O' [ a
, as ,
i it '. i, s E,;.. _ _ _ _ .
G ,
,. J' , ,
o % . - .- .3 i i . 0; 1 64 t
, )4
,. ie '
o( :g hf Reactor f l
[}W,o r : a
?
o "km m .s. m
- yes,e,
- k. .
b s
- h. n ll
+ , ,,
. l .
,. J,63. yg;,,!
c demyl,- p g fy i
% n gr x
, _. g i % Y
/ ?, .
\
- 6, '. h,-
] . $u x. @+
p c , ,e,
,. <,; , .o.
o ,
.i* .: ,, Wb 10 ,
v . , , ..
- @. .;ffnt: s,* ?l$*
..o. ~9
., gl *:t.
).
.h fc v d. m 3-
.E 'd H 1Lc lumE 'Oc,
. .=,- .
8
. ., s -- -
- '.f t o - . - - - h r
yo' c
j .. '
?..;,.
e ! '"" '
I$
^^
'0 5 9 b & Ye? % '.%
b k, %tbif,%%Y!EiNq,.,E i
i FIGURE 6.18 FLOWPATH FOR FISSION PRODUCT TRANSPORT AFTER VESSEL FAILURE - SEQUENCES TC AND TW 6-23 1
. __ . . . . _ _ _ . _ . . _ . . _ . _ . ~ . - , _ _ . _ . _ _ _ _ _ _ _ _ . . . _ . . _ _ , _ _ _ . . _ _ _ _ _ _ _ . . _ . _ _ _ _ . _ _ _ . _ _ . . _ _ _ _
A ma a O
. g C
d l 8 5
~
M 8 W
b N
- m g T4 a w
b N
w l
m -
' l E
/ d g E 8
$5 C 9
i
< g s
- g erm w
i 1
g O
8 0 0 $ $ $ $ $g
'VISd SunSS3Hd 6-24
140.0 PT D.W.
120.0 -
100.0 -
4 4
i r4 80.0-d M
D M 60.0-M
.m 5%
. t
, - l 40.0- ,
}, :' , ,,,
~ L i- . <
d i 20.0- '
- : .......................................-_...=_.
l- , , . .-
? ,
f ,
L O.0 , , , , ,, ; . . , , ,
., , _ C= 50.0 100.0 150.0 200.0 250.0e 300.0 350.0 400.0
~
TIME.- (MINUTE) -~
' s. ' . , r
}_ l l t a s-
-v FIGURE 6.20. ,DRYWELL (DW) AND REACTOR BUILDING (RB) PRESSURE DURING TCY SEQUENCE
/ ', ;
. 9 '
- // # y r
- p. , ' j f ,
e
$ f
7 7
g ,_ - .
hl -
m s
- s ontainment decreases substantially as
}' fail. The: steam release, rate to)th the core-uncovers and the pc.rer level drops to decay levels.
%,, .Theventilationsystemintherhactorbuildingwasassumedto
> 'centinue to operate af ter drywelPfailure(exhausting gases up the stack at a
' constant rate of 25,000 ft 3/ minute. It is important to note that the reactor-L building stays at.a positive pressure for a prolonged period of time even with
, the very largelleak area; this will result in significant leakage out of the f
butiding'inadditiontotidflowup,thestack. When.the pressure falls below j atNosphericpresside,thefansbegintodrawoutsideairintothereactorbuild- l S4 i ng .- Figure 6.21 shows the concentration of oxygen in the reactor building..
Following containment failure, essentially all of the' air (and oxygen) is
-exhausted from the byilding. Cont [r$ed operation of the fans has little.effect
, on'theair'(oxygen)concentrationungilabout175 minutes. After 175 minutes, the' relatively low gas generation ratehCring the concrete decomposition phase softhesequence:allowthereactorEbilo[rhpressuretofallbelowatmospheric pressure. Thecombinationofsteam(3ndensationonwallsandtheoperationof th fanssubsequentlydrawsoutside'aYrintothebuilding.
,, NoteinFigure6.21that$oflammablehydrogen-airmixturesare predicted in the reactor building for the TC sequence. When the hydrogen s) concentration peaks at abouti 14 percent, there is essentially no oxygen in the
% building. When the oxygen concentration increases above 5 to 6 per, cent, the hydrogen is generally below 3 to 4 percent. Such mixtures are marginally, b'ut not'ghnerally,consideredflammable. If the fire protection system in the react,or building had beeg assumed to operate, inflow of air would have occurred earlier in the sequence resulting in flavraable mixtures.
In the analysis of the reactor 6611 ding response the entire volume of the Duilding was considered to be available for the expansion of the gases and vapors released from the primary containment. Obviously a variety of paths
~
for gas (and f;ssion product) release from the primary containment, through' i,N the reactor building,'and out to the environment can be postulated in addition to the one explicitly considerer in this study. Alternate paths that are more directthantheoneconsideredwili)likelyinvolvehigherlocalpressuresand
+
, t
. shorter residence times within the building. The assumed availability of the entire reactor building volume for gas ^ expansion tends to enhance the potential
- 'for fission product deposition.
6-26 t
~
s
6.1.3 Sequence TW l In the TW sequence, the reactor shuts down, the emergency core cool- 1 ing system operates, but the heat removal system for the suppression pool-fails.
-This sequence is similar to TC in that the flow paths for fission product transport are the same. .The time scale for the TW sequence is much more protracted, however, due to the substantially lower power levels involved. As in TC, the suppression pool heats up and the containment fails prior to core meltdown. Containment failure is not predicted to occur until 28 hours3.240741e-4 days <br />0.00778 hours <br />4.62963e-5 weeks <br />1.0654e-5 months <br />, how-ever, as opposed to I hour in the TC sequence. In addition, it was assumed in
'the analysis of the TW sequence that the ope'rators would depressurize the primary system prior to core meltdown. Table 6.2 gives the timing of the key-events and Table 6.3 presents core and primary system conditions at key times for this sequence. Because of the low decay heat level, the emergency core cooling system is able to keep the primary system filled; this extends the time of core uncovery when the emergency core cooling system fails.
Figure 6.22 illustrates the heatup of selected regions of the core. The same control volume setup was used in the MERGE analysis as described previously for the TC sequence. Figures 6.23 and 6.24 show the gas and structure tempera-tures in the control volumes as a function of time. In Figures 6.23 and 6.24, time zero corresponds to 2,622.3 minutes into the accident. Flow paths in the containment are also the same as for the TC sequence. Figures 6.25 and 6.26 illustrate the time-dependent behavior of the pressure and temperature in the wetwell and drywell throughout the accident. Table 6.4 gives the containment conditions at key times and Table 6.5 summarizes the leak rates used in the fission product release calculations.
6.2 Radionuclide Sources 6.2.1 Source Within Pressure Vessel Inventory The fission product inventories used in the analyses in this report are based on results of the ORIGEN 2(6.1) code for the Browns Ferry Unit 1 6-27 s .
0 0
0 4
0 G
. 0 N I
. ,5 R 3 U D .
G
. N Dy .
- I .
Yx . D L
Ho .
. 0 I 0 U Bg
. ,0 B Rg 3 R O
T C
. A
. . E
. R 0 N I
- ,05 ) )
2E D Y
,, . T H i./ U (
/ N N E
G
/.'
0.
0MI O
R
.": / ., /
i0(
2 D
Y H
/ - D
' N A
E )
Y X
0.
0 I M 0
(
i5 T N 1
E G
Y X
O .
S N
^ S E
LY OC
. i;!!::i:I:ii i* ;:!i::iliiiiii:: :! i :' MT 0
,0 .
5 1 2
6
_ E R
U
- G I
0 F
. 0 0 8 6 4 2 0 6 6 4 0 2 1 1 1 1 1 0 0 0 M. 0 0 0 0 0 0 0 0 0 0 0 0 ZoEO4%% gaO5 py
5000 TR0(R.L) where R = Radial zone number TR0(5,6.5) l L = Distance above core bottom, ft MARCH Predictede I
l -- 2500 t
l 4000 --
-- 2000 3000 -.
u.
o 4 1500 @
4 E
.h. g c.
2
- 8. E B
F- ,
c, 2000 -
TR0(1,6.5) --
1000 m
TR01 9,11.5)
TR0(1,6.5) u TR0(5,6.5) 1000 { --
500 TR0(1,1)
TR0(5,6.5) f TR0(9,11.5) N. s f
TR0'(1,1 )
I 0 I I I 1 I I I 0 20 40 60 80 100 120 140 160 180 3
Time, 10 seconds l
FIGURE 6.22. TEMPERATURE OF SELECTED FUEL REGIONS AS A FUNCTION OF TIME - SEQUENCE TW 6-29 l
l
4500 Core Exit
-- -- --- Pl a te 4000 - - -
Head Separators
- . . . - . . . Dryers 3500 -
-- Upper Annulus
Lower Annulus
' ~ ~ ~ ~
Steam Line j
l ~'~--- Relief Line /
3000 -
,o ]
I s
i j
/ -
l l w '
- 2500 -
/ ./
- # /
- j ,/
o 4 /
4.J
- o D / ./
e
?'
i g- 2000 -
- / ./ j!
Q F-
,a* ./ : {
,/
/ -
/ : -
1500 -
~~.-
'./; , -. . /
.J x
,,, - ' i *
~~... .' :
t 2000 -
-A,,,, . _<#8N
_ . . . _ . . .~~_-._---- . . _ . -.G',W . _ . . . _ . _ b.-3.g*
300 _
m M
\
I I I I I
-0 7000 8000 9000 10,000 11,000 12,000 13,000 Time, seconds FIGURE 6.23. GAS TEMPERATURES IN RCS VOLUMES - SEQUENCE TW 6-30
- t
l l
[
4500
- - - - - Plate '
. ---- Head 4000- -
- -- Separators Dryers
- Upper Annulus 3500 _,
Lower Annulus
Steam Line
- - - - Relief Line 3000 -
u.
2500 -
,* .--}
- I L /
/ I 3- i s.
/ t b ' t E 2000 '
/ '
B ,
, l
.- i8
' ' 4}
r ni. .
1500 -
a *
- I %,
/ -
y%
..*,,,,e s s .s -.r, .//
i 1000 -
,. '/.- .
. . * . . . - ..s *
- _:./oap sv. s s. \.
f..~,,.....-....s p* .
s v.ve " , .= ,. . . . . = . * ,,,. / /,
}t ,,, %,, '
~ ~ ~ ~ - ~ ~ -"~""~~"~"*~' .
- 500 -
's, -
"J.W. .mmanmarsu.n. .............
.e..:.._. . . _...........
,_m._.:_M_ . . . . ,_.,
. . ,,g.W ,%
l l
l l
0 i i i i i 7000 8000 9000 10,000 11,000 12,000 13,000 Time, Seconds l
i FIGURE 6.24. STRUCTURE TEMPE"ATURES IN RCS VOLUMES - SEQUENCE TW
\
l l
6-31
0 M
0 b.
LL LL EE A
H WW Yf T
Rg 0 C E
Dw N E
b 0
. 0
. E S
S E
) M E U 0.
0 T L
O V
i00 U 2NI T N
E
( M I M
N
( A T
N
- O 0 C 0 N i05 E I 1 M I
E R
U T S S
4 E
R P
0 .
0 5 i0 2 0
1 6 E
R U
G I
F 0
0 d
5 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 2 1
0 1
8 8 4 2 1
+
4 4
n 4 i%mMP$mM%
+
E~
i e.
E 4
0 4
600.0 i
DRYWELL m i 500.0-(
i g f m
D3 400.0-P E*
4 yy & n 300.0-
g.
j f ' m ....,. : C, r 200.0- .-
100.0 - L
~
i i i i i ,
0.0 500.0 1000.0 1500.0 2000.0 2500.0 3000.0 3500.0 TIME - (MINUTE)
FIGURE 6.26. GAS TEt1PERATURES IN CONTAINMENT VOLUMES - SE0VENCE TW o
reactor. The computer runs were performed by Oak Ridge National Laboratory in the station blackout study for Browns Ferry Unit 1.(6.2) An actual core loading (Cycle 4). involving seven different types of fuel assemblies are described in Table 6.6. For each type of fuel, two ORIGEN 2 calculations were performed, for the maximum axial power factor and for the minimum axial power factor.
The fission product elemental. inventories from the high and low power runs were weighted and sunened to obtain the inventories for the average power for each type of fuel.
Inventories for each fuel type were summed to determine the total core inventory. In the subsequent analyses of fission product release, the core was treated as homogeneous, rather than ascribing different inventories to different radial and axial zones. The total core inventories of fission products and actinide elements are provided in Table 6.7. Masses of structural materials from the Peach Bottom FSAR are also tabulated.
Release from Fuel The rates of release from the fuel were computed using the C0RSOR code, which utilizes the MARCH predictions of core nodal temperatures as func-tions of time for each sequence. The MARCH predicted extent of r oxidation at each node as a function of time was used in evaluating the Te release from the melting core. The fractions of inventory released prior to vessel dryout and prior to vessel failure are given for each species and each sequence in Tables 6.8 and 6.9. Release from the melting core fragments is calculated up to the point of bottom head failure in these analyses. This results in each of these sequences having emission into a stagnant volume for a portion of each accident sequence, since MARCH predicts vessel dry out to occur prior to head failure by a significant margin. The differences between the values in these two tables indicate that portion of the emission which occurs while the RCS is essentially stagnant. The majority of the volatile species emissions l occurs during the period before vessel dryout, though the actual values differ among the sequences examined. The dashes in these tables indicate values less than 0.01. It is clear that only small fractions of the less volatile species 1
are released from the core during the in-vessel melting period of these sequences.
6-34_
1 Tabl'es 6.10 through 6.12 contain the mass release rates for cesium, iodine, tellurium, and aerosol particles for the AE, TC, and TW sequences at Peach Bottom. The aerosol particles are considered to be composed of the sum
.of all the nonfission products, along with fission products other than the noble gases, Cs, Te, and I.
The difference between the releases implied by Table 6.9 and the l ' initial core inventory represents the' melt composition as it enters the drywell cavity and becomes a source for fission products within the containment. This
' inventory', which is available for emissions into the drywell is presented for each of the sequences in Table 6.13. More discussion of this aspect of these analyses will be presented in the sections dealing with transport anc' aeposition in the primary system.
Regrouping of Released Species In order to track the Reactor Safety Study groups independently, an additional CORSOR run was performed for the AE sequence which produced release rates for all groups. A description of the makeup and methodology for release of each group follows.
e Group 1 (Xe, Kr) -- Xe and Kr releases were summed and a release rate computed. This group was not previously computed.
e Group 2 (I, Br) -- Br release was not considered due to an absence of data concerning Br release and its small inventory relative to I (1:16).
e Group 3 (Cs, Rb) -- Thermodynamic and physical properties of Rb justify treating it identically to Cs. As a result, the Rb inventory was lumped into the Cs inventory for treatment by C0RSOR.
Releases for I and Cs and Cs0H were combined to produce release rates for Csl which were the forms assumed to be transporting through the primary system. This assumption is based on the predicted temperatures and gas compositions com-bined with consideration of the 1jkely chemical thermodynamic equilibrium states.(6.3,6.4) e Group 4 (Te, Se, Sb) -- Se and Sb were not consi-dered based on their small inventory and lack of data concerning their behavior. Their inclusion 6-35
in this group would be further complicated by the dependence of Te release on Zircaloy oxidation. l o Group 5 (Ba, Sr) -- Ba and Sr were released )
separately and their releases sumed to form the l release rates for this group. Further, their releases were not included in the aerosol materials sum.
e Group 6 (Rh, Pd, Tc, Ru, Mo) -- Rh, Pd, and Tc inventories were added to the Ru inventory for purposes of release. The releases of Ru and Mo were then summed to produce release rates for this group. The aerosol materials sum does not include Mo or Ru releases.
e Group 7 (La, Y, Eu, Nd, Np, Sm, Pm, Pu, Zr, Ce, Nb, Pr) -- All members of Reactor Safety Study Group 7 with the exception of Zr were treated identically for purposes of release, using 002 release rate coefficients. Their release and the release of Zr were sumed to produce release rates for this group. Table 6.7.B lists initial inventories of Group 7 members not included in Table 6.7.
e Aerosol Materials (Fe, UO2 , Zr (cladding)),
-- The release rate for this group includes only nonfission products.
Table 6.7c lists initial inventories and final CORSOR releases for the Reactor Safety Study groups and compares the results with the "ungrouped" results.
It is necessary to select an initial particle size for those materials forming the aerosol species. It has been shown(5.5) that when significant agglomeration occurs, the initial aerosol size has a negligible effect on subse-quent aerosol behavior after agglomeration has proceeded for a very short time.
Nevertheless, initial particle sizes were chosen to correspond to the best available information. Numerous reviews of experimental mean aerosol sizes from vaporizing and condensing fuel indicate that the sizes will De from slightly below 0.01 um to about 0.1 um with the most likely size being about 0.05 u m.(6.6,6.7) A number median radius of 0.05 um and a geometric standard deviation of 1.7 were assumed for the primary particles in the current analyses, and a bulk density of 3 g/cc was assumed for the particles.
6-36
- _ -. . . - - . . . - - - . - ~ .. . .. -
N 6.2.2 Sources'Within the Containment P.
Release'into the'drywell region can come from,the primary circuit, l the wetwell, or.from. core-concrete interaction. The pathway from the wetwell Lhas as~its original source,' materials transported through the suppression pool but.which originated within the reactor vessel or resulted from core-concrete L interaction. ~The flow paths among compartments and the timing of events which-control the flows.were. discussed in Chapter 4 and in earlier sections of this chapter. The-source within the pressure vessel was discussed in Section 6.2.1.
Discussion of the release during the core-concrete interaction follows.
L Release from Core-Concrete Interaction, The VANESA code (described in Chapter 5) was used to make predictions i
- of aerosol and gas release rates and compositions as functions of. time'. Compo-sition of the core materials contacting the concrete was determined with the CORSOR code. These represent the materials remaining in the melt at the time of head melt-through. These compositions for the various sequences are given in Table 6.13. The concrete was taken to be a limestone concrete and the j initial temperature of the molten material was as calculated with the MARCH
!- code. The total release rates and composition of the release are given in
- Tables 6.14, 6.15, and 6.16. These rates and compositions define the source to the drywell after vessel head failure.
l l Source Term for Volatile lodides In a previous section it was noted that the thermodynamics of the
'~
cesium-iodine-hydrogen-oxygen system indicate that iodine will be present i primarily as a nonvolatile iodide in the primary coolant system.. After release j from the primary system, a small fraction of the iodine inventory in the con-tainment.is believed to be present as volatile iodides.(6.3) The presence of
. volatile iodide species in containment-type systems has been observed in experi-
- ments I6*8) and in the TMI-2 post-accident containment atmosphere.I6*9) At
.present, the mechanisms responsible for the generation of these volatile iodides.
are not well-understood. Since a theoretical model is not available, an 6-37 L
i
_ - - . _ _ _ _ _ . _ . - - . _ _ . _ . _ . _ _ . . . _ _ . - . - _ _ _ . . - _ _ . . _,. ._.. ,_~ ., , _
empirical approach has been selected for the formulation of a source term for volatile iodides. This source term consists of two components. One component represents the fraction of the containment iodine inventory which is present as volatile iodides before containment failure. The second component represents a generation rate for volatile iodides after containment failure. The contain-ment inventory of volatile iodides present prior to containment failure was estimated from levels observed in TMI-2(0*9) and from estimates of the probable detection limits in relevant experiments.(6.10) The volatile iodide generation rate was estimated from a conservative evaluation of the measurements of the airborne iodine levels in the TMI-2 containment over the time period from 100-2000 hours after reactor trip. Based on these estimates it has been assumed for this study that 0.05 percent of the containment iodine inventory will be present as volatile iodides prior to containment failure and after [
containment failure, additional volatile iodides will be generated at a rate
, of 2 x 10-7 fraction / hour of the containment iodine inventory.
Of this volatile iodine source, it is believed that a fraction of the iodine inventory in a reactor containment will be present as volatile organic iodides (predominantly CH 3 I).I
- I (Other volatile species may also be present.) Therefore, in the analysis of reactor accidents involving a radio-nuclide release from the reactor s) Pem and containment failure, formation in the containment and subsequent release of organic iodides should be consi-dered. Unfortunately, the mechanism responsible for the generation of organic iodides has not yet been elucidated. As a result, it is not yet possible.to establish a definitive source of organic iodides. Early estimates of the ,
organic iodine source terms were based on a conservative interpretation of i experimental systems studies.IO*0'0*II) Early thermodynamic studies predicted that organic iodides should be present in much smaller concentrations than observed in experiments.I
- I These calculations predicted that CH3I would comprise only 10-4 percent of the total gaseous iodine inventory modeled.
Experimental dataIO*0) and " chemical species specific" measurements of the TMI-2 airborne iodine inventory (6.9) imply that the concentration of organic iodides present in a reactor containment during and following an accident may be higher than the concentrations predicted by thermodynamic calculations for 1 an equilibrium system. Additionally, observations of the airborne iodine behavior at TMI-2(6.9) imply the presence of competing sources and sinks for 6-38 1
. volatile iodine species. In light'of these d'ata, a kinetic description may be
. required to adequately quantify the time _ dependence of the organic iodide con-centration in reactor containments during-and following re'ictor accidents.
Pending results of studies, such as those which are currently under way,(6.4) use of a general. source term for volatile iodides rather than separate source terms for CH3 I, I2, etc., has been assumed as noted above.
i f
a 6
6 l
l d
I 6-39
References (6.1) Groff, A. G., "0RIGEN 2 -- A Revisd and Updated Version of the Oak Ridge Isotope Generation and Depletion Code", ORNL-5621 (July, 1980).
. (6.2) Wichner, R. P., et al, " Station Blackout at Browns Ferry Unit One --
Iodine and Noble Gas Distribution and Release", NUREG/CR-2181 (August,1982).
(6.3) Technical Base for Estimating Fission Product Behavior During LWR Accidents,NUREG-0772,(June,1981).
(6.4) Torgerson, D. F., et al., Fission Product Chemistry Under Reactor Accident Conditions, presented at the International Meeting on Thermal Nuclear Reactor Safety, Chicago, Illinois, U.S.A.
(September, 1982).
(6.5) Jordan, H., Schumacher, P. M., and Gieseke, J. A., " Comparison of QUICK Predictions with Results of Selected, Recent Aerosol Behavior Experiments", NUREG/CR-2922, BMI-2089 (September, 1982).
(6.6) Gieseke, J. A., et al, " Aerosol Source Term for Fast Reactor Safety Analysis", BMI-X-637 (August 11,1972).
(6.7) Nuclear Aerosols in Reactor Safety, CSNI/ SOAR No. 1 (June, 1979).
J (6.8) Postma, A. K. and Zavadoski, R. W., Review of Organic Iodide Forma-tion Under Accident Conditions in Water-Cooled Reactors, WASH-1233 (1972).
(6.9) Pelletier, C. A., et al., Preliminary Radiciodine Source Term and Inventory Assessment for TMI-2, SAI-139-82-12-RV (September, 1982).
(6.10) Lin, C. C., Chemical Effecs of Gamma Radiation on Iodine in Aqueous Solutions, J. Inorg. Nucl. Chem., g , pp. 1101-1107 (1980).
('.11) 6 Reactor Safety Study: An Assessment of Accident Risks in U.S.
Commercial Nuclear Power Plants, WASH-1400 (1974).
(6.12) Barnes, R. H., Kircher, J. F., and Townley, C. W., Chemical 'Equili-brium Studies of Organic Iodide Formation Under Nuclear Reactor Acci-dent Conditions, BMI-1816 (1966).
6-40
- TABLE 6.1 REACTOR CHARACTERISTICS, CONTAINMENT PARAMETERS, AND MARCH OPTIONS FOR THE BWR MARK I CONTAINMENT l
l Weight Zircaloy in core: 144,382 lb (65,490 kg) l
! Weight otSer metal in core: 15,680 lb (7,112 kg)
Weight UO in core: 351,440 lb (159,410 kg) 2 Weight of support structures 106,238 lb (48,188 kg) included in debris:
. Weight of bottom head: 207, 500 lb (94,120 kg)
Bottom head diameter: 20.9 ft (6.4 m)
Bottom head thickness: 0.576 ft (17.6 cm)
Containment Parameters Number of compartments: 2 Compartment 1: Drywell Compartment 2: Wetwell
__ ciae Volume Initial Temperature Initial Pressure ft 3 ,3 *F C psia MPa Drywell 159,000 4,503 100 37.8 14.7 0.1 Wetwell 119,000 3,370 100 37.8 14.7 0.1 Thermal Conductivity Heat Capacity Density 3
Material Btu /hr ft F W/cm C Btu /lb F J/kg K lb/ft kg/m 3 Iron 25. 0.4325 0.113 473.1 487 7801.7 Concrete 0.8 0.01384 0.238 996.5 158 2531.2 i Heat Transfer Thickness Area ft m 7t2 mP Wetwell shell 0.053 1.61 5 17,060 1,584.9 Drywell concrete 1.0 0.305 7,238 672.4 Drywell concrete 1.5 0.457 2,262 210.1 Drywell concrete 3.0 0. 91 4 1,005 93.4 Drywell shell 0.083 2.530 17,700 1,644.3 Misc steel 0.042 1.280 39,180 3,639.8 i
6-41 l
TABLE 6.1 (Continued)
Steel / concrete interface coefficient: 100 Btu /hr ft 2F-(0.0568 W/cm2/C)
Initial slab temperature: 100F(37.8C)
Concrete Composition Weight fraction CACO : 0.8 3
Weight fraction Ca(OH)2: 0.15 Weight fraction SiO : 0.01 2
Weight _ fraction Al 0 : 0.01 23 Weight fraction free H 0: 0.03 2
Gm rebar per gm concrete: 0.135 r
Engineered Safety Systems Rated Flow- Pressure ECC Pump GPM 1/s psig MPa High Head 5,000 3,500 24.2 Low Head 40,000 20 0.24 Heat exchanger rated conditions:
Rated Capacity, Btu /hr 7. x 10 (2.05 x 108 y)
Primary flow, lb/ min 83,429 (630.7 kg/s)
Secondary flow, lb/ min 37,543 (283.8 kg/s)
Primary inlet, F 165 (73.9 C)
Secondary inlet, F 90(32.2C) 6-42 l
-.__-.-_-_____m-.- _ - - - - _ _ - - t- -* - - - - ' -
l l
TABLE 6.1 (Continued)-
i ECC Storage and Injection Tanks CST 6
Weight of water 3.108 x 106 lb 1.41 x 10 kg l
Initial pressure 14.7 psia 0.1 MPa Temperature 80 F 26.7 C Fractional value of CST to start ECC recirculation: 0.033 Large LOCA Blowdown Enthalpy Blowdown Rate Time, min Bru/lb J/kg lb/ min kg/s 6 3 0.0 545.5 261,029 1.0194 x 10 7.707 x 10 0 3 0.5 545.5 261,029 1.0194 x 10 7.707 x 10 Water remaining in vessel: 132,010 lb Calculational Model Input Core heatup section:
Number of radial zones: 10 Number of axial zones: 24 Meltdown model: BOIL Model A Core slumping starts when lowest node in region is molten.
Core collapse occurs when 75% of core has melted, or core support reaches melting temperature.
Zircaloy-water reaction: Urbanic-Heidrick reaction rate data, steam limited, continues for melted nodes, reaction of Zircaloy with water in bottom ha d calculated.
6-43
TABLE 6.1 (Continued)
Bottom head failure section:
Head melting temperature: 2800 F (1538 C)
Debris melting temperature: 4130 F (2277 C)
Heat loss from top of debris: none Debris thermal conductivity: 8 Btu /hr ft F (0.1384 W/cm C)
Tensile strength of vessel: o = min ( 80,000, 1.49 x 1016 ,
TEMP-3.9105),lb/in2 Reactor cavity processes, debris fragmentation:
Particle diameter: 0.125 inch (0.32 cm)
Particle thermal conductivity: 2 Btu /hr ft F (0.0346 W/cm C)
Reactor cavity processes, concrete decomposition:
Metal-concrete interface heat 2 transfer coefficient: HIM = 0.01 W/cm K 0xide-concrete interface heat 2 transfer coefficient: HIO = 0.01 W/cm g Top surface emissivity: E = 0.5 Heat to cover water: film boiling and radiation Containment section:
Atmosphere-wall heat transfer coefficient: h = bc (TSAT - WALL) + 0.19(T - WAL[/
(T - TWALL) hc = 0 if TSAT < WALL 2
2.0 < h" = Uchida data < 280 Btu /hr ft p 2 2 Containment break area: 7.0 ft for overoressure failure (0.65 m )
l' Failure of safety systems:
Containment failure fails ECR if sump is saturated.
i 6-44 l
TABLE 6.2 ACCIDENT EVENT TIMES Event Time, minutes Peach Bottom AEY i
Core Uncover 1.5 l Suppression Pool Cooling On 10.0
. Start Melt 11.5 Core Slump 2C.8 Containment Fail 33.9 Bottom Head Dry 40.0 Core Collapse 65.2 Bottom Head Fail 126.2 Reactor Cavity Dry 126.3 l Start Concrote Attack 126.3 End Calculation 727.0 i
Peach Bottom TCY Containment Heat Removal On 10.0 Containment Fail 58.1 ECC Recirculation On 72.4 ECC Off 72.6 Core Uncover 73.0
' Start Melt 93.6 Core Slump 124.6 Bottom Head Dry 136.6 Core Collapse 178.9 Bottom Head Fail 216.6 Reactor Cavity Dry 216.7 Start Concrete Attack 216.7 End Calculation 816.9 6-45
TABLE 6.2 (Continued)
Event Time, minutes Peach Bottom TWY '
l Containment Fail 1756.2 i Core Uncover 2619.6 Start Melt 2747.9 Start Slump 2817.1 Core Collapse 2818.9 Bottom Head Dry 2829.3 Bottom Head Fail 3055.2 Reactor Cavity Dry 3055.2 Start Concrete Attack 3055.2 End Calculation 3655.4 i
f' i
6-46
TABLE 6.3 CORE AND PRIMARY SYSTEM RESPONSE Primary Prinary System System Water Average Core Peak Core Fraction. Fraction Accident Time, Pressure, Inventory, Temperature, Temperature, Core Clad Event minutes psia lbm F F Melted Reacted Peach Bottom AEY Core Uncover 1.5 51.5 1.31 xld 1277 2110 0. O.
5 Start Melt 11.5 30.3 1.24 x 10 2233 4130 0.00 0.01 5 0.04 Start Slump 26.8 30.0 1.23 x 10 3229 4170 0.24.
4 Containment Fail 33.9 132.05 6.58 x 10 2558 4130 0.62 0.34 3
Bottom Head Dry 40.0 24.2 1.13 x 10 1674 5072 0.62 0.36 2
Core Collapse 65.2 14.8 9.81 x 10 2805 5995 0.75 0.36 p Bottom Head Fail 126.2 14.8 2.22 x 10 4 3784 --- ---
0.36 O
Peach Bottom TCY 5
Containment Fails 58.1 1237 2.40 x.10 727 922 0. O.
Core Uncover 73.0 1200 1.97 x 10 5 682 802 0. O.
Start Melt 93.6 1149 1.75 x 10 5 1461 4130 0.00- 0.01 5
Start Slump 124.6 1124 1.575 x 10 3400 4499 0.53 0.26 Bottom Head Dry 136.6 1126 1150 2401 4130 0.50 0.47 Core Collapse 178.9 1120 982.6 3073 4130 0.67 0.47 4
Bottom Head Fail 216.6 1120 2.39 x 10 3517 --- ---
0.47
TABLE'6.3(Continued)
Primary Primary System -
System Water Average Core Peak Core' ~ Fraction Fraction -
Accident Time, Pressure, Inventory, Temperature, Temperature, Core Clad Event minutes psia lbm 'F F ' Melted- Reacted Peach Bottom, TWY Containment Fail 1756.2 285.5 6.551 x 10 5 416 420 0; 0.
5 Core Uncover 2619.6 170.3 3.39 x.10 372 - 376- 0. O.
Start Melt 5 2747.9 170.1 2.31 x 10 1212 4130 0.01 0.01:
Start Slump 5 2817.1 172.0 1.90 x 10 3591 4130- 0.77 0.60 Core Collapse 5 2818.9 171.5 1.61 x 10 4130 4130. 0.82 0.61 Bottom Head Dry 2829.9 172.3 1765 1588 -- -
0.61
, Bottom Head Fail 3055.2 169.9 1191 3655 ---
---. .0.61 b
d c - _ _ _ _ _ - - - - . - - . _ _ _ - . - - - - - - - - - - - - - - - - -
TABLE 6.4 CONTAINMENT RESPONSE Reactor Compartment Compartment Suppression Pool Reactor ' Cavity' Steam Pressure. Temperature. RWST or CST Water Cavity Water 'Cond.
psia F Temp., Water Mass, Temp., .on Walls Accident Time, Water Mass,. Mass, ~
Event minutes 1 2 1 2 'I bm Ibm F lbm F' Ibm / min Peach Bottom, AEY 6 4 Core Uncover 1.5 35.0- 35.0 246 171 8.88 x 10 121' L2.37 x 10 241- 3308/0.
Spray On 10.0 30.3 30.3 231 122 8.93 x 10 6 123 2.38 x .104 232 160/25 4
Start Melt 11.5 30.3 30.3 231 122 8.95 x 10 6 122 2.39 x 10 230 153/22 Start Slump 26.8 30.4 30.4 235 121 8.95 x 106 121 2.39 x 104 '227 0/10 Containment Fail 33.9 131.8 131.8 2031 332 9.0 x 10 6 129 2.32' x 10 4 197 '0/0 Bottom Head Dry 40.0 15.4 15.4 466 124 9.0 x 10 6 128 2.26 x 104 172 0/0 Core Collapse 65.2 14.8 14.8 385 188 9.0 x 10 6 128 2.25. x 104 170 .0/0-Bottom Head Fail 126.2 15.4 15.4 194 202 9.0 x 10 6 128 2.25 x 10 4 169 0/0 Start Debris / 2.25- x 10 4 169 0/0 Water Interaction 126.3 29.3 29.3 490 276 6 0/0 Cavity Dryout 126.3 31.7 31.7 488 284 9.02 x 10 131 0. ---
Start Concrete 0/0 Attack 126.3 31.5 31.5 C7 285 9.02 x 106 131 0. ---
6 End Calculation 727.0 15 C 15.6 828 181 9.02 x 10 131 .O. ---
0/0 Peach Bottom, TCY Containment Heat 6 6 10.0 18.8 18.8 122 167 2.5 x 10 9.04 x.10 169 288 116 0/291 Removal On 5 0 .4 Containment Fail 58.1 132.6 132.6 326 346 6.11 x 10 11.2 x 10 348 2.41 x 10 207 1061/0 6
ECC Recirc. On 72.4 90.8 90.8 321 321. 11.02 x 10 333 2.41 x 10 4 207 0/0 4 6 ECC Off 72.6 90.5 90.5 321 321 9.94 x 10 11.01 x 10 333 2.41 x 10 207 174/0 Volume 1/ Volume 2 1
TABLE 6.4 (Continued)
Compartment Compartment Temperature, Suppression Pool Reactor Cavity ' . Steam Pressure. Wa ter Accident Time, psia F RWST or CST Cavity ' Water ~ Cond.
Water Mass, Mass.- ' Temp., Water Mass, Temp., on Walls Event minutes 1 2 1 2 lbm lbm F lbm- F ' Ibm / min Peach Bottom, TCY (Continued). '
Core Uncover 73.0 89.7 89.7 320 320 9.94 x 104' 10.94 x .106 326 2.41 x 10 4 207 0/0 Start Melt 93.6 48.8 48.8 280 288 9.94 x 104 10.44 x 10 6 275 2.41 x 10 4
207 0/0-Start Slump 124.6 21.8 21.8 230 244 9.94 x 10 4 9.89 x 10 6 229 4
2.41 x 10 . 207 0/0 Bottom Head Dry 136.6 25.9 25.9 242 260 9.94 x 10 4 9.85 x 10 6 236 2.41 x'10 4
207 0/0 Core Collapse 178.9 15.0 15.0 270 255 9.94 x 104 9.60 x 10 6 213 2.41 'x 10 4 207. 0/0 Bottom Head Fail 216.6 27.7 27.7 655 356 9.94 x 10 4 9.60 x 10 6 217 2.41 x 10 4 207 0/0 .
Start Debris /
Water Interaction 216.6 25.8 25.8 1231 347 9.94 x 10 4 9.60 x 10 6 217 2.41 x 10 4 207 -0/0 Cavity Dryout 216.7 29.9 29.9 496 366 9.94 x 10 4 9.61 x 106 216 0. '0/0 Start Concrete Attack 216.7 32.1 32.1 443 375 9.94 x 106 9.61 x 10 6 216 29. 254 0/0-End Calculation 816.9 15.3 15.3 746 272 9.94 x 10 6 9.54 x 106' 209 0. ----
0/0 Peach Bottom, TWY 5
Containment Fail 1756.2 130.1 130.1 324 347 1.025 x 10 11.02 x 10 6 348 2.45- x.10 4 158 39/13 Core Uncover 5 4 2619.6 15.3 15.3 242 225 1.025 x 10 2.45 x 10 158 Start Melt 2747.9 15.0 15.0 244 226 1.025 x 10 5 9.49 x 106~ 213 2.45 x 10 4
158 0/0 !
Start Slump 2817.1 16.8 16.8 225 220 +
Core Collapse 2818.9 16.3 6 16.3 223 218 9.44 x 10 211 Volume 1/ Volume 2
- - - _ _ = _ _ _ _ _ _
TABLE 6.4 (Continued)
Compartment Compartment Reactor Suppression Pool Reactor Cavity.' Steam Pressure, Temperature, psia F RWST or CST. Water Cavity- Water ' Cond.
' Accident Time, Water Mass, Mass... Temp., Water Mass, Temp., 'on Walls Event minutes 1 2 1 2 1hm lha F 1ha 'F lbm/ min Peach Bottom, TWY (Continued)
Bottom Head Dry 2829.3 18.2 18.2 234 223 Bottom Head Fail 3055.2 15.0 15.0 213 231 9.435 x 106 213 2.45 x 10 4 158 Start Debris /
4
. Water Interaction 3055.2 17.2 17.2 467 254 2.45 x 10 158 0/0*'
6 Cavity Dryout 3055.2 21.4 21.4 403 292 9.46 x 10 216- O.
, Start Concrete Attack 3055.2 15.2 397 363 236 5
End Calculation 3655.4 15.3 15.3 618 229 1.02 x 10 9.42 x 106 211- O. 0 0/0 ,
e
$1
- Volume 1/ Volume 2 k
9 n
i 4
Revised 7/12/83 TABLE 6.5 CONTAllWENT LEAK RATES Peach Bottom 2
Time. Leak Rate {a)
Drywell Pressure Drywell Temp. Wetwell Pressure Wetwell Temst Subsequence min v/ min Wa psia 'F 'C Wa psia *F 'C Remarks AEY 0.5 6.9 x 10' O.36 52 283 139 0.36 52 258 126 End of blatalown 0.5-33.9 6.9 x 10' O.44 64 1093 590 0.44 64 211 100 Core heating and melting 33.9 2.7 0.91 132 2031 1111 0.91 132 332 167 Containment fails 33.9-40.0 1.8 0.20 29 1170 632 0.20 29 126 52 Reactor vessel dryout 40.0-66.0 0.03 0.10 15 385 196 0.10 15 183 84 Vessel hr.atup 66.0-126.2 0.03 0.11 15.3 341 172 0.11 15.3 188 87 Vessel melting 126.2 0.35 0.11 15.4 194 90 0.11 15.4 202 94 - Bottom head fails 126.2-156.3 0.6 0.11 16.4 460 238 0.11 16.4 173 78 Initial concrete attack 156.3-727 0.06 0.11 15.6 796 - 425 0.11 15.6 180 82 Concrete decomposition TCY 58.2 1.5 0.92 133 326 163 0.92 133 347 175 Containment fails 58.2-124.5 1.5 0.46 67 294 146 0.46 67 296 147 Core heating and melting 124.5-136.6 1.2 0.17 24 245 118 0.17 24 247 119 Reactor vessel dryout 136.6-180.0 0.4 0.11 15.8 259 126 0.11 15.8 252 122 Vessel heatup m 180.0-216.6 1.8 0.17 24 1093 589 0.17 24 328 ~165 Vessel melting g 216.6 2.2 0.14 21 1307 708 0.14 21 320 161 Bottom head fails 216.6-218.3 1.2 0.15 22 346 174 0.15 22 285 141 Initial concrete attack 218.3-245.5 0.24 0.12 17 366 186 0.12 17 256 125 Concrete decomposition 245.5-818.0 0.12 0.11 15.4 648 342 0.11 15.4 274 135 Concrete decomposition TWY(w/ ADS) 1756 2.7 0.91 132 326 163 0.91 132 347 175 Containment fails 1756-2748 0.24 0.10 15.2 242 117 0.10 15.2 225 107 Core heating and melting 2748-2819 0.6 0.11 16.3 236 113 0.11 16.3 222 106 Core slumping 2819-3055 0.24 0.10 15.3 244 118 0.10 15.3 229 109 Vessel heating and melting 3053 0.25 0.10 15.0 213 101 0.10 15.0 218 103 Bottom head fatis 3055 3085 0.06 0.10 15.2 350 171 0.10 15.2 227 108 Initial concrete attack 3085-3365 0.03 0.11 15.3 409 209 0.11 15.3 230 110 Concrete decomposition 3365-3657 0.24 0.11 15.4 614 323 0.11 15.4 228 109 Concrete decomposition (a) Normalized to a drywell free volume of 1.59 x 105 gg 3. Units are volume fraction / min.
i
TABLE 6.6 FUEL CHARACTERISTICS IN CYCLE 4 AT BROWNS FERRY UNIT 1(6.2)'
- ~
Initial'U Initial Gd Approxima}e
.Burnup(ai
. Fuel Number of Cycle Array Loading Enrichment Loading Type Assemblies Inserted Size (kg) (%) (g) (mwd /MT)'
2 87 1 7x7 187.06 2.50 441 30,400<
3 127 1 7x7 186.93 2.50 547 23,800 4 140 2 8x8 183.361 2.74 292' 22,900 5 23 2 8.x 8 183.361 .2.74 442 24,000 6 87 3 8x8 182.52 2.65 355 .16,600.
7 68 3 8x8 -182.32 2.65 537 16,900
- .8 232 4 8x8 182.185 2.84 330 8,900
= . ,
'~ #
(a) Burnup calculated through 11 months of Cycle 4 operation... #
.j
y - a
_[ s, ,
~
~y l ;
y, . . . .,
g ? ,
, , , a. >> '
b
[
j' , . <
p s }~G" 's _
%~ ',
=<,a f
+
. - *
- j M f 9
- W n
9 I %.
S
= ..
.r s
{
gy TABLE 6.7. INVENTORIES OF RADIONUCLIOES AND STRUCTURAL MATERIALS
.c FOR PEACH BOTTOM DURING IN-VESSEL PERIOD OF MELTING Fission Products Actinides / Structural Element Mass (kg) Element Mass (kg)
Kr 25.7 U 138,000 Rb 23.3 Pu 743
., Sr 62.7 Y %~ 36.2. 4,140 Cr I
Zr , 267 Mn 432 Mo j- 237 Fe 15,150 Tc ', 58.8 Ni 2,560
?
Ru 172 i ;.' Zr 64,100 33.2 Rh -
Sn 1,050 Pd ,3 .
83.2 "c Gd 287 Te ', - <
34.9
' s s
I 3 16.6
, , Xe' , y 1
-I 387 Cs -( 207 Ba 105
' La : 98.3-
,r .
Ce J 208 s Pr V '
80.4 l, Nd A '
271 ,
Sm '
53.8 s
- ' ,o .
k
' ^
w lw.,
'N
\:
TI
,+\
6-54 /
.. 9
@\ '
(' _g ~, 1 7;
,-*----w
TABLE 6.7b. INITIAL INVENTORIES OF ADDITIONAL SPECIES INCLUDED IN GROUPED RELEASE CALCULATIONS i
-Element Mass (kg)
Eu 14.1 Nb 4.3 Np 41.2 Pm 11.5 t
TABLE 6.7c. CORSOR RESULTS WITH AND WITHOUT RSS GROUPINGS FOR THE AE SEQUENCE Without RSS Grouping With RSS Grouping Initial Initial Inventory Final CORSOR Inventory Final CORS0R Group (kg) Releases (kg) (kg) Releases (kg)
Xe 413 413 413 412 I 16.7 16.6* 16.7 16.7*
Cs 207 207* 230 230*
Te 34.8 7.7 34.8 6.9 Sr NA NA 168 22.6 Ru NA NA 584 29.9 La NA NA 2612 1.23 Aerosol 219144 890 218300 835 Not Grouped Grouped Csl 34.1 34.1 Cs0H 213 239 l
r
.*CORSOR releases these species in the form of Csl and Cs0H.
- c. 6-55 i
TABLE 6.8 CORS0R PREDICTIONS OF FRACTION OF INVENTORY EMITTED
. BY C0RE~ PRIOR T0 VESSEL DRYOUT FOR THE.THREE ACCI-DENT SEQUENCES ~FOR PEACH BOTTOM I
Core Inventory i Species- AE TC TW (kg)
'Xe 0.79' O.75 '0. 94 387 Kr 0.79 0.75 0.94 26 I 0.77 0.73 0.92 17 Cs 0.80 0.75 0.95 207 Te 0.10 0.17 0.41 35-Sr 0.01 0.04 0.10 63 Ba 0.06 0.10 0.21 105 Ru -- --
- 0. 01 172 Mo 0.03 0.04 0.09 237 Zr FP) -- -- --
267 U0 a) -- -- --
138000 Sn a) 0.22 0.30 0.52 1050 Zr(a) Clad -- -- -- 641 00 Fe(a) 0. 01. 0.01 0.02 15150 E IE I (a) Nonfission product species.
i l
6-56
(
L TABLE 6.9. CORSOR PREDICTIONS OF FRACTION OF INVENTORY EMITTED BY CORE PRIOR TO VESSEL FAILURE FOR THE THREE ACCI-DENT SEQUENCES FOR PEACH BOTTOM l
l l
! Core Inventory Species AE TC TW (kg)-
Xe 1.00 0.97 1.00 387 ,
Kr 1.00 0.97 :1.00 26 I 1.00 0.95 1.00 17 Cs 1.00 0.98 1.00 207 Te 0.22 0.25 0.48 35 Sr 0.07 0.06 0.12 63 Ba- 0.17 0.17 0.27 105 Ru 0.01 0.01 0. 01 172 Mo 0.12 0.10 0.16 237 Zr(FP) -- -- -- 267 UO (a) -- -- --
138000 Sn a) o,47 o,44 0.63 1050 Zr(a) Clad -- -- --
64100 i Fe(a) 0.02 0.02 0.03 15150 i.
(a) Nonfission pmduct species.
i-f 6-57 L
TABLE 6.10. CORE RELEASE RATES INTO PRIMARY SYSTEM PREDICTED BY CORSOR FOR AE SEQUENCE FOR THE PEACH BOTTOM PLANT Time Mass Release Rate (g/s)~
(s) Cs -I 2 Te Aerosol 0 0.0 0.0 0.0 0.0 51 0 50.4 1.9 0.1 5.3 1 690 24.3 6.1 0.6 80.6 930 114 9.4 - 1.3 111 1170 112 9.3 2.1 274 1410 93.7 7.8 2.7 342 1650 1 94 16.1 2.4 3 06 1890 24.2 2.0 1.0 117 2130 3.8 0.3 3.8 144 2370 20.0 1.7 0.8 81.0 2610 9.2 0.8 0.4 122-2850 8.8 0.7 0.5 124 3870 5.6 0.5 0.2 33.9 5070 14.8 1.2 0.7 60.4 5670 17.4 1.7 1.4 146 6450 0.1 --
2.2 267
-6990 0.0 0.0 1.7 236 6-58
_ ,._y - _
r.r--- t - i-* -
7-- w- - -- - -- - -
TABLE 6.11. CORE RELEASE RATES INTO PRIMARY SYSTEM l PREDICTED BY CORS0R FOR TC SEQUENCE FOR THE PEACH BOTTOM PLANT Time Mass Release Rate'_(q/s) I (s)_ Cs I Te Aerosol 0 176 6.2 0.2 11.6 120 27.9 2.1 0.2 22.2 360 46.3 3.8 0.5 67.5 600 59.2 4.9 1.0 122 840 74.6 6.0 1.5 184 1080 74.5 6.2 2.0 244
-1320 81.8 6.8 2.6 312 1560 79.5 6.6 3.2 375 1800 125.4 10.4 6.3 590 2040 34.2 2.9 2.4 125 2280 0.8 --
2.3 23.2 2520 3.3 0.3 ' O.9 26.3 2760 4.2 0.3 0.7 29.5 3000 0.8 --
0.2 4.0 3240 0.0 --
0.2 1.3 3480 0.0 0.0 0.2 2.0 3720 0.0 --
0.3 2.8 4380 0.5 --
0.2 10.4 5460 2.2 0.2 0.2 29.1 7380 5.0 0.4 0.7 81.1 9900 23.8 2.1 0.6 45.4 l-l l 6-59 l
l
, = . .. . .. , . - ~ - .
I 4
4 TABLE 6.12. CORSOR PREDICTIONS OF CORE RELEASE RATES INTO PRIMARY l SYSTEM FOR THE TW SEQUENCE.FOR THE PEACH BOTTOM PLANT ~
Time Cs Ig . Te Aerosol (s) (g/s)' (g/s) '(g/s) -(g/s) 0 83.5 ,3.5 0.1 14.5
.180 13.4 0.7 0.1 12.2 540 14.0 1.2 0.2 28.9 900 22.6 1.7 0.4 '50.1 1260 24.1 2.0 0.7 80.9 1620 30.6 2.4 1.0 112-1980 33.1 2.7 'l .2 139
! 2340 33.1 2.6 1.5 168 2700 45.8 3.6 1.8 207 3060 55.8 4.6 2.8 2 71 3420 88.1 7.3 9.0 481 3780 29.7 2.5 9.4 773-4140 1 31 10.9 11.7 549 4500 7.8 0.6 0.4 37.2 4860 0.0 0.0 0.0 0.3 8040 0.1 --
0.0 0.7 11940 0.7 0.1 0.1 4.6 l
13740 1.6 0.1 0.2 13.2 16560 1.6 0.1 0.5 59.9 18590 0.0 0.0 0.6 97.1 1
i 6-60
TABLE 6.13. MELT COMPOSITION AT TIME OF BOTTOM HEAD FAILURE, l EXCLUSIVE OF' BOTTOM HEAD MOLTEN MASS, DETERMINED FROM CORSOR PREDICTIONS l-Inventory After RPV Failure (kg)
Species AE . TC TW Cs .1 5.1 0.1 I2 -- 0.8 --
Xe 0.5 10.0 0.6 Kr -- 0.7 --
Te 29.2 26.3 17.9 Ba 86.7 87.6 76.3 Sr 58.7 59.0 55.7 Mo 209 21 3 199 Zr 267 267 267 Ru 171 171 170 Sn 554 5 91 384 Fe 14918 14919 14759 Zr(Clad) 64090 64090 64083 00 137899 137947 137836 2
i 6-61 l
l
TABLE 6.14. AEROSOL COMPOSITION AND TOTAL RELEASE RATE FOR PEACH BOTTOM, AE Species. Time,sec 0 1200 2400 3600 4800 6000 7200 8400 9600 1
5.47 11.54 12.26 12.84 13.37 14.18 15.58 16.32 Fe0 0.12 2.62 0.19 0.13 0.11 0.096 0.082 0.071 0.059 Cr23 0 -
0 0 0 0 0
- . M1 0.07 0.03 0 0
-6 -6
-6 -6 4 x 10 -6
~0 Mo 3 x 10 -8 6 x 10-6 4 x 10 -6 3 x 10 3 x 10 2 x 10 2 x 10 1 x 10
-5 -5 -5 -5 -5 Ru 2.4 x 10
-I 4 x 10-5 3 x 10-5 3 x 10 2.6 x 10-5 2 x 10 2 x 10 1 x 10 1 x 10 2.89 3.73 3.41 ?.09 2.78 2.48 2.27 2.06 Sn 0.91 0.98 1.32 1.06 0.86 0.71 0.60 0.52 0.45 Te 0.76 11.35 12.06 12.63 13.15 11.90 10.15 8.63 m 4.63 5.38 6.04 12.75 13.82 14.55 15.12 15.89 '17.24 17.89 Ca0 --
4.M 4.52 5.05 4.97 4 .51 3.86 2 x 10 2 x 10 A1 0 --
23 0.14 0.26 0.32 0.33 0.34 0.35 0.39 0.38 cn Na 0 --
, 2 13.52 14.35 15.75 16.45 O K0 2
-- 6.36 11.44 12.30 12.95
-- 11.78 24.83 26.39 27.62 28.76 30.52 33.52 35.12 510 2
1.46 1.29 1.03 0.83 0.66 0.51 0.41 0.33 00 0.89 2
Zr0 2 6 x 10-3 0.54 0.074 0.063 0.053 0.043- 0.034 0.007 0.008 0 0 0 0 0 0 0 0 Cs20 31.55 13.68 2.61 1.85 1.40 1.10 0.90 0.77 0.65 Ba0 34.02 10.15 2.60 1.43 0.90 0.607 0.437 0.34 0.26
, Sr0 22.%
La23 0 1.8 x 10 10.53 3.57 2.M 2.M 1. M3 1,46 1.13 1xd 7.87 5.88 4.45 3.34 2.46 1.85 1.39 l Ce0 4.22 16.88 2
Mb 0 9x 0 5xd 9x d 7xM 6xd 5xd 3xd 3x d 2xd 25 0 0 0 0 0 0 0 Csl 0.61 6 0.062 Source 230 212 189 176 201 613 336 283 253 Rate.
g/sec
__ _ _ - ~_. _ _ _ , ._ _ ___ -
TABLE 6.14. (Continued)
Species. Time, sec '~
1 10B00 12000 13200 14400 15600- 16800 18000 Fe0 16.74 17.05 17.33 17.59' 19.15 21.08 1.02 Cr 0 0.05 0.M 0.M 0.M 0.035 0.033 0.38 23 Ni 0 0 0 0 0 0 0 Mo 1 x 10 -6 8 x 10 ~7 7 x 10'I 6 x 10'I 6 x 10~7 5.6 x 10 -7 0.002 Ru 8 x 10-6 7 x 10-6 6 x 10 -6 5 x 10-6 5 x 10-6 4.6 x 10-6 2 x 10-5 Sn 1.88 1.75 1.63 1.53 1.53 1.55 27.44 Te 0.38 0.33 0.29 0.25 0.24 0.23 1.08 m 7.44 6.54 5.79 5.14 5.00 4.85 22.98 Ca0 18.22 18.49 18.72 18.92 20.51 22.50 7.41 A1 0 23 2 x 10 2 x 10 2 x 10 2 x 10 2 x 10 3 x.10 1.5 x JO Na 0 0.39 0.39 0.40 0.39 0.43 0.48 0.13 2
i K0 16.81 17.03 17.17 17.25 18.51 19.91 28.63 g 2 510 36.01 36.69 37.29 37.85 33.76 28.71 5.80 2
U0 2
0.272 0.233 0.202 0.175 0.163 0.154- 5.06 8 x 10-3 8 x 10'3 -3 Zr0 2
8 x 10-3 8 x 10 9 x 10-3 0.0% 0.050 Cs 0 0 0 0 0 0 0 0 2
Ba0 0.54 0.45 0.36 0.29 0.23 0.16 0.018 SH) 0.20 0.16 0.12 -4 0.094 0.073 0.048 'd.8 x 10 La23 0 1 x 10 1 x 10 1 x 10 9 x 10 1 x 10 l 'x 10 5 x 10 4 Ce0 2 1.06 0.83 0.64 0.47 0.35 0.22 1 x 10-3
-5 -5 -5 -5 -3 Nb 0 25 1.7 x 10 1.4 x 10-5 1.3 x 10 1.2 x 10 1.2 x 10 1.3 x 10-5 2.3 x 10 Cs! 0 0 0 0 0 0 0 Source Rate, 169 164 159 154 134 116 22 g/sec a
1 -
s TABLE 6.15. AEROSOL COMPOSITION AND TOTAL RELEASE RATE FOR PEACH BOTTOM. TC 4
Species. Time, see 1 0 1200 2400 3600 4800 6000 7200 8400 5600 Fe0 6.03 4.82 8.64 1G.38 15.88 16.36 16.86 17.47 17.87 Cr230 - 3.M 0.17 0.27 0.22 0.18 0.15 0.12 0. m Ni 0.08 0.84 0.21 1.6 x 10-6 ,, ,, ,, , , . ,,
4 -6 8 x 10'I 6 x 10' 4 x 10'I No -- 1.7 x 10-6 1.1 x 10-6 1.6 x 10 1.3 x 10-6 1.1 x 10 4 1 x 10-5 8.2 x 10-6 6.1 x 10-6 4.4 x 10-6 3.3 x 10
Au -- 1.3 x 10-5 8.3 x 10 1.2 x 10-5 Sn 0.14 0.65 0.75 1.49 1.13 1.04 0.92 0.80 0.72 Te 0.22 0.28 0.38 0.61 0.57 0.52 0.57 .0.44 0.41.
- 1.41 4.74 5.58 8.66 7.80 6.97 6.10 5.34 4.76 Ca0 -- 5.26 9.46 17.04 17.65 18.12 18.50 18.% 19.24 A1 0 23 4.40 7.89 1.70 1.63 1.44 1.18 1.4 x 10-4 1.5 x 10~4 Na 0 --
0.13 0.20 0.40 0.41 0.42 0.42 0.43 0.42 2
, K0 2
-- 6.65 11.32 14.56 15.11 15.60 16.08 16.64 16.98 h $10 2
-- 10.39 18.63 33.14 34.22 35.26 36.32 37.62 38'47 0.08 1.12 0.84 1.10 0.85 0.66 0.43 0.36 0.28
) UO 2
Zr0 2
6 x 10-3 3.33 0.59 0.043 0.035 0.028 0.022 6.7 x 10'3 6.7 x 10-3 Cs 0 25.92 -- -- -- -- -- -- -- --
2 8a0 20.65 13.26 0.30 0.01 2 9 x 10'3 7 x 10~3 6 x 10-3 5 x 10-3 4 x 10'3
-5 Sr0 17.23 8.94' 9 x 10 5
l*2 03 1.5 x 10-5 14.05 8.28 2.43 1.93 1.50 1.11 0.80 8.5 x 10 l Ce0 2.06 14.95 26.56 3.26 2.39 1.75 1.23 0.86 0.63 2
Nb 0 25 8 x 10 1.6 x 10 1.8 x 10 1.1 x 10 8.7 x 10 6.8xid 4.9 x 10 4 3.4 x 10 2.6 x 10 4 Csl 15.46 1.78 6 x 10'I4 -- -- -- -- -- --
< Source l Rate. 219 696 449 225 204 188 178 168 160 g/sec I
TABLE 6.15. (Continued)
Species, Time, see 1 10800 12000 13200 14400 15600 16800 18000 19200- 20400 Feo 18.05 18.19 10.32 18.44 18.56 19.86 0.69 0.94 1.22 Cr 0 0.073 0.068 0.M0 0.053 0.046 0.N2 0.30 0.70 0.78 23 Ni - - -- -- - -- -- -- --
~I 3.6 x 10-4
~7 ~
~4 2 3.5 x 10-7 2.9 x 10 2.5 x 10 2.1 x 10-7 1.8 x 10~7 1.6 x 10'I 4.2 x 10~4 3.2 x 10 Ru 2.7 x 10 -6 2.3 x 10 -6 1.9 x 10-6 1.6 x 10-6 1.4 x 10-6 1.3 x 10-6 5 x 10-6 3.6 x 10-6 2.3 x 10
-6 Sn 0.66 0.60 0.57 0.53 0.50 0.49 6.82 7.00 7.15 Te 0.377 0.352 0.331 0.311 0.292 0.297 1.29 1.49 1.72 m 4.30 3.96 3.67 3.40 3.20 3.16 13.60 14.22 14.64 Ca0 19.33 19.40 19.47 1 9.54 19.59 20.90 4.68 5.42 5.46
~4 -4 -4 A1 0 23 1.6 x 10 1.6 x 10 1.7 x 10 1.7 x 10 1.8 x 10~4 1.9 x 10-4 9 x 10~4 1.2 x 10~3 1.7 x 10~3 Na 0 2
0.42 0.42 0.40 0.41 0.41 0.44 0.12 0.23 0.43 m K0 2
17.11 17.17 17.19 17.18 17.12 18.03 21.40 26.74 35.36
$ 510 2
38.86 39.17 39.44 39.70 39.95 36.51 47.21 39.72 30.06 00 2
0.230 0.194 0.165 0.141 0.121 0.111 2.98 2.66 2.20
~3 -3 ~3 ~3 l 2r02 6.6 x 10 6.6 x 10~3 6.5 x 10-3 6.5 x 10 6.5 x 10 6.8 x 10 0.031 0.039 0.048 Cs 0 -- -- -- -- -- -- -- -- --
2
-4 -4 Ba0 3.4 x 10~3 2.8 x 10 ~3 2.3 x 10-3 1.8 x 10 ~3 1.4 x 10 ~3 9.1 x 10 1.6 x 10"* 1.9 x 10 ~4 1.5 x 10 Sr0 -- -- -- -- -- -- -- -- --
La230 7.8 x 10-5 7.4 x 10-5 7.4 x 10 -5 6.6 x 10-5 6.3 x 10-5 6.4 x 10 -5 2.8 x 10 3.5 x 10-4 .4.4 x 10-4
~4 Ce0 2
0.469 0.357 0.270 0.1 % 0.133 0.081 6.6 x 10 6.3 x 10-4 6.1 x 10-4 2.1 x 100 1.6 x 10 1.4 x 104 1.3 x 104 1.4 x 10 4 1.6 x 10 1.3 x 10 2.7 x 10 Mb 0 25 1.8 x 10 Cs! -- -- -- -- -- -- -- -- --
Source Rate. 157 1 54 151 147 138 123 26.5 36.4 24.8 g/sec
m TABLE 6.16. AEROSOL COMPOSITION AND TOTAL RELEASE RATE FOR PEACH BOTTOM. TW .
Species. Time, sec
% 0 1200 2400 3600 4800 6000 7200 8400 9600 Fe0 2.14 5.17 9.15 14.21 14.49 14.76 15.61 17.80 18.12 Cr23 0 - 3.33 0.m 0.152 0.127 0.108 0.m 0.085 0.06 9 MI 0.140 1.00 0.330 -- -- -- -- -- --
Mo 1 x 10-8 1.6 x 10-6 1 x 10-6 1.3 x 10-6 1 x 10-6 8 x 10'I 6 x 10'I 5 x 10'I 4 x 10'I Ru 1 x 10'I 1.3 x 10 -5 8.4 x 10 4
1.1 x 10
-5 9 x 10-6 7 x 10 -6 5 x 10-6 4 x 10-6 3 x 10-6 Sn 0.113 0.365 0.410 0.584 0.537 0.486 0.440 0.425 0.378 Te 0.232 0.184 0.252 0.352 0.323 0.295 0.277 0.281 0.259 Mn 2.29 4.93 5.66 7.69 6.85 6.06 5.45 5.26 4.67 Ca0 20.01 7.51 11.73 17.83 17.81 17.75 18.29 20.36 20.37 AI23 0 4 x 10-3 4.72 8.36 12.97 13.23 13.48 10.52 0.003' O.003 Ma 0 2
11.14 0.136 0.207 0.370 7.365 0.353 0.368 0.417 0.400 m
& K0 2
-- 6.46 10.44 12.75 13.05 13.32 14.09 16.05 16.28 m
510 2
-- 11.13 19.71 30.59 31.19 31.78 33.60 38.32 39.01 UO 0.029 0.302 0.254 0.329 0.277 0.228 0.186 0.160 0.130 2
Zr02 4 x 10~3 1.36 0.349 0.015 0.012 0. 01 0 0.008 0.004 0.004 Cs20 0.346 - - - - - - - -
Ba0 6.13 14.20 1.96 0.056 0.048 0.041 0.036 0.033 0.026 Sr0 5.69 9.59 0.672 0.013 0.011 0.009 0.007 0.006 0.005 La23 0 6.6 x 10-5 12.27 5.60 1.15 0.931 0.730 0.554 0.437 4 x 10 -5 Ce0 2
0.752 15.% 24.67 0.78 0.602 0.452 0.3 31 0.248 0.166 Mb25 0 3 x 10-7 8 x 10 4 8 x 10-6 6 x 10-5 5 x 10-5 4 x 10-5 3 x 10-5 3 x 10 -5 2 x 10-5 Cs! -- - -- -- - -- -- -- --
Source Rate, 128 649 424 244 224 209 192 165 158 g/sec
TABLE 6.16. (Continued)
Species. Time sec
. 1 luovu Icuuv luvu 14445 13e00 leauu 15Uuo Iwvu ZOGEN Fe0 1 9.53 22.79 0.745 1.03 1.27 1.48 1,64 1.71 1,72 Cr230 0.061 0.048 1 .001 1 .31 1.43 1.46 1.43 1.22 0.90 N1 .. .. .. .. .. - .. .. ..
No 3 x 10'7 3x10-7 8 x 10'4 8 x 10 8 x 10-4 9 x 10-4 1 x 10'3 9 x 10 8 x 10
-6 -6 b 3 x 10'4 3 x 104 1 x 10-5 9 x 10-6 9x10 -6 8 x 10-6 7 x 104 5 x 10 3 x 10 Sn 0.367 0.397 5.72 5.86 6.04 6.27 6.52 6.51 6.65 Te 0.256 0.278 1.173 1.181 1.194 1.213 1.242 1.377 1.533 Mn 4.51 4.81 19.74 19.63 19.88 19.90 19.98 20.19 20.08 Cao 2!.68 25.10 10.10 9.95 9.83 9.78 9.71 10.78 10.20 A1 0 23 3 x 104 3 x 104 0.015 0.015 0.01 6 0. 01 7 0.018 0.W2 0.W7 Ma 0 0.420 0.352 0.185 0.247 0.296 0.340 0.379 0.448 0.573 2
i K0 2
1 7.71 IS.86 17.86 19.23 20.58 22.07 23.59 28.37 35.31 .,
510 2
35.41 26.09 38.84 36.95 34.86 32.87 30,83 25.00 18.33 UO 0.117 0.120 3.312 3.06 3.00 2.97 2.96 2.65 2.31 3
-3 ~3 Zr02 4 x 10 5 x 10 0.022 0.023 0.024 0.025 0.026 0.032 0.040 Cs 0 - -- - - -- - -- - --
2 I Sec 0.019 0.007 0.004 0.004 0.004 0.004 0.004 0.004 0.004 Sr0 3 x 10-3 1 x 10'3 2 x 10-5 2 x 10-5 2 x 10-5 2 x 10 ~7 3 x 10 ~7 3 x 10'I 4 x 10-7 La230 4 x 104 5 x 104 2 x 10 2 x 10 2 x 10 2 x 10 3 x 10 3 x 10 4 4 x 10 4 4
6 x 10 ~4 5 x 10 -4
~4 Ce0 2
0.108 0.034 7 x 10 6 x 10 5 x 10 5 x 10 5 x 10'4 Nb 0 25 2 x 104 4 x 104 2 x 10 2 x 10d 2 x 10 2 x 104 2x d 2 x 10 2 x 104 Cs1 -- -- ~~ -- - -- -- -- -
Source Aate. 145 123 27.3 24.9 21.9 19.3 17.9 25.5 17.9 g/sec 4
n . - - . - . . - .. - .. . - - .. .
- 7. RESULTS AND DISCUSSION p
i 7.1 Introduction f
Results of calculations for the transport and deposition of radio-
, nuclides are presented and discussed in this section. The plants and sequences-l :
selected for' consideration were discussed in Chapter 4, the-analytical and l calculational methods were described in Chapter 5 and the assumptions and bases for the calculations were described in Chapter 6. Results presented in i this chapter include the deposition and release from the reactor coolant system of radionuclides leaving the core region. These results are based on TRAP- ,
j- MELT code calculations. 'Also included as results are the masses of radionu- l
- clides airborne and deposited in the containment and suppression pool, as well as the airborne materials leaked to the environment. These results are based on SPARC calculations for retention in the suppression pool and NAUA-4 cal-j - culations for transport in the containment and reactor buildings. ,
- .Three accident sequences are considered in the analyses
- AE, TC, and TW. In each sequence failure of the primary containment was assumed to l.
result in a direct pathway for the release of radioactivity to the environment.
i This is the most severe failure mode (referred to as y') from the viewpoint of ,
- the magnitude of fission product source term to the environment. Additionally l for the TC sequence, the potential influence of the reactor building was ,
j examined by assuming the reactor building maintains its integrity following
- failure of the primary containment (y failure mode). Retention in the reactor
! building is also a possibility for the AE and TW system sequences. No attempt is made in this report to estimate the relative likelihoods of the y' and y f failure modes.
l 4
l 7.2 Transport and Deposition in Reactor Coolant S.ystem (RCS) i j The analyses of the transport and deposition within the RCS of .
l materials released from the melting core have been performed using the TRAP-
- j. HELT 2 code. The time frame of interest in the RCS for core meltdown accidents l such as those considered here spans the period of time starting with the onset t of core melting and ending with failure of the bottom head of the RPV. For l accidents involving only minor fuel damage, the gap release term, which occurs j 7-1 i
i-i
.-m-,-.-,-+.---,e--e,+.--mi,-ewe,e-t--,yvv, .wre------*ew--~-we-,-y.-,---, -e,==w.,-r.- - -
y 4
prior:to melting of fuel, may be the major release.and require careful
~
consideration. For the accidents examined here, however, this release term is insignificant in comparison with the melt release and the period imediately prior _to-the onset of core melting;is not considered. Rather,'the gap releases
--calculated by.CORSOR are added to the initial material emitted'by the melting core.
Releases from the melting core and their behavior in the RCS are j evaluated beyond the time of core collapse in these analyses.. This differs
, from the analyses presented in Volume I'of this report wherein the source to j the RCS was assumed to go to zero when the core slumped into the residual water- I
- in the lower plenum.. In the analyses of the Peach Bottom sequences presented.
l here, the slumping of the core has tMn simulated in MARCH lon a nodal basis as ;
l discussed earifer. This change and the presence of much greater below core f structures in the BWR pressure vessel dictate that emission from nodes which l have left the core be included in the analyses.
Another significant difference in the RCS phenomerology for the BWR sequences is the prediction of a significant time period between RPV dry out and RPV failure. During this period, the melting core emits fission products l into an essentially stagnant RCS, since there is no water available to provide l the steam as a carrier. The duration of time between vessel dryout and bottom j head failure is predicted by MARCH to be 78, 104, and 225 minutes for the AE,
! TC, and TW sequences, respectively. This residence time will be seen to result
) in significant attenuation of the aerosol.
} It should be recognized that the uncertainties in the behavior of l the molten core in the lower plenum region are quite large. The analyses j performed by the MARCH 2 code for this phase of the accident are greatly simpli-i fled. The rates of steam production as molten core material enters the lower
- plenum and the duration of time to failure of the reactor vessel have large f associated uncertainties. While the MARCH code is well suited to the explora-l tion of the effects of various assumptions on the overall results, practical l
- considerations have limited the extent of such exploration.
$ The heatup of the RCS structures downstream of the core is not modeled following dryout of the vessel. This is not expected to be a signifi-cant source of error since the surface temperatures in the core region are too high to permit condensation of the volatile species considered in these 7-2
d
. analyses prior to this period and there.is little subsequent emission of these species from the core.
At the time of bottom head failure, the materials suspended in the RCS are assumed to exit the primary system through the bottom head without further attenuation. This results in a short duration " puff" release of aged material into the drywell at the time of vessel failure. Re-entrainment of previously deposited particles or vapors is not considered to occur during this process.
One further aspect of the time frame of the primary system analyses l
which should be noted is that the primary system is not considered in the
! analyses after the molten core has left the RPV. Air ingress into the RPV and dIposition of materials evolved during the core-concrete interaction is not considered, nor is the primary system considered as a potential source of fis-2 ion prodLets due to reevolution of previously deposited materials.
The analysee presented in the following section are subject to a number of uncertainties. Principal among these are the source rates of materials emitted by the core, and the details of the flow patterns in the RCS. An additional aspect of the flow or, more precisely, the lack of flow during these sequences which represents a source of uncertainty is the duration of the " stagnant" phase of in-vessel portion of the melt down.
The results of the oaalyses of the AE, TC, and TW sequences for the i Peach Bottom plant are discussed separately in the following sections. The flow paths, associated geo;aetry, and the timing of the core-melt period in the RCS can be found in Chapter 6.
7.2.1 RCS Transport and Deposition for Sequerce AE
! This sequence resembles, in several respects, the AB sequence analyzed for the Surry plant in Volume I of this report. The principal similarity between these two is that, due to the location of the assumed large break, the materials released from the melting core transit only a very limited portion of the primary system during their release to the " containment". Here, the notion of " containment" is a bit more involved than was the case for the PWR l and this will be discussed at length later in this document. The primary system
- components of interest here are the core region, the steam separators, and the f lower annulus to the intake of the recirculation loop.
i 7-3 i
l-Table 7.1 presents the total masses (Total) of the RSS species groups of interest which have been emitted into the RCS as a function of time during I
- the in-vessel portion of the melt. These values, of course, increase monotoni-
[ cally with time, though not linearly due to the nature of the release process.
It is interesting to note that, on this time scale, vessel dryout occurs at
~
, .t = 1710 s. Thus, about_20 percent of the CsI'and Cs0H are emitted into a stagnant RCS,-and over 50 percent of the Te and aerosol are. emitted during 1
,. this period. This table also indicates that the mass of vapor species retained in the RCS first increases, then decreases as surfaces in the system heat up
- due to the high temperature steam flow out through the break. The aerosol retention does not exhibit any similar behavior since no resuspension mechanisms are included in the TRAP-MELT code. It is interesting to note the fairly drama-t tic increase in aerosol retention which occurs during the stagnant portion of
{ this sequence. This is due to the agglomeration of the particles to form larger -
sizes and subsequently enhanced settling of the particles in the RCS. Of course, there is no similar mechanism for the vapor species and their retained amount remains essentially unchanged during this period. t
,t Table 7.2 expresses these results in-terms of retention factors (RF) j which are simply defined as Ret / Total in Table 7.1. These values pertain to j the entire RCS and are integral in nature, so that the final valve (i.e., at I
t = 6360 s, which is the time at which bottom head failure occurs) can be taken i i to be the RCS retention factor which characterizes the sequence. The values
~ '
listed under Lower Annulus for CsI, Cs0H, and Te indicate the retention factor
) for that control volume. These values are also based on the total mass of the j species emitted from the core and therefore do not represent true efficiencies j of retention for the control volume. Figures 7.1 through 7.4 illustrate the
?
masses of each species retained in key volumes of the RCS, and the total amounts !
I emitted by the core as functions of time during the in-vessel period of core !
I j melting. ;
l The retention factors for the aerosol in the core and in the steam l l separators illustrate the basic features of the aerosol behavior. The reten-
- tion in the core region is well under 10 percent until after vessel dryout ;
j eccurs. Then, as the aerosol ages and the core continues to inject aerosol j mass into the stagnant system, the removal of aerosol mass in the core becomes-
!- increasingly effective. This is also indicated by the increase in the aerosol mass median diameter which is calculated from the discrete aerosol size l
lf -
t 7-4 '
{.
i-i
,--..---r.,,,,-m- ~ . . , ,----,,-,--vnr-.wnwnwn-~~-m--. ,c--nc- .-_m,,-,.,---w--- - w, r _ , vn , - , - , ., - - .r,.,,4
l l
l i l TABLE 7.1. CORSOR PREDICTIONS 0F MASSES OF SPECIES RELEASED FROM
~THE CORE (TOTAL) AND TRAP-MELT PREDICTIONS OF MASSES .
RETAINED IN THE RCS (RET) DURING THE AE SEQUENCE' FOS THE PEACH BOTTOM PLANT. FOR THE RSS SPECIES GROUPS (a) n (Times ' Measured from Start of Core Melting)
Cs! Cs0H Te Aerosol
, Time- Ret Total Ret. Total Ret Total. Ret Total (s) (kg) (kg) (kg) (kg) (kg) (kg) (kg) (kg)
L 500 0.0 1.7 0.4 . 21 .3 0.0 0.1 ~ 0.1 7.4 l
1000 2.7 8.3 20.4 66.2 0.2 0.4 20.7 62.2 i 1498 9.3 16.9 65.4 124 0.7 1.1 102 189
$ 1997 6.2 25.7 44.4 183 0.6 2.3 110 295 2497 6.3 26.9 42.5 1 92 0.7 3.1 112 351 3997 6.3 28.8 42.7 204 0.7 3.6 237 '465 4994 6.3 30.5 43.3 21 5 0.7 4.0 264 496
! 5994 6.3 33.1 44.3 233 0.8 5.1 359 606 64 91 6.3 33.9 45.0 238 0.8 6.0 467 71 9 j 6990 6.3 34.1 45.6 240 0.9 6.9 584 834 i
(a).SeeTable6.7c.
i l
l i 7-5
-,......,e --.-,.c4..,.._-.r.v. . . . . . . , e , . - . , , _ , _ , - _,m.- . . - , , , , . . , . . _ , - ,, , , ~ - - - , .
TABLE 7.2. TRAP-MELT PREDICTIONS OF PRIMARY SYSTEM RETENTION FACTORS (RF) AND VOLUME SPECIFIC RETENTION FACTORS AS FUNCTIONS OF TIME FOR THE AE SEQUENCE FOR THE PEACH BOTTOM PLANT' Cs! Cs0H Te Aerosol l Time Lower Lower Lower Steam l (s) RF Annulus RF Annulus RF Annulus RF Core Sep j l
500 .01 .01 .02 .02 .20 .01 .02 . 01 .0 1000 .32 .20 .31 .20 .44 .14 .33 .01 .14 1498 .55 .27 .53 .27 .69 .21 .54 .06 .26 1997 .24 .24 .24 .24 .26 .09 .37 .04 .18 2497 .23 .23 .22 .22 .21 .08 .32 .04 .15 3997 .22 .22 .21 .21 .19 .07 .51 .29 .12 4994 .21 .21 .19 .19 .18 .06 .53 .33 .11 5994 .19 .19 .19 .18 .15 .05 .59 .43 .09 6491 .19 .19 .19 .18 .13 .04 .65 .51 .08 6990 .19 .19 .19 .17 .13 .04 .70 .58 .0/
7-6
40
~
i I Sgetid m.
6 M VOL 2 -
I VOL 2+3 2 m.
- 8 I z I
b E
10 - *
- '~
l
\' ' ~ _
0
- -A' ,
\ , , , , ,
0 1000 2000 3000 4000 5000 6000 7000 TlWE (sec)
FIGURE 7.1. MASSES OF Csl EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF l
TIME FOR THE AE SEQUENCE (Vol 1 = Core, Vol 2 = Steam Separators, Vol 3 = Lower Annulus,
- Vol 4 = Drywell). Times Measured from Start of Core Melting.
i i
250
! Legend TOTAL i
VOL 2 -
3, v
.V.O..L.-----
2+3 to 15 0 -
I 2 i
! w o
kJ l do E 10 0 -
y a / ,
\
j 50- / ,
j -------------------------------------------------------------------.
/ A
'/ N 0 5 I y 3 3 g 0 1000 2000 3000 4000 5000 6000 7000 TlWE (sec)
I i FIGURE 7.2.
i MASSES OF Cs0H EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL V0 LUES AS FUNCTIONS OF i
TIE FOR THE AE SEQUENCE (Vol 1 = Core, Vol 2 = Steam Separators, Vol 3 = Lower Annulus, Vol 4 = Drywell). Times Measured from Start of Core Melting.
i i
8 i
l 6-
! m Legend
- a 3 TOTAL M VOL 2 m -
4 2 4- VOL 2+3 0
l 7 W e Z_
h w
E 2-
, ~~ ~ .------ ------------------------------------------------ ----~~- C
, y ~~~ ~ -
O i i i i i i ,
0 1000 2000 3000 4000 5000 6000 7000
- TIME (sec)
)' FIGURE 7.3. MASSES OF Te EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF TIME FOR THE AE SEQUENCE (Vol 1 = Core Vol 2 = Steam Separators, Vol 3 = Lower Annulus, Vol 4 = Drywell). Times Measured from Start of Core Melting.
i i
6 o
m a 5
g, ME 2 ". i o
o o 2I Q'
\,
E" >n N '
>g
's' o g
e Ct: a
~
s,\ s sE
., uns mic
, e-h s ,
3 o
o sae
-s
,! I _S mEE
\t ~M8
!i1 t
$"t n"a
\ \, o s-s
\ u23*
-8 w
] o .
s\\ ,
s w
ag!
uu*
onv
\, o 3 z"g u
\ .o 5 3 Oo$
j\
I i \
\,
'E2 O
w w vi 8
Z i
t .i .o o EWs w cr
!. $ _Ju o .
L. 8 m
\ '
- e i sC
, V,. Ni
+e f M' '
s, .O o ci a:
I w J .-j o on m6 Oh J Jl J N
- S O O! O 3 i I I s o 4 o o o o o o o o o o u g = . % a o i I
2 (SM) SSYN 03NIY13W l
7-10 ;
\
.-,g-y---y, y-?w-wyw,-g--ywey m rr - ---,p-w ,ew - wm.,ye---w-----m ----,,m-- --=-,-r--,, - - - y,--a
distribution followed through the TRAP-MELT calculations. At t = 1800 s, the l mass median diameter is 0.188 um, and by t = 4800 s, it has increased to 2.76 um, which, of course, leads to an enhanced removal rate for the aerosol via gravitational settling. A further note regarding the aerosol retention values l
in this table concerns the values of the retention in the steam separator.
The decrease evident after t = 1770 should not be misinterpreted to indicate a resuspension of previously deposited aerosol. Rather, it is merely a conse-quence of the fact that the total aerosol mass emitted by the core has continued to increase while the amount retained in the steam separators is fixed, due to the absence of flow.
The source to the drywell for this sequence can be characterized as a fairly steady injection rate up to the time of vessel dryout. During this period, approximately 30 percent of the Csl, Cs0H, and Te passing through the RCS is retained, as is just over half of the aerosol. There is then no source until the bottom head fails, at which time the suspended materials in the RCS are injected into the drywell. No further emissions from the RCS are considered in these analyses.
The simulation of the AE case was performed using TRAP-MELT to track the RCS retention of the WASH-1400 radionuclide groups as described in Chapter 6, and the results are sumarized in Table 2a. The retention of groups I, Cs, and Te is consistent with that previously noted for CsI, Cs0H, and Te. The Sn and Ru groups closely parallel the retention of the total aerosol throughout the sequence. At various times during the sequence, the La group varies es much as 19 percent from the retention of inert aerosol. However, at the time of bottom head failure, the retention of La is very nearly the same as the i other aerosols. This interesting behavior is primarily the result of the timing of releases of various groups with respect to the varying gas flow rate through the RCS.
7.2.2 RCS Transport and Depo,sition for Sequence TC In this postulated accident sequence, the RCS is essentially at operating pressure up until bottom head failure, making this sequence somewhat analogous to the TMLB' sequence studied in Volume I for the Surry PWR. During the in-vessel phase of the meltdown the flow pathway exits the RCS through the relief lines which transmit the gases and fission products to the suppression 7-11
l 1
-TABLE 7.2a. CORSOR PREDICTIONS OF RELEASE FROM CORE AND TRAP-MELT PREDICTIONS OF PRIMARY SYSTEM RETENTION OF WASH-1400 GROUPS FOR AE SEQUENCE Released Retained Group (kg) (kg)
I 16.7 3.1 Cs 230.9 43.8 Te' 6.9 0.9 Sr 22.6 16.1 Ru 30.0 23.3 La 1.19 0.89
- Groups are defined in Chapter 6.
{
1
\
I 7-12 l
pool. The flow pathway and relevant geometry were presented in an earlier 4
section, so it is_ sufficient to simply restate that the' flow passes through l the steam separators and then splits, with about 855 considered to pass ;
-through the steam dryers, and the remainder bypassing the dryers via the outer j annulus. These flows then merge at the steam lines and enter the relief lines to the pool.
Table 7.3 presents the total masses of the various species emitted ;
'fror,the core (Total) as functions of time during this sequence, and the mass l retained in the RCS at any time. Following from the CORSOR results, these
{. sources to the RCS are only on the order of half what the values for the AE l
sequence were predicted to be. Vessel dryout for this sequence occurs at t = 3720 on this time scale, by which time the Cs! and Cs0H emissions are nearly .l completed. The masses retained in this sequence are considerably less than
!- for AE, partly due to the reduction in the source term for the RCS and partly ,
, due to the differing system thermal hydraulics. An interesting feature of ;
i this table is the larger retention predicted for Cs0H, in comparison with the i retention of Cs!. The cause of this apparent discrepancy is that the meetinism of Cs0H retention is chemical reaction with surfaces, which is not a viable I l
j process fo- Cs!. This difference between Cs! and Cs0H retention is emphasized l in Table 7.4, which contains the RCS retention factors, along with several
! volume specific values as defined above. Figures 7.5 and 7.6 illustrate j retained masses of these species.
! The relatively large surface area of the steam dryers is reflected ,
l in their large contribution to the reaction of the Cs0H and Te. These vapors l which react are considered to be irreversibly bound to the surface in all l sequences examined here. The location of retained Te is indicated in Figure 7.7. The aerosol retention in this sequence exhibits similar behavior l ,
i to that observed in the AE results, but on a less impressive scale as is shown l l in Figure 7.8. The increase in aerosol RF from 0.43 just before vessel dryout l to 0.51 at the time of bottom head failure is, again, due to the aging of the I particles in the stagnant core region of the RCS. That the increase in the RF i is not greater than what appears in this table is due to the lower rate of i
- aerosol generation predicted for this sequence. ['
i .
l
- 7-13
, 1 1-t
y 5 ; *. _
s n~
s ,
$ [ ;' i l' ;
( ,
TABLE 7.3. CORSOR PREDICTIONS 0F MASSES OF, SPECIES RELEASED FROM y, '
THE CORE (TOTAL) AND TRAP-met.T PREDICTIONS OF MASSES RETAINED IN THE RCS (RET) DURING THE TC SEQUENCE FOR -
THE PEACH BOTTOM PLANT s (Times Measured from the Start of Core Melting) 1N Cs! Cs0H Te Aerosol ^
Time det Total Ret . Total Ret Total Rep 4 Total (s) (kg) (kg) (kg) (kg) (kg) (kg) (kg)' *(kg)
\
370 0.4 3.3 4.3 27.3 --
< 0.1 1.5 13.1 j 740 3.8 8.1 28.2 56.4 0.2 0.4 22.5 55.3 ,(
1100 9.3 13.5 62.9 90.0 0.7 0.9 81.6 129 ,f 1480 15.2 18.9 99.0 123 1.4 1.6 175 236
~
2210 1.9 26.8 40.0 171 10.6 11.0 232 648 ;
2580 2.0 27.6 43.0 175 11.4 3 1118 240 770 y 2950 2.0 27.9 44.4 177 11.7 1211 288 875 ;
3320 2.1 28.0 45.3 178 11.94 12.2 346' , 978 /
3690 2.1 28.2 45.9 179 12.0 12.3 40$ r1081 4430 2.1 28.5 45.9 181 12.0 12.5N 624 .a 1284 \
1, ,
5900 2.0 29.6 45.6 188 12.0 s 10.3 ,,833 4454 %-
s ,. .
7380 2.0 32.2 45.5 203 12.0 '15.1 " 887 1520 s
j 'A
, n :,'$. h I g 4 ! ,
g a l 7-1Ib A' g
a .
_ ,e ?$ !
.- >g -
m -
3
~
h
-,- , s 7 ,
i j - 3:
I TABLE 7.4. TRAP-ELT PREDICTIONS OF PRIMARY SYSTEM RETENTION FACTORS (RF) AND VOLUME SPECIFIC RETENTION FACTORS AS FUNCTIONS OF TIME F]R THE TC SEQUENCE FOR THE PEACH BOTTOM PLANT x
n----
? . Cs! Cs0H Te Aerosol l
Lower -
Steam Steam 5 team
! .Timei'j7 (s) RF.,
,Annulus RF Dryers RF Dryers RF Core Dryers 376, .13 .01 .16 .06 .46 .01 .11 .03 .02 e740[ .47 .02 .50 .20 .61 .04 .41 .04 .11 F ,1 !? l' 1100 .69 ' .04
.70 .22 .76 .05 .63 .07 .14 4
1480 .80' ; .06 .81 .23 .88 .04 .74 .08 .14 2210 .07 6 07 .23 .15 .97 .45 .36 .05 .08
^
2580 .07 .07 .24 .16 .97 .43 .31 .05 .07
^
2950 .07 .07 ,.25 .16 .97 .43 .33 .04 .09 3320 .07
- 07,
. 25 .17 .97 .42 .35 .04 .11
- )
3690 . 0 7, . 07 ' .26 .17 .97 .42 .37 .04 .12 4430 . 07, e ' [37 < .25 .17 .96 .41 .49 .19 .11
, 3 5900 f.07 ~ .07 .24 .16 .91 .39 .57 .30 .11 t ,
7380 . 0G, .06 .22 .15 .80 .34 .58 .33 .10 r
f j
(:.
j p' .
r ,
7-15
!g y s
Msi -y.
1.: .'% '
,4 ~
\
- i. % % .
. k; ~*h *
- w ,. __v.
1 40 Legend VOL 2 ____
VOL 2 + VOL 3 ,-
y Cn 30- _v_0_L_s_2_ _ _4_ _ _ _.
U ~___
EMITT_ED (n /
1 20- l
~ Q /
5 w
- z
_ /,
l<-- / -
w 10 - I' , ,
cr j ,
/ ',
i i
j,' ._____________________________________________________
0 , , ,
0 2000 4000 6000 8000 TIME (sec)
FIGURE 7.5. MASSES OF Csl EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF TIME FOR THE TC SEQUENCE (Vol 1 = Core, Vol 2 = Steam Separators, Vol 3 = Steam Dryers, Vol 4 = Lower Annulus, Vol 5 = Steam Line, Vol 6 = Containment). Times Measured from Start of Core Melting.
250 Legend VOL 2 ____
200- VOL 2 + VOL_3 ,-
CD _V_O_L_S_2_ _ _4_ _ _ _ .
y ' _ __ _
v eri_te __ f m
m 15 0 - /
s /
7o /
Gy 100- f ,,,
Q n
hJ
/ /; '\
CC /
n e, 50- i,_____ ____________________________________________.
/ ,
f ,e / \
i i
O i i i 0 2000. 4000 6000 8000 TIME (sec)
FIGURE 7.6. MASSES OF Cs0H EMITTED FRCM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF TIME FOR THE TC SEQUENCE (Vol 1 = Core, Vol 2 = Steam Separators, Vol 3 = Steam Dryers, Vol 4 = Lower Annulus, Vol 5 = Steam Line, Vol 6 = Containment). Times Measured from Start of Core Melting. ,
20 Legend VOL 2 _ _ _ _ _
VOL 2 + VO_L 3 y
m 15 - yoy_2 _4_ ___, ,-
v em_a _ '
m __ _
g _
~
2 jo _
5 o
= w i E $
p< 'p___ ___
f W 5-f' I
s'
~'
/
0 , , ,
0 2000 4000 6000 8000 TIME (sec)
FIGURE 7.7. MASSES OF Te EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF TIME FOR THE TC SEQUENCE (Vol 1 = Core Vol 2 = Steam Separators Vol 3 = Steam Dryers, Vol 4 = Lower Annulus, Vol 5 = Steam Line, Vol 6 = Containment). Times Measured from Start of Core Melting.
2000 Legend VOL1 VOL 1 + VO,L_2 -
9 1500_ vas :-s _ ' - ~ ~ ~ -
f 19M.' 4.... , ,
v0Ls :-s _
f EMITTED y 4 /
2 1000- /
Ya /
G Ld / ,, ,C ~- ------- n T.~
Z -
/ / ,- f ~
Q W
uj 500-
/ /,4 , ,
g
/
/
-f / /,g,-l
- i
/
/ f __.
=e2-: J ,/
g -
I 1 I o 2000 4000 sooo 8000 TIME (sec)
FIGURE 7.8. MASSES OF AEROSOL EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF TIE FOR THE TC SEQUENCE (Vol 1 ~= Core, Vol 2 = Steam Separators, Vol 3 = Steam Dryers, l Vol 4 = Lower Annulus, Vol 5 = Steam Line, Vol 6 = Containment). Times Measured from Start of Core Melting.
l 7'.2.3 RCS Transport and Deposition for TW Sequence Tables 7.5 and 7.6 present the masses emitted, masses retained and
)
RFs for the various species in the TW sequence. On the time scale given in these tables,~ vessel dryout occurs at t = 5010 s and bottom head failure is predicted by MARCH to occur at t = 18540 s.
The masses of Csl and Cs0H retained in the primary system are seen in Figures 7.9 and 7.10 to increase steadily until t = 3900 s. At some point between this time and 4900 s in these tables the surface temperatures in the volumes where condensation of the vapors has occurred increase sufficiently to drive off the condensed material and permit it to be transported. While cooler surfaces are encountered by these vapors in transit to the " containment", the surface areas and residence times are not large enough to significantly attenu-ate the flowing vapors. The principal retention mechanism for the Cs1 and Cs0H is the retention of particles upon which'these vapors have condensed.
The slightly larger retention factor presented in Table 7.6 for Cs0H is due to the influence of reactions with the system surfaces--which do not occur for CsI.
Tellurium behavior as exhibited in Figure 7.11 is similar to what has been presented for the other sequences discussed above. Only a portion of the core inventory is released, and of this material, over 80 percent is predicted to be retained in the RCS due to chemical interaction with system surfaces.
Aerosol retention illustrated in Figure 7.12 is again approximately 70 percent for this sequence. Since this is a more prolonged sequence than the others, there is a greater mass of aerosol emitted and a longer residence time is available to the material released from the core region into the RCS.
The influence of the long residence time during the stagnant portion of the accident is seen in the increase in the fractional deposition of aerosol parti-cle mass in the core region. Nearly all of the mass released by the melting core during this period is retained in the RCS. There remains, however, a portion of the aerosol emission which is suspended in the RCS at the time of bottom head failure and released as a puff from the RCS.
7-20
- - u - ,
TABLE 7.5. CORSOR' PREDIr.TIONS OF MASSES RELEASED FROM THE CORE AND TRAP-MELT PREDILi!ONS OF MASSES RETAINED IN THE RCS DURING THE TW SEQUENCE FOR THE PEACH BOTTOM PLANT l- (Times Measured from Start of Core Melting) )
Cs! Cs0H Te Aerosol
, Time Ret Total Ret Total Ret Total Ret Total l (s) (kg) (kg) (kg) (kg) (kg) (kg) (kg) (kg) 490 0.1 1.4 0.7 14.1 0.1 0.1 1.8 8.2 4
1960 2.8 7.1 18.6 50.9 0.9 1.0 56.5 120 2940 8.2 13.3 53.3 89.9 2.2 2.6 214 295 3420 13.0 18.8 82.9 123 4.5 5.0 344 462 3910 17.6 23.5 111 152 8.8 9.5 588 781 4890 4.7 31.8 32.8 202 14.0 14.3 693 1038 6360 4.7 31.8 32.8 202 14.0 14.3 694 1039 7830 4.7 31.8 32.8 202 14.0 14.3 694 1040 10760 4.7 31.9 32.8 203 14.0 14.4 698 1045 13700 4.7 32.4 32.8 206 14.0 14.6 714 1065 15170 4.7 32.8 32.8 208 14.0 15.0 745 1100 16630 4.7 33.2 32.8 211 14.0 15.7 811 1170 18540 4.7 33.5 32.8 212 14.0 16.8 960 1323 l
l l
i
! 7-21
TABLE 7.6. TRAP-MELT PREDICTIONS OF PRIMARY SYSTEM RETENTION FACTORS (RF) AND VOLUME SPECIFIC RETEN-TION FACTORS AS FUNCTION OF TIME FOR THE TW SEQUENCE FOR THE PEACH BOTTOM PLANT (Times Measured from Start of Core Melting)
Csl Cs0H Te Aerosol
- Time steam Relief Steam Relief Steam Steam Relief (s) RF Dryers Lines RF Dryers Lines RF Sep RF Core Dryers Lines 490 .05 .01 .03 .05 .02 .03 .81 .73 .21 .19 .01 .02 1960 .39 .22 .07 .36 .20 .07 .85 .79 .47 .11 .22 .06 2940 .62 .32 .06 .59 .30 .06 .87 .82 .72 .15 .29 .04 3420 .69 .36 .06 .67 .34 .05 .92 .85 .74 .12 .33 .05 3910 .75 .38 .06 .73 .37 .06 .92 .85 .75 .16 .34 .05 7
N 4890 .15 --
.05 .16 .01 .05 .98 .73 .67 .08 .30 .05 -
N 6360 .15 --
.05 .16 .01 .05 .98 .73 .67 .08 .30 .05 7830 .15 --
.05 .16 .01 .05 .98 .73 .67 .08 .30 .05 10760 .15 --
.05 .16 .01 .05 .97 .73 .67 .09 .30 .05 13700 .15 -- .05 .16 .01 .05 .95 .72 .67 .10 .30 .05 15170 .14 -- .05 .16 .01 .05 .93 .70 .68 .13 .29 .05 16630 .14 -- .05 .16 .01 .05 .89 .67 .69 .17 .27 .05 18540 .14 -- .05 .15 .01 .05 .83 .62 .73 .27 .24 .04
40 Legend TOTAL /
p 30-VOL 2 ,
15; X9L 213,,
d VOL 2-4_ _
g 2 20-7' O i 0 z E 'b l
10 - l \s\,
'l
. \.
a' \s
\\ _ _ _ _ _ _ _ _ _
0
// i' 0 5000 10$00 15$00 20000 TIME'(sec)
FIGURE 7.9. MASSES OF CsI EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF TIME FOR THE TW SEQUENCE (Vol 1 = Core, Vol 2 = Steam Separators, Vol 3 = Steam Dryers, Vol 4 = Lower Annulus, Vol 5 = Steam Line, Vol 6 = Containment). Times Measured from Start of Core Melting.
250 Legend TOTAL -
200-VOL 2 .
^
On .V.O..L. 2.+.3...
O VOL 2-4 --
M 150-2 i 8 3 10 0 -
h w ,I ,
E !\'
o
/5 .
50- :l
- \.
\t II
, ,/ / g 0 . -""""""""""",""""*""'"""---------------------
0 5000 10000 15000 20000 TIME (sec)
FIGURE 7.10. MASSES OF Cs0H EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF TIME FOR THE TW SEQUENCE (Vol 1 = Core, Vol 2 = Steam Separators, Vol 3 = Steam Dryers, Vol 4 = Lower Annulus, Vol 5 = Steam Line, Vol 6 = Containment). Times Measured frcm Start of Core Melting.
I
20 15 -
m Os ________________r__________!__________!____________________!____.
6 m
.i m 4 ,
y 2 10 - /
h, Q ,A y w I Legend j TOTAL
- t VOL 2 5- ,
_V_O__L_ _2_+ 3_ _.
VOL 2 l l 0 i i i l 0 5000 10000 15000 20000 l TIME (sec)
( FIGURE 7.11. MASSES OF Te EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF TIME FOR THE TW SEQUENCE (Vol 1 = Core, Vol 2 = Steam Separators, Vol 3 = Steam Dryers, Vol 4 = Lower Annulus, Vol 5 = Steam Line, Vol 6 = Containment). Times Measured from
, Start of Core Melting.
1500 Legend TOTAL VOL 1 ,
m .V.O..L.1.+. 2...
[ 1000- VOL1-3_.
(A VOL1-4 VOL1-5
- e /
4 - - -- _j D 500- ,! ','
a: ,,*
, ,....................................------ -- ~~....../
/
\
s
/
_/
/./
', g 7 ~~~~
0 , i i 0 5000 10000 15000 20000
, TIME (sec) l FIGURE 7.12. MASSES OF AEROSOL EMITTED FROM CORE AND RETAINED IN THE RCS CONTROL VOLUMES AS FUNCTIONS OF TIME FOR THE TW SEQUENCE (Vol 1 = Core, Vol 2 = Steam Separators, Vol 3 = Steam Dryers, Vol 4 = Lower Annulus, Vol 5 = Steam Line, Vol 6 = Containment). Times Measured from Start of Core Melting.
- - __ _ + _ m_ .____._-_
l A
- -7.3 Transport of. Fission Products Through Containment
'Results are presented.in this section for analyses performed for the transport and retention of various fission products that' leave the reactor l
_ . coolant system. Thelvarious compartments of the reactor considered for these L. analyses include the suppression pool, the wetwell, the drywell, and the reattor
[ containment building. The NAUA code that calculates transport of fission products in particulate form was utilized for the mentioned compartments except that the SPARC code was utilized for calculating the retention of fission product in the-suppression pool.
Results from the MARCH, TRAP-MELT, and VANESA calculations.were.
. - utilized to provide the required input for the NAVA and SPARC calculations.
I
{ Three species CsI, Cs0H, and Te, were distinguished in the calculations and
- ~ all the other species were treated as'one group. All' species in this additional
]- group were assumed to exist in the particulate form'once they escape the reactor L coolant system. <
' Dimensions of the compartment volumes were based on geometric data l provided in'the Peach Bottom 2 FSAR. The largest cross sectional area of a i
volume was used to' estimate the floor area. This area directly affects the removal rate for particle sedimentation. For. example, the cross sectional j area of the spherical section based on a diameter of 67 ft (20.4 m) was used 4 for the drywell calculation'.
5-Three different accident sequences, AE, TC, and TW, were considered i
j in the present calculations.
) 7.3.1 AEy' Sequence 1
1 i
This sequence involves a large pipe break accident resulting in loss of reactor coolant. The fission products released from the reactor coolant system enter the drywell through the lower annulus section and subsequently-j- are allowed to enter the suppression pool to reach the wetwell. As the drywell j fails due to the pressure built.up by the release of gases from the reactor
- coolant system, the fission products released to the drywell no longer enter
- the suppression pool but are released to the failed reactor building or environ-
{ ment. Due to this event-dependent flow path of the fission product involving l several physically separate compartments, the analyses for the AE-scquence 7-27 a <+-,,---.*-r < - . , - - +-vvv , - , - - - , . . , . r .- - -
t were performed sequentially as depicted by Figure 7.13 and this is briefly L
described as follows.
First, the NAUA code was utilized to calculate the behavior of i particulates in the drywell. The calculation was suspended as the containment
-failure time was reached. The results of this NAUA calculation were then used with the SPARC code for calculating the scrubbing of fission products in the suppression pool. A separate NAUA calculation was performed to describe the behavior of'the fission products in the wetwell air space. The input for this was the output of the SPARC calculation. To cover the accident time beyond I the containment failure time, the NAVA calculation that had been suspended was i reactivated for the drywell. The source terms used for this resumed calcula-tion are the TRAP-MELT calculation results, the separate NAVA calculation results for the wetwell, and the VANESA calculation results. Figure 7.14 shows the suspended mass concentration of particulates in the drywell. It is seen that the amount of the suspended particulate increases rapidly as the core starts melting at 12 minutes and the highest concentration results at a time of 34 minutes when the drywell fails. It should be realized that during this time period, a considerable amount of particulate enters the suppression pool and is captured. Distribution of the total particulate at various locations and at various times is listed in Table 7.7. As the drywell fails, the gas in the dry well is directed toward the environment and the particulate concentra-tion in the drywell decreases rapidly due to this new leakage flow path. As l
the time elapses further and the head finally fails, the particles released through the bottom head cause the airborne particulate concentration to increase again and the amount leaked to the environment builds up (see Table 7.7). At a time of 1200 minutes (20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br />), the particulate released during the accident is located in the drywell, the suppression pool, and in the environment each with the amounts listed in Table 7.7.
The particle size distribution of aerosol suspended in the drywell at selected times is shown in Figure 7.15. It is noted that the mass median particle size of the aerosol ranges from 0.1 to 1.2 um before the drywell fails.
As will be discussed shortly, scrubbing of particulates by the suppression pool depends substantially on the size of particles that enter the pool.
Combined with a particle density of 3 g/cc which was used in the present analy-sis, an aerosol with the above range of mass median diameter is found to be removed rather effectively by the suppression pool.
7-28
_ _ - _ _ _ _ = _ _ _ _ _ _ _ -__ . . _ - .___
. - - -- - , ., . -. .- . - . . . __ .~ . . - _ . ._ _ _- -.
! TRAP-MELT VANESA
- -NAUA (DRYWELL) > NAUA (DRYWELL) >- Envimnment i
J v
i~
SPARC (POOL) i I y I i
- e l l
' I NAUA (WETWELL)
I I i 1 I I I I Containment Vessel' Failure Failu re i
i FIGURE 7.13. NAUA AND SPARC CALCULATION PATHS USED FOR PERFORMING ANALYSES OF AE SEQUENCE 4
0
RIRBORNE b_
b_
b_
co .
m mb-x:
r _
t 1
1 3
x -
H ~
o .
H b_
b_
b_
, . ' ' - i ' ' '
, ,,,,,, 2 3*10' 10 1O b= 10' TIME, MIN FIGURE 7.14. AIRBORNE CONCENTRATION IN THE ORYWELL AS A FUNCTION OF TIME, AE 7-30
_, , _ . . . _ , ~ . + - - - - ~ ~ ' ' ' * ' ' ~ ' -
TABLE 7.7. DISTRIBUTION OF TOTAL SOLID PARTICULATE MASS (KG) AT VARIOUS TIMES.-AE Time Drywell Pool (min) Event Suspended Deposited Captured Wetwell Environment
+
12 Melt start 20 17.0 11.9 0.09 0.001 0 34 Cont fails 46.3 22.1 218 '18.1 0 35 16.1 22.5 218 14.4 49.4 40 Vessel dry 1.,
50 0.25 23.9 218 10.9 -129 y
58 Core collapse
$ 80 0.15 24.2 218 10.4 129.3 118 Head fails 120 28 40.9 228(a). -0(*) 157.6
, 420 51 1458 228 -0 2682 i
1200 End of accident -0 1485 228 -0 2729-(a) Wetwell mass being added to pool captured mass i-i e
- O'
....i H R . --
l l 7 0-
- 2.6 HR.
5.4 HR.
-f '
! - : 8. 3 HR . ----------[/
~
v .
i o_
.-e :
l E l O'o o~i ~. ,
1 5 , ,
e : , 's
/ ,
zT .
/ ',
o-o ,/
- - i \
s : ,
i, g -
o x -
/ ,
Ho - o i ze ,i i .
wo i a _2 ,/ ,
z : i a - ,
i o -
- - ,/
or ,
z o_
E-5 '
< i, i
o : - 8 o -
\
E -
i .
a - 8 C7 - '
o_ i
-5 i i
i
~ '
l i \
c- i
'o _ .
i \
~! / ', \
- i
. ./
=- / -
'o _
~5
/
10 10 10 10 4*10' RRDIUS, MICRONS FIGURE 7.15. PARTICLE SIZE DISTRIBUTION OF AEROSOL SUSPENDED IN THE DRYWELL, AE 7-32
_ _ _1 _ - ______ ___. _ _ _ _ _ ___ . - ._. . . _ . _ . .- . _
Since the release timing.of individual fission product species may be different than that for the total. particulate, the time-dependent amounts of CsI, Cs03, and Te were distinguished in the present calculation. Table 7.8 shows the calculated fractions of the core inventory of each species found to reach various locations at the end of the accident. It is interesting to see
-that significant fractions of Csl and Cs0H are captured by the suppression pool. As a result, 4 bout 30 percent is seen to escape the drywell to reach the environment.
The MARCH calculation results indicated'that between the core melting period and the containment failure, the thermal hydraulic conditions are such that there is little supersaturation of water vapor in the drywell. The NAUA calculation also shows that the diffusiophoresis of fission products onto surfaces is not significant.
Collection in the suppression pool as calculated by the SPARC code-indicated that the pool is indeed an effective means for scrubbing particulates.
The SPARC code considers removal of particulates by sedimentation, inertial deposition, diffusion, and diffusiophoresis inside bubbles. In addition, particulates are allowed to grow by steam condensation, thus enhancing gravita-tional sedimentation and the inertial deposition. It has been speculated that particulates would be removed by an additional impaction mechanism just after the gas is discharged from a vent pipe for AE or from quencher holes for TC and TW into the pool water forming an impaction regime. For modeling this type mechanism, an existing theory for an impactor was added to the SPARC code.
The SPARC code calculation results showed that depending upon the l size of particles, decontamination factors over a wide range can be obtained.
l Typical decontamination factors obtained from SPARC calculations are listed in Table 7.9.
7.3.2 TCY' Sequence This accident represents the sequence in which the containment fails first and the fission products released from the relief valve enter the suppres-sion pool through the T quencher, pass through the wetwell and reach the drywell before being released to the environment. As the reactor vessel fails, the fission products enter directly into the drywell and are then released to the failed reactor building or environment. The sequential use of various computer 7-33 P
TABLE 7.8. DISTRIBUTION OF SPECIES AT 20 HOURS AFTER ACCIDENT, AE M EI EI IE Fraction of Core Inventory Species RC5 Pool Drywell Wetwell Environment Cs! 0.19 0.35 0.12 0 0.34 Cs0H 0.19 0.34 0.14 0 0.33 Te 2.9 x 10-2 3.2 x 10 -3 0.32 0 0.65
{
l 7-34 4
TABLE 7.9. DECONTAMINATION FACTORS CALCULATED AS A FUNCTION OF
- PARTICLE SIZE AND OF TIME FOR AE SEQUENCE 1
I
. Time Particle Diameter, um DF Based on.
l (min) 0.1 0.7 1.2 5 8.4 Total Mass 5 5 14.3 1.2 3.3 x 10 2 5 10(a) 10 10 1504 5 5 2.9 x 10 2 5
18.9 1.2 10 10 10 1400 27.4 1.2 51 2.5 x 10 4 10 5
10 5
25 5 5 33.3 1.3 5.4 69 10 ' 10 4,j (a) A decontamination factor larger than 105is assumed to be 105, i Pool depth: 4 ft Bubble diameter: 0.75 cm Aspect ratio: 1:3 l
l l
7-35 l
- _ _ . .~. _ . - . . ___ _____ _ ____ -
codes'for calculating the transport of fission products in this accident sequence is shown in Figure 7.16.
Since the first volume fission products encounter after being released from the RCS for the TC sequence.is the suppression pool, the SPARC code was utilized to calculate.the retention of-fission products by the pool. The I
-1 decontamination factors calculated are listed-in Table 7.10. This table shows j a particle _ size dependency that is similar to that'shown previously for the AE l sequence. Time variations of the calculated decontamination factor for 0.1 and 0.7 um particles are, of course, due to the difference in the thermal hydraulic condition that prevail at the corresponding times.
Figure 7.17 is the suspended aerosol mass as a function of time for the drywell. Sharp increases in tne mass concentration as shown in Figure 7.17 represent the initial core melting time and the bottom head failure time, respectively. Figure 7.18 is the accumulated mass leaked outside the drywell shown by species. Figure 7.19 is the geometric number median radius of the aerosol mass suspended in the drywell. It is interesting to note that due to the presence of the suppression pool, which removes predominantly large parti-cles, the particle size during the time up to bottom head failure is rela-tively small, while the size becomes large thereafter.
Table 7.11 is the calculated distribution of the CsI, Cs0H, and Te that are located in the RCS, the suppression pool, the drywell, and the wetwell.
The last column indicates that fractions of core inventory of these species between 0.21 to 0.37 are released to the environment. It is also noted that fractions of 0.69 for Csl and 0.56 for Cs0H are captured in the suppression pool. The reason why the amount of Te captured by the suppression pool is low
. is because it is released predominantly during the core-concrete interaction phase of the accident and thus does not pass through the pool.
7.3.3 TCy Sequence This accident sequence is similar to the TCy' sequence except that the fission products released from the drywell are assumed to pass through the secondary containment (reactor building). Therefore, an additional calculation 1
has been made simulating fission product behavior in the secondary containment. l The flow path of fission products from the secondary containment is partly directly to the environment and partly through the Standby Gas Treatment System 7-36 1
I
.-l
TRAP-MELT I
! (
'SPARC.(P0OL) y NAUA(WETIlELL) i VANESA I
I I
, y I v
' ' i NAUA (DRYWELL) NAUA (DRYWELL) i I
I I
I I
Vessel Failure t
FIGURE 7.16. NAUA AND SPARC CALCULATION PATHS USED FOR PERFORMING ANALYSES OF TC AND TW SEQUENCES 1
7-37 l
TABLE 7.10. DECONTAMINATION FACTOR CALCVLATED AS A FUNCTION OF PARTICLE SIZE AND OF TIME FOR TC Time Particle Diameter, um DF Based on (min) 0.1 0.7 1.2 5.1 8.4 Total Mass:
3 96.2 1.3 1.08 x 10 105 (a) 10 5
10 5
3690 99.2 5 5 5 5.2 98 10 10 10 2850 5 5 5 104 3.0 45 10 10 10 2166 5
121.7 1.1 15.8 1.87 x 10 3 10 10 5
7,7 131.5 5 5 1.2 4.5 41 10 10 298 5 5 156.3 1.2 4.0 32 10 10 600 (a) A decontamination factor larger than 105 is assumed to be 105 ,
Pool depth: 6.5 ft (198 cm)
Bubble diameter: 0.75 cm Aspect ratio: 1:3 i
l 7-38
~~
RIRBORNE .
h.: LEAKED - -
,f
b_ -= -
~
l :
~
I
- r. . .
/
l b.-; {'
1 "o _
-2 C.D -
, - 1
$b.,
c -' =: l r .
_.2 C-6 o o.=
t-e * =
4
- o _ -
- i i
l -
r o_
.-e 5 M
'o _
~5 10' 7*10*
TIME, MIN FIGURE 7.17. AIRBORNE AND LEAKED AEROSOL MASS FOR THE ORYWELL AS A FUNCTION OF TIME, TC-y' 7-39
- ~~~.______.
CSI ...
b .= CSOH
,- E TE -- ,
- OTHERS -----------
= ,
l
-i . . . . . .__ l
,,/./ l l
a ,.
"i . ,
l -
/~~
l.
b~1 I
cs -
Rb.
c -!
r :
. l
.a -
, c -
s a o.
s; .
b-: ./,
i l
'O ., l y . ,
O. ,
-E l 1 . 4 i , , , , ,
10' 7*10' ;
TIME, MIN l 1
l FIGURE 7.18. ACCUMULATED MASS LEAKED INTO ENVIRONMENT FOR VARIOUS SPECIES TC-y' 7-40
l l
l l
\
m z
O i x -
O e--e r
~^
g ,
's ,
a H
- a x
x l
l i
'o 10' 7*10' TIME, MIN FIGURE 7.19. GEOMETRIC MEAN PARTICLE RADIUS IN DRYWELL, TC-y' 7-41
j t *
^
4.1 - i t' 1
~
% ) \
<'e .~* \f t
. TABLE.7.11. ' DISTRIBUTION OF SPECIES AT 20 HOURS AFTER ACCIDENT, TC-Y y .I Fraction of Core #' Inventory Species RCS Pool Drywell Wetwell Environment Csl 0.06 0.69 1.5 x 10 -2 0 0.24 Cs0H 0.22 0.56 1.4 x 10-2^ 0 0.21 4qi e$
~
-3 0' Te 0.34 7.9 x 10 0.29* ,. O.37 *
. . . . . . . . t .
- This includes.a fraction of 0.13 for Tc that is found not to be
. released from the core-concrete interaction. .
s e
a
-. ,g
~
y
,. %g %.
.g.
I a
4 [
T
+ %
A I
4 1
3 e
' g. '
\
4 4 q e
, ~f 'f .J (
r ,
); f } n 4 ?! ,
y1 i ... s r*
l%
Z
' .h ) ~% "
- s \
iy l
's 7-42
- 1
~
- l. k e
)
. ., ._. _ , .- r - - - - . - . _ _ . , . - _ _ ,, , . _ , . - - .-
.,4 4
xM g (SGTS) which consists of High Efficiency Particulate Air (HEPA) filters. The cp; ration pnd the effectiveness of the SGTS are described in Appendix A. It wasashJme in the }.Nse'n$ calculation that a constant volumetric flow rate of b25,000cfMt$roughthe'SGTSismaintainedduringtheaccident. A collection p qfficiency of 99.99 percent, regardless of, particle size, was utilized.
Howevek,thefilterbankswereat,sumed.tcfailafter108kgofparticulate catter was collected as 'speciflied in the analyses in Appendix A.
, Figure 7.20isthy%airbornemassplottedasafunctionoftimefor th3 secondary cont,ainment.4 As expected, a time dependency of the airborne mass very similar'to that for the drywell as depicted by Figure 7.17 is seen to occur in the secondary containment.
, Table 7.12shows'thediskributionoftheCsI,Cs0H,andTeascalcu-t lated % tie,NAUA code. It is interesting to find that compared with Table 7.11 a fraction eo'f ' core inventory ranging from 0.12 for Te to 0.14 for CsI is retained in the secondery containment and in the filter.
s ,. i 7.3.4 TWV Sequence 1 mf ,
< t -
s7 3 in m This seqde ce is similar to the TCr ' sequence in that the flow paths for fission productk are the'saIe and that the containment fails before core i ^
meltdown takes piace. Hgwever, the tinie scale for the present sequence is d much more ext $rded than hat for the IOY' sequence. The key accident events and containment conditions used in the TW analysis are given in Section 6.1.3.
Generally a callulation procedure similar to that adopted for the TC sequence was taken. The sequential use of various computer codes depending upon the accident events is shown in Figure 7.16. The decontamination factors calculated by the SPARC coc'e are listed in Table 7.13. It is shown in the a;
table t'at e the overall decontamination factor of the suppression pool ranges p from 326 at 2818 minutes to 105 at 2827 minutes after the accident commences.
The calculated decontamination factor depends not only upon the thermal hydrau-lic conditions in the suppression pool but upon the particle size distribution 4
of tho aerosol entering the prol. For example, at 2815 minutes, the mass median partf;1e size enterin$the suppression pool is found to be about 0.66 um causing 3 ,
the g%al,1 decontamination factor at;that time to become as high as 352.
+ ' Figure 7.21' is the calculated airborne mass as a function of time for t$e drywell and Figure 7.22 is accumulated mass leaked out to the atmosphere.
N ,
f ;T- 7 43 9 : y I .,-
0 -
s m, l ; a 4 ;\
1is 1
~
RIRBORNE
% LERKED
- 7 #.
l l
b_-= -
/
l :
I 1
.r-
/
~
o_
.-4 : L i
t
. t
""o _ i s .-* :
m : '
w -
c -
r -
I
__J eb s- j o : t E-* -
- .J
- I
~
o_
t
- 1
_ f o_ i
~! ~
i
.\
_ f
'o _ b
~E J i
\
i , , , , , , , , ,
10' 10' 2*10' TIME, MIN ,
FIGURE 7.20. AIRBORNE MASS FOR THE SECONDARY CONTAINMENT AS A FUNCTION OF TIME TC-y 7-44
J 1
TABLE ' 7.12. DISTRIBUTION 0FSPECIESAT'20HOURSAFTERACCIDENT,TCY Fraction of Core Inventory Reactor- . w Species -RCS Pool Drywell Wetwell B1 dg ..SGTS E'nvironment Csl 0.06 0.69 1.5 x 10 -2 0 6.'9 x 10
-2 6.8 x 10 -2 ' 0.10 -
-2 Cs0H 0.22 0.56 1.4 x 10 -2 0 6.1 x 10 5.8 x 10 -2 9.l ' x 10 7.9 x 10 -3 1.3 x 10 ~
Te- 0.34 0.29* 0 '0.11 0.25-
' *This includes a fraction of 0.13 for Te which is found not to be . released from the core-concrete interaction.
?
g ,
9
.- - L'--_l_._--.._-
W N TABLE 7.13. DECONTAMINATION FACTORS CALCULATED AS A FUNCTION OF PARTICLE SIZE AND OF TIME FOR TW SEQUENCE l
Time Particle Diameter, um DF Based on-(min) 0.2 05 1.0 4 10 Total Mass l 4
2756 -1.9 18 1.17 x 10 105 (a) 10 5
257 2777 1.4 9.4 2.51 x 10 3 10 5
10 5
576-2801 1.2 10.5 3.8 x 10 3 10 5
10 5
408 2811 1.1 10.4 2.2 x 10 3 10 5
10 5
865 2815 1.1 11.7 3.5 x 10 3 10 5
10 5
352 2818 10.3 60 9.9 x 10 3' 5 5 10 10 326 2820 2.3 x 10 3 7.2 x 10 3 1.7 x 104 110 5
10 5
1336 5 5 5 5 5 5 2827 10 10 10 10 10 10 (a) A decontamination factor larger than 105is assumed to be 105 ,
Pool depth: 6.5 ft (198 cm)
Bubble diameter: 0.75 cm Aspect ratio: 1:3 7-46
E l RIRBORNE L. -b_ LERKED ~ .._:__-
y, l
~
l b_-, if --
\
b._:
1 co b
_5 s : ,
m -
m _
x .
r
_ab- -
x :
H -
o -
e -
~_a
~5 o_
-2 T*
o_
~ ::
i :
~
~
'o
~-
3*10' 4*10'
' TIME, MIN FIGURE 7.21. AIRBORNE AND LEAKED AEROSOL MASS FOR THE DRYWELL AS A FUNCTION OF TIME, TW-y' 7-47
x CSI ,
CSOH - -
TE- --
OTHERS - . _ m ,. =
b_
4
. .i i 1 i
-_ i l
i l
- 8 i
- i i
i i
e !
s i sb_
c-_
l r : i g:
H o
l7
.i./
H -
lf "a _ Y,
. i i
b_
3*10' 4*10' TIME, MIN FIGURE 7.22. ACCUMULATED MASS LEAKED INTO ENVIRONMENT FOR VARIOUS SPECIES, TW-y' )
7-48 i
1 4
A
.-- , 4
' Table;7.14'is the distribution of solid' particulate mass at various. containment
. locations and at-various times.. Recognizing that the flow path'of-fission
-products for an accident ~ time up to the vessel failure time (3055 minutes) is from RCS~through the suppression pool and wetwell and finally to the drywell,- l approximately 1.2:kg of particulate is calculated to be released:to the environ-
. ment during this period of time while about 540 kg is retained in the suppres-sion pool. After the_ reactor vessel fails, fission products are released
_directly into the drywell and transported to the'_ atmosphere. From Table 7.14 sit isLseen that about 2000 kg of particulate is retained in the drywell
- primarily by the gravitational settling coupled with the particulate growth
- ' mechanism while 1350 kg reaches the atmosphere.
- Table 7.15 is the locational distribution of various species;as frac-F tion of core inventory. -It should be noted that a very small-amount of core
'~
inventory fraction was found to be retained in the wetwell at the. time the vessel failed. Considering that the assumed " floor" area of the wetwell is the water surface of the suppression pool, the' amount calculated.to settle in the wetwell was=added to the amount retained in the s'sppression pool. For t'eis reason, distribution of particulates beyond the vessel failure time is i
shown to be nil both in Tables 7.13 and 7.14. In general, about one-fifth of' Csl and Cs0H is seen to be released to the environment in Table 7.15 when compared with the results for TCy'3(Table 7.11). Nearly one-half of Te escapes from the reactor in TWT' cc:npared to TCy'.
I t
7.3.5 Pesults for Release qf
~
l Reactor 5afety Study Groupss An additional calculation for the AEY' sequence wcs Carried out to obtain infc mation for transport ands retention of fission prcducts in the l
containment for WASH-1400 fission product groups. As was discussed in Chapter 6, l
six' groups.were identified: I (Group 2), Cs (Group 3), Te (Group 4), Sr
! '(Group 5), Ru (Group 6), and La (Group 7). Xe a'd Kr, which are members of Grcup 1, do not appear in the containment calculation because they are not deposited in the containment and are in gaseous form and not treated in the aerosol codes. For Groups 2 through 7, results from TRAP-MELT calculations were adopted as the sources to the containment in addition to the core-concrete interaction source from VANESA. Other than Cs, I, and Te, VANESA calculations 7-49
1 TABLE 7.14. DISTRIBUTION OF TOTAL SOLID PARTICULATE MASS (KG) AT VARIOUS TIMES, TW ud Time Drywell Pool (min). Event Suspended Deposited Captured Wetwell Environment 2737 Melt start 2786 0.15(a) -0I ") 55 0.01- -0(a)
Slump starts 0.42I ") - -0I ") I
-0")
I 2815 203 .0.131.
l 2822 Core collapse 1.18I *) -0I ") 491 0.148 - 0I" ).
- 2825 1.21I ") -0(a) 511 0.032 --0(a) 2827 Vessel dry 2900 -0 OI ") 540 -0 11.22I ")
Head fails
? 3055 o --3062 25.8 3.8 540 -0 56.3 -;
3120 116 677 540 -0 835' j 3400 16.3 1990 540 -0 1325 4
3600 -0 1997 540 -0 1338 i 3660 Accident ends p (a) Estimated.
l i
l TABLE 7.15. DISTRIBUTION OF SPECIES AT 60 HOURS AFTER ACCIDENT, TWy
E ll I EI II Fraction' of Core Inventory Species RE3 Pool Drywell Wetwell Environment Cs! 0.14 0.80 5.4 x 10-3 0 4.8 x 10 -2 Cs0H 0.15 0.79 5.0 x 10 -3 0 4.5 x 10 -2 Te 0.40 ' 8.6 x 10 -3 0.40* 0 0.19
- This includes a fraction of 0.20 for Te that is found not to be released from the core-concrete interaction.
7-51 s
n
[ supplied _therelease~ratesforMo,'Ru(Group 6),Ba,Sr(Group 5),La,Ce,'and
~
h
'Nb (Group 7). Using Cs to represent Group 3, Mo and'Ru to represent Group 6, L 'and La, Ce, and Nb to represent Group 7, scaling factors were used to obtain " '
[ the respective total group' release rates. . Scaling factors are defined as the elemental inventory for each entire group divided by the sum of the inventories.
of_the elements representing the group.
Table 7.16 shows the fraction of fission products released to the ;
i environment for this calculation of release by group.- Compared to Table 7.8,
, the release fraction of Group 2 is the same as the Cs1 release fraction, and the release fraction of Group 3 is the same as the Cs0H release fraction. This is because all of the iodine is included in Cs! and the major portion of Cs is included in Cs0H.- For.the. release rate of Group 4, nothing is modified from the previously calculated release rates for Te due to the small inventoriesc for Se and Sb, which are two of the three members in Group 4. Therefore the release fraction for Te in Table 7.8 is identical to the release fraction for t
Group 4 in Table 7.16. Since the scaling factor for Group 7 is 5.9, the uncer-tainty in the release fraction for Group 7 could be considerable. The release fraction for Group 6 is the smallest because the low fraction released from "
4 the fuel for this group.
7.4 Discussion l In WASH-1400, the TW and TC system sequences dominated the predicted l risk to the public for this reactor design. This dominance was a result of 1
the estimated likelihood for these sequences as well as their consequences.
I The estimatcd mean core melt frequency for each sequence was 1 x 10-5 yr-1 This was compared with a median frequency of 5 x 10-7 yr-1 for the TQUV system sequence which had the next highest frequency. The estimated median frequency l for the AE sequence was 1 x 10-7 yr-1 Thus, the TW and TC system sequences were more than an order of magnitude higher in estimated frequency than the next most likely sequence and were two orders of magnitude more likely than the background of remaining sequences. This perspective on the importance of l the TW and TC sequences has not changed substantially in the intervening years and is reflected in the results of the Accident Sequence Evaluation Program.
For the BWR sequences only four fission product release categories
- were used for core meltdown sequences as opposed to the seven categories used 7-52 J
---, - - . - ,-,--r.-r. -- - , - . , - . . - - - . - - , . . , , . - - - ~ , r,, - - - - ~.--er---.-~m.,~ ,r---w .-
I b
i 1
l i
TABLE 7.16. FRACTION OF CORE INVENTORY RELEASED TO THE ATMOSPHERE-FOR GROUPS OF REACTOR SAFETY STUDY-(AE)
Time I Cs Te Sr Ru La (hr) Group 2 Group 3 Group _4 Group 5 Group 6 Group 7 0.5 0 0 0 0 0 0
-3 1 0.19 0.19 3.6 x 10-2 1.2 x 10 -2 2.7 x.10 9.9 x 10-5
-3 2 0.251 0.24 6.6 x 10-2 1.3 x 10-2 3.6 x 10 1.0 x 10 -4 4 0.34 0.33 0.51 0.64 4.6 x 10-3 0.44 7 0.34 0.33 0.64 0.68 4.6 x 10 -3 0.49 10 0.34 0.33 0.65 0.68 4.6 x 10 -3 0.49 15 0.34 0.33 0.65 0.68 4.6 x 10 -3 0.49 7-53
+.
I
for the PWR. The time-integrated release fractions are shown in Table 7.17 for each of the release categories. Categories BWR 1 and 2 involve very large fractions for the volatile fission products and are comparable in severity.
The comparatively.large release of ruthenium in the BWR 1 sequence is the result of fuel fragmentation, dispersal, and oxidation following a steam explosion that f*'Is the primary containment.
All of the in-vessel steam explosion failure sequences, a, were assigned to the BWR 1 category. Because of the low probability of steam explo-sions leading to containment failure (10-2), this failure mode was not a major risk contributor in WASH-1400.
In general, the overpressure failure sequences with direct release to the environment, y', were assigned to Category BWR 2 and the overpressure failure sequences with deposition in the reactor building, y, were assigned to Category BWR 3. This category is less severe than BWR 2, but would still result in major public health consequences. Since 25 percent of overpressure sequences were assigned to y', the relative probabilities of tne sequences in Category 2 and Category 3 in WASH-1400 were approximatelj in the ratio 1:3. An important exception to the assignment of failure modes to release categories was the treatment of the system sequence TC. The analyses performed for TC in WASH-1400 indicated that core meltdown would precede containment failure and that some retention of fission products would occur in the suppression pool. For this reason, TCy and TCy' were both assigned to Category 3, increasing its relative importance.
The final core melt release category, BWR 4, involves significantly '
reduced release fractions. At this level of in-plant attenuation, the actual values for the release fractions for the volatile fission products are not particularly important because the public health consequences would be dominated by the noble gases. The principal sequences in this release category involved isolation failures in which the small leakage rate from the primary containment could be treated by the SGTS prior to release to the environment. The likeli-hood of this mode of containment isolatio1 failure was estimated to be small.
The combination of low probability and low consequences made the contribution j of Category BWR 4 to the public risk negligible. Thus, Categories BWR 2 and BWR 3 were the important risk contributors.
In Table 7.18 the release fractions for sequences TWy', TCy', and AEy' are compared with the WASH-1400 values. The lower release fractions are 7-54
.s TABLE 7.17. WASH-1400 RELEASE CATEGORIES Release Fraction of Core Inventory Released Category Ke-Kr I Cs-Rb Te-Sb Ba-Sr Ru La BWR 1 1.0 0.4 0.4 0.7 0.05 0.5 5 x 10-3 BWR 2 1.0 0.9 0.5 0.3 0.1 0.03 4 x 10 -3 BWR 3 1.0 0.1 0.1 0.3 0.01 0.02 3 x 10-3 l BWt 4 0.6 8 x 10-4 5 x 10 -3 4 x 10 -3 6 x 10-4 ' 6.x 10
-4 1 x 10
-4 '
7 l
l
\
- a i
i i
i i
e 1
TABLE 7.18. COMPARISON OF SEQUENCES ASSIGNED TO BWR 2 l
Fractic., of Core Inventory Released Sequence Xe-Kr I Cs-Rb Te-5b BWR 2 1.0 0.9 0.5 0.3 AEY' 1.0 0.2 0.2 0.7 TCY' 1.0 0.3 0.3 0.3 TWY' 1.0 0.3 0.3 0.2 i
1 7-56
Cie result of predicted retention in the reactor coolant system and in the suppression pool. Because of the high temperatures in the reactor coolant system, the predicted retention for the volatile species is limited. The retention of the less volatile species as aerosols is more effective. In the transport of fission products through the suppression pool, it is assumed that all radionuclides other than the noble gases are in the form of aerosols. For the AEy' sequence, the decontamination factor is rather large for material carried into the pool prior to containment failure. After containment failure th2 flow path would no longer include the suppression pool.
In the TCy' and TWy' sequences, the release pathway is through the pool until the time of reactor vessel melt-through. The pool of water is saturated during this time period so the amount of decontamination is not as great as for the AE system sequence. As discussed earlier, the decontamination factor for the saturated pool is very sensitive to the size distribution of the aerosols and the resulting uncertainty in the release fractions in Table 7.18 is large.
The TCy sequence was analyzed in order to examine the potential effectiveness of the reactor building and SGTS in further mitigating the conse-quences of an accident when the structural integrity of the reactor building is maintained following failure of the primary containment. The results for this sequence are compared with the release fractions for Category BWR 3 in Table 7.19. The effect of the reactor building systems can be determined by comparing the results for the TCy and TCy' sequences directly. The quantities of fission products, other than noble gases, that do escape to the environment are primarily the result of outleakage from the reactor building. The TCy sequence analyzed did not consider the effect of the fire sprinkler system on l further reducing the fission product source term if it were activated by the l hot steam during blowdown of the primary containment. The sprinkler system would be expected to lead to a significant additional reduction in the source term both due to washout of airborne aerosols and due to steam condensation which would result in a net air in-leakage to the building. Eventually the mixing of air and hydrogen in this sequence would probably result in flammable conditions with the potential of destroying the reactor building. The conse-quences of the sequence would be very sensitive to the timing of the hydrogen burning event.
7-57
l i
- TABLE 7.19. COMPARISON OF SEQUENCES ASSIGNED TO BWR 3 ,
, i l
l - . _ . ._ .. _
Fraction of Core Inventory Released l Sequence Xe-Kr I Cs-Rb Te-Sb BWR 3 1.0 0.1 0.1 0.3 TCY 1.0 0.2 0.2 0.2 j
i h
7-58
.i
The analyses performed for the Peach Bottom 2 reactor, indicate that the WASH-1400 release fractions in the high consequence BWR 2 category are overestimated for iodine, cesium, and'the involatile fission products. The specific scenario analyzed to compre to the BWR 3 category resulted in compar-able but slightly higher releases than in WASH-1400. Slight variations in
~
modeling assumptions would have led to substantial additional retention, however.
The implications of these results to the WASH-1400 perspective on risk depend on the relative likelihoods of the y and y' failure modes. If the reactor building were expected to withstand blowdown forces,from'the failure of the primary containment with a high degree of confioence, the risk could be reduced substantially. The uncertainties in the thermal, hydraulic, and structural response of the primary containment structure and the reactor building are very large, however. The scope of the current effort was oriented at develop-ing an understanding of the consequences of given accident scenarios rather than estimating the likelihoods of alternative scenarios.
The results presented in this report are referred to as best-estimate results. Sensitivity studies will be performed later in the program to provide perspective on the ranges of the uncertainties associated with the best-estimate values. In each phase of release and transport, important uncertainties exist.
Changes in assumptions and models could affect the prediction of the environ-mental source term substantially.
In calculating the initial period of core heatup, the MARCH 2 code has been demonstrated to have reasonable accuracy. After fuel melting and slumping begins, the MARCH 2 code models very complex phenomena with very simple models. The maximum fuel temperature is probably underpredicted by MARCH 2 and the time at temperature is quite uncertain. As a result, the timing of release of volatile fission products and the magnitude and timing of release of less volatile fission products from the fuel could be affected by thermal hydraulic mcdeling deficiencies.
Thermal hydraulic modeling uncertainties also have a major impact on uncertainties in the transport of fission products in the reactor coolant system, primary containment and the reactor building. The temperatures and flows in the reactor coolant system directly influence the extent of retention.
Indeed, the MARCH 2/ MERGE results indicate temperatures that are just high enough to limit the retention of the volatile species Csl and Cs0H. The predicted time, location, and mode of failure of the primary containment 7-59
- ~. . _. _ _ ~_ -_ _ - ___ _ - 1
t building also have a major influence on the predicted environmental source term. :The predicted failure of the primary containment prior to pressure j vessel melt-through, coupled.with failure in the drywell region, resulted in a large fraction of the fission products bypassing.the' suppression pool in the L AE sequence. The rate of' generation.of hydrogen, failure pressure for.the- l primary containment, and location'of failure (drywell versus wetwell) are
.sufficiently uncertain that the actual' accident scenariocould follow a path-way that could result in substantially greater attenuation in the suppression L pool. Similarly for the TC and TW sequences, failure of'the primary containment in the wetwell region could result in-su'bstantially less retention in the suppression pool.
-The decontamination factor for the pool is itself quite uncertain ,
even with well specified input thermal hydraulic conditions. The decontamina-tion factor predicted by SPARC is very sensitive to aerosol size. A small shift in the predicted size distribution could change the results of the calcu-lated aerosol retentions substantially.
Finally, the reactor building potentially has the capability to significantly reduce the environmental source term depending on its response to failure of the primary containment. The associated technical questions will be extremely difficult to resolve.
In conclusion, the reader should be careful not to assign greater significance to the single-valued best-estimate source terms that are provided in this report than warranted by the uncertainties in the underlying assumptions and methods of analysis.
7-60 4
1 I
i l
APPENDIX A
. STANDBY GAS TREATMENT SYSTEM OPERATION AND
! EFFECTIVENESS UNDER SEVERE ACCIDENT CONDITIONS:
PEACH BUIIOM AND GRAND GULF NUCLEAR STATIONS I
l
r:
APPENDIX A STANDBY GAS TREATMENT SYSTEM OPERATION AND EFFECTIVENESS UNDER SEVERE ACCIDENT CONDITIONS:
PEACH BUIIOM AND GRAND GULF NUCLEAR STATIONS Introduction In the unlikely event that a severe accident should actually occur at a Boiling Water Reactor (BWR) facility, the primary contain- -
ment would eventually fail. Fission products would pass from the pri-mary containment into the secondary containment and from there to the atmosphere by way of the Standby Gas Treatment System (SGTS). It is the function of the SGTS to mitigate the fission product release by means of exposure of the exhaust flow to filters and activated char-coal. The operation and effectiveness of the SGTS at the Peach Bottom and Grand Gulf nuclear stations under severe accident conditions are.
discussed in this report. Most of the design information has been obtainedfromtherespectiveFinalSafetyAnalysisReports(FSARs).
Information concerning filter train effectiveness has been taken from
" Analysis of a Small-Break LOCA Outside Containment at Browns Ferry Unit One --Iodine, Cesium, and Noble Gas Distribution and Release",
NUREG/CR-2672 Volume 2 (to be published).
Peach Bottom System Design and Description In the BWR MK-I containment design employed for Units 2 and 3 at the Peach Bottom Station, the primary containment comprises the drywell and wetwell. In each unit the primary containment is surrounded by a reactor building which includes two ventilation zones, the areas above and below the refueling floor. During routine operation, the
A-2
. reactor-building is kept at a pressure just below atmospheric by the normal building ventilation system.
Under accident conditions *, the normal ventilation system for the affected reactor building is automatically shut down and the SGTS is automatically started. The SGTS is designed to minimize the escape to the environment of radioactive material released to the
]
reactor building of either unit from its primary containment. The SGTS filters the exhaust, exposes it to charcoal absorption, and i provides an elevated release through the 500-ft plant stack. A schematic drawing of the system is provided in Figure A-1.
The SGTS comprises suction ducts, two parallel filter trains, three full-capacity (10,500 cfm) exhaust blowers, and the exhaust ducting to the plant stack. The filter trains and blowers are located in an underground shielded room in the basement of the radwaste building. The suction ducts connect with the normal ventilation system exhaust piping and duct work of each of the reactor buildings. The normal ventilation system exhaust ducts are provided with fast-acting (3-5 sec) cylinder-operated butterfly valves for rapid isolation when an accident signal occurs. The isolation signal also actuates the SGTS; both parallel filter trains go into operation with suction from all three blowers.
The volumetric flow from the affected reactor building would be about 25,000 cfm with all three blowers running, but the operator would be expected to shut off two of the blowers to reduce the flow to the SGTS design value of 10,500 cfm. A vacuum relief system is provided to prevent the differential pressure between the outside atmosphere and the interior of the reactor building from exceeding l 1/2-inch H20. The total flow into the SGTS is taken approximately equally from the two ventilation zones above and below the refueling floor.
- Signaled for an affected unit by high radiation in the exhaust duct, low reactor vessel water level, or high drywell pressure.
1 1
l ens.-e n-assa 4 -
am n
_ =
l _
_ =
l
- T emme nam FLOW mn to sovs l l SYsTDs i
l'\ /'\
) -
sors nota
?
w a
, inuus i r (2) 1 r i, U, ,
M = .- J j i L m
e ,.
v anv G
! .;:: s s s s sss:--- --- :::N6+99+2 l
2.il d d;:.:.:.--- b {
sois stowots (3) i i FIGURE A-1. SCHEMATIC DRAWING 0F PEACH BOTTOM STANDBY GAS TREATMENT SYSTEM. AS SHOWN, PROVISION IS MADE FOR EXHAUST FROM THE PRIMARY CONTAINMENT DURING NITROGEN INERTING OR UNDER ACCIDENT CONDITIONS.
i
~ . . - - . _. _ - . - ....
1 1 :A l Each SGTS . train comprises six elements. -Th'e first is a mois-
[ ture separator designed to remove 99.9 percent of the_ moisture particles l
2 microns or larger and'30 percent of the 1-micron particles. The
- separated moisture _is' collected in a sump and pumped.to-the radwaste building.. -
-Just after,the moisture separator,'an electric heater (ener-gized with the startup of the SGTS)'is~provided to reduce the relative humidity of the mixture of steam and other gases in the flow. The heater will cut off automatically on low flow'o- if the temperature of the flow exceeds a set value. Thus, the relative humidity of a flow at a temperature greater than the set value would not be reduced.
A water resistant'prefilter_(roughing filter) designed to-
~
- j. remove large particulates and thereby protect the downstream High Effi-4 ciency Particulate Assembly (HEPA) filters is installed just after the moisture separators and heater. The prefilter actually serves as a ,
l backup to the moisture separator, which also acts as a good prefilter.
The fourth element in each filter train is a bank of nine 2-ft2 High Efficiency Particulate Assembly (HEPA) 1000 cfm filter cells, t
- - which will remove 99. 97 percent of particulates 0.3 micron and larger.
The HEPA cells have water retardant glass fiber media, aluminum separa-tors, and cadmium-plated steel frames with rubber base adhesives and neoprene gaskets and have a design temperature capability of 250 F (200 F for continuous service).
Exposure to radiation during a severe accident might reduce
, the efficacy of the HEPA filters. Radiation testing of HEPA filters *
! coordinated at the Naval Research Laboratory for exposures as high as l 108 rads has shown some reduction in the tensile strength of the filter
! media and some deterioration of the rubber adhesive and the neoprene gasket. The gasket, although brittle, continued to seal. Thus the
- filter media, adhesives, and gaskets should remain satisfactory for
- _ continued filtration ard normal pressure losses with exposures as high as 108 rads.
The fifth element in each filter train is an activated char-L coal bed containing a uinimum of 1,320 lb of charcoal. The bed consists-I L r
A-5 of individual rigid, welded, leaktight, stainless steel rectangular canisters. The canisters are filled with impregnated charcoal with an ignition temperature of'640 F. Each canister contains'about 50 lb df charcoal. The charcoal canisters are mounted in dual tray module drawers.
Each train has 27 drawers arranged in a single bank of nine horizontal rows with each row being three drawers wide. The flow is vertical through a 2-inch thick layer of charcoal, and is equally distributed across all of the canisters. The canisters are sealed to the bulkhead frame with continuous gaskets and are held in place by individual clamps.
Water sprays are provided to prevent overheating of the charcoal filter if the air flow should be totally lost while a large radioactive iodine inventory is present.
According to the Peach Bottom FSAR, the carbon absorber unit is capable of removing at least 85 percent of methyl iodide (CH31) and 99.9 percent of elemental iodine (12 ) under entering conditions of 70 percent relative humidity of 190 F. It should be noted that the Browns Ferry FSAR claims a 95 percent removal efficiency for methyl iodide for a similar charcoal bed.
The sixth element in each filter train is a second HEPA fil-ter, identical in characteristics to the first HEPA filter.
l l
HEPA Filter Behavior HEPA Filter Efficiency.' Individual 1000 cfm HEPA filter units usually have an efficiency of 99.99 percent when tested with dioctyl-pthalate(DOP) smoke. An assembly of these filters might show a lower average efficiency because of gasket leakage or minor damage to the filter medium during installation. Filtration of accident-generated aerosols should occur with efficiencies similar to those found for 00P aerosols. When both HEPA filter banks (in front of and behind the charcoal absorber units) are intact, the combined efficiency can safely be assumed to be 99.99 percent or greater.
, . - = .
1 1
j A-6 HEPA Filter Plunaine and Ruoture. Large amounts of aerosol )
.will be collected in the SGTS units during a severe accident sequence.
I I
- Calculations show that the amount collected can result in a pressure-
- drop across the HEPA filters' sufficient to cause rupture of the filters. ,
, The pressure differential required for filter failure is uncertain, ,
but it is reasonable to assume that a differential pressure equivalent [
. to 12-inch water will cause rupture. With this assumption, the flow !
- conditions at the time of filter failure are illustrated in Figure A-
- 2, which is a pressure drop-flow diagram for a single 9000 cfm SGTS train which has nine filters in-parallel in each bank.- According to ,
(. this diagram, the initial clean system flow is 9000 ft /3 min with a ,
f pressure drop across the system of 13.3-inch water. (EachcleanHEPA l filter bank has a pressure differential of 1.0-inch water.)
l As aerosol particles are collected on the nine filters'in I the first HEPA bank, the pressure drop across these filters increases ;
j and the system flow decreases following the characteristic curve of l I
the blower. When the first HEPA filter bank reaches a pressure drop' f of 12-inch water, the system flow will have decreased to 3300 ft 3/ min [
and it is reasonable to assume that one or more of the nine HEPA filter units in this bank would fail with openings large enough to decrease' r i both pressure drop and filtration efficiency to zero. In addition, it :
! is reasonable to assume that the second HEPA filter bank is not damaged, j and that the aerosol particles passing through the charcoal absorber
, unit do not affect the adsorption of iodine species, aranular charcoal l beds are poor filters of particulates, and it can be assumed that no ,
j aerosol deposition occurs in the charcoal beds. l l The behavior of the second HEPA filter bank can be assumed !
l to duplicate that of the first with respect to buildup of pressure
} differential and rupture at 12-inch water. The overall filtration !
f efficiency would drop to 99.9 percent since only one filter bank is in )
l, operation.
1 As previously stated, the exact conditions that result in i massive filter failure are unknown. New filters are required to with-t I
- stand 10-inch water while the filters are exposed to a moist I V ,
~-~r , - , - - --w -w mm -
ew, nm,,.w, e,--- ,mn,,--r,+, n - ,,-m+--nn,--,,--,---,,.,~--r- .~ , ~ .,, m - ~ wr
A-7 onese.-asse es-mse era f l I I f Less M884STAseC4 Of 8418 Cees-I A44de#e4 44EAVsLY LOADES seafA 4
74 LIGA DAesetNO.488704876LTEA Atritmet AT af a sala.WA184 f
' f L Oev A ssesT Aasca e6 - of 84iscues- _
TA4eee##4 CidAss j 48EfAfitIGAS i
CLE A40 SGIS i E ACLOO4864 fIMST 8487A l SANK i s /
- a /
8 /
i j
j
/ CelAllACIGH45IIC i 6 -
/ Cl44VE of etOsttil/ -
'. E i
g !
8 Mme inone as.ooe ILOW atATE ele3panel l
l FIGURE A-2. FLOW CHARACTERISTICS OF A SINGLE SGTS TRAIN 1
i t
l l i
. , , . . . _ -_...-.__ _ -, ,, ,_,,. . ,.en...n -,n.,-_n.. , , , , , . - _ , _ _ . ~ . , _ - , . _ . _ . , . _ _ . , . , , _ . , , , _ . - , - , _ - - . _ _ _ _ _ _ . .0__._.
'A-8 i
l atmosphere.(A-1) The flow conditions cor esponding to the assumption of filter failure at other pressure differentials can be obtained from Figure A-2. .For 10-inch water, the system flow would be 4600 ft 3/ min; for 12-inch, 3300 ft3/ min; for 15-inch, 800 ft3/ min; and for failure threshold above 16-inch water the filters would not fail.
No data are available for pressure buildup on HEPA filters ,
caused by accident-produced aerosols. In severe accident sequences, most of the aerosol would result from core melt-concrete interaction.
Pressure buildup data in the literature are mainly from natural atmos-pheric aerosols, and these tests are usually terminated when the pres-sure drop reaches 3- or 4-inch water. An average for seven such tests is 4-inch water pressure drop for an accumulation of 1.2-kg atmospheric aerosol particles on a 1000 ft3/ min HEPA filter.(A-2) This is equivalent to 10.8 kg on a nine filter SGTS bank. Similar behavior was observed with five tests of various HEPA filter types when cperated at design flow conditions.(A-3)
A best estimate of pressure drop across the first bank HEPA filters as a function of accident-produced aerosol collected in one 9-filter SGTS train is shown in Figure A-3. The mass shown is that col-lected in all parts of the SGTS: ductwork, demisters, prefilters, and HEPA filters. It should be emphasized that the pressure drop shown in Figure A-3 is for the full flow conditions; reduced flow and pressure drop would occur in a real SGTS system as suggested in Figure A-2.
As an illustration, the HEPA filter behavior assumed for the computer simulation of the Browns Ferry Scram Discharge Volume Break (SDV) accident sequence is shown in Table A-1. Flow reduction through the SGTS resulting from aerosol collection was not considered. This results in faster coilection of the amount of acrosol required for obstruction and consequent rupture of the HEPA filters. However, the time difference would not be very great since the high pressure drop j and flow reduction occurs to a significant extent only with the last l 10 or 20 percent of the aerosol collected. i If we had chosen 15-inch water as the rupture criterion, the flow reduction would probably have been enough to allow the reactor a
- s I
. 1 i
l A-9
?.3
~
J %'
i j
! (
?
l easet-espe es-etet 370 I I l 1 l i .
1 1
,M - -
I s
e i
M - -
g le - <
y.
(
't .
?s ':
, > L q~ ,< . 9
' * .c 4
e
' I I I l e s se is 2e 1s 2e 4 i
MASS OP AGR080L COLLECT 80 iM ONE SGTS TRAIN thgl d
d' "y g.. ,
FIGURE A-3. PRESSURE DROP A^ROSS A HEPA FILTER BANK t
d l
l :t d
-S
_ _ .. . .. . . . .. - . . . . . . . . . . . . . . . . , . . . ~ . .
/\
- j. % /
- s. ,
?
A-10
- s. ,
r ,. i
- i. -
l4 T' TABLE A-1. 'hEPA FILTER BEHAVIOR FOR THE SDV BREAK ACCIDENT CALCULATION FOR BROWNS FERRY t .
I Interval ., Aerosol
? (Determined Aercsol Load Filtrati on by Aerosol . in SGTS Efficiency
! (kg) Flow Rate
- Load) (%)
k .
- . . First 0-27 per train 99.99 Normal
- no reduction p ;J (0-81 per 3 trains) -
5 , Second 27-54 per train 99.9 Normal: no reduction ^
- (81-162 per 3 trains)
E Third 54 per train 0 Normal
- (>l62 per 3 trains)
F
. i; V ,.
E E ,
E E
a i
P s
h-u _
P ,
m ;
E
.. ~ . . . . . _ . __
A-ll building to become positive in pressure so that some of the aerosol and iodine vapor species would have leaked out without filtration.
Leakage directly from the building would also occur if no HEPA filter rupture had been assumed. The low flow rates concomitant with the 15-incn water rupture and no-rupture scenarios would result in elevated charcoal bed temperatures, enhanced iodine desorption, and possible charcoal ignition.
Charcoal Adsorber Efficiency The tray type charcoal adsorber system described previously provides a face (superficial) velocity through the bed of 40 ft/ min which is typical for American nuclear reactor adsorber systems. The 2-inch deep beds result in a stay-time of -0.25 s.
Charcoal Adsorber System Efficiencies -- General. Charcoal adsorbers in nuclear reactor ventilation systems are designed to trap two chemical forms of radioactive iodine: molecular iodine (12 ), and organic iodides (mostly CH 3 1). Molecular iodine is easily adsorbed from gases by activated charcoal, but the charcoal must be treated with special additives (impregnants) in order to efficiently trap and retain radioactive iodine in the organic iodide form. Even with very high quality impregnants, charcoals are normally significantly lower in efficiency for collecting organic iodides. Once the radioiodine from the organic iodide is trapped, either by isotopic exchange with the excess of nonradioactive iodine in the impregnant (as with a KI or KI3 type impregnant) or by direct chemical reaction with the impregnant (triethylenediamine, TEDA, for example), the radioactive iodine from the organic iodida becomes indistinguishable from that originating as radioactive 12 . No significant desorption of trapped radioiodine will occur unless the charcoal temperature rises above -150 F.
Charcoal efficiencies are usually determined by subjecting small representative quantities of a comercial lot to laboratory tests as specified in ANSI N510 and ASTM D3803. If the charcoal meets these
- . . . . , . . . . . . - . . . . ., ~ .. . ..
b 9-- _
"E
-2 A-12 $
requirements and is installed in a properly built and leak-tested sys-tem, it is assumed to be able to function with 12 and organic iodide 7 trapping efficiencies somewhat lower than the test results show.
Adsorption efficiencies and desorption rates (at elevated
[
f-temperatures) can be calculated for nonstandard conditions by using T_
conventional correlations for gas-to-solid mass transfer in packed -j beds. An example is the correlation of Chu, et al.(A-4-A-6) The 2 parameters of importance are particle size, void fraction, gas velo- 2 city, density, viscosity, diffusion coefficient (I 2 or CH3 I in the bulk gas), temperature, and bed depth. At elevated temperatures $
desorption must be accounted for by including an absorption / desorption i isotherm that specifies the pressure or concentration of iodine in the gas phase in equilibrium with iodine sorbed on the charcoal. J Tests results from the standard tests must be used to adjust i _
the mass transfer correlation for the inherent properties of the char- -p coal. This can be done by means of a term that specifies what fraction &
of the charcoal grain surface is available for active adsorption, iso- Ivr m
topic exchange, or chemical reaction. "Psisoning" of the charcoal =_
surface by adsorbed or reacted contaminants, or the presence of adsorbed [
water will reduce the surface area available for trapping 12 and CH3I trapping. The effective surface arca available for CH3 I trapping is _g less than that for I2 sorption because the impregnants are not perfect y in their surface coverage and chemical action. 4 l Nx Radiation Effects on Charcoal Adsorber Behavior. The first @
radiation effect investigated was the temperature increase that would occur from the decay of adsorbed radioiodine(A-7,A-8). It was deter- y mined that normal air flow rates would supply sufficient cooling, g Tests using only heat from the decay of highly radioactive iodine (up '
to 2000 C1 1301/cm2 adsorber face area) showed that charcoal ignition h from decay heat could occur but that the heat released by oxidation of Y the charcoal as the temperature rose past -392 F would exceed the decay (
heat.(A-4,A-5) g
=
k D
dE-
_ _ _ . -s
i A-13 Radiation effects on the retention of iodine sorbed on char-coal were fi-s'. conducted using 60C o gamma rays.(A-9) Organic iodides were formed that were released unless impregnants were present to trap them. A small decrease in iodine retention was observed even with good quality impregnated charcoal.(A-10) A radiation intensity satura-tion effect was observed at about 107 rad /h 60Co gammas. The opposite effect was also observed. It was necessary to add the iodine to the charcoal immediately after placing the apparatus in the 60 C o facility because preliminary irradiation of the charcoal improved its perform-ance.(A-11)
Tests with highly radioactive iodine as the radiation source (A-4,A-5) (<200 Ci 130I/cm2 adsorber bed face area) showed a low desorption rate <110 C and that even at high temperatures desorption was slow enough at reduced air flow rates that several hours time would be available to correct the abnormal air flow occurrence. Tests with service-aged charcoals were included.
Recent tests with 60C o and accelerator irradiations showed that irradiation of old, poisoned charcoals before adding iodine reju-venated the charcoal.(A-12) Tests with 60C o irridiations of charcoal containing sorbed iodine have been initiated.
In sumary, both beneficial and detrimental effects of radia-tion have been observed. The magnitude of the effects were not large when compared to others. Exposure of iodine on charcoal at 80 C to 60C o gamma irradiation for 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> resulted in an iodine fractional release rate of 4/h. The released iodine was easily trapped by impregnated iodine on charcoal in moist air released iodine at a rate of -5 x 10-6 fraction /h. The released iodine was not collected effi-ciently by any of the impregnated charcoals or silver-exchanged zeolites tested.
Operating Conditions During the Browns Ferry SDV Break Acci-dent. Calculations for the analysis of a hypothetical accident sequence at Browns Ferry Unit One will be described for illustrative purposes.
There are three SGTS filter trains at Browns Ferry, each similar to
A-14 the filter trains used at Peach Bottom. A heating system with a humidity sensor built into the charcoal adsorber system assures that the charcoal will be initially at equilibrium with air containing a maximum of 70 percent relative humidity. For the accident sequence studied, the SGTS began operating at the time of the SDV break. By- .
the time that radioactive material began being transported to the SGTS, -
the air inlet temperature to the charcoal bed was 180 F because of the combined effect of warm air entering the SGTS and temperature con-trol by the heaters in the SGTS. The relative humidity of the air entering the charcoal bed oscillated between 70 and 100 percent.
Following the failure of the drywell and actuation of the reactor build-ing fire protection sprinkler system, the relt.tive humidity of the air entering the SGTS was 100 percent, but the lowered temperature of this air,-140 F, enabled the preheaters to warm the air entering the charcoal bed to 154 F so that the relative humidity in the adsorber system was 70 percent.
Collection Efficiency for Radioactive Iodinc During the SDV Accident. Typical iodine collection efficiencies expected for the Browns Ferry charcoal bed are listed in Table A-2. As described in the previous section, the relative humidity was above 70 percent ini-tially and decreased to 70 percent following start of the fire protec- -'.'
tion sprays. The efficiencies shown in Table A-3 are those deemed to be most representatives of the accident conditions. The delayed change to the higher efficiency is a result of the slow dryout of moisture ..
fromcharcoal.(A-6) ..{ {.e..
.lf Desorption of Radioiodine from the Charcoal Bed. A large ))]
amount of iodine will desorb if the charcoal temperature increases and 3Y air flow is maintained. Since the SGTS was expected to continue opera- Nib tion throughout the SDV break accident sequence (the HEPA filters were r3-9, p assumed to plug and then immediately tear), the charcoal temperature NN3 was not expected to increase significantly during this sequence. 42,)g(;
- ' g _
- h
$jj
U A-15 T
TABLE A-2. TYPICAL ADSORPTION EFFICIENCIES FOR 2-INCH DEEP CHARCOAL BEDS Reg Guide 1.52(a)
Relative New Charcoal New Old Browns Humidity State of Art Charcoal Charcoal Ferry x Species (%) (%) (%) (%) FSAR I 95 99.99 99.9 99 99 2
CH3 I 95 99 95 90 95 :
I2 70 99.999 CH 1 3
70 99.9 99.5 99 .
(a) Charcoal tested in the laboratory at design temperature and velocity.
u
.1 7
M 9
A-16 TABLE A-3. BEST-ESTIMATE EFFICIENCIES FOR BROWNS FERRY SOY ACCIDENT CHARC0AL BED Radioactive Iodine Collection Efficiency (kg)
Iodine 2 h after Fire Sprays Species Before Sprays + 3Fir hg 9a) Start and ThereafterlD) 1 99.9 99.95 2
CH I 95 99.0 3
(a) Charcoal temperature 180 F (82.2 C), relative humidity 70 to 100 percent.
(b) Charcoal temperature 154 F (68 C), relative humidity 70 percent.
I P
4 6.m.um-
A-17 Ordinarily, the air flow through a charcoal bed must be reduced well below normal to cause a significant increase in charcoal temperature since air flow is the cooling medium that removes decay heat. At high temperatures, the captured radioiodine will move slowly toward the outlet end of the bed at a speed which increases with bed temperature and air flow rate, much as with material in a gas chromato-graph. In tests with highly radioactive iodine on charcoal,(A-4,A-5) this movement of iodine was monitored continuously. A computer program named DES 0RB was used to convert the rate of movement into linear adsorp-tion isotherm coefficients as a function of charcoal temperature. The linear coefficient assumed that the iodine partial pressure in equili-brium with the iodine sorbed on the charcoal was directly proportional to the concentration of sorbed iodine plus impregnant iodine. This model was in good agreement with data from the five different types of charcoal tested.
Although the model for desorption is simple, it requires a cmoputer program to follow desorption since the radioactive iodine is distributed nonuniformly through the bed. A second version of DES 0RB was used to calculate desorption that would occur during the Browns Ferry SDV break accident sequence. It was found that the desorption rate would be negligible because of the low charcoal temperatures.
The continued supply of methyl iodide would result in more penetration of methyl iodide during adscrption than the desorption of both I2 and methyl iodide.
Figure A-4 shows the result of DES 0RB calculations for vari-ous combinations of gas velocity, flow time, and bed temperatures.
(These combinations did ..ot occur in the case of the Browns Ferry SDV break accident sequence but are intended to show what is required for massive desorption to occur.) Note that the penetration experienced at the time of initial adsorption is calculated separately.
The plotted line in Figure A-4 for 70 C and 20 cm/s is close to that expected for the charcoal bed conditions following initiation of the fire protection spray system. Extrapolation of that line to 1-week operation would indicate -1.5 percent desorption. Note that the
l A-18 Osusk-ows as-te o etc
'# 1 6 6 6 6 1 ,6- i e i s -
,/ -
E /
}. ,... -
f / -
Is _/
8
/
e'
,1 2 -
i i -
/
l 10-8 / /
/
e' I
.rd'p ,,,' -
. $ ,,. O /s.,i 4' ,
l= ..-=1 -
,i , -.
/ / / _
s A ,/ }'
'S -.
/
+ y _.
gg-e 'l/ l I I I t t io' s s io' 2 s to 8 s io: a s ta' ossonmon ni.cw rius e, .
~
FIGURE A-4. DES 0RPTION FROM A 2-INCH DEEP CHARC0AL SED
h m
A-19
^.
_- t 7
Y approximately 10-4 fraction per hour release rate measured at 80 C in a 60C o gamma field extrapolates to approximately 1.7 percent desorption 7
^
in 1 week. This good agreement is fortuiticus sincf both calculations involve extrapolation of short-term test data. The true amounts could m differ as much as by a factor of 10. '
The calculation results shown in Figure A-4 are for a bed with an initial adsorption efficiency of -99 percent. A bed with higher adsorption efficiency would desorb at a lower rate, especially initially. _
These desorption curves are calculations for radioactive iodine existing on the bed at time zero. Iodine adsorbed later as will occur with q CH3 1 must be accounted for separately. The program DES 0RB is written for simultaneous adsorption and desorption. m r-Conclusions Regarding SGTS Behavior in the Browns Ferry SDV Break Accident Sequence. With the reasoning and the assumptions previ-ously discussed, the HEPA filters in the SGTS trains were predicted to '_
plug with aerosols, then intnediately tear shortly af ter drywell failure. -
~
The ability to remove aerosols by filtration was assumed lost after the filtars were torn, but the blowers were assumed to continue to .
operate so that charcoal bed iodine adsorption efficiency would be ,
maintained.
Other SGTS failure modes are conceivable and the choice of an alternate scenario might have a major impact on the calculated fis-sion product transport. For example, if the HEPA filters did not tear after plugging, flow through the SGTS would cease, the secondary con- h tainment would pressure, and the only pathway from the interiors of ;
the reactor building and refueling bay to the surrounding atmosphere would be via exfiltration through the exterior walls. In addition, the SGTS charcoal beds might overheat, and plastic and organic sealants i of the ducting in the portions of this system not exposed to the reactor _
building fire protection system sprays might fail by overtemperature.
The estimated SGTS train aerosol filtration efficiencies for (
the period when both the upstream and downstream HEPA filters are effec- 9 tive and for the period when the upstream filter has tori, and only the h-
~,- .
m.
A-20 4'
downstream filter is effective are listed in Table A-3. The absorption efficiencies for 12 and organic iodide actually used in the Browns Ferry SDV fission product release calculation are those listed under
" Browns Ferry FSAR" in Table A-2 (99 percent for 12 and 95 percent for organic iodide). These efficiencies are assumed to be effective for the entire sequence. .
Summary of Recommended Reasonable Assumptions Concerning SGTS Operation During Severe Accident Sequences at Peach Bottnm ..
^
With the exception of Station Blackout, the SGTS should oper-ate continuously during at severe accident sequence at Peach Bottom. .
Although positive pressure pulses just after reactor vessel bottom head failure or drywell failure might cause rupture of the reactor building blowout panels, the SGTS blowers are of sufficient capacity to maintain the building pressure at a slight vacuum during most of the accident sequence. Fission product transport would be from the primary containment into the lower portion of the reactor building and ,.
from there directly into the SGTS or up into the volume above the J refueling floor and from there into the SGTS.
It seems reasonable to expect that sufficient aerosols would -
be generated from the core-concrete reaction on the drywell floor to ..
plug the HEPA filters in each of the twso SGTS filter trains. This -
should result in filter destruction by tearing when about 3 kg of aero-sol has~ accumulated on each 1000 cfm filter face or 27 kg for each bank of nine in-parallel filters. The upstream filter bank would tear first, and because the charcoal bed is not effective in removing aero-sols, the downstream HEPA filter bank would subsequently fail. Since ;
tne differential pressure between the atmosphere and the reactor build-l ing would not significantly change during the accident sequence, the -
/
flow through the filter train should remain at a constant value between 10,500 cfm and 25,000 cfm depending on the number of blowers operated.
After the upstream and downstream HEPA filter banks have torn, the o efficiency for particulate removal should be about zero, but a s.
i r
A-21 significant efficiency for the removal of iodines by the charcoal bed would remain.
Grand Gulf System Design and Description The BWR 6 - Mark III containment design employed at the Grand Gulf nuclear station provides that the drywell and pressure suppression pool are located in a primary containment. The primary containment structure consists of a reinforced concrete cylinder and hemispherical dome, lined with welded steei plates to form a continuous intenral membrane.
The secondary containment at Grand Gulf consists of an auxiliary building and an enclosure building. The auxiliary building is a reinforced concrete structure which completely surrounds the lower cylindrical portion of the primary containment with a free volume of about 3 x 106 ft3 . The enclosure building is a metal-sliding structure with a free volume of about 6 x 10 5 ft3 which completely surrounds the uppermost cylindrical portion and the hemispherical dome portion of the primary containment above the auxiliary building roofline. The secondary containment is at atmospheric pressure during normal reactor operation.
The SGTS does not operate during normal plant operation but is automatically started in response to an accident signal.* The SGTS is designed to limit the release to the environment of fission products that might leak from the primary containment, the fuel handling area, ECCS systems, main steam isolation valve leakage control systems, and other sources to the secondary containment region under accident conditions. The system recirculates the atmosphere within the secondarycontainment while continually exhausting a relatively small portion of the recirculation flow to the outside atmosphere through a charcoal and HEPA filter train.
- Signaled for an affected unit by high radiation in the exhaust ducts, low reactor vessel water level, or high drywell pressure.
_1 u
r E
E r
- A-22 E
i Each Grand Gulf unit has its own SGTS. Each SGTS comprises E two completely redundant (including ductwork) systems. The components y of each system include: a 17,000 cfm enclosure building recirculation l fan, a charcoal filter train with a 4000 cfm centrifugal blower, and
! the associated sets of ductwork, dampers, and controls.* The enclosure building recirculation fans are located at elevation 208' .
f and the charcoal filter trains are located at elevation 138' in the i
auxiliary building. Each system is designed to maintain a 1/4-inch wg I negative pressure in the secondary containment under accident
=
conditions, with a minimum mixing ratio of enclosure building air to auxiuliary building air of 8:1 for the design long-term exhaust flow
! of 2300 cfm. The recirculation fan draws air from the auxiliary w
building and the enclosure building, mixes this air by the turbulent f flow conditions in the ductwork, and returns most of the mixed air to .
[ the enclosure building. A portion of the recirculation fan discharge .,
is drawn into the charcoal and HEPA filter train and exhausted via the ..
- centrifugal blower to the atmosphere. A schematic drawing of one of a
7 e
the two redundant systems is provided in Figure A-5.
g The elements of the charcoal and HEPA filter trains are
- similar to those previously discussed for the Peach Bottom SGTS h trains. The first two elements are a demister to remove entrained
[ water followed by an electric heater, sized to reduce the humidity of
{ 4000 cfm of air from 100 percent to 70 percent. The third element is a prefilter whose function is to remove large particulate matter from .
h the flow. The fourth element is a set of parallel HEPA filters to -
remove fine particulate matter. The fifth element is the charcoal o adsorber to remove gaseous elemental iodine and organic iodides, and g the sixth element is a second HEPA filter bank to collect any charcoal g fines carried over from the charcoal adsorber bed. The design
[ features of these components are provided in Tables A-4, A-5, and A-6.
- lg f *It should be noted that the Grand Gulf SGTS incorporates a e
recirculation system and thereby differs significantly from the Peach g Bottom SGTS.
- v e e p k.
g ;{y m . -< .
yc M
lll jl 1l pO !
M' " M r
t t
3' 9
7 3
7 l 6 SS
_ ; IN 2 h O 1
3
W D
"I m
2 SL R
E
- EG L HN N TI R OD O EI L
R G I FN LO U N X I UI S
O D I
U U GT L L A B DR A
M
- 0 N TE AT SY MS S
I A E -
$T R E
T R
g YG SL SN I
L I TS F N m L EN 7
A MW u O C
TO AD R EW
> kCH
AT GS D
0 O 0 YP 3 B 2 DR y NO AF T
SE R
n EA .
O H
, TS m ,
T T 1 s E ni F R U 2 c A O O ND o0 S H NS OE n02 T 2 n R L O
L L En '
FA OC T ww I .
ER R GDT E MM J '-
L I 1 IE NNE IIE E S S W G
N TS 'M-- AE RHN F
I 0 0 0 0 0 -
- DTI L 4 1 1 D -
- C N N A Q ,~- I .E A E ! TRV AAI H R MLG y L
E A>o > > y > >
EI HME CIR SSA U w 0
2 0
6 2 0 0 5 F 1 4 1 3 5
A T E N R Y E U G
RM I A N M A I F
I R T P O N %
C l lllil I
A-24 TABLE A-4. ELEENTS IN THE GRAND GULF SGTS CHARC0AL FILTER TRAIN Units _ Unit Units Required Component Installed Capacity for Operation Filter units 2 1 Fans, per filter unit: 1 1 capacity ft /3 min, drawdown /long term 4300/2300 total pressure, in, wg. 13.0 motor, hp 20 Prefilters, per filter unit: 4 4 capacity, ft3/ min 1,000 HEPA filters, per unit: 8 8 capacity, f tJ/ min 1,000 Demisters, per filter unit: 4 4 capacity, ft /3min 1.000 Electric heaters, per unit: 1 1 capacity, kW 18 Charcoal (8") per filter 2746 2746 unit, lb ,
/
A-25 TABLE A-5. GRAND GULF SGTS CHARC0AL FILTER TRAIN COMPONENT DESCRIPTION Demister (each train)
Type Woven mesh pads -
Capacity, cfm 4000 per bank Pressure drop, clean, 0.5 in, wg.
Electric Heater (each train)
Type Electric, finned tube Quantity 1 Capacity, kW 18 Prefilters (each train)
Type Dry Quantity 1 bank Capacity, cfm 4000 per bar.k Efficiency, % 90 ASHRAE 52 Pressure drop, clean, 0.5 in. wg.
HEPA Filters (each train)
Type High efficiency Quantity 2 banks Capacity, cfm 4000 per bank Efficiency, % 99.97, DOP '
Pressure drop, clean, 1.0 in, wg.
Cnarcoal Adsorber (each train)
Type Deep b(d Capacity, cfm 4000 Media Impregnated coconut shell Efficiency credit, % 99, elemental iodine and organic iodide Depth of bed, in. 8 nominal Charcoal volume, ft3 70.0 minimum Face velocity, fpm 40 Residence time, see 1.0 i ,
A-26 TABLE A-5. (Continued)
Charcoal Adsorber (each train)
(Cont'd)
Pressure drop, clean, 4.5 in, wg.
1 Iodine desorption temp 250-300 range, F Charcoal ignition temp, F 640 approx Charcoal density, gm/cc 0.3 to 0.55 Impregnant content, % by 5 maximum weight Charcoal size distribution 8 x 16 Tyler mesh nominal Charcoal surface area, m2/gm 1000 minimum Charcoal moisture content 3 maximum efficiency, %
Charcoal ash content, % 6 maximum Exhaust Fan (each train)
Type Centrifugal SWSI Quantity 1 Capacity, cfm. 4300/2300 drawdown /long term
A-27 TABLE A-6.- MATERIALS USED IN GRAND GULF SGTS CHARCOAL FILTER TRAIN COMPONENTS Demister Media Fi berglass ~
Casing 304 stainless steel Heater Sheath Steel Prefil ter Media Fiberglass Casing Chromized steel Separators Aluminum HEPA Filters Media Fiberglass Casing Chromized steel Face guard Galvanized steel Mounting frame 304 stainless steel Charcoal Filters Media Impregnated, activated coconut shell charcoal Casing 304 SS Frame 304 SS Unit Housing Shell Carbon steel, A36 -
Mobil Val-Chem Hi-build epoxy, 89 Series, 5-7.0 mils DFT, over Mobilizinc 7,1-3 mils DFT Exterior coating Mobil Val-Chem Hi-build epoxy Series 89, 5-7 mils DFT, over Mobil 13-R-56,1-3 mils DFT, underside coated with Koppers Bitumastic Black Solution or Gaco Neoprene Asphalt NA-62, by Gates Engineering Co.
Door gaskets Neoprene
A-28 The Grand Gulf SGTS uses approximately four times as much -
charcoal per volumetric flow rate as do the Peach Bottom and Browns Ferry systems. This improves the charcoal filter efficiency to a marked degree. For a given form of iodine such as CH3 1, the absorp-tion efficiency might increase from 95 percent to greater than 99.99 percent. Such an increase is no usually observed in practice because of gasket leaks and the presence on formation of other more penetrating forms of iodine.
SGTS Operation Following an accident signal (high drywell pressure, low reactor vessel water level, or high radiation in the secondary con- =
tainment exhaust ducting), both enclosure building recirculation fans and both charcoal filter train centrifugal blowers automatically start and all normal ventilation penetrations through the secondary contain- .
ment are automatically closed.
l SGTS' flows are controlled by modulating inlet vanes on the !
charcoal and HEPA filter train centrifugal blowers and by two-position flow control dampers installed in the recirculation fan suction ducts from the various regions of the secondary containment. Upon system ' '
statup, the inlet vanes and control dampers are fully open and the centrifugal blowers operate at full capacity (4300 cfm) until the .
secondary containment has been drawn down to the design negative pres-sure of 1/4-inch wg. This requires 101 seccnds if only one train is in operation and somewhat less time in the more likely event that both ~
trains are operable. _
After the secondary containment has been drawn down, two-position motor-operated dampers located in the SGTS ductwork automat-ically throttle the exhaust flow so that the volume of air being
~
exhausted is equal to the volume of air infiltrated at the 1/4-inch wg negative pressure. Subsequent flow through the charcoal and HEPA fil- _
ter train is controlled by the centrifugal blower motor-operated inlet vanes, which respond automatically to a pressure controller as necessary -
1 . . -.
I A-29 to maintain a 1/4-inch wg negative pressure in the secondary contain-ment for the long-term operation. The required long-term flow is expected to be 2300 cfm; however, if necessary, the modulating inlet vanes will automatically open to establish a system flow up to 4000 cfm (per train) as needed to maintain a negative pressure of 1/4-inch wg.
When the desired negative pressure has been achieved in the secondary containment, the operator is alerted that one recirculation fan and one charcoal filter train centrifugal blower can be shut down and placed in a standby mode. The system in a standby mode will auto-matically restart in response to a low flow signal from the operating enclosure building recirculation fan or from the charcoal and HEPA filter train centrifugal blower, or in response to an insufficient negative pressure in the secondary containment.
The air flow to the suction of the enclosure building recir-culation fans is taken from corridors and general areas within the auxiliary building rather than from specific rooms. These areas com-municate through stairwells, an elevator shaft, and several equipment hatches. For long-term operation with one SGTS in operation, a flow of 1640 cfm is taken from the auxiliary building and mixed with 160 cfm from the fuel handling area and 15,200 cfm from the enclosure build-ing in the suction ducts to the enclosure building recirculation fans.*
Of the 17,000 cfm passing through the recirc':lation fans, 2300 cfm is drawn into the suction ducting for the charcoal and HEPA filter train and 14,700 cfm is released back into the enclosure building atmos-phere. ,
- With the negative pressure maintained by the SGTS, the enclosure build-ing, by design, has a much higher in-leakage rate from the atmosphere than does the auxiliary building.
s
A-30 Recommended Reasonable Assumptions Concerning Severe Accident Sequences at Grand Gulf Each system of the SGTS is completely redundant and provided with a 100 percent capacity filter' train, centrifugal exhaust blower, recirculation fan, ductwork, and associated instrumentation and controls.
Any accident signal automatically starts both of the SGTS systems.
During long-term operation, one of the systems is manually placed in a standby mode. The standby system would automatically start in the event of a major component failure, degradation of performance of the operating system, or failure of electrical power to the operating sys-tem. Component malfunctions that result in startup of the standby system are summarized in Table A-7.
In certain severe accident seq 3ences, sufficient core-concrete reaction aerosols might reach the secondry containment to plug the HEPA filters in the charcoal filter trains. However, this is less likely in the Grand Gulf Mark III containment design than for the Peach Bottom Mark I design because the Mark III design provides that flow from the drywell must pass through the pressure suppression pool on its way to the secondary containment. From the information in Table A-4, there are four 1000 f t3 / min capacity HEPA filters in parallel in [f the HEPA filter bank upstream of the charcoal bed and a similar arrange-S Nv..
ment for the downstream HEPA filter bank. As noted in Table A-7, the yd}i operator would be alerted to a high differential pressure across the l' N
.,4.v upstream HEPA filter bank and would probably switch the standby SGTS p.
system. M 'b n%
In the Peach Bottom SGTS filter train, it was assumed that f"J the HEPA filters would tear at a differential pressure of 12 inches of
!] l water. Available experimental data indicate that this would occur kj vhen 3 kg of aerosol had entered the filter train for each HEPA filter bc.J '7 arranged in parallel across the flow. Since there are nine 1000 cfm hs n HEPA filters in parallel for a 9000 cfm flow, a total accumulation of ?f 27 kg was estimated necessary to cause filter tearing at Peach Bottom. .
w I, ' )"
1 y ej VJ +
- a ?,;- ,b.
a-I TABLE A-7. GRAND GULF SGTS FAILURE MODES AND EFFECTS ANALYSIS Component Mal function Coment Enclosure building Fan failure resulting Low flow switch will automatically shift to redundant recirculation fan in loss of air flow train and fan.
! Filter train Fan failure resulting Low flow switch will automatically shift to redundant exhaust fan in loss of air flow train and fan.
Electric heater Heater failure result- High humidity will give alarm in control room, operator ing in inability to may switch to redundant train.
reduce air stream to 70 percent RH Heater failure result- Heater is equipped with thermal overload cutout switches.
ing in overheating Filter train Failure resulting in High differential pressure is alarmed in control room. d high differential Operator may switch to redundant train, pressure across HEPA or charcoal sections Failure resulting in High temperature is alarmed in control room. High-high high temperature in temperatere is alarmed in control room and initiates charcoal bed deluge system. Initiation of deluge system is alarmed in control room, and operator may switch to redundant train.
Ductwork fan inlet Duct failure: Redundant ductwork is provided, vanes Fail closed Low air flow automatically starts redundant train.
Fail open High air flow causes excessive negative pressure in building which actuates an alarm. Upon alarm actuation, the control room operator can determine which system has failed (A or B) and remove it from service, allowing the other train to continue to operate.
hyk
TABLE A-7. (Continued)
Component Mal function Conment Pressure sensor / Instrument failure Redundant instruments provided. l l
transmitter /
controller instru-ments Positioning dampers fail closed Redundant train achieves negative pressure. If second 1 train is in standby mode when failure occurs, loss of air flow or low negative pressure starts redundant train.
Loss of low negative pressure signal opens filter train Fail open fan inlet vanes.
2 N
A-33 At Grand Gulf, each filter train is designed for 4000 cfm and there are four 1000 cfm HEPA filters arranged in parallel. There-fore, for 4000 cfm flow the required accumulation to cause tearing would be 4/9 x 27 or 12 kg of aerosols. However, as described previ-ously, the train is operated at 2300 cfm and therefore a larger aero-sol deposition would be required in order to develop 12 inches of water pressure drop at the lower flow. Thus, a reasonable estimate might be 16 kg of accumulated aerosol within a filter train to cause rupture of the upstream HEPA filter bank and an additional equal amount for rupture of the downstream filter bar.k.
For Grand Gulf, the presence of aerosols in the secondary containment during the course of a severe accident sequence should affect the operation of the SGTS as follows. Aerosols would deposit on the upstream HEPA filter bank of the operating SGTS system. When about 9 kg have deposited, a high differential pressure alarm would sound in the control room and the operator would shift to the standby system. When about 16 kg has accumulated on the upstream HEPA filter bank of this system, the upstream filter bank would tear and when an additional 16 kg of aerosols has entered the system and deposited on the downstream HEPA filter ban, it also would tear. After this, the efficacy for particulate removal by the system would be about zero but a significant efficiency for the removal of iodines by the charcoal bed would remain.
As in the case of Peach Bottom and Browns Ferry, other SGTS failure modes are conceivable and the choice of an alternate scenario could very well have a major impact on the calculated fission product transport. At Grand Gulf, the SGTS charcoal and HEPA filter trains and the recirculation fans are located within the auxiliary building and therefore are subject to the hazards of an accident environment within the building.
If the HEPA filters plugged did not tear, flow through the operable SGTS charcoal and HEPA filter train to the atmosphere would cease, the building pressure would rise above atmospheric and fission products would be released to the surrounding atmosphere only by way
- _ _ _ _ _ _ _ _ _ _ _ - - _ __ x --
A-34 of exfiltration through the walls of the auxiliary building within the building.
The efficacy of the SGTS at Grand Gulf must be considered to be accident-sequence dependent. The system should not be simply ignored, yet there are accident sequences in which its beneficial effect would be diminished. The SGTS should be recognized as a very important fis-sion product release mitigation system whose function or loss should be evaluated in every accident sequence, i.e., on a case-by-case basis.
A-35 References (A-1) Military Specification MIL-F-51068, Filter, Particulate, High Efficiency, Fire Resistant.
(A-2) First, M. W. and Price, J. M., " Performance of 1000- and 1800-cfm HEPA Filters on Long Exposure to Low Atmospheric Duct Loadings, III", Proceedings of the 17th DOE Nuclear Air Q_g4 e ~
Cleaning Conference (in preparation).
(A-3) Gunn, C. A. and McDonough, J. B., " Survey of Loading Perform- 4.+- Y ance of Currently Available Types of HEPA Filters Under In- Lp%E Service Conditions", Proceedings of the 16th DOE Nuclear Air ?4 Cleaning Conference, CONF-801038, Vol 1, 667-668 (February, "j ~ (,
1981).
8%
)
(A-4) Lorenz, R. A., Martin, W. J., and Nagao, H., "The Behavior f ., c, of Highly Radioactive Iodine on Charcoal", in Proceedings of 5.. a the 13th AEC Air Cleaning Conference, CONF-740807, 707-735 n. J '
(March,1975). ;
(A-5) Lorenz, R. A., Manning, S. R., and Martin, W. J., "The Behav- '
cD.
ior of Highly Radioactive Iodine on Chercoal in Moist Air", b in Proceedings of the 14th ERDA Air Cleaning Conference, O CONF-760822,323-352(February,1977). g; (A-6) Shiomi, H., et al, "A Parametric Study on Removal Efficiency G.i of Impregnated Activated Charcoal and Silver Zeolite for u Radioactive Methyl Iodide", presented at the 17th DOE Nuclear Air Cleaning Conference, August 2-5, Denver, Colorado.
k!(
"D
~ rh__
(A-7) Shields, R. P., " Ignition of Charcoal Adsorbers by Fission I Product Decay Heat", Proceedings of the lith AEC Air Cleaning I' O Conference, Hanford, Washington, August 31-September 3, 1970, CONF-700816.
?.h Y ':
Shields, R. P. and Siman-Tov, M., "The Effect of Iodine Decay NV (A-8) ':S.
Heat on Charcoal Adsorbers", US AEC Report ORNL-4602 (April, 3 1971). .'
s : ,
0'L Evans, A. G. and Jones, L. R., " Confinement of Airborne Radio-(A-9) activity", progress report, January-June, 1971, DP-1280 hh J (October,1971). iE
}'-
(A-10) Evans, A. G., "Effect of Intense Gamma Radiation on Radicio-dine Retention by Activated Carbon", in Proceedings of the "/-
J 12th AEC Air Cleaning Conference, CONF-720823, Vol 1, 401-416 F%
(January,1973). M;;
}l #
(A-11) Personal communication from A. G. Evans, March, 1973.
W a g. ,
Mk' D
_ _ _ . _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ . _ _ _ - - - - - - - - _ --- ^ ~ - - - ~ - ' '
s A-36 (A-12)
Deitz, Victor R., " Charcoal Performance Under Simulated Acci-dent Conditions", NRL Memorandum Report 4760 (NUREG/CR-2660)
(June,1982).
..... . n mm .ma , t, u-ew.oe.
e
, 's e UNITED STATES '
i FouaTMCLAS3mMt '
NUCLEAR REGULATORY COMMISSION Postass a Fers ewo WASHINGTON, D.C. 20555 ,$5","l c j 3
l> Pt Auer me in CFFICIAL SUS:f4ESS ,
e'ENALTY FOR PR'VATE USE, s300
- . ,- 1
-+ )!
.e -
'A
'h y i
4 db F
9
'k pF -s
/ ,
f 5
_g
'( ,
'I. ,i vf r
/
4 - JJ i, 4
'(
-f .g i
s k
)
i