ML20069J499
| ML20069J499 | |
| Person / Time | |
|---|---|
| Site: | Grand Gulf  |
| Issue date: | 03/31/1983 |
| From: | Burnham B, Byers R, Camp A, Jamarl Cummings, Sherman M, Tomasko D, Wester M SANDIA NATIONAL LABORATORIES |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| CON-FIN-A-1246, CON-FIN-A-1308 NUREG-CR-2530, NUDOCS 8304220626 | |
| Download: ML20069J499 (225) | |
Text
NUREG/CR-2530 SAND 82-0218 Review of the Grand Gulf Hydrogen Igniter System Prepared by J. C. Cummings, A. L. Camp, M. P. Sherman, M. J. Wester, D. Tomasko, R. K. Byers, B. W. Burnham S:ndia National Laboratories Pr:psred for U.S. Nuclear Regulatory Commission I
hDR O OOO 6 P PDR
NOTICE ,
l l
This report was prepared as an account of work sponsored by an egency of the United States Government. Neither the United States Government nor any agency thereof, or any of their employees, makes any warranty, expressed or implied, or assumes any legal liability of re-sponsibility for any third party's use, or the results of such use, of any information, apparatus, product or process disclo:3d in this report, or represents that its use by such third party would not infringe privately owned rights.
Availability of Reference Materials Cited in NRC Publications Most documents cited in NRC publications will be available from one of the following sources:
- 1. The NRC Public Document Room,1717 H Street, N.W.
Washington, DC 20555 ,
2 The NRC/GPO Sales Program, U.S. Nuclear Regulatory Commission, Washington, DC 20555
- 3. The National Technical Information Service Springfield, VA 22161 Although the listing that follows represents the majority of documents cited in NRC publications, it is not intended to be exhaustive.
Referenced documents available for inspection and copying for a fee from the NRC Public Docu-ment Room include NRC correspondence and ir ternal NRC memoranda; NRC Office of Inspection and Enforcement bulletins, circulars, information notices, inspection and investigation notices; Licensee Event Reports, vendor reports and correspondence; Commission papers;and applicant and licensee documents and correspondence.
The following documents in the NUREG series are available for purchase from the NRC/GPO Sales Program: formal NRC staff and contractor reports, NRC sponsored conference proceedings, and N RC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code of Federal Regulations, and Nuclear Regulatory Commission issuances.
Documents available from the National Technical Information Service irriude NUREG series reports and technical reports prepared by other federal agencies and reports prepared by the Atomic Energy Commission, forerunner agency to the Nuclear Regulatory Commission.
Documents available from public and special technical libraries include all open literature items, such as books, journal and periodical articles, and transactions. Federal Register notices, federal and state legislation, and congressional reports can usually be obtained from these libraries.
Documents such as theses, dissertations, foreign reports and translations, and non-NRC conference proceedings are available for purchase from the organization sponsoring the publication cited, Single copies of NRC draft reports are available free upon written request to the Division of Tech-nical Information and Document Control, U S. Nuclear Regulatory Commission, Washington, DC 20555. J
! Copies of industry codes and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, and are available there for reference use by the public. Codes and standards are usually copyrighted and may be purchased from the originating organization or, if they are American National Standards, from the
- American National Standards Institute.1430 Broadway, New York, NY 10018. l GPO Pr.nted copy once $8.50
l
' NUREG/CR-2530 SAND 82-0218 R2 Review of the Grand Gulf Hydrogen Igniter System Manuscript Completed: February 1983 Date Published: March 1983 Prepared by J. C. Cummings, A. L. Camp, M. P. Sherman, M. J. Wester, D. Tomasko,3. K. Byers, B. W. Burnham Sandia National Laboratories Albuquerque, NM 87185 Prepared for Office of Nuclear Reactor Regulation Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, D.C. 20555 NRC FINS A1308, A1246 l
j Abstract The Mississippi Power and Light Company has proposed installation of a Hydrogen Igniter System (HIS) at the Grand Gulf Nuclear Station (BWR Mark III) to burn hydrogen generated during accidents more severe than the design-basis accidents. Sandia National Laboratories, under a contract with the US Nuclear Regulatory Commission, has performed a technical evaluation of the adequacy of the proposed HIS to meet the threat posed by hydrogen combustion. Areas considered in this review include: HIS design and testing; location and distribution of igniters; containment pressure and temperature response calculations; detonations; containment atmosphere mixing' mechanisms; actuation criteria for the HIS; and the spectrum of hydrogen-generating accidents.
I i
k i
I l
5 i
i l
Acknowledgments A number of individuals provided help, assistance, support, and review, Without their assistance this report could not have been produced.
Lawrence D. Buxton and Michael L. Corradini were involved early in the project with the RALOC and MARCH computer codes.
Susan Dingman and Peter Prassinos were instrumental in prcviding assistance with the HECTR and MARCH codes when difficulties arose well into the project.
Marshall Berman provided leadership to the project and painstakingly reviewed several drafts of this report.
His high standards of excellence and commitment of time are deeply appreciated by the authors.
We are forever indebted to Pat Rosario, Jan Frey, and Carol Henry. Without their dedicated efforts in typing the manuscript,we never would have produced the first draft of this document on schedule. Finally, we would like to thank E. W. Shepherd and the Publication Services Division for the high quality of the format of this document.
{
6
Contents 1.0 Summary . .. . ............ . . . . . . . . . .. 9 1.1 Introduction . . . . . _ _ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - . . . . . . . . . . . 9 1.2 Preliminary (Task 1) Report . . . . . . . . . . . . . . .. . . . . . . . . . . - . . . . . . . . 9 1.3 Final (Task 2) Report Summary... ... . . . .................10 1.4 Referenee . . .. .... . ...... . ..................13 2.0 HECTR Calculations for the Grand Gulf Nuclear Station- . . . . . . . . . . . . . . . . 15 2.1 Description of the HECTR Computer Program - ........... 15 2.2 Grand Gulf Analysis . . . - . . . . . . . .. ..... .. . 17 3.0 Grand Gulf Accident Calculations Using the MARCH Code... . . ... ... .. . . . . . . . . . . . 97 3.1 Summary and Conclusions...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... .. ... . . . 97 3.2 Introduction ..... . . . . . . . . . . . . . . . . . . . . . . . . . . , ... ..... ... 98 3.3 Analysis of MARCH Code - . . . . ...... . . . . . .101 3.4 Analysis of MARCH Calculations for Configuration B.. ..... . ..... .. . . . . . ..110 3.5 Analysis of MARCH Calculations for Configuration C. ..... ... .. . 122 3.6 Analysis of MARCH Calculations for Configuration D.. .... . . . . . . . . . . . . . . . . . . . .... . 129 3.7 Analysis of MARCH Calculations for Configuration E'. .... . . . . . . . . . . . .. . ... .. . .. .. 138 3.8 The Effect of Connections to the Drywell Froni the Wetwell/ Containment.. ....... . ... ... 148 3.9 References....... . . . . . . . . . . . . . - . . . . . . . . . . . . - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 4.0 RALOC Hydrogen Transport Calculations for the Grand Gulf Nuclear Station.... ... 171 4.1 Summary-. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . 171 4.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... ... ... 171 4.3 R A L O C .......- ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... . .. 172 4.4 RALOC Grand Gulf Models . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . ........................................172 4.5 Grand Gulf Hydrogen Source Models.. .-. . . . . . . . ..................~..............174 1 4.6 Results of a Typical RALOC Calculation- . . . . . . . . . . . . . .. ....... .... 174 4.7 Code Reliability Calculations ...... ...-.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......... 174 4.8 Physical Parameter Sensitivity Study.. .... . . . . , . . . ... .. .. .. 179 4.9 Conclusions- ..._... . . . . .... ............... . ..... .. 185 4.10 References._ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . .. . .. . ... .. 18 5 5.0 Dynamic Combustions and Impulsive Loading . .. . . . . . . . .. ....... .... 187 5.1 Introduction . . . . . . . . . . . . - - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......................187 5.2 Necessary Conditions for Detonation. .. ..... ...............................................................187 5.3 Detonations, Quasi. Detonations, Transitions to Detonation, and Accelerated Flames....,. ...187 i 5.4 Local Detonation Calculation for Grand Gulf .. . . . . . . . .............................188 5.5 Likelihood of Detonation at the Grand Gulf Nuclear Station ..... .. .... . ...... ..... ............ . ....... 194 5.6 R e fe r e n ces .... . .............. ........-...... ... ... .. . ......... .... .. ...........................................194 6.0 Assesament Results . . . . . . . . . . . . . . . . . . . . . . . . . . .......................... ........................195 6.1 Adequacy of the HIS Design . .......................................... . . . . . . . ....... 195 l 6.2 Adequacy of the Actuation Criteria... .. ....................... ........................................196 6.3 Adequacy of the HIS Testing... ........................................................................ -196 6.4 Adequacy of the Spectrum of Accidents Considered in the Applicant's (MP&L) Evaluation ...197 6.5 References........................................................................ ...... . .. . ... . . -.. . 197 '
l Appendix A - Excerpt From Task 1 Report .........-.............................................. ............. .... ... ........ 199 A ppendix B - H ECTR Description ........... .. ............ ................ .................. ..... ..........................211 Appendix C-The MARCH Code.. .................................................................................217 78 l
'l
)
1 i
l i
l 1
1 l i
- l. 1 l
l i
f
('.
a
- _ __ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -as. -m--w - 1
1 Review of the Grand Gulf Hydrogen Igniter System 1.0 Summary 1.1 Introduction As a result of severalinvestigations of the accident 6. To assess the adequacy and completeness of at the Three Mile Island Nuclear Plant (TMI-2) in the spectrum of accidents considered in the March of 1979. the US Nuclear Regulatory Commis- applicant's evaluation.
sion (NRC) has required that boiling water reactor As necessary, we were to provide recommendations to (BWR) nuclear plants employing the Mark III design the NRC for corrective action or design changes.
he equipped with additional systems to control large quantities of hydrogen. In this regard, the Mississippi Power and Light Company (MP&L) has proposed 1.2 Preliminary (Task 1) Report mstallation of the liydrogen Igniter System (111S) at the Grand Gulf Nuclear Station to burn the hydrogen At the end of August 1981, we submitted a Task I generated during accidents more severe than the Report to the NRC. This report summarized our design-basis accidents. The NRC must now evaluate preliminary evaluation, discussed our technical ap-the effectiveness and reliability of the proposed 111S proach for the Task 2 (draft report) evaluation, and as part of the Grand Gulf licensing activities. requested additional information in a number of In August of 1981, Sandia National Laboratories areas. Our preliminary evaluation was based primarily (SNL) agreed to assist the NRC by making a technical n the following items:
evaluation of the adequacy of the 111S to meet the 1. Review of the MP&L document Preliminary threat posed by hydrogen combustion at the Grand Evaluation of Additional Hydrogen Control Gulf Nuclear Station. The technical goals of our eval- Measures for the Grand Gulf NucIcar Station uation of the design and testing of the lilS were: (April 9,1981)
- 1. To assess the adequacy of the location and 2. Review of the MP&L document Final Report distribution of igniters, on the Grand Gull Nuclear Station Hydrogen
- 2. To perform independent calculations of the Igniter System (June 19, 1981) containment atmosphere pressure and tem- 3. A tour (and associated photographic records) of perature response to hydrogen combustion and the Grand Gulf Nuclear Station on August 6, compare these to the calculations performed by 1981 by J. C. Cummings, M. P. Sherman, and the applicant (MP&L has used the CLASIX-3 A. L. Camp of SNL computer code exclusively), 4. Discussions (and associated SNL notes) with
- 3. To determine the likelihood (qualitative as- S. Ilobbs and J. Richardson, MP&L, on August sessment) and consequences of detonations (in 6,1981 terms of calculating impulsive loads, not in 5. MARCil computer code calculations of acci-terms of estimating structural failure), dent scenarios in the Grand Gulf plant
- 4. To examine and estimate containment (includ- 6. Engineering judgments made by the Sar.dia -
ing wetwell) atmosphere mixing, staff during review of the above five items
- 5. To assess the adequacy of the actuation criteria In our Task 1 Report, we indicated that our for the IIIS, and evaluation of the adequacy of the lilS would depend 9
1
l heavily on the value of containment failure pressure or hydrogen coinbustion. The German code RALOC was
" ultimate capacity of the containment." An NRC anal- used to investigate mixing effectiveness by calculating ysis indicated that 56 psig (71 psia,4.8 atm) would be the spatici and temporal behavior of the hydrogen an acceptable value. MP&L has reported calculated concentration (preburn conditions only). The Sandia values from 47 to 67 psig for the ultimate capacity of code CSQ was used to calculate the impulsive loads the containment. Unfortunately, this uncertainty that would accompany a h> cal detonation.
band overlaps with the expected uncertainty in In addition to the items listed in Section 1.2, we hydrogen-burn pressure associated with the transition have reviewed other documents as part of our Task 2 from incomplete to complete combustion. For exam- evaluation (most of these were not received until late ple, the NRC estimate of failure pressure could be December 1981):
exceeded if a single, complete, adiabatic, isochoric 1. Report on the Grand Gulf Nuclear Station burn were to take place with 8% to 9% hydrogen flydrogen Ignition System, M P& L ( August 31, distributed homogeneously in the containment air.* 393g)
Our preliminary conclusion, therefore, was that the 2. Letter and attachments from L. F. Dale, uncertainties of the HIS performance overlap with he MP&L, to H. R. Denton, USNRC (September uncertainties of the calculated failure pressure. Con- 11, 1981) sequently, we stated that more accurate and detailed 3. The CLASIX Computer Program for the analyses would be required before an assessment of Analysis of Reactor Plant Containment Re-the HIS adequacy could be made. Our goal was to sponse to Hydrogen Release and Dc/lagration, reduce the uncertainty of the estimates of HIS perfor- Off-Shore Power Systems Rept No. 36A31, mance during various accident scenarios, but we noted nonproprietary version (October 1981) that uncertainties would still remain even after we had 4. CLASIX-3 Containment Response Sensiticity completed this evaluation. We also recommended that Analysis for the hiissiaippi Pou:er & Light the NRC obtain independent estimates of contain- Grand Gulf Nuclear Station, Off-Shore Power ment failure pressure for Grand Gulf. Systems Rept No. 37A15 (December,1981)
The request for additional information that was 5. Letter and attachments from L. F. Dale, included in our Task 1 Report is included at the end of MP&L, to H. R. Denton, USNRC (December this report as Appendix A. As part of that request, we 21,1981) recommended a significant change to the compart-mont model used by MP&L in their CLASIX-3 calcu- In summarizing the results of our Task 2 assess-lations and a number of additional CLASIX-3 calcula- ment, we must repeat several statements that were tions to examine variations in several key parameters made in the prehm, m ary (Task 1) report. First, we (hydrogen source term, flame speed, etc). We also have used the NRC value of 56 psig for the contain-requested answers to specific questions that we felt ment failure pressure. If this value were to increase or were important in conducting our evaluation of the decrease by 5 to 10 psi, it could have a significant design and testing of the Grand Gulf HIS. effect n ur evaluatmn and recommendations. Sec-ond, after several months of intensive study and hun-dreds of computer calculations, we are still left with 1.3 Final (Task 2) Report uncertainties about the adequacy of the HIS. None of the available computer codes (MARCH, CLASIX-3, MmmaG or HECTR) is sophisticated enough at present to Our techm. cal approach has m.volved the use of allow us to predict peak pressures with high precision, several computer codes combmed with the engineer-and all the codes are very sensitive to certain input mg judgment of the SNL staff. The Sandia code parameters and models of phenomena that are simply HECTR (Hydrogen Event: Contamment Transient not known very well.
Response) and the MARCH code (written by Battelle It is also important to note that we have not Columbus Laboratory) were used extensively as coun-analyzed in any significant detail a number of phe-terparts to CLASIX-3 to calculate the contamment nomena associated with hydrogen combustion. These atmosphere pressure and temperature response to include accelerated flames, missile generation, spatial and temporal nonhomogeneities in gas mixtures, mul-
- The exact hydrogen concentration required to attain 56 tiple and sequential ignitions, and equipment failures.
psig depends on the preburn pressure (e.g., the addition of Theoretical models and computer codes are not pres-tha drywell non-condensable gases to the containment at. ently available as calculational tools to analyze these mosphere can increase the initial pressure by 19"o). phenomena. In addition, experimental data are 10
sparse, especially under nuclear plant accident condi- significant hydrogen generation prior to complete core tions (very large geometrical scales, three-dimensional uncovery. Our present understanding of core behavior obstacles and flame paths, concentration gradients in during loss-of-coolant accidents (LOCAs) is quite lim-the gas mixture, etc). ited, primarily due to a lack of knowledge and data The net result of our evaluation, including uncer- concerning core-slump phenomena. MARCH calcula-tainties, can be stated as follows. In our judgment, the tions for TPE* and S(* BWR accidents indicate that currently proposed HIS is " marginally adequate" to little or no hydrogen would be produced before com-meet the threat posed by hydrogen releases within the plete uncovery of the core Consequently, the bulk _of containment building. By marginally adequate, we the hydrogen generation and release is predicted to mean the following: In order to calculate the combus- occur when the core slumps into the vessel lower tion behavior of hydrogen inside the Grand Gulf con- plenum.
tainment, we must specify a number of parameters for The processes governing production of hydrogen the simple models presently in our computer codes. If during a degraded-core accident are not well known,
- each of these parameters is at the optimistic end ofits and in some respects, are expected to be indetermi-
- "best-estimate" range, the computer codes predict nate. The hydrogen generation rate will depend on the
( combustion-generated pressures well below the speci- rate of coolant injection from the Emergency Core
! fied value for failure. If each of these parameters is at Cooling System (ECCS) and the time that the injec-
! the pessimistic end of its best-estimate range, the tion begins. The degraded-core models in MARCH j computer codes predict pressures in excess of the can predict very high to very low hydrogen yields, assumed failure value. Midrange values for these pa- because the predictions are dominated by input as-rameters will generally result in calculated pressures sumptions. Until better data and models are pro-which are high, but below the specified failure pres- duced, we must either accept the (perhaps artificial)
. sure. conservative upper bound or select another (equally Later in this report, we make recommendations arbitrary) model.
for changes to the proposed HIS design which we feel MP&L used the MARCH code predictions up to 4
would improve its ability to meet the hydrogen threat the point of core slump (# hen the rate increases j (Chapter 6.0). Several " advanced concepts" that we enormously) and then assumed a constant rate equal 4 believe might be desirable alternatives or adjuncts to to the largest previous value. Since the hydrogen I
an igniter system for BWR Mark III plants are dis- release rate is such an important parameter, we rec-cussed in another report." ommend further study of the MP&L approach to Several things cause us to judge the HIS to be assess its validity.
marginally adequate. There are significant uncertain- The degree and speed of mixing within the ties in the hydrogen release rate, the degree and rate of wetwell/ containment region are very difficult to esti-1 mixing within the wetwell/ containment, and the igni- mate. Based on engineering judgment, we believe that i tion and propagation behavior of burns. In addition, if containment sprays are operating, the upper dome -
for some accident scenarios the containment spray region will be well mixed. If they are not operating, the
, system is not available. These items tend to dominate degree of mixing will be uncertain. Independent of the the results of our calculations, and consequently, reli- question of spray operation are the degree and rate of able conclusions are difficult to draw. We do know mixing in the annular region. The computer codes
! that if conservative upper bounds or assumptions are MARCH, CLASIX-3, and HECTR essentially ac-
! made regarding these key items, our calculations indi- count for only pressure-driven flow between idealized cate the generation of pressures in excess of the as- compartments. These codes do not realistically model sumed failure value. the postburn flow and mixing, or natural convective Hydrogen release rates for accidents in nuclear mixing. Consequently, when the codes predict a large plants cannot be predicted with confidence at present. number of burns in one region and no burns elsewhere,
- This is true for low-pressure acridents in either a
~
or oxygen depletion in one region, or a volume filled PWR or a BWR. Low-pressure (large-break) acci- with hydrogen existing stably below a volume filled i
dents tend not to produce significant hydrogen prior to total core uncovery due to high steam flow rates which maintain the core temperature below that re-
. Accident that is initiated by a transient event, followed by l quired for significant metal-water reaction. On the failure of a safety / relief valve to reclose and failure of l other hand, high-pressure (e.g., small-break) acci- emergency core cooling.
dents tend to have considerably lower steam flow "Small pipe (with an equivalent diameter of ~2 to 6 in.)
rates, which results in higher core temperatures and break and failure of emergency core cooling. -
11 i
with air, we think the models are too crude to provide an artifact of the compartment model and input as-refined answers. sumptions. Our engineering judgment dictates that, if The ignition and propagation of burns in the possible, the sprays should be initiated before signifi-containment are obviously phenomena that must be cant hydrogen production occurs and remain on until correctly modelled in order to calculate realistic pres- the accident is brought under control.
sure and temperature transients. The computer codes Comparisons between the various computer codes treat these phenomena very simplistically. Conse- is difficult because we know little of the calculational quently, we are forced to specify values for ignition models within CLASIX-3 and even MARCH. We have criteria, propagation criteria, combustion complete- attempted to carry out calculations with MARCH and ness, and flame speed. The values we select can have HECTR that are nearly identical to several tremendous leverage on the calculations. The combus- CLASIX-3 cases. Unfortunately, we invariably end up tion processes that would actually occur in contain- comparing " apples" to " oranges." HECTR has no ment for our postulated accident scenarios are proba- drywell compartment model; MARCH has a drywell, bly much different than the idealized models. Spatial but only zero-flow-resistance connections to and from and temporal gradients in hydrogen, oxygen, and it. Since it is not an input for MARCH calculations, steam concentrations, coupled with discrete igniter the MARCH hydrogen release rate is not the same as locations, would tend to produce a very complex series that used in HECTR and CLASIX-3. Despite these oflocal burns which might result in smaller loads than and other differences, the three codes predict the same our calculated " compartment" burns. It is well beyond general kind of wetwell/ containment behavior. (As our present capability to model such behavior. acknowledged in the CLASIX-3 sensitivity study, the l'inally, the issue of whether or not containment calculational results are sensitive to input parameters sprays operate is very important. It is beyond the for the baseline compartment configuration-our "B" scope of this report to determine the quantitative configuration.) Specific differences in calculated pres-likelihood or probability of spray availability. We can, pures and temperatures from the three codes can however, observe the effect of spray availability on our probably be attributed to differences in modelling and calculations. We conclude that spray operation is very input assumptions for heat and mass transfer. We important to plant safety for the accident scenarios have some concern regarding the CLASIX-3 calcula-(hydrogen release rates in particular) examined in our tional models for flow through the drywell vacuum study. The benefits of containment sprays are three- breakers and flow into the drywell through the sup-fold: the peak pressure of a single burn will be reduced pression pool. The CLASIX-3 treatment of heat by spray evaporation if the burn duration is sufficient. transfer to and evaporation of the spray droplets is ly long; the baseline pressure for multiple burns will be another area of concern.
reduced if the time between burns is sufficently long; Regarding detonations inside the Grand Gulf con-and the thermal environment to which plant equip- tainment building, two conditions must be satisifed ment will be exposed will be less severe. simultaneously: a detonable mixture must exist, and Calculations of the containment atmosphere pres- an ignition source or mechanism must exist. In our sure and temperature response to hydrogen combus- judgment, the likelihood of forming detonable mix-tion are a very important part of this HIS assessment. tures is small, especially for regions of significant But due to the uncertainties discussed above, total extent. We also currently believe that the likelihood of reliance on the code calculations is unwarranted. At detonation occurring, either by direct initiation or by present, engineering judgment may be as importent to transition from deflagration (flame acceleration), is the assessment as calculations even though it might very small. However, it is our belief that some degree appear to be more subjective. A good example of this is of flame acceleration, upward through the wetwell the CLASIX-3 sensitivity study of containment obstacles, is highly likely. The degree of acceleration sprays. The CLASIX-3 calculations indicate that will depend on the effectiveness of the HIS in burning burns without sprays result in lower peak pressures very lean mixtures and on the spatial concentration than burns with full spray operation. This result is gradients of hydrogen and air.
exactly opposite to our intuition, and in our opinion, is 12
Details of our calculations and assessments are ,
contained in the chapters that follow Chapter 2.0 ;
discusses the HECTR calculations while Chapter 3.0 covers the MARCH calculations. Where possible, we have made comparisons to CLASIX-3 calculations.
Chapter 4.0 presents RALOC code calculations of hydrogen mixing for a number of scenarios. Chapter 5.0 discusses a CSQ code calculation of the effects of a h>calized detonation as well as the likelihood of deto-nations or detonation-like phenomena. In Chapter 6.0, we present our assessment results and recommen-dations.
1.4 Reference UM. Berman, Light Water Reactor Safety Research Program Quarterly Report, Apdf-September 1981, Sandia National Laboratories, SAND 82-0006, NUREG/CR-2481, February 1982.
i i
s 13-14
2.0 HECTR Calculations for the Grand Gulf Nuclear Station 2.1 Description of the HECTR
! Computer Program j In order to model hydrogen burns in nuclear criteria are met in those compartments. These criteria i power plant containment buildings, we have devel- are based upon the hydrogen concentration, and typi-I oped a computer code called HECTR (Hydrogen cal values are > 4.1% hydrogen by volume for upward l Event: Containment Transient Response). HECTR propagation, > 6% for horizontal propagation, and I
can be used to analyze as many as ten separate con- > 9% for downward propagation.
tainment compartments (at present HECTR does not The radiative heat transfer model breaks up the have a drywell compartment model). It includes mod- emission from steam into seven spectral bands and els for hydrogen burns, radiative and conveciive heat calculates an emissivity for each band. The wall sur.
transfer, heat transfer to sprays, and wall heat con- faces are assumed to be gray and have constant emis-duction. siv' ties.
Flows between compartments are pressure driven, Convective heat transfer is modelled for both and the compartment interconnections are modelled vertical slabs and horizontal water pool surfaces. The as orifices. Gases are instantaneously mixed within vertical slab model allows for free or forced convec-each compartment. Steam and hydrogen source terms tion, either with or without condensation. The pool are specified by the user. model assumes laminar flow over a horizontal flat The hydrogen burn model in HECTR allows com- plate.
bustion to be initiated in a compartment whenever a The spray model treats both condensation on and specified hydrogen concentration is exceeded, provid- evaporation from spray drops. The spray drops aie ed that the compartment is not inerted due to a assumed to fall through a homogeneous atmosphere, shortage of oxygen or an excess of steam. Typical none of whose properties changes during the fall time values used to determine whether or not a gas mixture of the drop. Heat and mass transfer rates are deter-is inert are oxygen < 5% and/or steam > 55%. mined from the size and temperature of the drops The compartment burn time is calculated by di- when they reach the bottom of the compartment. A viding a characteristic compartment length by the distribution of droplets including up to ten different flame speed.The flame speed, v, is determined in most drop sizes can be modelled.
cases from the following experimental correlation: Walls are treated either as slabs, using standard finite-difference techniques, or as lumped masses.
v - 59.2X + 1.792 m/s (2.1) The heat flux to each wall surface is the sum of the radiative and convective heat fluxes. Up to 30 differ'-
where X is the hydrogen mole fraction present at the ent wall surfaces can be analyzed.
beginning of the burn (v is he!d constant during the Each of the various models in HECTR is dis-burn). The above correlation was derived from turbu- cussed in more detail in Appendix B. < '
lent flame experiments in the Sandia VGES 16 ft tank. (Note that this was an essentially open, clean-walled vessel; hence, flame acceleration due to obsta-cles did not occur. See Figure 2.1 and Reference 2.1 for 2.1.1 Summary and Conclusions more details.) Other values of the flame speed can be A wide variety of cases were run using HECTR, specified, if desired. Combustion is carried out either including cases to examine compartmentalization, to a given completeness (user specified) during the flame speed, completeness of combustion, propaga-calculated burn time or until the compartment runs tion limits, sprays and the hydrogen source term.
out of oxygen. Based upon the HFCTR analysis and a study of the The burn can propagate into adjacent compart. CLASIX-3 analysis presented in Reference 2.3, we ments after a specified time delay if the propagation have reached the conclusions discussed below.
15
)
12.0 , , , , ,
11.0 - """ O ~ -
g 10.0 - '"-O--MM W EMNS .
E
- 9.0 - -
0 4 4
8.0 -
, 4 -
" 7.0 - -
E 4 j 6.0 -
g."!92 -
l' $$
5.0 y g* 59 I -
3 4.0 -
4 -
I
- 3.0 -
- 9 ~~;
2.0 -
~~-h e _,,,0....
.5 KH2+ 0.s45 1.0 - Vg- q . . -o --
S -
0.0 0.04 0.05 0.06 0.07 0.06 0.09 0.10 MOLE FRACTION HYDROGEN XH2 Figure 2.1. Experimental Flame Speed (From Ref 2.1)
In most cases the HECTR analysis predicts peak cooling rate could be due to assumptions regarding 1 pressures that are much higher than those predicted passive heat sinks, heat transfer models, or the treat- I by the CLASIX-3 analysis.* The CLASIX-3 analysis ment of containment sprays. Because a small differ-divided the containment into two compartments: the ence in the models or input assumptions can have such wetwell region just above the suppression pool and the a dramatic impact on the results, the problem is rest of the containment, including the dome. The "ill-conditioned." Despite the differences in the re-CLASIX-3 analysis generally predicts a large number sults, we feel that the behavior of CLASIX-3 and of burns in the wetwell resulting in small pressure HECTR is similar. Both codes predict eventual rises and no burns in the dome. HECTR generally wetwell inerting, just at different times; and, as will be predicts at least one burn in the dome resulting in high discussed later, some of the CLASIX-3 cases were pressures, due to the large volume of the dome relative close to ignition in the dome at the end of the run.
to the wetwell. The reason that burns are predicted in Differences in the heat transfer models appear to be th'e dome is that HECTR predicts inerting in the small.
wetwell due to oxygen depletion at an earlier time A significant uncertainty in the analysis involves than CLASIX-3 does. Once t'ae wetwell inerts, hydro- mixing of hydrogen and oxygen in the containment, gen is transferred from the wetwell to the dome at a and neither HECTR nor CLASIX-3 properly ad-higher rate, and burns in the dome occur soon after- dresses this issue. Both codes include pressure-driven ward. Both CLASIX-3 and HECTR show that, after a masa transfer, but not convection or diffusion. There-burn in the wetwell, oxygen is drawn back into the fore, both codes underestimate the rate of mixing wetwell from the dome as the wetwell cools down. between compartments. In contrast, the rate of mixing HECTR predicts a slightly lower rate of cooling in the within a compartment is over-estimated, since it is wetwell, and thus, inerting occurs a few burns earlier assumed to occur instantaneously, than in the CLASIX-3 analysis. This difference in Based upon experimental data discussed in the next section, we feel that the flame speed used in the CLASIX-3 analysis is a factor of three or more low.
- Including the drywell compartment volume might reduce The lower flame speeds result in more time for heat ths pressures calculated with HECTR. tramfer during a burn, particularly if the sprays are 16
+
on, and correspondingly lower temperatures and pres- 2.2 Grand Gulf Analysis sures. The difference in peak pressure using the CLASIX-3 flame speed as opposed to the HECTR flame speed can be 5?6 to 15Fo depending upon the 2.2.1 Case Descriptions and input assumptions. Flame speeds higher than those HECTR Inputs used in llECTR (due, for example, to flame accelera.
Forty-four separate cases were run using IIECTR.
tion) could result in an additional increase of 2?; to Six cases were run for comparison with the CLASIX-3 8?6 in peak pressure.
cases presented in Reference 2.3, and the remaining 38 When varying other burn parameters we found cases were selected to address what we felt were that incomplete combustion results in decreased pres-important questions regarding the CLASIX.3 analy-sure rises; however, the benefit is not always as large as expected due to the fact that burns are more numer- sis.The factors that needed more evaluation included
- 1. Compartmentalization ous and closer together in time. Additionally, we found that burns propagating from the wetwell into 2. Flame speed hydrogen concentrations of 4 ro to 6?; in the dome will 3. Completeness of combustion probably not produce pressures in excess of the as- 4. Propagation limits numed failure pressure. Burns propagating into 8?6 to 5. The hydrogen source term 10?; hydrogen concentrations in the dome or burns In our Task 1 Report, we stated that the way the originating in the dome might produce pressures that containment was compartmentalized would be very could threaten containment. important, particularly in the way that mixing would Containment sprays almost always result in a he affected. Both CLASIX-3 and HECTR have significant reduction in temperature and pressure pressure-driven mass transfer between compartments during a burn. Sprays may have a slight negative and instantaneous mixing within compartments.
effect early in the accident if they enter the contain- There is considerable uncertainty as to how mixing ment at a higher temperature than the containment should be modelled. Because of this uncertainty, we atmosphere and result in a small elevation in pressure, felt it was prudent to examine five different compart.
However, this effect is small, if in fact it is real, and we ment configurations, shown in Figures 2.2 to 2.4 and would recomnxnd that the sprays be turned on as designated as A, H, C, D, and E.
early as possible in the accident and remain on until Configuration A treats the containment as one all danger of a hydrogen burn has passed. large compartment and allows us to evaluate the Additional efforts should be undertaken to better effects of global burns. Configurations B and C add a define or bound the hydrogen source term. Very slow second compartment in the annular region surround-release rates allow more time for mixing and make ing the drywell. Configuration H was used in most of global burns more likely. Very fast release rates result the CLASIX-3 cases and divides the compartments at in rapid inerting in the wetwell and also a significant the 135 ft level, while configuration C puts the entire addition of hydrogen to a burn in progress. annular region into compartment 2. The region below As expected, we found that relatively small the 135 ft level is open, while the region between the changes in initial pressure can produce significant 135 ft level and the 209 ft level contains large amounts changes in observed peak pressure. This occurs be- of equipment and structures. Configuration D divides cause, for any given hydrogen concentration, the ratio the containment into three compartments with divi-of peak pressure to initial pressure is almost constant sions occurring at the 135 ft level and the 209 ft level, for an adiabatic burn. Therefore, a loro change in Configuration E divides the containment into five initial pressure results in about a 1076 change in the compartments, with four of the compartments located peak pressure. Because the results are fairly sensitive in the annular region surrounding the drywell. Com-to assumptions regarding the initial conditions, one partments 1,2, and 5 are relatively open, while com-should use caution when examining the quantitative partments 3 and 4 contain large amounts of equip-results. ment and structures.
Finally, we observed that hydrogen burns may The flame speed of 1.83 m/s used in the occur well after hydrogen production has stopped due CLASIX 3 analysis appears to be significantly low.
to condensation of steam and the corresponding in. This speed may be appropriate for a flame in a quies-creases of the hydrogen and oxygen mole fractions. cent atmosphere (Figure 2.1), but is probably not Several of the HECTR cases and some of the appropriate for the environment expected in a reactor CLASIX-3 cases produced final conditions indicating containment. Higher flame speeds result in less time that such an occurrence was likely. for heat transfer, and subsequently, higher pressure.
17
The flame speed used in most ilECTR cases was term was constant over a time interval, while in fact, determined from Eq (2.1). CLASIX-3 uses linear interpolation between data The next area of concern involves the complete- points. This results in a difference of ~ 8fL in the ness of combustion. CLASIX-3 cases 3,5, and 6 from total amount of hydrogen released (the llECTit Reference 2.3 had ignition occurring at 8"6 hydrogen source term results in less hydrogen released). Case with a burn completeness of 85"F. We did not feel that B-1 was altered to use a source term more in agree-there was sufficient evidence available to conclude ment with the CLASIX-3 source term (Figure 2.6).
that co nbustion would be incomplete in all cases, and Compartment volumes and i.urface areas were therefore, conservatively assumed 100G combustion estimated based upon available drawings, or in some for most of the llECTR cases. cases were taken from the CLASIX-3 inputs. One Another question involves propagation limits. In difference between the input parameters for llECTR all of the CLASIX-3 runs, the propagation limits were and those for CLASIX-3 is that the upper pool was set equal to the ignition limits. In CLASIX-3 case 2, assumed to remain in place during the llECTR analy-for example, ignition occurs in the wetwell region with sis, while during the CLASIX-3 analysis, a portion of 10% hydrogen present, but propagation into the dome the upper pool was drained into the suppression pool.
region does not occur even when the hydrogen concen- Additionally,in the IIECTR analysis, the suppression tration in the dome is above 8?6. This is clearly pool was kept at a constant temperature of 358.15 K unrealistic. For most of the llECTR cases, we have set (185*F), and the hydrogen was assumed to enter the upward propagation limits at 4.1% hydrogan and containment at that temperature. In all cases, the downward propagation limits at 9.0% hydrogen.2 containment atmosphere was initially assumed to be Using realistic propegation limits generally causes air saturated with water vapor at 299.82 K (80*F).
upward propagating burns to occur, which limit the These values were obtained from Reference 2.3.
buildup of hydrogen in upper compartments.
The final area of concern involves the hydrogen source term. The CLASIX-3 source term in Reference 2.3 comes from a MARCH" calculation for a degraded. core accident modified by an arbitrary ex-trapolation to a specified quantity of hydrogen (con-sistent with 75?h metal-water reaction excluding the ,,_
channel boxes). There is a great deal of controversy surrounding the MARCH hydrogen source term.
Therefore, we felt it was prudent to make some
2"'
HECTR calculations with alternative release rate his- - -
tories (Figure 2.5). These runs considered the possibil- h h " " " " " ' " " '
ity of a very rapid release of hydrogen, although we m nn m .m. .
have no estimates of the probabilities of such release rates. Detailed descriptions of five of the CLASIX-3 l cases and the 44 HECTR cases are presented in
)
Tables 2.1 to 2.6. In all cases, the hydrogen is assumed to enter the containment through the suppression pool.
Inputs to IIECTR are described in Tables 2.7 to "" "
l 2.14. The CLASIX-3 hydrogen source terms from l Heference 2.3 are presented in Table 2.7. The time, for su - - -
l all cases, is referenced to 3295 s into the accident, n.in. _
l which corresponds to the time when significant hydro-gen production begins.
h h - = = = = =
l One problem was encountered in interpreting Reference 2.3. We assumed that the hydrogen source Figure 2.2. Compartment Configurations A and B 18
i I
l l
Comp.rtment 1 CoeIPAATRIENT 1 EL 209'- EL 200' -
DAYWELL {""" y g EL 1st' - a 3 CoasPAnTastuT 2 C :2 e -4 EL 136' - -_
- t.
gg g g g..
g g _ E ._ P_
L,.. -
g -
g, C_
CoesPAafnAENT CoNFloWII AT'oel C f . :3 CoasPARTIAENT 1 EL 209'-
Drywell
- CoeseARTtBENT 2
- ^
82 EL IS S* =-
- CoasPARTtIEN" 4 EL 111* -
g g _ P.E ,_ P.
. CotePAPTasENT CoesFlounATioes D % .4,g.
Figure 2.3. Compartment Configurations C and D Fleure 2.4. Compartment Configuration E
_jsooo , , , ,
G
- = - _
g .ig.rj;;;;"-----
l *~ -
lll i.g -
- / 4 aLL ODGER casse encen mi 8vooo -
1 .00 J-i g o ' ' '
o tooo sooo sooo 4000 sooo nose t .a41 Figurs 2.5. Hydrogen Release Into Containment for flECTR Cases B8, B9, B10, and All Others (Except Case B1) 19
3000 , , , ,
n,,00 .
9 f3000 -
. , s..
1s00 -
.s' b ALL OTHERS 8 1000 - .'
, s00 -
t g --
- O 1000 2000 3000 4000 8000 fMAE (setende) !
I Figure 2.6. Hydrogen Release Into Containment for HECTR Case B1 (Approximately the CI.ASIX-3 Release Ilistory) and All Others (Except B8,119, and HIO)
Tcble 2.1. CLASIX-3 Case Descriptionsa I i i I l Temp at Pressure I l Ignition ' Propagation Combustion l Flame 3973 at 3295 l l Case l Compartment l Limit ll Limite l Completeness Speed l Secogos Secongs l l l Number 1 Configuration ! 4H 2 l 1 % l (m/s) i (K) (atm) l Spray Cl
%H2 i l I l 1 l l l l I I I I I I I ll 1 1 N l 10 1 10 l 100 1.8288 1 299.82 l 1.157 l Auto l i l i I I I I 2 l F l 10 1 10 l 100 l 1.8288 l 299.92 l 1.0 Auto l I I I I l I I I 3 i r I 8 I 8 l 85 l 1.8288 l 299.82 l 1.0 Auto I I i l I l i l i I I I 5* I r 1 8 I 8 I 85 i 1.8288 1 299.82 1 1.0 1 Off I I i 1 I I I I I I I l 6 I r 1 8 I 8 I 85 1 1.8288 l 299.82 l 1.0 l off I l i I I I l l 1 I l
- These descriptions were taken from Reference 2.3 i h 3295 seconds is the time when hydrogen production becomes significant. This corresponds to time - 0 in the llECTR cases.
Conditions at this time are taken from the plots in Reference 2.3 C
For Auto, the sprays were turned on after the first burn and remained on for the rest of the run d Similar to configuration B, Fig 2.2, except that the drywell is included as an additional compartment
' Ignition was prevented in the wetwell but a burn could propagate into the wetwell with 1% He in it and there burn to 100%
completion i
i i
l l 20 L
l Table 2.2. HECTR Case Descriptions for CLASIX-3 Comparison ,
i i I ,
I i l l I I !
I Ignition Propagation Combustion l Plame Initial Initial l
.I Case Compartment Limit Limite Completeness i Speed Temp Pressure l l l Number Configuration %H2 %H2 %
(m/s) (K) l (atm) l Spray
- I 1 I I I I I i i I l B-1 B I 10 l 10 100 1.8288 1 299.82 1 1.157 l Auto l l l l l l B-2 B l 10 l 10 100 1.0288 299.82 1 1.0 l Auto l i l I i l
! B-2' l B l 10 1 10 100 1.8288 299.82 l 1.0 Auto l I I I 1
i l i I I I i l I i l I l I l l 1 i B-3 I a f 8 '
8 I 85 1.8288 l 299.82 1 1.0 Auto l I I I I l A-1 A I 8 8 85 l 1.8288 299.82 1 1.0 Off I I I I l s-4 I o I 8 1 8 85 l 1.8288 .
299.82 l 1.0 l off I I I I I I l I l
- Sprays: Off--off during entire run Auto-turned on after ternperature or pressure setpoints exceeded and left on for the rest of the run Table 2.3. Single Compartment (Configuration A) HECTR Case Description i i I i i I I l l l I I Ignition l Propagation I Combustion Flamea i Initial l Initial l l Case Compartment ,
Limit l Limits l Completeness l' Speed ' Temp ! Pressure ! l
' Number Configuration l %H2 l 4H2 l % (m/s) (K) l (atml I SPrayb l l i I I I I I I I I I I A-2 A 8 l --
l Ida-- f F(Z H2) 299.82 l 1.0 l Off I I l i I l l A-3 A 8 l --
l 100 l F(% Hy ) i 299.82 l 1.0 l On l I I I l l l ,
I I I A-4 .
A 8 I --
l 100 i F(% Hy ) I 299.82 l 1.0 I Auto l I l 1 1 I I I I I I I I I I I I I A-5 i A l 8 -- :1 100 I F(% H2) 299.82 1 1.157 i Auto I l l I I I I A-6 A
I ll 10 l --
100 l F(2 H2 ) l 299.82 1 1.0 l Off l I I I I I I I
- I A-7 , A . 10 1 --
100 i F(% Hy ) 299.82 l 1.0 1 On I I I I I I I I I i l l 5 I I I I A-8 i A '
10 1 -- l 100 i F(% Hy ) I 299.82 l 1.0 l Acto l I I I I I I I I i F(% H 2 ) l 299.82 1 A-9 A 10 l -- l 100 1.157 l Auto l I I I I I I I A-10 A I 4 l --
100 l F(% Hy ) ' 249.82 1 1.0 l On j I I l l I I I I I l l 1 l i I 7 A-11 A l 7 --
l 100 i F(% H2 ) I 299.82 l 1.0 l On i l' I l l l l l l l A-12 A I 9 l --
l 100 l F(I H2 ) I 299.82 l 1.0 l On i l I I I I I I I
" Flame speed is calculated from V = 59.2 Xu, + 1.792 m/s where Xu, is the initial hydrogen mole fraction b Off--off during entire run Sprays:
On--on during entire run Auto--turned on after temperature or pressure setpoints exceeded and left on for the rest of the run 21
i I
Tcbie 2.4. Additional Multicompartment HECTR Case Descriptions i
I I i i I l l l l l l l Ignition Propagation Combustion Flame 1 2nitial l Initial I I Case Comipartment ,
Limit Limitse Completeness Speedhl Temp i Pressure 'l I Nulber Configuration l SH2 %H2 % (m/s) l (K) (atm) SprayC l
. I l l l i I I . I B-5 B l 8 l C(2 Hy ) 100 F(2 Hg ) l 299.82 1 1.0 ll Auto I i
l I i 1 1 i l I I B-6 8 10 C(% H2 ) 100 l F(% Hg ) 299.R2 1.0 l On l 1 i I 8-6' B 10 G(1 H2 ) 100 F(% H2 ) 299.92 1.0 On I ll l I E-7 B 10 C(% H2 ) 100 F(% H2 ) 299.82 1 1.0 Auto l
]
l a 1 i ; i
. I l C-1 C 8 g C(1 H2 ) 100 F(2 Hy ) l 299.82 1.0 l' Auto l t l l C-2 C 10 C(1 H2 ) 100 i F(1 H2 ) I 299.82 1.0 On l l l C-3 C '
10 C(% Hy ) I 100 F(2 Hg ) l 299.82 1.0 Auto I j l l l l I l C-4 C 8 C(% Hy ) l 100 F(2 H2 ) 299.82 l 1.0 Off l ,
1 I l l D-1 D 8 G(% Hy ) .
100 l F(1 Hy ) l 299.82 l 1.0 I Auto I !
l l I D-2 D 1 10 l C(% H2 ) 100 F(%yH ) l 299.82 l 1.0 Auto l i I i 1 I i i i 3 D-3 D l 8 l G(% Hy ) 100 1 F(1 H2 ) 299.82 1.0 Off l E-1 E 10 G(% H2 ) 100 F(2 Hy ) 299.82 1.0 l Auto l 1
" Propagation limits: Upward - 4.1 %
Horizontal - 6.07.
Downward - 9.0%
b Flame speed is calculated from V - 59.2 XH, + 1.792 m/s where Xu, is the initial hydrogen mole fraction CSprays: Off -off during entire run On - on during entire run Auto- turned on after temperature or pressure setpointa exceeded and left on for the rest of the run l
- l l
l I
l l
l 22 ;
l l
t I -
Table 2.5. Single-Compartment Flame Speed Sensitivity Case Descriptions l l i' i' I I I Ignition Propagation *l Combustion Flame Initial Initial , l Case Compartment Limit . Limite Completeness Speed Temp Pressure l l l Number Configuration SH2 l til2 % (m/s) (K) .
(atm) Sprayb l l i A-13 A 10 C(2 H2 ) 100 I 42.37 299.82 1.0 Off A-14 A 10 C(t H2 ) 100 8.47 299.82 1.0 Off A-15 A 10 C(% H2 ) 100 . 4.24 299.82 '
1.0 ,
Off l .
A-16 A 10 C(% H2 ) 100 1.69 299.82 1.0 ll Off i
l A-17 A 10 C(2 Hg ) 100 0.67 299.82 1.0 Off A-18 A 10 l C(% Hy ) 100 l 0.42 299.82 1.0 Off A-19 A 10 C(% H2 ) 100 f 42.37 299.R2 1.0 l On i I
! A-20 A 10 C(2 H2 ) 100 8.47 '
299.82 1.0 on l
A-21 A 10 C(% H2 ) 100 4.24 299.82 1.0 On A-22 A 10 c(g g2 ) 100 l 1.69 299.82 1.0 l' On l l l l A-23 A '
10 C(% H2 ) 100 0.67 ' 299.82 1.0 l' On A-24 A 10 C(% H1 ) 100 0.42 299.82 1.0 On l l l 1 l l l l aPropagation limits: Upward - 4.1 r.
Horizontal - 6.0%
Downward -9.0%
b Sprays: Off-off during entire run On - on during entire run l
23
Tcble 2.6. Source Term Sensitivity Case Descriptions i i i i i 7 I I l' I I I llignition a Combustion IFlameb Initial Initial l Maximum H2 I C ee Compartment Limit (Propagation l Limite Completeness ISpeed Temp Pressure l Release Ratel NumberlConfigurationi lH2 %H2 1 4 l(m/s) (K) (atm) '
Spray C l (Moles /s) l l l l l l I I I I I B-8 8 l 8 G(% H2 ) I 100 l F(1 H,)! 299.R2 1.0 Auto l 93R.4 l l l'
- l l l l I I i B-9 B l 8 '
G(% H2 ) I 100 1.0 Auto 3753.6 l l
' F(1 H )l 299.82 l l l ! l l B-10 l B [ 8 G(% H2 ) I 100 lF(% gH )! 299.82 l 1.0 Auto ! 15014.4 i l l I I I I f I aPropagation limita: Upward - 4.1 re l Horizontal - 6.0% '
Downward - 9.0To b
Flame speed is calculated from V = 59.2 Xu, + 1.792 m/s where Xu, is the initial hydrogen mole fraction C
Sprays: Auto-sprays turned on after temperature or pressure setpointa exceeded and left on for the rest of the run d
htzximum H2 release rate for all other cases is 234.6 moles /s Tcble 2.7. Hydrogen Source Terms Used in HECTR Calculations l 1 I i Case B1 1 Case B8 l Case B9 l Case B10 All Other Cases l l 1 I I I I I i 1 l Time ! Release Rate Time ! Release Rate l Time i Release Rate ! Time l Release Rate Time Release Rate (s) l (Moles /s) (a) l (Moles /s) l (e) l (Moles /s) l (s) l (Moles /s) (s) (Moles /s) l I i 0 1 0 I l 0 0 0
.8671 l .1351 l .
.1351 .1351 I I .1351 306 l 306 l l 306 1 306 l 306 i 1.802 l l 1.599 l l 1.599 l 1.599 l 1.599 336 l l 336 1 336 l 'i36 336 l l 6.397 l l 2.004 1 1 2.004 2.004 1 2.004 606 l l 606 l l 606 1 1 606 '
! 606 l l l 10.87 I l 10.79 l l 10.79 l l 10.79 l l 10.79 906 l l 906 l l 906 I 906 l l 906 l i
l 41.36 l l 10.95 l i 10.95 l 10.93 1 10.95 1246 l l 1246 l l 1246 l 1246 l l 1246
( l 153.2 I l 71.76 l l 71.76 1 1 71.76 1 'I 71.76 1563 I i 1563 l l 1563 I i 1563 1 1563 1 172.6 l l 234.6 l l 234.6 l l 234.6 l 234.6 1863 l 1863 1 1863 l 1863 1 1963 1 63.03 l 110.5 l l 110.5 110.5 l 110.5 2163 l l 2163 l 2163 1 2163 2163 1 122.4 l 15.56 l l 15.56 15.56 l 15.56 2463 l 2463 2463 l 2463 2463 l
, l 120.9 l 229.2 1 229.2 229.2 1 229.2 2763 l l 2763 2763 2763 l l 2763 l 105.3 l 12.52 1 12.52 1 12.52 1 12.52 3063 I l 3063 3063 3063 I l 3063 l 216.4 l 198.1 '
198.1 l 19R.1 L 198.1 3064 l l 3064 3064 1 3064 3064 l 234.6 l 938.4 '
3753.6 l 15014.4 234.6 4512 l 1 3426 :l 3155 l' l 30R7 l 4512 l l 0 l 0 1 0 l l 0 l ll 0 i 4525 l l 4000 l 4000 l l 4000 l l 4525 1 24
l Table 2.8. Compartment Volumes l
Compartment Compartment Volume (m')
Configuration 1 2 3 4 5 A 37886 - - - -
B 33545 4341 - - -
C 23683 14203 - - -
D 23683 9862 4341 - -
E 23683 1654 4480 3278 4341 Table 2.9. Compartment Flame Propagation Lengths Compartment. Compartment Flame Propagation Length (m)
Configuration 1 2 3 4 5 A 25.4 - - - -
B 21.9 3.7 - - -
C 19.0 11.1 - - -
D 19.0 11.1 3.7 - -
E 19.0 11.I 14.3 7.9 7.3 Table 2.10. Interconnection Areas and Flow Coefficients t
Compartment Interconnection Configuration or Area Flow Connections (m') Coefficient Two Compartment Models )
B 206.8 1.2 C 226.4 1.2 Three-Compartment Model (D) 1-2 226.4 1.2 2-3 206.8 1.2 Five-Compartment Model (E) 1-2 74.3 0.75 1-3 152.1 1.5 2-3 135.64 1.5 2-4 157.0 1.5 2-5 62.4 0.75 3-4 228.8 1.5 1 4-5 144.4 1.5 l 25
Table 2.11. Spray Data for All Cases Data Measurement Inlet Temperature 330.37 K Flow Rate (One Spray Train) 11300 gpm Fall Height
- 19.5 m Number of Drop Sizes 2 Frequency of First Drop Size 0.95 Frequency of Second Drop Size 0.05 Diameter of First Drop 309 microns Diameter of Second Drop CIO microns Containment Temperature Setpoint" 358.15 K Containment Pressure Setpoint** 1.6124 atm Time Delay for Spray Actuation 120 s
- For Case B2', in which the spray was allowed to carry over into the wetwell, the wetwell fall height was assumed to be 7.315 m, and 10% of the spray was assumed to carry over from Compartment 1 into the wetwell.
" Sprays on Auto are turned on when either of these setpoints is exceeded in any Compartment and remain on throughout the rest of the run.
Table 2.12. Surface Descriptions Surface Number surface Area (m2) Material
- 1. Concrete rioor 297 Concrete
- 2. steel Pool walle 128 steel
- 3. Upper Pool 396 water
- 4. Crane 1188 steel
- 5. Dome 3291 steel
- 6. Wetwell wall 393 steel
- 7. Grating at 209' t. eve l 575 steel
- 8. Crating at 185' tavel 358 steel
- 9. Equipment 3027 steel
- 10. Concrete 2103 Concrete
- 11. Wetwell wall 950 steel
- 12. Drywell wall 202 Concrete
- 13. Grating at 161' Level 745 steet
- 14. Equipment 1334 steel
- 15. Concrete 396 Concrete i 16. Wetwell wall 102 steel
- 11. Drywell wall 487 Concrete
- 18. Crating at 135' I4 vel 792 steel
- 19. Concrete 308 Concrete
- 20. Grating at 121' 14 vel 335 steel
- 21. Suppreselon Pool 619 water
- 22. Wetwell well 832 steel
- 23. Drywell wall 557 Concrete 26
Table 2.13. Location of Surfaces Compartment Surface from Table 2.12 Contained in Configuration Compartment 1 2 3 4 5 l
A all -- -- -- --
B l-17 18-23 -- -- --
l C 1-5 6-23 -- -- --
D 1-5 6-17 18-23 -- --
1 E 1-5 6 7-12 13-17 18-23 l l Table 2.14. Wall Properties All Cases Steel Concrete Water Thermal Conductivity 43.27 .l.385 ---
W/m-K Thermal Diffusivity 1.17x10-5 5.5x10-7 ___
m2/s Specific Heat 460.5 879. ---
J/kg-K f smissivity 0 .7
- 0 .9 0.94
- Assumes all steel surfaces are painted l
27
2.2.2 Results case 2 reveals that the wetwell is inert (Xo, = 0.049 <
Results for the five CLASIX-3 cases are summa. 0.05) near the end of the run (Table 2.15). In the rized in Table 2.15. Results for the 44 HECTR cases HECTR cases, wetwell inerting occurred just soon are summarized in Tables 2.16 to 2.21. The number of enough to get a compartment 1 burn before the end of burns in each compartment is presented along with the run, and in CLASIX-3 case 2, wetwell inerting the final gas compositions and the maximum tempera, occurred just late enough not to get a compartment 1 i tura and pressure encountered during the run. Each of burn. Note that there was 8.5% hydrogen in compart- l the HECTR cases is discussed briefly below, and ment 1 at the end of CLASIX-3 case 2.
saven cases (B2', A6, A8, B5, C3, D2, and E1) have been selected for presentation in more detail. For many two-or-more-compartment cases only a single Cases B3, A1, and B4-HECTR cases B3, A1, I pressure plot has been presented. These can be con- and B4 (Figures 2.22,2.23, and 2.24) were meant to sidtred to be effectively the containment pressure model CLASIX-3 cases 3,5, and 6. Each of these runs plots since the pressures in individual compartments involved ignition at an 8Te hydrogen concentration
, follow each other closely. In the following discussions, and combustion that was 85% complete. HECTR d
the containment failure preuure is assumed to be 4.8 cases B3 and B4 involved compartment configuration I
atm" (all pressures are absolate). B, while case Al involved compartment configuration i A. CLASIX-3 cases 3 and 6 predicted a large number Ctes B1, B2, and B2'-HECTR case B1 was of wetwell burns and no burns in the dome, while the i meant to model CLASIX-3 case 1, and HECTR cases corresponding HECTR cases B3 and B4 predicted B2 and B2' were meant to model CLASIX-3 case 2. In eventual inerting in the wetwell followed by a burn in CLASIX-3 cases 1 and 2, large numbers of burns the dome. However, containment failure was not pre-occurred in compartment 2 (wetwell), and no burns dicted in either case B3 or B4, although case B4 occurred in the dome. In the HECTR runs, oxygen predicted pressures within 0.2 atm of the assumed s inerting in the wetwell occurs earlier than in the fail-re pressure. HECTR case Al and CLASIX-3 case CLASIX-3 runs, leading to more hydrogen trans- 5 produced similar results, with HECTR predicting a ferred to compartment 1 (containment). The result is higher pressure, but not one close to failure.
that HECTR cases B1, B2 and B2', contrary to CLASIX-3, predict a burn in compartment 1 and a Cases A2 Through A12-Cases A2 through A12 peak pressure in excess of the 4.8 atm failure pressure. were run to examine the effects of global burns under a The only difference between HECTR cases B2 variety of conditions. As discussed earlier, combustion and B2'is that in case B2' the sprays were allowed to is assumed to be 100% complete and the flame speed fall from compartment 1 into the wetwell. Case B2' (6.5 m/s for 8% burn and 7.7 m/s for 10To burn) is was added to see if the treatment of sprays in the about a factor of 4 higher than that used in CLA-wetwell was responsible for the differences between SIX-3.
HECTR cases B1 and B2 and CLASIX-3 cases 1 and Cases A2 and A6 examine the effects of having no
- 2. The results were that in case B2', wetwell inerting containment sprays. As expected, high pressures are was delayed with respect to case B2, but the end achieved. A pressure plot for case A2 is shown in results were similar. Pressure plots for cases B1 and Figure 2.25. Case A6, with a maximum pressure well in B2 are shown in Figures 2.7 and 2.8, and case B2' is excess of the failure pressure,is presented in detailin presented in detail in Figures 2.9 to 2.20. Figure 2.19 Figures 2.26 to 2.31.
shows oxygen inerting at approximately 3600 s and Cases A3, A7, A10, All, and A12 (Figures 2.32 Figures 2.14 and 2.20 show large increases in the through 2.36) were run with the sprays on from the hydrogen mole fraction after this time. Note that beginning of the accident, and significant reductions when ignition finally occurs in compartment 1, oxygen in maximum pressures are observed, although in case is pushed back into compartme it 2 and a burn occurs A7 the maximum pressure is very close to failure.
there also. The wetwell temperature plot for HECTR These cases also show the effect of varying the ignition case B2 is presented in Figure 2.21 for comparison point from 6% hydrogen to 10% hydrogen. Case A10, with case B2' (Figure 2.15). which ignites hydrogen at 6%, has a peak pressure of Despite the fact that the CLASIX-3 case 1 and 2 3.2 atm, and case A7, which ignites hydrogen at 10%,
results and the HECTR cases B1, B2, and B2' results has a peak pressure of 4.8 atm.
appear to be very different, we feel that the codes are In cases A4 and A8, the sprays come on shortly behaving similarly. Close examination of CLASIX-3 after the first burn, and the results are similar to the 23
results of cases A3 and A7. A pressure plot for case A4 expected that, as the containment cools down, the is shown in Figure 2.37, and case A8 is presented in steam will condense, thus raising the hydrogen mole detail in Figures 2.38 through 2.43. It should be point- fraction and resulting in a burn at a later time. Case ed out that in case AG, the first burn did not fail B6' is an exact duplicate of case B6, except that a containment, so the main difference between the re- longer run time is used, and demonstrates this point.
sults of case A6 and case A8 is the difference in In case B6', a compartment 1 burn that fails contain-pressure rise after the second and third burns. The ment occurs about 2 min after the hydrogen source reason that cases A3 and A7 resulted in higher pres- becomes zero.
sures than cases A4 and A8 is that the sprays contrib-uted to heating up the containment slightly before the Casas C1 Through C4-Cases C1 through C4 first burn. This was due to our choice of spray tem- (Figures 2.61 throu;;h 2.75) involve compartment con-perature which was taken from Reference 2.3.
figuration C. In each of these cases, upper compart-Cases A5 and A9 (Figures 2.44 and 2.45) show the ment burns occur; however, they occur just above the effect of a slightly elevated pressure at the start of the propagation limits, and the pressure rises do not result run. Marked increases in maximum pressure over in containment failure.
cases A4 and A8 are observed, with case A9 well above the failure pressure. These elevated initial pressures would be characteristic of a drywell break, in which air Cases D1 Through D3---Cases D1 through D3 (Figures 2.76 through 2.89) represent the three-com-is pushed from the drywell into the containment. It partment model. In each case, most of the burns occur should be noted that reverse flow, from the contain-in compartment 3 (wetwell). Compartment 3 eventu-ment into the drywell (through the suppression pool),
ally inerts, leading to burns in compartment 2 and one would afford some measure of pressure relief during a burn. burn propagating into compartment 1 (dome). The predicted peak pressures do not threaten contain-ment.
Case E1-Case El represents the five-compartment Cases B5, B6, B6' and B7-Cc.ses B5, B6, B6' model and is presented in Figures 2.90 through 2.109.
and B7 (Figures 2.46 thrcugh 2.60) all use compart- There are several burns observed in compartment 5 ment configuration B, which is very similar to that (wetwell) before it finally inerts due to oxygen deple-used in the CLASIX-3 analysis. However, we have tion. After cor,partment 5 inerts, most of the burns assumed complete combustion and higher flame occur in compartment 2. The burns occur in compart-speeds. Additionally, the propagation limits discussed ment 2 rather than in compartment 4 due to a lower earlier were in effect. flow coefficient between compartments 5 and 2 than in case B6, a global burn was predicted, but the between compartments 5 and 4. This results in hydro-pressure rise was small due to the fact that the hydro- gen moving preferentially into compartment 2. Two gen concentration in compartment I was not far above burns are observed in compartment 1 (dome); howev-the 4.1% upward propagation limit when the burn er, they occur at hydrogen concentrations just above occurred. Cases B5 and B7 resulted in containment the propagation limits and do nat result in contain-failure due to the fact that compartment 2 eventually ment failure.
inerted, resulting in compartment 1 burns with hydro-gen concentrations at the ignition limits rather than at Cases A13 Through A24-Cases A13 through the propagation limits. A24 (Figures 2.110 through 2.121) show the effects of Most of the cases were arbitrarily stopped shortly varying the flame speed. The flame speeds used result after the hydrogen source went to zero. However, it in burn times ranging from 0.01 to 1.0 min. Cases A13 was noted that in some of the cases (particularly A2, to A18 assumed no sprays, and the differences in A3, A4, and B6) the hydrogen concentration in the pressure rise were small except for very long burn dome region at the end of the run was very close to the times. Cases A19 through A24 assumed the sprays ignition point.This was also true of CLASIX-3 cases 3 were on, and the differences in pressure rise were and 6 and, to a lesser extent, CLASIX-3 cases 1 and 2. much more pronounced. This was to be expected In all of these cases, there is a significant amount of because the sprays are usually the dominant heat steam present at the end of the run. It is to be transfer mechanisms.
29
C:ses B8 Through B10-Cases B8 through B10 there is no time to Cool down and draw oxygen back (Figures 2.122 through 2.124) show the effects of into the wetwell from the dome between burns. So<m injecting hydrogen very rapidly into the wetwell. In after the wetwell inerts, a burn in the dome follows, each case, the results are similar. Once the hydrogen resulting in very high pressures (even with sprays on).
starts coming in rapidly, the wetwell inerts after one Usually this burn in the dome pushes oxygen back or two burns. The hydrogen is released so fast that into the wetwell, Causing a burn there also.
Table 2.15. Results for CLASIX-3 Runsa I i I l I l l 1 Final Mole Fractions I l l 1 Number of Burns ! l l l l l l 1 Maximum 1 Maximum I l Case l l l l Temperature l Pressurel INumberl l l Compartment 1 (Containment) l Compartment 2 (Wetwell) 1 (K)b l g b l l l l 1 1 I I I l l Comp. 1 I Comp. 21 i i l l l I l l N2 Og H2 HO 2 l N2 02 H2 HO2 I 1 l t 1 I I I i i I i I I l I I I I I I I I i 1 1 0 1 18 i .64 I .135 I .075 I .15 1.50 I .11 I .05 ' 26 1
. 1085.4 1 1.755 :
1 1 1 1 I I I I I I I I I I 2 I o i 43 I .64 I .oss t .085 ! .22 1.59 1 .o49 I .111 .25 l 1072.6 1 1.504 I I I I I I I I I I I i l i I 3 I o I se I .66 i .os2 1 .078 1 .21 1.61s 1 .04s I .11 l' . 2 3 I e45.4 l 1.456 i i l l I I I I ! I I I I I I i s 1 4 1 4 I .655 I .o3s ! .o3 I .2a f.64 I .o4 I .oss 1.2651 1379.3 1 3.361 I I I I I I I I I I I I I I I I 6 1 0 1 68 i .63 I .os I .075 I .245 1.603 I .049 1 .112 1.236] 1130.4 l 1.701 I I I l ! I I I I I I I I i 1 aThese results were taken from Reference 2.3 b
Maximum of the wetwell and containment values Tcble 2.16. Results of HECTR Runs for CLASIX-3 Comparison I I I I I l l l Final Mole Fractions l l l 1 Number of Burns l 1 l l l l l l Maximum Maximum l l Case I l l l Temperature ll Pressurel lNumberI I l Compartment i I Compartment 2 I (K) l (atm) l I I I I I I I l l IComp. 1 l Comp. 21 1 1 l 1 I l N2 O2 H2 HO 2 l N2 O2 'I2 HO2 l l
! ! 1 1 l 1 1 1 l I i l l I i i i i IB-1 1 1 l 21 f.4890 1.0462 .0112 l .4535 1' .3759 1.0306 1.2871 1.30641 1298 1 5.600 l l I I I I I B-2 l 1 24 1.5715 1.0468 .0192 I .3625 .3424 l.0185 1.3273 .3118 1323 4.845 l i I I I I 1 I I I
[B - 2*l 1 1 21 1.4201 1.0291 ! .0060 I .5448 . .3601 1.0144 l.2110
.41451 1263 1 5.042 l i I I I I I I I I I I I I I I I I I I I I B-3 1 1 32 .7013 1.0778 .Os71 .1638 .2749 l.0210 1 .4770 1 .2271 1298 3.830 '
I I I I A-1 1 4 l --
1.6315 1.0403 l .0330 .2951 -- I -- -- 1 -- 1053 1 4.090 1 1 I I I ll l 1 l .1 1 B-4 1 1 30 1.6759 1.0766 I .0409 I .2067 1.2635 1.0180 1.5073 1.21121 1315 I 4.630 1 I I I I I I I I I I l 'I 30
Table 2.17. Results for Single-Compartment HECTR Cases i I i I i l l Final Mole Fractions l l l I ! Number of Burns l 1 l l l Maximum Maximum l Case 1 l Temperature Pressurel' l Number i I Compartment 1 Compartment 2 l (K) l l 'l (atm) I Comp. 1 I Comp. 2 I N2 og n2 noa N2 02 n2 no a
1 T I i i i i A-2 3 --
.6231 .0591 .0712 .2466 --
-- l -- -- 1178 4.414 i
A-3. 3 -- .7056 .0630 .0727 l .1586 --
-- I -- -- 968 3.982 I i I 1 1 I I A-4 3 i --
H .6953 .0652 .0781 .1614 -- -- -- -- 1122 3.902 1 I I
I I A-5 3 --
f.6773 .0651 l .0358 .2218 -- -- -- --
1126 4.514 I I I :l I
lA-6l 3 l --
1.6362 1.0266 l .0052 .3319 .
-- l -- 1 -- -- 1353 l' 5.278 I I I I I I ! ! I la - 7 I 3 1 --
1.5057 1.0208 1 .0034 .4701 l -- I -- I -- --
1156 4.774 I I I i l I I I I I i i I i lA-8 l 3 ! --
1.5306 1.0245 1 .0089 l .4360 -- . -- 1 -- ' --
1304 4.647 l I l I l I i l I i 1 ;I I IA - 9 1 2 1 --
1.6894 1.0849 I .0737 I .1520 '
-- I -- 1 -- -- l 1308 1 5.326 1 I I I I I I i 1 1 I ! I la - 101 5 l --
1.5807 1.0265 I .0086 I .3763 I -- I -- 1 -- - -- ! 767 I 3.177 '
I I I I I I I I I I I I I I I I i i I i i i I I i I i IA- til I I 4 1 I
1.6396 I I 1.0375I I .0269 I .2960 I -- I -- 1 -- I -- l 866 1 3.599 I I l t t t t I I lA - 121 3 l --
1.6240 1.0389 I .0308 l .3063 l --
! -- l -- l -- l 1067 1 4.441 l I I i I I i i ! I I i l I i Table 2.18. Results for Additional HECTR Two-Compartment Cases I I I I I I l l l Final Mole Fractions 1 l l Number of Burns l 1 I
- I I I I l Case l i
Maximum ! Maximum l Temperature 1 Pressurel INumberi I Compartment 1 Compartment 2 I I I I
(K) 1 (atm) l I I I l Comp. 1 l Comp. 21 I l l
l l l N2 C2 H2 HO 2 l N2 02 H2 HO 2 I I I l l l I I I I I I B-5 1 1 30 1.6944 1.0672 .0434 .1950 l.3304 s.0292 .4137 1.2267 1305 4.877 l l l 1 I I I I IB - 6 1 18 .6710 1.1145 .0968 i .1177 1.1051 .0136 .7872 .0940 1305 1 2.467 I I I I I is - 6' 2 19 1 .7557 l.0510 .0410 .1523 .6517 .0424 .0717 .2331 1410 1 5.693 I I
B -7 2 21 .5100 i .0332 .0043 .4526 .4566 .0263
.2684 .2487 1384 5.502 I
I I i C-1 1 9 l.7451 1.1003 .0278 .1267 .4675 .0308 .2601 .2416 1088 2.694 I I I I l
'C - 2 1 1 7 l.7293 1.1029 .0411 1 .1268 1.4878 .0307 .2430 .2384 1212 2.898 I I I l l l l l l 1 IC - 3 1 1 l 7 f.7277 1.1046 I .0417 .1260 1.4828 1.0317 .2512 1.2343' 1211 1 2.85a l i I I I I "
1 I I I I I I I i l I I I lC-4 l 1 l 9 f.6786 .0883 l .0263 I .2069 .5145 1.0300 n.1832 1.2723 l 1111 l I I I I 1 3.310 l I I I I I I 31
Tcble 2.19. Results for HECTR Three- and Four-Compartment Cases Maximum Maximum Final Mole Fraction Temperature Pressure Number of Case Compartment Burns N: 02 H: HO2 (K) (atm)
D-1 1 1 0.7333 0.0921 0.0078 0.1668 1195 2.630 2 8 0.5468 0.0458 0.1394 0.2630 3 17 0.2365 0.0205 0.6006 0.1423 D-2 1 1 0.7431 0.1071 0.0168 0.1330 1310 2.800 2 6 0.5179 0.0400 0.1568 0.2853 3 13 0.1957 0.0195 0.6611 0.1237 D-3 1 1 0.6918 0.0854 0.0092 0.2136 1207 3.270 2 8 0.6377 0.0314 0.0293 0.3017 3 19 0.3258 0.0274 0.4781 0.1688 E-1 1 2 0.7185 0.0774 0.0466 0.1574 1477 2.579 2 24 0.5069 0.0496 0.2669 0.1766 3 1 0.7113 0.0901 0.0371 0.1615 4 6 0.6941 0.0932 0.0519 0.1608 5 18 0.2377 0.0221 0.5906 0.1496 Table 2.20. Results of Flame Speed Sensitivity Analysis Maximum Maximum Case Flame Speed Temperature Pressure Number * (m/s) (K) (atm)
A - 13 42.37 1322 4.652 A - 14 8.47 1301 4.574 A - 15 4.24 1274 4.482 A - 16 1.69 1207 4.253 A - 17 0.67 1084 3.828 A - 18 0.42 1004 3.555 A - 19 42.37 1287 5.106 A - 20 8.47 1167 4.838 a A - 21 4.24 1043 4.555 A - 22 1.69 756 3.76G A - 23 0.67 456 2.658 A - 24 0.42 405 2.394
'These runs were stopped after the first burn.
32
Table 2.21. Results of H, Source Term Ser.sitivity Analysis I I i l i i l l I Final Mole Fractions l l l l Number of Burne l I l l l 1 i i Maximum 1 Maximum 1 Case l l l Temperature I Pressurel Numberl l l Compartment 1 l Compartment 2 (K) 1 (atm) l l I I I l l lOomp. 1 l Comp. 2I I 'l I
I l l N2 O2 H2 HO 2 ' N2 02 H2 H2O l 1
l I
1 1 i ! I i i I i l i l I la - 8 l 1 22 l.6930 1.0748 l .0740 1 .1582 .4194 f.0430 .2887 1.24901 1362 5.438 1 I I I I I I l i I I IB - 9 l 1 21 l.7020 1.0723 l .0768 l .1489 .4658 1.0451 .2194 1.26971 1345 5.762 1 1 I I I I l 1 1 I I 1.7455 1.0417 I .0251 I .1876 .5373 1.0324 1.0775 IB- 10 2 21 .35291 1342 '
5.271 I I i l I l I I I i I l
33
Ign6 tion Limit - 10% H,
, Propogetion Limit - 10 % H, Combustion Completeness - 100 %
Flame Speed - 1.8288 m/s S'speays - AUTO e.s -
4- a b
b 3.s -
g
-o
. 1-E
! 2.s -
a-1.s -
t-o.5 - ..- ..-- .- .-- . ___. . .___.
o 400 000 4200 IEG) aEC ..2g00 2s00 3J00 Mao ._
4000 .,-.
440C 4aan T6 to.cond.1 Figure 2.7. Case Ill, Compartment 1 Pressure l
Ignition Limit - 10 % H, Propegation Limit - 10 % H, Combustion Completeness - 100 %
Flame Speed - 1.8288 m/s S"Spreys - AUTO 9.s -
- 4-3.s -
g
.o.
e 3' b
l a.s -
a-i.s -
t o.s .- .._. . _ _...__.. _,___ _
o too ano.. nao non an_.__.._
am anno nao Mao tono .,4ao._ 4eno T. t .condet Figure 2.8. Case B2, Compartment 1 Pressure 34
isso , , , ,
Ignition Limit - 10% H, Propegotion Limit - 10 % H, Con.bustion Completeness - 100 %
'*" ' Flame Speed - 1.8288 m/s Spreys - AUTO
_ SM <
.5 2
g isse.
_I l ses -
e..
M*
aaa a samt a sansenna
- 2. .
e ens eso sano sans asse see Mo uns uso eens esso es.
Time toecondet Figure 2.9. Case BT, Compartment 1 Temperature Ignition Limit - 10 % H, g,. Propagation Limit - 10% H, Combustion Completeness - 100 %
Flame Speed - 1.8288 m/s Spreys - AUTO e.$ -
4-3.s - .
-o e :- -
h e
s.s *-
A s-e.s -
W I
i t
as o
one ens name esos asas aos asso nos aso esas esso esso i Time leecondet Figure 2.10. Case BT, Compartment 1 Pressure 35
I Ignition Limit - 10% H.
Propagation Limit - 10 % H,
" ' Combustion Completeness - 100%
Flame Speed - 1.8288 m/s o.s . Spreys - AUTO o.7 -
o.s -
n e o.s -
2 l o.e -
a o.3 -
o.2 -
- 0. 0 -
o - - ,-- .---.-- .-- .-- .-- 0-- .-- .-
o one eso.- ime isso ano 2ess amo mes 3mso eene eeso gene Time leecondel Figure 2.11. Case B2', Compartment 1 Steam Mole Fraction 8 . . . .
o.9 - .
o.e -
o.? -
5 f o.s -
A f 0.5-o.e -
A o.3 -
1 c.2 - Ignition Limit - 10 % H, Propagation Limit - 10% H,
_ Combustion Completeness - 100 %
Flame Speed - 1.8288 m/s Spreys - AUTO o - - . . .- - . .. .. .. . .. --- - --
Time to.condei Figure 2.12. Cane B2', Compartment 1 Nitrogen Mole Fraction 36
i . . . . . .
Ignition Limit - 10 % H, Propagation Limit - 10 % H, 8 ' Combustion Completeness - 100 % '
Flame Speed - 1.8288 m/s o.s , . Sprays - AUTO o., -
.$ o.s -
U e
3 e.s . -
t k e.4 - -
8 e.3 -
c.3 - -
o - - - - . - --- - - --- --- -- -- - - -
Ti e teecondel Figure 2.13. Case 112', Compartment i Oxygen Mole Fraction i . . . . . . .
'gnition Limit - 10 % H, Propagation Limit - 10 % H,
' ' ' " Combustion Completeness - 100 %
j Flame Speed - 1.8288 m/s o.s .. Sprays - AUTO o.7 -
.; o.s .
e o.s -
r o .e -
z o.3 -
c.a -
- 8. i -
o - - - - - - - --- --- --
f --- --
Ti to.c e i Figure 2.14. Case 112', Compartment I liydrogen Mole Fraction 37
isso . . . . .
lenition Umit - 10 % H, Propagation Us-It - 10 % H, Combustion Completeness - 100%
'*' ' Fleme Speed - 1.8288 m/s Spreys - AUTO
_ iJeo -
5 3
i j iano -
_h
{ m- -
o m.
eso - l ww%J u ouu ano - -
e eco eso.- saco.- in 200o anco ammo saco inao soon teco
-.. esoo Ts e leecond.1 ,
Figure 2.15. Case B2', Compartment 2 Temperature e .
i lenition Umit - 10 % H, l
, , , , , Propagation Umit - 10 % H, .
Combustion Completeness - 100 %
Flame Speed - 1.8288 m/s
' Spreys - AUTO i
4.s -
4-3.s -
.a.
e l' h
e i ! 2.5 -
a.
t i a-i.s -
e.s . - .
. - . - . 4. . .. --4--
n i.econd i
--4 .--
! Figure 2.16. Case B2', Compartment 2 Pressure 38
e . . . . . ...
lenition Limit - 10 % H, Propagation Limit - 10 % H,
" Combustion Completeness - 100 %
Flame Speed - 1.3258 m/s o.. . Spreys - AUTO o.7 -
5 e.. -
5 3 a.s -
t l o.e -
0 h e.s -- \
\
o.2 -
- 0. 4 - Q o - - . . --- . - --
Time I condel
! Figure 2.17. Case 112', Compartment 2 Steam Mole Fraction Ismtiosa Limit - 10 % H, Propagation Limit - 10 % H,
"' ~ Combustion Completeness - 100 %
Flame Speed - 1.8288 m/s e.e . Spreys - AUTO 0.7 -
5, p .-
j o.s-fe-A l l 0.5 -
0.2 -
- e. .
e,. . . . .. -
Time f econden j Figure 2.18. Case B2', Compartment 2 Nitrogen Mole Fraction 39
i '
I . . . . -
Ignition Umit - 10 % H, Propagation Limit - 10% H, Combustion Completeness - 100%
Flame Speed - 1.8288 m/s e,. . . Spreys - AUTO o.r -
! 5
! j o.s -
8 6
3 a.s --
l 8 k e.. .
1 a e.3 -
l l o.a .
l o.i .
l o - - - - .-- - - - -- -- --- -- --- --- -
L e leecondel Figure 2.19. Case B2', Compartment 2 Oxygen Mole Fraction i . . .
j Ignition Limit - 10 % H, Propagetion Limit - 10% H, 8 ' Combustion Completeness - 100 % ~
l Flame Speed - 1.8288 m/s c.s - Spreys - AUTO -
)
i 1
a.r - -
1 5
c.s - - '
t 2
j o.s-c .e -
t r I o.s .
o.a - .
- o. i -
! o -
! o eos ,eso sano.- isso anno -.anos anon.- une isso esos .- eeoo esos b e leecondel Figure 2.20. Case B2', Compartment 2 Hydrogen Mole Fraction 40 l
l
1 l
iam . . .
Ignition Limit - . 10 % H, Propagation Limit - 10 % H, Combustion Completeness - 100 %
- * * ' Flame Speed - 1.8288 m/s Sprays - AUTO iano -
_.5 3
i j mo-
_I
$ am.
1
.m- \ ,
o ,
snoo___,-. ,- eson o eco,. eco saco ison ,...,___,___,20oo acon ao __,saac ecos eeno Ti.e (seconden Figure 2.11. Case B2, Compartment 2 Temperature e . . . . . .
Ignition Umit - 8 % H,
"" Propagation Umit - 8 % H, j Combustion Completeness - 85 %
Flame Speed - 1.8288 m/s 5' Sprays - AUTO e.s -
e g-e 1
- s.s -
E t 5
. 3-h 4
2 a.s -
a-t.s -
os .._.,-__,_._,.-_,--_.
la00 1400 2000 ae00 ae00 320D 1600 _,_
O e00 000.. 4000 seco... esce T wo is.conds:
Figure 2.22. Case B3, Compartment 1 Pressure 41
lenition Umit - 8 % H,
,,,, , Propogetion Limit - 8 % H, Combustion Completeness - 85 %
Flame Speed - 1.8288 m/s S-'Spreys - OFF 4.s =
o g.
e 3.s -
g
-5
. s-h t a.s -
n.
,}
i a-I.s -
1 0.s --- - --
T6=e (second I Figure 2.23. Case AI, Compartment 1 Pressure e . . .
Ignition Umit - 8 % H,
, , , . ,Propagetion Limit - 8 % H, Combustion Completeness - 85 %
Flame Speed - 1.8288 m/s 5"Spreys - OFF 4.s -
- 4-s.s -
g
-5 e S" h
t 2.s -
e.
g.
i.s -
l 5
as _.- :- ..- .-
0 400 GOD,- 130D les 200D ..2400 250_.___,__
3300 M.,___....,_ 4810 4400 4GD0 Time ts.condet Figure 2.24. Case B4, Compartment 1 Pressure 42
e u . s y T Ignition Limit - 8 % H,
, Combustion Completeness - 100 %
Flame Speed - F(% H,)
Sprays - OFF s.
4.s -
S q.
e 3.s -
g
_5 e 1' s
a.s -
L i
a-1.s -
t as .
..- ..-..l-- ....l-o ..eco soo saco.- ison acoo agoo asao iaos isoo acco eeoo ,- econ i m iseconden Figure 2.25. Case A2, Compartment 1 Pressure inao . , , . ,
Ignition Limit - 10 % H, Combustion Completenees - 100 %
Flame Speed - F(% H,)
l'E' Sprays - OFF
_ 12.o -
.5 3
i j tono-
_h
{ ago -
5
- e.o .
eco -
ago ,---l...l-
.. . .- - . -l. .--.l-- .-.
o eco eso taco seco.- anno no aseo saco isso eeno seco econ Time Isecondel Figure 2.26. Case A6, Compartment 1 Temperature 43
- l
4 l e . . .
t Ignition Limit - 10 % H,
' ' ' ' ' combustion Coms.Ntoness - 100 %
Flame Speed - F(%H,)
j
- Spreys - OFF i 5-i
..s .
g s.s -
, -5 l e 1-t s.s -
N.
l 4-i.s - .
'~
g-l i * o - ,6a- eaa 6& sW aka a 6~k & ~ seca h te.coads:
Figure 2.27. Case A6 Compartment 1 Pressure 1
Ignition Limit - 10 $ H, Combustion Completenese - 100 %
8 Flame Speed - F(% H,)
Spreys - OFF o.e -
c.r -
5 3 o.s -
8 e
3 o.s -
2 I a., -
5 o.3 -
o.a -
- o. i-I o -
g--- g---g --g44 ,
n Esecondel Figure 2.28. Case A6, Compartment 1 Steam Mole Fraction :
1 44 i
1 i . . .
i Ignition Limit - 10 % H, 1 Combustion Comptetoness - 100 % .
- * ' " Flame Speed - Fl% H,)
Spreys - OFF o.e -
o.r -
5 4
y o.s-e j o.s -
c.e -
t a
o.3 -
o.2 -
- o. I -
o 'edo 85o tJG3 16iIl Jtbo 29ilo 250 1[oo Seco toIlo etil0 esoo T6 to.conden Figure 2.29. Case A6, Compartment 1 Nitrogen Mole Fraction I . . . . , , ,
Ignition Limit - 10 % H, Combustion Completeness - 100 %
" Flame Speed - F(% Hz)
Spreys - OFF o.s -
c.s -
5
's o.s -
8 2
3 o.s - -
t h o.e -
0 I o.3 -
o.2 - --
l
- o. i - 3 L
'o eco eio iso ne'an 2mo 24o aeIoo sso is' ao~ so'on esiin esoo T6me (secondsl Figure 2.30. Case A6, Compartment 1 Oxygen Mole Fraction 45
i , , , , , , ,
Ignition Limit - 10% H.
Combustion Completeness - 100 %
8 '
' Flame Speed - F( % H,)
Spreys - OFF o.S -
c.? -
k y o.s -
e j o.s.
o t..
I o.3 -
o.a -
c.1 - j o
o - _.- _- ,
_ _ ... _ D_ -_ .- _.- _.._..- _
T6 teecoadst Figure 2.31. Case A6, Compartment 1 Hydrogen Mole Fraction s , , .
lenition Limit - 8 % Ha
,,,, , Combustion Completeness - 100 %
Flame Speed - F(% H,)
Spreys - ON s-e.s -
- 4 s.s -
g
.o.
. 3" h
? s.s -
S-a- {
i.s -
g-c.s -
o <as eso.- -.-uno seno amo aes asso mas neo .eas e a.
Ts te.condet Figure 2.32. Case A3, Compartment 1 Pressure 46
s-- . , . .
Ignition Limit - 10 % H,
,,,, , Combustion Completeness - 100 %
Flame Speed - F(% H,)
Sprevs - ON s-4.3 g s.s-5
. 3--
h l
L t 2.s -
4 a-t.s -
~
-I o.s . ..- -
4 400 000 1200.- IGG) 2000 2400 JE00 1J00 M00 4000.. 4400.- 4000 Ti.e Isecondes Figure 2.33. Case A7, Compartment 1 Pressure e . . . .
Ignition Limit - 6 % H, Combustion Completeness - 100%
, Fla.no Speed - F( % H3 )
Spreys - ON s-4.5 - -
g 3.5 -
-5
, 1-L 22.s-a-
i .s -
r o.s - -.- .,- ...-..-- .....-...- .- .. ..-
0 900 000 t200.- 1600 2000 2400 2000 3J00 M00 4000 4400 4000 T - Iseconds Figure 2.34. Case A10, Compartment 1 Pressure 47
lenition Limit - 7 % H,
,,,, , Combustion Completeness - 100 %
Flame Speed - F(%H,)
Sprays - ON s-e.s -
~
e .-
1 -
3.5 -
p 5
. 3-h 6
i 2.s -
J-1.5 -
s- _-
r C.s - - . - --,-- ,-- .- 4000 -,--
4400 4000 0 900 000 - 0 00,- 16 33... 2000 2400.-- .2E00 1200 16o0 T .. Is conds Figure 2.35. Case All, Compartment 1 Pressure Ignition Limit - 9 % H,
, Combustion Completenese - 100 %
Flame Speed - F(% H,)
Spreys - ON s-1.5 -
- 4-g 3.s -
_o
. s-h A
22.5-I s.
i.s -
-I 0.5 - -
geoo.- ee0s 0 -.400 se0.- u00 ison anon aeon aene saco is00 ecco T6 i.ecoadet Figure 2.36. Case A12, Compartment 1 Pressure 48
. . . . . . . . . .. . m
s . . , , ,
, Ignition Limit - 8 % H,
,,,. Combustion Completeness - 100 %
Flame Speed - F(% H,)
Spreys - AUTO s-4.s -
. 4-g z.s -
e l' h
L i 2.s -
a-
)
s.s -
E s I
as -
0 400 000,- ,1200 12 2000 2400 2000 3200 % 00 4000 4400 4000 h.. is.condil Figure 2.37. Case A4, Compartment 1 Pressure im , , . , , , , ,
Ignition Limit - 10 % H, Combustion Completeness - 100%
Flame Speed - F(% H3) 5*- Spreys - AUTO uno -
_.5 3
i j inso-h
{ sno -
5 1
Q*
ano - . . -
.-g -g . -- --- - -- -
.g n te.condel l Figure 2.38. Case A8, Compartment 1 Temperature 49 I
4 . . . .
lenition Limit - 10% H,
,,,, , Combustion Completeness - 100%
Flame Speed - F(%H,)
Spreys - AUTO g..
9.S -
4-k .
g 3.s-
.t e 3'
? a.s -
a-
\
1.5 -
1 0s,- ,;,-
,g, ;g .
T,=e to.cond i Figure 2.39. Case A8, Compartment 1 Pressure i . . . .
Ignition Limit - 10% H, Combustion Completeness - 100 %
8*" Flome Speed - F(%H,)
Spreys - AUTO 0.6 -
0.7 -
5 30.s-8
.l:
e 0.s -
I l o.e -
5 0.3 - t o.a -
(
- 0. 5 -
8,- ,;, - ,j,- ;j, g g.g. . . - _ _
T6-e ie.cond.:
Figure 2.40. Case A8, Compartment 1 Steam Mole Fraction 50
I l
1 ,
1 l
8 . . . . . .
Ignition Limit - 10 % H, ;
Combustion Completeness - 100 %
' Flame Speed - F( % H,)
Sprays - AUTO o.e -
o.F - -
- j o.s -
d j o.s- [
i- e.4 -
h 4
o.3 -
i 0.3 -
I
- o. I -
o -
o aos ano.- im.. iaan sono asas asno .__ .sano mo .ano siao... eeno T i.econdei Figure 2.81. Case A8, Compartment 1 Nitrogen Mole Fraction I
i
( . .
lenition Limit - 10 % H, Combustion Completeness - 100%
8'
" F1sme Speed - F( % H,)
Sprays - AUTO o.e -
o.f -
5 j o.s -
8 4
3 o.s -
t k o,e--
5 o.3 --
o.a - -
- o. I -
o - - - - . - . _ .. __. _
Time isecor. dei Figure 2.42. Case A8, Compartment 1 Otygen Mole Fraction 51
i . . .
Ignition Limit - 10 % H, Combustion Completeness - 100%
8'" Flame Speed - F(% H,)
Spreys - AUTO o.e -
c.7 -
.y as-6 j o.s -
- c. < -
t I
a.3 -
e.a -
- e. t-o -:- .---. .-..
..:ao'esoo o en eso.- uan iean amo a<m amo.- non maa eace T (..cond i Figure 2.43. Case A8, Compartment 1 Hydrogen Mole Fraction s . . . .
Ignition Limit - 8 % H,
,,,, , Combustion Completeness - 100 %
Flame Speed - F(% H,)
Spreys - AUTO s-
..s -
1,,.s .
~o l
, s-h L
i s.s -
4 a-a.s - v l-c.s - ..- - .
o eco eso.- uoo.- ison,-- amo.- a<m--..amo nos maa econ.- ..oo esao T6 't .eondei Figure 2.44. Case A5, Compartment 1 Pressure 52
l l
s . , , , ,
Ignition Limit - 10 % H, Combustion Completeness - 100 %
, Flame Speed - F( % H,)
Spreys - AUTO s-4.s -
- 4 1,.-
5 e 1*
h 22.s-0.
l 2-i.s - N - -
t-0.5 -
0 400 000,. 1200 5500,-- 2000, - .--
2400 2e)3.-- 1200 -,360L. 4000 4400 4000 Ts e fsecondst Figure 2.45. Case A9, Compartment 1 Pressure mio . . . . , ,
Ignition Limit - 8 % H, Propagation Limit - G(% H,)
Combustion Completeness - 100 %
- 4a0 - Flame Speed - F(% H 28
- Spreys - AUTO
_ saa0 -
5 i
j t000 -
-h l m- -
o I
too -
i rw, . , ,,a Q
ae0 - ..- -.. .... - ....._...-__,_ . .. .: _
o 400 s00 tJ00 1500 250 2000 2000 3200 3000 4GD0 .4400
. . 4000 Time (seconden Figure 2.46. Case B5, Compartment 1 Temperature 53
e . .
lenition Limit - 8 % H,
, , , . . Propegation Limit . G(% H,)
Combustion Completeness - 100%
Flame Speed - F(% H,)
" Spreys - AUTO e.s -
7 4 s.s -
g 5
e 3*
h L
! 2.s -
a-t.s - .
I o.s . - -
Tre (..condet Figure 2.47. Case B5, Compartment 1 Pressure i . . . . .
Ignition Limit - 8 % H, Propegation Limit - Gl% H,)
' Combustion Completeness - 100 %
Flame Speed - F(% H,)
o,. .. Spreys - AUTO ,
o.r - .
5 a.s - ;
8 d '
e o.s .
l o.e .
n o.3 - .
o.2 -
i *
- o. i -
I e -
,eos eso.- - 2o00 m Jean -
O saco- .seco uno-- .1sso sono em esoo Ts t econdet Figure 2.48. Case B5, Compartment 1 Steam Mole Fraction l 54 l l
)
i . . .
Ignition Umst - 8 % H, Propogotion Umit - G(% H,)
" Combustion Completenese - 100 % ;
Flame Speed - F(%H,)
o.e Spreys - AUTO o.r .
.j o.s .
d e
j o.s --
f a.e.
z o.3 -
o.2 -
- o. 8 -
o eso eio siao iam aabo ~24o ~amo sabo m'o a oo'on 44o esoo T6=e feecondel Figure 2.49. Case B5, Compartment 1 Nitrogen Mole Fraction i . . . .
Ignition Umit - 8 % H, Propogetion Umit - G(% H,)
8' ' Combustion Completeness - 100 %
Flame Speed - F(% H,)
,,,. .Spreys - AUTO o.r -
5 4 a.s .
8 6
3 a.s -
t k o.e .
4 o.3 o.2 -
- o. s -
f o,. -f- - - .- -. -- -.- - ___ _. __
T6 reecond.1 Figure 2.50. Case B5, Compartment 1 Oxygen Mole Fraction l
l l
- 55 l
l l
i . . , .
Ignition Limit - 8 % H, Propagation Limit - O(% Hal
' Combustion Completeness - 100%
Flame Speed - F(% H,)
o.e -. Spreys - AUTO r
0.r -
.y o.s -
e e
j o.s -
r r
o .4 -
0.3 -
0.3 -
- o. t -
i o ,- ..-- C -_ ,_ _,__ ..._ _..:_
0 900 ..000 1300 1GGI 2000 2900 2000 3J00 3600 4000 4400 4000
' Time fsecondst l Figure 2.51. Case B5, Compartment 1 Hydrogen Mole Fraction ita . . . ,
Ignition Limit - 8 % H, Propagation Limit - G(% H,)
{ Combustion Completeness - 100 %
" " " Flame Speed - F(% H,)
l Spreys - AUTO i
! iJoO -
C
'i 2 ,e
! j ion 0 - ,
f
$ so0 - ,
5
- 1
\ l
~ ~~
.i t \
\\\ '
~
a- -
i as0 - - - - - -. -- ... -- --. -. -. . -..
T6=e (secondel Figure 2.52. Case B5, Compartment 2 Temperature 56 l
_u -- ' ~ w"
e , , , ,
Ignition Limit - 8 % H, i l
, , , , , Propagation Limit - G(%H,)
Combustion Completeness - 100 % l Flame Speed - F(% H,)
' Sprays - AUTO i.s -
4-1 g 3.s -
-o e '-
h I I A
t 2.s -
s-i.s -
i .
o.s - - - - - -- --- -- -- -- - - --
Ts i.econdit Figure 2.53. Case H5, Compartment 2 Pressure i .
Ignition Limit - 8 % H, Propagation Limit - G(% H,)
'd "
Combustion Completeness - 100 %
Flame Speed - F(% H,)
o.s .. Sprays - AUTO o.7 -
5 o.s -
8 A
e o.s -
2 l o.e -
5 lll' AkT(lQ o.a - i
- o. i -
q ,
i o I- - ---
- - g --
-g- - - - - .g - l Ti.e to.cond 1 Figure 2.54. Case B5, Compartment 2 Steam Mole Fraction 57
i Ignition Umit - 8 % H,
- o. . Propegation Umit - Gl% H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,)
88' ' Spreys - AUTO o.7 - -
- o.s -
e o.s - -
f o.4 -
z o.3 -
o.3 -
- o. s - -
o - - - - - - - --- --
Ti.e (seconds)
Figure 2.55. Case B5, Compartment 2 Nitrogen Mole Fraction i . , , , , ,
Ignition Umit - 8 % H, Propogetion Umit - G(% H,)
88' Combustion Completeniess - 100 %
Fleme Speed - F(% H,)
, , , . . Spreys - AUTO o.7 -
5
'4 o.s -
8 6
3 o.s .
I o.e - -
c3 o.s -
c.a -
- c. e -
o -
1200 1000 o 400 eso.- 3010 Mos asco 1300-- inoo ecos egeo,-- esos Time teecondel i
t Figure 2.56. Case B5, Compartment 2 Oxygen Mole Fraction 58
l l
Ignition Limit - 8 % H, Propagetion Limit - G(% H,)
" Combustion Compfsteness - 100 %
Flame Speed - F(% H,) '
o.s .. Sprays - AUTO o.7 -
y o.s -
4 j o.s -
o.e -
r t
o.3 -
o.2 -
- o. i -
N N o,- - - - - - - -
g _-
T .. to.condst Figure 2.57. Case 115, Compartment 2 Ilydrogen Mole Fraction
~
Ignition Limit - 10N
"' Propagation Limit - Gl% H,)
' Combustion Completeness - 100 %
Flame Speed - F(%H,)
5" Spreys - ON t.s -
T. ,,
s.s -
~
o
, s-h 4.
22.s-a-
- s. ?
o.s - - - . .. -- . . -. .
Ti.e leecondel Figure 2.58. Case 116, Compartment 1 Pressure 59
s ,
ignition Limit - 10 %
s., . Propogetion Limit - G(%H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,)
Spreys - ON t.s -
q.
o.
3.s -
g
-5 e 1-4
! a.s-E a-1.s -
i .N o.s ,--- -- -;g;g g g g g g g g 4 g g m T6=e teecondel Figure 2.59. Case B6', Compartment 2 Pressure ignition Nmit - 10 %
Propegation Limit - Gl% H,)
' Combustion Completeness - 100 %
Flame Speed - F( % H,)
s- Spreys - AUTO 1.s -
e e.
e 1
3.s -
.t.
. 2-h L
t a.s -
/
a-t.s -
6-o.s g- g- - g g- g g g g-Ts t ecoadei Figure 2.60. Case B7, Compartment 2 Pressure 60
1 l
l l
6 , . ,
Ignition Limit - 8 % H,
,,,, Propagation Limit - G(% H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,)
5'
' Sprays - AUTO 4.s -
I g 4-
)
f 3.s -
, 1-L i 2.s -
t-E 0.s - - - - - - - - - -- --- --- - - - - -
T... ts.coadst Figure 2.61. Case Cl, Compartment 1 Pressure Ignition Limit - 10 % H,
, , , , , Propagation Limit - G(% H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,) {
'" Sprays - ON l l
4.s -
- i. .-
g 3.s -
5
. 3-A t a.s -
t-
,. c o.s -
.. ..-..l-.. ... __.: .. l- .. ._ -l 0 400 .000 tJ00.- 16G) 203) 2400 2000 3200 % 00 4000 4400 4000 Ts-e isecoadsl Figure 2.62. Case C2, Compartment 1 Pressure 61
l 16 3 , ., , , i
' 9 lenition Limit - 10% H, Propegation Limit - G(% H,)
Combustion Completonses - 100%
""" Flame Speed - F(%H,)
Spreys - AU10
_im-5 i
{ v.u-b
_4
$ eco-n L
h e
W gg .
9 m
- " a ,6s eio iion isc aim ~as ~asino sioo ~ W W ~ ~ek ' ~ om s
T6 e toecondst Figure 2.63. Case C3, Compartment 1 Temperature s , , , , , , ,
Ignition Limit - 10 % H,
, , , , ,Propogetion Limit - G(% H2)
Combustion Completeness - 100 %
Flame Speed - F(%H,)
5" Spreys - AUTO 4.s -
- 4-e s.s .
g
-t o I' h
,t a.s -
a-o.s , - - - - ;g g g- g- g g- g g ,
Time leecondel Figure 2.64. Case C3, Compartment 1 Pressure 62
. W Ignition Umit - 10 % H.
Propogetion Limit - Gl%H )
" " Combustion Completenese2 - 100 %
Flame Speed - F(%Hz) c.e - Spreys - AUTO c.7 -
6
- a.s -
8 e
e o.s -
a s
l o. -
s 0.3 -
o.2 -- .
o - - - - . . - - ... - .. _. _- ...
Ti== teeconds!
Figure 2.65. Case C3, Compartment 1 Steam Mole Fraction Ignition Umit - 10% H Propogetion Umit - G(%Hal
' Combustion Completenese - 100 %
Flemo Speed - F(%H,)
c.e -.Spreys - AUTO o.r .
b Y y a.s .
& G.s =
s
, o. < -
1 x
0.3 - ,
0.2 -
0.1 -
a - - -. .-. --
Time (seconden Figure 2.66. Case C3, Compartment 1 Nitrogen Mole Fraction 63
i . . , .
Ignition Limit - 10 % H, Propogetion Limit - Gl%H,)
- * ' " Combustion Completeness - 100%
Flame Speed - F(%H,)
o.s . . Spreys - AUTO
- o. , --
5
$ 3.5 -
i 8 e
.5 o.s -
I' E
g o.. -
5
%3-J a.a -
~
- o. i -
1 a ~ eso ado iabo tehe aabo abo sabo ubo nbo iobo .ebo .soo Tia. Isecowei Figure 2.67. Case C3, Compartment 1 Oxygen Mole Fraction i . . . .
Ignition Limit - 10 % H, Prcpegetion Limit - G(% H,)
" Combustion Completeness - 100%
Flame Speed - F(%H,)
o.e . . Spreys - AUTO o.' --
5 e
y o.s -
A f o.s <
o.e -
t r
o.3 -
o.3 =
o.1 -
o
' b #
o aio aio imho ieio acon abe ambo sabo nbo uso i.bo . eon Ts e teecondet Figure 2.68. Case C3, Compartment 1 Hydrogen Mole Fraction 64
isoo ,
lenition Limit - 10 % H, Propagation Limit - G(% H,)
Combustion Completeness - 100 %
"'"~
~ Fleme Speed - Ft % H,)
Spreys - AUTO
_ n 200 -
5 3
i j icoo. .
b, 4
$ soo-5 k
.ao .
I
.0o -
m, -
g- - - - -- -
Ts Is ,condst Figure 2.69. Case C3, Compartment 2 Temperature Ignition Limit - 10% H,
,,,.. Propegation Limit - G( % H,)
Combustion Completeness - 100%
Flame Speed - Fl% H,)
Spreys - AUTO 4.s -
~*
e <-
1 g 3.s -
e 3' h
- s.s-1 2 I
~
0.5 - - - - - - - -- - - --- -
Time f.econden Figure 2.70. Case C3, Compartment 2 Pressure 65
a . . -
Ignition Limit - 10 % H, Propegation Lani: - G(% H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,)
o.e . .Spreys - AUTO o.1 -
5
- e.s -
8 i
e o.s -
t k o.4 -
a
- o. s .
o.2 -
x( x
- o. e -
(
o - -
g- g- g g- --g 4- - g g 4 A T..e i s.cond.I Figure 2.71. Case C3 Compartment 2 Steam Mole Fraction a . . . . . . .
lenition Limit - 10 % H, Propogetion Limit - G(% H,1
' ' ' Combustion Completenese - 100%
Flame Speed - F(%H3) o.. . .Spreye - AUTO
- 0. 7 -
k g o.s .
A f o.s o.e -
b 4
o.3 -
o.2 -
c.1 -
o -
Time leeco Wol Figure 2.72. Case C3, Compar* ment 2 Nitrogen Mole Fraction 66
lenition Umit - 10 % H.
Propagetion Limit - G(% H,)
' Combustion Completences - 100 %
Flame Speed - F(% H,)
0.. -.Spraye - AUTO 0.7 --
8' 0.t. -
8 e
3 e.s -
1 C
h 0.4 -
! 5
- 0. s -
0.2 -
- 0. t -
o - - - - - - - --- - -- - ---
Time leecondel Figure 2.73. Case C3, Compartment 2 Oxygen Mole Fraction n ,
Ignition Umit - 10 % H.
Propegation Umit - Gl% H,)
" Combustion Completeness - 100%
Flema Speed - F(% H,)
0.e -.Spreye - AUTO -
0.r -
5 J
g o.a .
t j o.s -
b 0.. -
sz 0.3 -
0.2 -
- 0. i -
j 0
/ '
/' ' '
O 450 e50 h$lo leiD ~ 20il0 2eim ~ JIIOD 32hD 3800 4Qbo oth0 4a00 Time te.condei Figure 2.74. Case C3, Compartment 2 Hyurogen Mole Fraction 67
lenition Limit - 8 % H,
,,,, , Propogetion Limit - G(% H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,)
S' ' Spreys - OFF 9.S -
e 4 e
1,,,.
_o e 3" h
A 2 a.s -
a-t.S a t
o.5 , - - - - - - - - - g---g-- -- g---g- g- g- ,
T6 e (secondel Figure 2.75. Case C4, Compartment 1 Pressure s . , ,
Ignition Limit - 8 % H,
,,,, , Propagetion Limit - G(% H,)
Combustion Completeness - 100%
Flame Speed - F(% H,)
S'
' Spreys - AUTO
- 9.S -
1 g s.s --
~5
. 3 l
e.
t s.s -
I a-t.S -
0.5 - - - - - - - - --- --- --- --- --- --- - - ---
Itso Isocondel Figure 2.76. Case D1, Compartment 1 Pressure 68
Ignition Limit - 10 % H, Propagation Limit - G(%H,)
Combustion Completeness - 100%
Flame Speed - F(% H,)
Sprays - AUTO ianc -
_5 2
j icoo-h
_4
{ eco -
1
.. . ..e $
aan - ..- :-- . .-.
o eco .eco taco,- isco,-- aaao
,. - am - ,--amo.-- was .- isao eaan giao esoo Ts-e teeconds Figure 2.77. Case D2, Compartment 1 Temperature e
j Ignition Limit - 10 % H,
,,,. . Propagation Limit - Gl% H,)
Combustion Completeness - 100 %
'" Flame Speed - F(%H,)
Sprays - AUTO 1.s -
3.s .
. 2-4 L
i 2.s -
t-i.s - O i
0.5 - - -
T6.. te.conds:
Figure 2.78. Case D2, Compartment 1 Pressure 69
I , . . .
Ignition Umit - 10 % H, Propogetion Umit - G(% H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,)
o.e -. Spreys - AUTO o.1 -
5'
's o.s --
8 a
3 o.s -
I h e.< -
c5 o.3 -
o.a -
! o. : --
i a -
o eco ano sano.. ...-
isoo aceo.--.._.
m asco .saac- ,___..
zoo uno._ .,4ao._ esco Time isecoadst Figure 2.7g. Case D2, Compartment i Oxygen Mole Fraction i . . . . . .
lenition Umit - 10 % H, Propogetion Umit - G(% H,)
Combustion Completeness - 100%
l Flame Speed - F(% H 3) o.e -. Spreys - AUTO o.r -
.y o.. .
e 3 o.s - l I 1 0 .4 -
I l
o.3 -
o.a .
- o. I --
o . . .. . .
,.-r...
T6=e (secondet Figure 2.80. Case D2, Compartment 1 Hydrogen Mole Fraction 1
I 70
isso . . . .
lenition Limit - 10% H, Propogetion Limit - Gl%H,)
Combustion Completenese - 100 %
888 " Flame Speed - F(% H,)
Spreys - AUTO
_ saso -
.5 3
i i _ j
_t
$ eso-o I.
400 -
200 - - - - - - --- --- --- --- --- --- --- - - --
n t .cond.i Figure 2.81. Case D2, Compartment 2 Temperature Ignition Limit - 10 % H,
, , , , , Propogetion Limit - G(%H,)
Combustion Completenose - 100%
Flame Speed - F(% H,)
Spreys - AUTO e.s -
o q.
3.5 -
-5
, 3-h E
2 a.s -
a-i.s - O 0.5 - - - - - - --- --- --- -- --- --- --- - --
n te.condet Figure 2.82. Case D2, Compartment 2 Pressure j 71
e . . . .
Ignition Limit - 10 % H, Propagetion Limit - G( % H,)
' Combustion Completeness - 100 %
Flame Speed - F(% H,)
o.s .. Spreys - AUTO o.7 -
5
- o.s -
8 e
3 a.s -
t k o.e -
0 c.s --
e.2 -
- o. 5 -
o g-- f--- g- g---f- ,g-T W te.coadet Figure 2.83. Case D2, Compartment 2 Oxygen Mole Fraction i . . . . . .
Ignition Limit - 10 % H, Propagetion Limit - Gl% H3)
" Combustion Ccmpleteness - 100 %
Flame Speed - F(% H,)
, , . . Sprays - AUTO o.r -
j o.s --
e j o.s-f o.e .
t x
o.3 - (
1 o.3 -
- o. i -
j
- ~
o one aio iloo isho ~ aso ~ W ~ aeIno W inao eene es seco T. i..cond.1 Figure 2.84. Case D2, Compartment 2 Hydrogen Mole Fraction 72
isso . . . . .
lenition Umit - 10 % H, Propogetion Umit - G(% H,)
Combustion Completeness - 100 %
8*' ' Flame Speed - F(% H,)
Spreys - AUTO
,, tase .
.5
- t
.s g teso -
I E ess-g ano , - - - - - --- -- --- --- --
---f---f---4--
Ti te.condei Figure 2.85. Case D2, Compartment 3 Temperature s . . . .
lenition Umit - 10 % H,
,Propogation Umit - G(% H,)
Combustion Completeness - 100 %
Flame Spee<* - F(% H,)
'" Spreys - AUTO t.s -
"is ,.
s.s -
. s-L
! 2.s -
I i g- I a.s -
e.g - _ - - - - - - --- --- -- - --- --- --- --
Time teocondel Figure 2.86. Case D2, Compartment 3 Pressure 73 l
l
e . . . . . . . . .
lenition Limit - 10% H, Propogetion Limit - G(% H,) ~
8 Combustion Completenees - 100 %
Flame Speed - F(% H,)
e,s ..Spreys - AUTO o.r -
.k. o.s - -
3 e.s -
I k o.e-5 e.3 -
o.a -
- 0. l -
e,- -
g--,g g --g---g --g - g g-- g - -, '
Time toecondel Figure 2.87. Case D2, Compartment 3 Oxygen Mole Fraction lenition Limit - 10 %' H, Propegation Limit - G(%H,)
8 Combustion Completeness - 100 %
Flame Speed - F(% H,)
u- . Spreys - AUTO o.r .
! e.o- ///
i -- -
f z
e..a o.3 -
e.a -
- 0. l -
o o
ese eso isso use . ..-- asse w
-Sb .--
-- .- ....+esoesso m nos isso esas Time I % m l Figure 2.88. Cr.se D2, Compartment 3 Hydrogen Mole Fraction 74
i l
l s . , , , , ,
Ignition Limit - 8 % h,
, Propagation Limit - Gl% H 3)
Combustion Completeness - 100%
Flame Speed - F( % H,)
5' Sprays - OFF 1.5 -
~
e e-ig 3.s -
-5 I;
. v h l
- l 22.5-a.
{
i a-i l \ '
i.s - \
l t
o.s - - -
o a son in,_ ..m ao_,_ __, _ mean_....,non a ao man eaos ..oo,_ eson Ti I..cond Figure 2.89. Case D3, Compartment 1 Pressure isoo . . . .
Ignition Limit - 10 % H, Propagation Limit - Gl% H,)
Combustion Completeness - 100 %
l'" ' Flame Speed - F(% H 3)
- Sprays - AUTO
_ iano - ,
.5 3
i j iono-
_h
$ eaa -
5 400 - ,
._ g ':' C C :C.'. ,, .':; M%
200
.- ..- aacoo ,aeos __,_ __:_mao .,_ eooo
.,_ _es_ .6 soo,--
o a saco anco nas Ti (..cond.:
Figure 2.90. Case Et, Compartment 1 Temperature 75 l
e . . . . .
lenition Limit - 10 % H, Propegation Limit - Gl% H,)
Combustion Completeness - 100%
Flame Speed - Fl% H,)
S- Sprays - AUTO e.s -
o q-3.s --
g
_5
, 3-h E
i s.s -
a-o.s - - - -
g-g- g- g g - g- g- 4- 4- A T .e fsecondel Fiqure 2.91. Case El, Compartment 1 Pressure i . . . . . .
f Ignition LIrtit - 10 % H, Propagation Limit - G(% H,)
0'" Combustion Completsness - 100 %
Flame Speed - F( % H,)
c.e - Spreys - AUTO 0.7 -
5 j o.s -
8 4
3 o.s -
2 k o.4 -
8 o.3 -
c.a -
~
- o. i -
v o,- - - - ;g g- g- g- g- g- g- g ,
Time Isocoadel Figure 2.92. Case Et, Compartment 1 Oxygen Mole Fraction 76
r-I - . .. . .
Ignition Limit - 10 % H, Propogetion Limit - G(% H2) 8' " Combustion Completeness - 100 %
Flame Speed - F(% H,)
o.s . Spreys - AUTO
. J a.1 --
.y a.s -
e o.s -
t .-
o z
o.3 -
o.a -
- o. a -
o ~ ~
o ~e do~ edo~ tion ~
is ho mobo aebo asbo sion abo ~ echo ee'o c geoo T6.. Is condst Figure 2.93. Case Et, Compartment 1 Hydrogen Mole Fraction ism . . . .
Ignition Limit - 10 % H, Propegation Limit - G(% H2)
Combustien Completeness - 100%
8+'D " Flame Speed - F( % H,) I "
Spreys - AUTO l
_ i200 -
.5
.5 i
j ioco-
.I I
l-
$ eac-f h
- se-( I 1
LI em. W d1 (
s J(
>= .- .---:-- ....:_ ..__.:- .:-
o 400 000.- ...
sacotsco aano aeon asco saco woo emo eeoo.- esoo T6 e Isoconds!
Figure 2.94. Case Et, Compartment 2 Temperature i
- 77
(
l l
S . . .
lenition Limit - 10% H,
,,,, , Propagetion Limit - G(% H,) ,
Combustion Completeness - 100 %
Flame Speed - F(% H,)
- 5"Spreys - AUTO 4.5 -
4-k -
g 3.s .
5
, s-h a.s -
L a-
+
a 1
o.s , - ,-- , - . .-, . ,-
200o.,---,.-
o (Do 800 i200 1s(D 2400 200o.,- 3200 150o 4000 4400 400 T6=. (seconds:
Figure 2.95. Case El, Compartment 2 Pressure i . . . . ,
Ignition Limit - 10 % H, Propagation Limit - G(% H,)
8 Combustion Completeness - 100 %
Flame Speed - F(% Hz) o.. . Spreys - AUTO o.7 -
5 d o.s -
8 e
3 a.s .
f h o.4 -
x a.3 -
e.a -
~
- e. i -
o a ~ 46o soo iian inbo aiuo 2.be ambo ~ 1h ' mbo ' eabo 44bo 4eco n is.conds:
Figure 2.96. Case Et, Compartment 2 Oxygen Mole Fraction 78
i . . . -
Ignition Limit - 10 % H, Propogotion Limit - G(% H,)
0 Combustion Completeness - 100 %
Flame Speed - F(% H,)
g,. . Spreys - AUTO o.7 -
e.s-
.f A
o.s -
c.e -
)
n.3 -
o.a --
- c. -
l o 46o em iie isha aabo as aeIoo sabo 3si. ea'ao * *m' sem T6.. (..cond.1 Figure 2.97. Case Et, Compartment 2 Hydrogen Mole Fraction ism . . . .
Ignition Umit - 10% H, Propagetion Umit - G(% H,)
Combustion Completeness - 100%
HW- Flame Speed - F(% H,)
Spreys - AUTO
_ saco -
5 3
i j icoo-
_h
$ aco -
o s..
.ao -
ano - - - - - - - - -- - - - --- -- -- -
T6 t .conds:
Figure 2.98. Case El, Compartment 3 Temperature 79
s , , ,
lenition Limit - 10 % H,
"" Propegation Limit - G(% H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,)
5" Sprays - AUTO e.s -
o q-e 3.s -
g
.l.i e 1" h
L 2 a.s -
a-k o.s -
o 400 000,. 1200,- 1600,--..200o 2400 200o 3200 1600 4000 440o,. 4400 Ts.e Is conds:
Figure 2.99. Case Et, Compartment 3 Pressure t , , . , ,
Ignition Limit - 10 % M, Propogetion Limit - G(% H,)
"" Combustion Completeness - 100 %
Flame Speed - F(% H,)
o.s - Sprays - AUTO o.7 -
5 j a.s - l 8
e 3 a.s -
I i k o.4 -
8 o.3 -
o.3 -
- o. t -
o.- -,.
o 400 000 iaco - ..._
isao 2000,_ _,_.2 0c.aeDo la00..,_
x00 _,_ .Doo _,_..ao esoo T6 (seconds F!gure 2.100. Case Et, Compartment 3 Oxygen Mole Fraction 80
l Ignition Limit - 10 % H, Propagation Limit - G(% H,)
0'
" Combustion Completeness - 100 %
Flame Speed - F(% H3) 0.e - Sprays - AUTO g ..
5a 0.8 -
e 0.5 -
g 0.4 -
1 0.3 -
0.2 -
0.1 -
0 -
, . ~_ _ _
0 0 .400 000 t200 lu Igo _ ho Jabo 12bo %ho' sche ag '94'o0 tu Ts.. to.cono.1 Figure 2.101. Case El, Compartment 3 Hydrogen Mole Fractwn isca . , . ,
Ignmon Limit - 10 % H, Propagation Limit - Gl% H,)
Combustion Completeness - 100 %
t*oo -
Flame Speed - F(% H,)
Sprays - AUTO iso -
_5 3
i
- m-
=
b
-.4
$ soo-v i
- m. ,
.. y1 l 200 :- _ ,
_ ,_ isco 0 *00 800 1300 2000_,_ agho atho who who iobo e,ho im T6 e Isecondst Figure 2.102. Case Et, Compartment 4 Temperature 81
e . . . . ,
lenition Limit - 10% H,
,,,, , Propagation Umit - Gl % H,) .
Combustion Completeness - 100 %
Flame Speed - F(% H 2) 5"Spreys - AUTO 1 I
..s -
I g 3.s -
5
. s- ,
l h I
- a.s-t L
l a-
%QO %
"'o edo odo sabo isho sabe aebo ~ ambo ~ ubo is'ao esoo **bo seco T+. is. cow s1 Figure 2.103. Case El, Compartment 4 Pressure ,
i i , ,
i Ignition Limit - 10 % H, Propegation Limit - G(% H3 )
Combustion Completeness - 100 %
Flame Speed - F(% H,)
o.s -.Spreys - AUTO o.r -
5 3 c.s -
8 e
3 o.s -
l t
o.e -
a l
l o.s -
l i
c.2 -
i i
- o. i -
o 450 d tailo idI) albo acIIe aiX) $ ' 35' o c4doo 44Ilo 4000 Ti toecondst Figure 2.104. Case El, Compartment 4 Oxygen Mole Fraction 82 i
l l
t , , , .
Ignition Limit - 10 % H, Propagation Limit - G(% H,)
"" Combustion Completeness - 100 %
Flame Speed - F(% H,)
o.. .. Sprays - AUTO 0.1 -
y 0.s -
t 0.s -
0 .4 -
t r
0.3 --
0.2 -
0.1 -
$r5 ON -
0 400 000 1200 le 2000 2400 2000 3200 1500 4000 4400 4000 Ts i.econdet Figure 2.105. Case El, Compartment 4 Hydrogen Mole Fraction isao , , .
Ignition Limit - 10 % H, Propagation Limit - G(% H,1 Combustion Completeness - 100 %
' ' " " Flame Speed - F(% H,)
Sprays - AUTO
, 1200 -
S 3
i 1000 -
_I
{ eoo-o L
~
-- L!
l 400 =
200 -
0 ..400 000 t200,- .-
IEal...2000 - 2400
,.- Ja)0
,. 3200 ., 3600 4000,. .,.
4400 4400 Ti.e f.econden Figure 2.106. Case El, Compartment 5 Temperature 83
l Ignition Limit - 10 % H,
- 3. .
. Propagation Limit - G(%H2) ,
Combustion Completeness - 100%
Flame Speed - F(% H,)
S"Spreys - AUTO t.s -
4-s.s -
g 5
. 3- -
t 2.s -
A a-- -
o.s -
,- _,_ _,. l_.., ._,___,. _, _
o eco eco iJoo .,ison acoo agoo asco saco inco 4000_,. .,_
egoo esoo T6.. Iseconden Figure 2.107. Case E1, Compartment 5 Pressure Ignition Limit - 10 % H, Propegation Limit - G(%H,)
Combustion Completenese - 100%
Flame Speed - F(%H 2) o.s ..Spreys - AUTO .
l o.7 -
5 d o.s -
l 8 i e 3 o.s -
2 l k o.4 -
i 5 c.3 -
c.3 -
0.1 -
o '4Eo gio IJilo he'G 3 2Glo 29ilo N M N to'oo 44I)o ecco Ti. is.conds:
Figure 2.108. Case Et, Compartment 5 Oxygen Mole Fraction 84
i , , , ,
Ignition Limit - 10 % H, Propegation Limit - G(% H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,)
...Spreys - AUTO o.r -
y o.. -
4 j o.s-c.e .
t I
o.3 --
l o.2 -
o 46o eso iiao ison aabo ~ Nm asis ~ saco iao sabo e.bo seco Time f.econa.1 Figure 2.109. Case E1, Compartment 5 Hydrogen Mole Fraction s , , ,
Ignition Limit - 10 % H,
,,,. . Propagation Limit - G(% Hal .
Combustion Completeness - 100 %
Flame Speed - 42.37 m/s
' Sprays - OFF 1.5 -
4-1, ... .
5 e 3*
h O.
t a.s - -
2-i.s -
i o.5 - - - - - --- -- -- --
- g- ,j Ts.e I.econdst Figure 2.110. Case A13, Compartment 1 Pressure 85
a , , , , .
Ignition Limit - 10 % H, t,. . Propagation Limit - G( % H,)
Combustion Completeness - 100 %
Flame Speed - 8.47 m/s
' Spreys . OFF 4.s -
- 4-s.s -
g
_t
, s- -
h t
I L
! 2.s .
i.s -
g-o.s - ..- ..-- .- ..-- .. ..--.: ....
o too Goo tXII 1600 Joao atos amD 32oo 1600 .- ...tono.... esoo 4400 Time Isecondel Figure 2.111. Case A14, Compartment 1 Pressure s , , , ,
Ignition Limit - 10 % H,
, , , , , Propegation Limit - G(% H,)
Combustion Completeness - 100 %
Fleme Speed - 4.24 m/s
Spreys - OFF 4.s -
's. ,.
1 g ts-
_5 e 3' h
L
! 2.s --
a-1.s - -
o.5 -
.- .. -l -.: -- -- .--..-- .-- .-
o too eso ...tano teco acao seco anno saco inao tone 44o0.-. seco Time to condel Figure 2.112. Case A15, Compartment 1 Pressure 86
1 I
s . . , ,
Ignition Umit - 10 % H,
,,,, ,Propogetion Limit - G(% H,)
Combustion Completenees - 100 %
Flame Speed - 1.89 m/s
~
"Spreys - OFF 4.s -
4-s.s --
g 5
e 3--
h 4
? 2.s -
2-1.s - -
as -
0 400 800.- 1200,- le 2000 2430 2mo 3200 3600 .-
4000 ..-
4400 4000 T6ae Iseconds Figure 2.113. Case A16, Compartment 1 Pressure e . . . , , , ,
Ignition Limit - 10 % H,
, , , , .Propogetion Limit - G( % H3)
Combustion Completeness - 100%
Flame Speed - 0.87 m/o S
"Spreys - OFF 4.s -
e ,.
e g s,s -
5 e 3*
h i 2.s -
l 2-1.s - -
t 0.s .
0 400 ,000 tJ00,- _.--
IGCC 2G10 ,-- ,2400 JE0.--.3200 1500 ...40004400
. .,- 4000 Tsae (secondst l
Figure 2.114. Case A17, Compartment 1 Pressure
?
i l
- 87 l
s .
Ignmon Limit - 10 % H,
,,, ;, Propagetion Limit - G(% Hal Combustion Completeness - 100 %
Flame Speed - 0.42 m/s
'" Spreys - OFF 9.5 -
e 1,,.
5 2
e 3" h
2 a.s -
e.
a-1.5 -
t o.5 - - - - - -
T6.e is.conds:
Figure 2.115. Case A18, Compartment 1 Pressure s , ,
Ignition Limit - 10 % H,
"" Propogetion Limit - Gl% H,)
Combustion Completeness - 100 %
Flame Speed - 42.37 m/s 5" Spreys - ON 9.5 -
. 4 e
1*
3.$ -
8 e 1-I
? 2.s -
G.
a-i.s -
~
le o.s -
o con ano taco.- isan-- .20o0 .-. asco aen .-- us.-- ,-taco isoo -- eeoo
-.- eson I6me (seconds)
Figure 2.116. Case A19, Compartment 1 Pressure 88
l l
b . v Ignition Limit - 10 % H,
, , , , , Propagation Limit - G(% H,)
Combustion Completeness - 100 %
Flame Speed - 8.47 m/s 5" Sprays - ON t.s -
%e .-
i, , , -
-5
, 3-L
! 2.s -
a-i.s - .
l-.
r o.s .. _,_
0 400 000 1200,_ 1600 2000 2400 2500 3200 1 10 4000 4400,_ 4a00 T6=. Is conosi Figure 2.117. Case A20, Compartment 1 Pressure s , , , . .
Ignition Limit - 10 % H,
' ' ' " Propagation Limit - G( % H2) .
Combustion Completeness - 100 %
Flame Speed - 4.24 m/s 5 " Sprays - ON ~
4.5 -
%. 4 Ij 3.5 -
_E e 3- -
L
? 2.s - .
3-t.s . \
I--
r -
0.s 1 ,_ _.. _,___,_
_,_ 4000 0 400 000,_ _,8200 1600 2000 2400 2e00 __,__ 3200 E10,. 4000,. 4400 T6-e Iseconost Figure 2.118. Case A21, Compartment 1 Pressure 89
e ,
lenition Limit - 10 % H,
, , , .. Propogetion Limit - G{ % H,)
Comimetion C:- .;L:_. ::: - 100 %
Flame Speed - 1.89 m/s
Spreys - ON 4.s -
g 3.s -
-.5
. 2-h E
i s.s -
a-i.s -
r 0.% - -
-l- .-
, - ,---l - ,
o <m soO iaoD isa0 2o00 a<m ..--isoD asco uoo ,- ecos
.,. .,-e.a0 ee00 hoe (secondel Figure 2.119. Case A22, Compartment 1 Pressure s , ,
lenition Limit - 10 % H,
, , , . . Propagetion Limit - G(% H,)
Combustion Completeness - 100 %
Flame Speed - 0.87 m/s
" Spreys - ON 4.5 - a e ..
e 3.5 -
5 e 1-h t :
l
[ 2.5 -
a-1.5 -
~
g..
F 0.5 -
,- ,- ,---l-- ,---l-IGGI,-aGIO
- ,-- 2900 2000 3a00 1600 0 C 800,-- 1200 .000 4400,- 4400 n e leocondit Figure 2.120. Case A23, Compartment 1 Pressure 90
e-- , , , ,
Ignition Limit - 10 % H,
,,,. . Propagation Limit - Gl % H,)
Combustion Completeness - 100 %
Flame Speed - 0.42 m/s
Sprays - ON t.s -
e 1
g 3.s -
5 e 3' I
A t 2.s -
a-i.s -
g-I o.s o
a a. - . ...
azoo sa acco..sa
..... ,2em ma uaa<a... uaa
... sa T6.. leecondst Figure 2.121. Case A24, Compartment 1 Pressure n , ,
Ignition Limit - 8 % H,
,,,, , Propagatica L.mit - G( % H,)
Combustion Completeness - 100 %
Flame Speed - F1% H 3)
'" Sprays - AUTO 1.5 -
4- a
)
I s .s -
t 1
o 3" h
L i s.s .
a-i.s - ,
0.s , - - . - - - . . .. -- . --- --- .--
n teecondet Figure 2.122. Case B8, Compartment 1 Pressure 91
e , ,
lenition Limit - 8 % H, s.. . . Proosgetion Limit - G(% H 2)
Combustion Completeness - 100%
Flame Speed - F(% H,)
Spreys - AUTO e.s -
q.
e.
ts-g 5
. I' h
2 s.s -
A g.
t.s -
e o.s - - - - - - - -- - -
--g---- -- ,
n te.condei Figure 2.123. Case B9, Compartment 1 Pressure )
l s . ,
Ignition Limit - 8 % H,
,,,. , Propogetion Limit - G(%H,)
Combustion Completeness - 100 %
Flame Speed - F(% H,)
'" Sprays - AUTO e.s -
g e- -
3.s -
g 5
. 3--
h
- e a.s -
(
a-1.s -
e o.s
---4-4-4--
- --4--
n i..cond.1 Figure 2.124. Case BIO, Compartment 1 Pressure i
1 l
l 92
2.2.3 Analysis can occur, particularly if the sprays are on. Our opin-In Section 2.2.1, we identified several areas of ion is that the CLASIX-3 flame speed may be too low concern regarding the CLASIX-3 analysis. These con- and, therefore, the precicted pressure rises may be cerns, alon th some additional considerations, are I **' '" C " t'
- 22 ts e resul of cases A13 through A24 m a different format to illustrate the effect of burn time on the pressure rise. The burnout Compartmentalization and Mixing-The 44 time in Figure 2.125 is equal to the flame propagation HECTR cases show clearly that high pressures can length divided by the flame speed. The flame spe(ds
, result ir global burns occur. Whether or not global used in most of the other HECTR cases are about a l burns occur depends on how rap,dly i hydrogen and
' factor of 3 or 4 higher than the flame speed of 1.83 m/s oxygen are mixed withm containment. There are three used in HECTR cases Al and B1 through B4, and in mechamsms to consider when addressmg m,xmg. i the CLASIX-3 cases. As described previously in sec-They are tion 2.1.3, these speeds are based upon experimental
- 1. Pressure-driven mass transfer data. It should be pointed out that even the higher l 2. Diffusion speeds do not take into account the possibility of l 3. Convection flame acceleration due to the presence of obstacles in l HECTR, and apparently CLASIX-3, ignore diffusion the annular region above the suppression pool. There-
[ and convection. Thus, in both cases, the mixing rate fore, even the higher speeds may not be conservative.
between compartments is underestimated; mixing Increasing the flame speed above the values used in within a compartment, however, is instantaneous; HECTR will result in an increase in the pressure rise.
The two-compartment HECTR cases tend to The pressure increases observed are already a signifi-show that the wetwell region eventually inerta due to a cant fraction (greater than 90% in most cases) of the lack of oxygen, and burns occur in the dome region adiabatic pressure rise and the net increase would be soon afterward. This inerting in the wetwell may not ~ 10% or less, which may be important for some cases.
be realistic. After a burn in the wetwell, the tempera-ture there may be higher than the temperature in the i dome, and significant convective mixing might result.
However, the same convective mixing mechanisms Completeness of Combustion-Incomplete that will bring oxygen into the wetwell between burns comb astion results in lower pressure rises than com-will also move hydrogen into the dome region. There- plete combustion; however, the pressure rises are not fore, the analyses may also underestimate the amount always as much lower as one would expect. For exam-of hydrogen that escapes from the wetwell region ple, the only differences between HECTR cases Al unburned. and A2 are flame speed and completeness of combus-As far as compartmentalization is concerned, the tion. Surprisingly, case A1, with a lower flame speed one-compartment cases would be typical of very rapid and incomplete (85%) combustion, has a peak pres-mixing, and the three- and five-compartment models sure close (93%) to that of case A2. The reason is that would be typical of slower mixing. For both CLASIX- case Al has one more burn than case A2. Therefore, 3 and HECTR, the transport and mixing of hydrogen while the pressure rise for any single burn is larger for are controlled more by arbitrary rnodelling assump- Case A2, the cumulative results are similar. Whether tions than by physics. HECTR will be modified to or not this will hold true in other cases will depend c:Iculate convection in the near future. A more de- upon the time between burns and the pressure and tailed analysis of mixing is presented in Chapter 4.0 of temperature decrease between burns (cases Al and A2 this report. had no sprays). Typically, incomplete combustion results in more burns that occur closer together in Flame Speed-The CLASIX-3 flame speed of 1.83 time and, therefore, tend to be more additive. Com-m/s results in i> urns that last about 12 s in the dome plete combuuian results in fewer burns with larger region. In 12 s a significant amount of heat transfer pressure rises accompanying each burn.
I I
93 i
l
Propagation Limits-In cases B5 through B7, Cl degree of reduction depends upon the burn time (time through C4, D1 through D3, and E1, burns were available for heat transfer; see Figure 2.125). The only allowed to propagate from one compartment to anoth- time that sprays have a minor negative effect is when er with a lower hydrogen concentration. It was no*ed they are turned on before the first burn and cause a previously that in some of the CLASIX-3 cases, signif- slight increase in pressure (due to the spray tempera-icant hydrogen concentrations were present in the ture exceeding the gas temperature) before the first dome while burns were occurring in the wetwell. The burn. However, this slight increase will be more than effects of including propagation vary. In some cases offset by heat transfer for slow burns, and in any case, (B5, B6, and B7) the end result is still eventual it is clearly better to turn the sprays on too early than wetwellinerting, fallowed by a burn in the dome at the to turn them on too late.
ignition limits. However, in other cases (C1 through The effect of adding sprays in the wetwell (case C3, D1 through D3, and EI) the only burns in the B2') was to delay inerting there. This effect appears to dome occurred just above the 4.1rc upward propaga- be due to increased cooling in the wetwell, which leads tion limit and resulted in relatively small pressure to a lower pressure and more mass transfer from the rises. oxygen. rich dome region back into the wetwell.
Burning at high concentrations in the wetwell and propagating into concentrations of 4Tc to 6Fe hydro-gen in the dome would cause few problems. However, Initial Conditions-For an adiabatic hydrogen propagatmg upward into 8Fe to 10Fo hydrogen would burn, the ratio of final pressure to initial pressure is produce results sirmlar to the results for cases A2 approximately constant for any particular hydrogen through A12 (one compartment model). Clearly, the mole fraction. Thus, a 10To change in initial pressure key question once again is mixing, and neither yields a 10% change in final pressure. While the burns HECTR nor CLASIX-3 adequately addresses this.
in the HECTR analysis are not adiabatic, the basic idea still holds. Comparing cases A8 and A9, for Hydrogen Source Term-The hydrogen source example, we find that a 0.157 atm difference in initial term used in the HECTR analysis came from Refer- pressure results in a 0.679 atm difference in peak ence 2.3, as discussed earlier. This source term was pressure. In both cases, the peak pressure occurred at apparently produced from a combination of the end of the first burn, and the ratio of peak pressure MARCH" results and hand calculations. Cases B8 to initial pressure was about the same.
through B10 were run to examine the effects of very Elevated initial pressures may result from drywell rapid hydrogen injection. There appear to be two air being puahed into containment and/or heating up major effects of rapid injection. First, the wetwell of the containment atmosphere during the accident.
inerts very rapidly because there is less time between Because relatively small increases in initial pressure burns to bring oxygen back ia. This effect results in can have such a significant impact on peak pressure, burns in the dome region soon afterward. Second, a caution should be used when evaluating the quantita-significant amount of hydrogen enters the contain- tive results produced by any of the analysis codes.
ment during a burn, thus making en 8% bur:. look more like a 9Fe to 10"o burn. Long-Term Considerations-It vcas noted that, In contrast to very rapid hydrogen injection, slow in several cases, the hydrogen concentrations were injection would result in a long time between burns close to the ignition point at the end of the run. As and little hydrogen injection during the burn. Howev- demonstrated in case B6', burns may occur at later er. it should be noted that slow injection rates result in times due to removal of steam from the atmosphere more time available for mixing, making global burns and the associated rise in hydrogen mole fraction.
more likely. A much better understanding of the Additionally, due to inerting, some of the cases show phenomenology of hydrogen production is needed be- very high hydrogen concentrations (40To to 80Fo) in fore definitive statements can be made about this the wetwell region at the end of the run. It cannot be issue. expected that this hydrogen would remain in the wetwell. Rather, it will eventually mix with the con-Centainment Sprays-Containment sprays pro- tainment atmosphere and additional burns could re-duce significant reductions in peak pressures. The sult.
94
2.3 References
- 'M. Berman, Light Water Reactor Safety Research Program Semiannual Report. April-September 1981, San- " Report on the Grand Gull Nuclear Station Hydrogen dia National Laboratories, NUREG/CR-2481, SAND 82- Ignition System, Mississippi Power & Light Company, Au-0006, February 1982. gust 31,1981.
M. P. Sherman et al, The Behavior of Hydrogen **R. O. Wooton and H.1. Avci, March (Meltdown During Accidents in Light Water Reactors, Sandia Nation- Accident Response Characteristics) Code Description and al Laboratories, NUREG/CR-1561, SAND 80-1495, August Users Manual, Battelle Columbus Laboratories, BMI-2064, 1980. NUREG/CR-1711, October 1980.
1.0 , .
0.9 -
0.8 - '
)
o 0.7 5
- a. e 0.6 -
O NO SPRAYS 0.5 -
O SPRAYS >
0.4 -
O.3 O.01 0.05 0.10 1.00 0.50 BURNOUT TIME (minutes)
Figure 2.125. Effect of Burnout Time on Peak Pressure l l
95-96
l 3.0 Grand Gulf Accident Calculations Using the MARCH Code 3.1 Summary and Conclusions The actual number, timing, location, and peak pression pool. The inerting of the wetwell by oxygen pressures that would be associated with hydrogen depletion occurred in all cases and led to the buildup burns during a reactor accident depend on the local of large hydrogen mole fractions in the wetwell and {
concentrations of the gases and the mixing processes combustible mole fractions in the upper compart-which produced those concentrations (prior burns, ments. In all cases considered but one, burns occurred natural and forced convection, H injection rates, etc). in the upper containment and led to significant pres-The h1 ARCH, HECTR, and CLASIX-3 computer sure peaks. Most of the cases predicted pressure peaks codes, however, introduce an artificial mixing of the below the NRC estimate of containment failure pres-containment atmosphere. The gas entering a compart- sure,71 psia (4.8 atm), but some were higher. The ment is assumed instantly mixed with the contents of results emphasize the importance of those failure the compartment, and hence each compartment at- estimates to conclusions drawn from this work.
mosphere is homogeneous. Real mixing processes such The compartment models we used are shown in as natural or forced convection are not considered.* Figures 2.2 through 2.4. Configuration A considers the The results of combustion, as predicted by these entire wetwell/ containment as one volume. Results codes, therefore depend strongly on the compartment from this simple compartmentalization will be realis-model used. The number and arrangement of com- tic if the time for mixing is shorter than the time partments is arbitrary. The most " realistic" model will between burns and also shorter than a characteristic be the one that best approximates the real mixing time for hydrogen release (since the mixture will then behavior in containment. We will first discuss some be nearly homogeneous). Ccnfiguraticn B divides this general conclusions from our work and then discuss volume into two compartments: a small wetwell vol-the results obtained from the individual compartment ume (from the top of the suppression pool to the 135 ft model studies. elevation) and a much larger upper containment. Con-If containment sprays are inactive, both MARCH figuration C also divides the volume into two compart-and HECTR predict a relatively small decrease in ments with the division being at the 209 ft elevation peak pressure as the burn duration is increased. Since (the top being the open volume under the contain-h1 ARCH does not include radiation heat transfer ment dome). Configuration D divides the volume into from the hot steam after a hydrogen burn (the most three compartments: a wetwell up to 135 ft, an inter-important heat transfer mechanism i:. HECTR when mediate annular region up to 209 ft, and an upper sprays are i:. active), the agreement is surprising. containment. Configuration E' differs from configura-Iamger burn time corresponds to slower flame speed. tion E used in the HECTR analysis since MARCH Experimental work at SNL would predict more rapid cannot analyze parallel flow paths. Configuration E' flame speeds, typically 8 m/s, than those used in the divides the volume into four compartments, similar to CLASIX-3 calculations, 2 m/s.84 With containment configuration D except for one more division at the sprays operating, MARCH and HECTR are not in 165 ft elevation. The more compartments used in the good agreement. MARCH predicted much greater model, the slower the calculated mixing. Without reductions in peak pressure due to hydrogen burns more knowledge of the mixing process, it is not clear than did HECTR for the same initial conditions and whin model gives the most realistic results.
same burn times. The results are shown in Figure 3.1. Our MARCH results using configuration A agree For multicompartment models, MARCH predict- fairly well with results obtained with HECTR. The ed a series of burns in the wetwell, the compartment number of burns predicted by the two codes for vari-into which hydrogen was introduced from the sup- ous cases either is the same or differs by one. The peak pressurcs predicted without sprays are very high and the two codes agree well. With sprays, the peak pres-
- MP&L did consider a small amount of forced convection in sures are lower, MARCH giving lower results than one CLASIX-3 sensitivity case. HECTR. Given the known differences in models and 97
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ l
i inputs, the agreement between MARCH and HECTR MARCH and HECTR for configuration C is fairly is probably fortuitous. good.
The majority of the CLASIX-3 results were ob- In configuration D, we predict gradual inerting of tained using configuration B. We have found the exact the lower sections of the wetwell/ containment and history of the accident to be very sensitive to the input burning higher in the containment. The peak pres-assumptions used with this configuration. In particu- sures are moderately high. MAtiCH and HECTR are lar, the number of wetwell burns is highly variable. in fairly good agreement.
However, in all MARCH and HECTR calculations The results for configuration E' are cimilar to upper containment burns were predicted. This is in configuration D. The E' configuration used in contrast to the CLASIX-3 calculations in which upper MARCH is sufficiently different from the E configu-containment burns were not predicted for most cases. ration used in HECTR to preclude direct comparison.
Depending on the assumptions made, the peak pres- For configurations B through E', high pressures sures for the MARCH and HECTR calculations range correspond to burns in the upper containment. In from moderately high to above the estimated failure nearly all cases, we get upper containment burns, and pressure for the containment. Sprays help reduce the hence the peak pressures are high. For upper contain-peak pressures, ment burns, the peak pressure depends on the as-For configuration C, in one MARCH calculation sumed volume of the upper compartment, the as-with a connected drywell there was no upper contain- sumed hydrogen mole fraction required for ignition, ment burn. For all other runs, one or more upper and the amount of heat tranferred from the gas during containment burns were predicted. The peak pres- the burn.
sures were moderately high. The agreement between 1.0 * ' ' * , . , , ,,,,
j k - . . E - _ '. _ , ',
os -
,,' g .
.,g-o.s -
'a ,g o.7 - -
o.s - --o- . HEcTR. No sphAYs
-- 0-- HECTR, SPRAyg h
- 0.s -
--0-- MARCH, No sprays -
I
--dr-- MARCH, sprays ,
o.4 -
A. -
f , , a) b gg f , t E 1 3 3 ,f , , f
- 0.01 c.os 0.10 0.30 1.oc sURN TNE (mhwtes)
Figure 3.1 Effect of Hydrogen Corabustion Burn Time on the Predicted Single
- Compartment Peak Pressure by HECTR and MARCH Codes l
l of the results. A recent study showed that MARCH 3.2 Introduct. ion contains numerous limitations and some errors."
, The MARCH computer code is a simple, widely Nevertheless, MARCH continues to play an impor-I used computer program for modelling a wide variety tant role in reactor accident analysis. Both CLASIX-3 of nuclear reactor accident scenarios." because com- and HECTR use the output of a comparable MARCH puter run time is short, studies involving many runs run to determine the conditions in the reactor coolant are feasible and inexpensive. Reservations have been system, and in particular the hydrogen and steam expressed by some MARCH users abcut the accuracy generation rates. Both codes replace the MARCH 98
containment-response subroutines with better ones. A major problem in comparing MARCH calcula-In this section we give a brief introduction to the tions to HECTR calculations is that HECTR uses a features of the MARCH program. A more detailed more sophisticated criterion for name propagation discussion can be found in Appendix C. from compartment to compartment. In MARCH, All three computer codes (MARCH, HECTR, and HECTR, or CLASIX-3 one can require each compart-CLASIX-3) treat the wetwell/ containment volume as ment to meet the conditions for ignition in order to if it were composed of one or more discrete compart- have combustion. Additionally, in HECTR and ments. In each compartment tha gases are assumed to CLASIX-3, one can also allow a name to propagate be uniformly mixed. The compartments are connected from one compartment to an adjacent one if the by user-specified, one-way, zero-flow-resistance con- hydrogen mole fraction in the second compartment is nections. Usually, one interconnects compartments above some specified propagation limit, a number with two such connections, one for How in each direc- usually lower than the ignition limit. In MARCH, tion. In MARCH, Rows between interconnected com- however, one can only use an option in which the partments are created so that the pressure in each name propagates from one compartment to an adja-compartment is equalized. If the required nows trans- cent compartment if the hydrogen mole fraction in the port too great an energy, more than one time step may second compartment is above a user-specified mini-be required to equalize pressure between compart- mum value for combustion to exist. If one specifies ments. This happens brieDy only at times such as complete combustion (i.e., a zero burnout value) in hydrogen burns. MARCH cannot cimulate connec- any compartment which has initiation of a flame, then tions with finite Gow resistance, nor can it analyze MARCH will propagate a flame into all connected parallel How paths through the various compart- compartments containing any hydrogen and com-ments. pletely consume all the hydrogen if there is sufficient For the cases considered in sections 3.3 to 3.7, oxygen present. It will do this in compartments con-MARCH was run with a connection from the drywell taining very small fractions of hydrogen, below the to the wetwell, but with no connection back to the Hammability limit. The MARCH option is much less drywell. These cases were run with steam and hydro- flexible than those available in HECTR or CL ASIX-3.
gen from the reactor coolant system directly injected MARCH contains provision for heat transfei (in-into the suppression pool and hence out to the cNding condensation) from the containment gases to wetwell/ containment. (This simulates a TPE acci- passive heat sinks, walls, gratings, etc. It also models dent, a transient followed by a stuck-open relief valve containment sprays, with cooling of the containment and failure of emergency core evoling.) With these gases resulting from water drop vaporization. From assumptions, the temperature, pressure, and gas com- the rapid temperature drops in our multicompart-positions of the drywell stayed fired. In particular, no ment models,it is clear that MARCH simulates water hydrogen entered the drywell. The drywell was effec- spray cooling in all compartments, but it is not clear tively isolated, and we have ignored it in sections 3.3 to how it distributes the sprays between compartments.
3.7. The drywell was isolated so that MARCH calcula- This point is discussed in more detail in Appendix C.
tions could be compared to HECTR calculations, MARCH does not consider radiative heat transfer in HECTR having no drywell compartment at this time. the containment, an important omission, since radia-For comparison, cases were considered in section 3.8 tion heat transfer from hot steam is the main heat with a connection to the drywell from the wetwell or transfer mechanism after a hydrogen burn if contain-from higher up in the containment to simulate now ment sprays are not operable.
from the vacuum breakers. With now possible to and Both HECTR and CLASIX-3 calculations were from the drywell, MARCH tends to equalize pressures performed with approximately the same hydrogen between the drywell and the rest of containment. release rate input, HECTR using an approximation of Hydrogen can enter the drywell, so that drywell burns the CLASIX-3 rates. Our MARCH calculations were are possible. Accidents involving a break in the reactor carried out using a different model for the reactor coolant system inside the drywell, a small (intermedi- core. Because of the interaction of the reactor coolant ate) break LOCA similar to an S2(S )i accident, can be system and the containment response in MARCH, the so modelled (instead of modelling the reactor steam same model for the reactor core produced slightly and hydrogen as entering directly into the suppression different hydrogen release rates for the various cases pool and tne original air in the drywell as being pushed considered by MARCH. However, these differences out into the wetwell; this approach was used for were negligible compared to the differences between HECTR cases B-1, A-5, and A-9). the hydrogen release rates predicted by MARCH and 99 ;
{
j
those used in HECTR and CLASIX-3. We produced Section 3.3 discusses results obtained using the about 4000 lb (1800 kg) of hydrogen by the end of the configuration A compartment model; section 3.4 dis-cecident as compared to 2605 lb (1185 kg) used in the cusses results obtained using the configuration B com-CLASIX-3 study (Figure 3.2). The combustion of partment model, etc. Diagrams of the various configu-about 3312 lb (1505 kg) of hydrogen is required to rations can be found in section 2.2.1, Figures 2.2,2-3, burn up all the oxygen in containment, including the and 2.4. Briefly, configuration A consists of a single drywell.The amount of hydrogen that must be burned compartment of the wetwell/ containment, Configura-to just inert containme:st by oxygen depletion will be tion B, the one used for most of the CLASIX-3 stud-lower: about 2484 lb (1130 kg) for the containment ies, consists of a small wetwell and a large upper including drywell and 2160 lb (982 kg) for the contain- containment, the division being at the 135 ft elevation.
ment excluding drywell. Configuration C also divides the wetwell/ containment h1 ARCH normally uses a value of 0.065 oxygen into two compartments, but puts the division at the mole fraction as the inerting limit. We have altered top of the annular region (the 209 ft elevation). Con-the h1 ARCH code to change this to 0.050, in agree- figuration D divides the volume into three compart-ment with HECTR and CLASIX-3. In section 3.4 we ments: a small wetwell, the intermediate annular re-present some results using the 0.065 limit and com- gion, and the upper containment above the 209 ft pare them to those using the 0.050 limit. We have also elevation. Configuration E' differs somewhat from introduced into the h1 ARCH code the possibility of configuration E used in the HECTR calculations. It steam inerting. Our MARCH code version will inert divides the volume into four compartments: a wetwell, compartments if the steam mole fraction goes above a lower annular region, an upper annular region, and 0.56,in agreement with HECTR.This did not occur in the upper containment (the divisions occurring at j our calculations, except in the drywell for periods elevations of 135,165, and 209 ft).
after a break into the drywell compartment from the reactor coolant system.
5000.0 1
4 0 4000.0-0 0
% 3000.0-Q
>+
lIl Iz.
2000.0 -
CD (n
d=
1000.0 -
OD - , , , , , , , , , ,
0.0 20.0 40.0 00.0 00.0 100.0 12 0.0 140.0 100.0 100.0 200.0 2E0.0 TIME - (MINUTE)
Figure 3.2. Hydrogen Release Into Containment,Ib, for Case A-1 Without Burns. Approximately true for all MARCH cases 100
Three containment spray options are considered would be increased if sprays were activated and other in the cases to follow: forced convection were present. This is discussed in
. No sprays Chapter 4.0.
. All sprays on very early in the accident and The case descriptions and summary of results for remain on the configuration A computer runs can be found in
. " Auto," all sprays on after the first hydrogen Table 3.1. The graphical results for case A-1 can be burn and remain on found in Figures 3.3 through 3.7. For the other config-uration A runs, only graphs of pressure vs time are Apparently, the automatic option is the one planned given (Figures 3.8 through 3.15). The cases correspond for use at Grand Gulf.
as closely as possible to the corresponding ones used in The time to burn up hydrogen in a compartment Chapter 2.0 is based on the expected flame speed and the compart-In case A-1 there were no containment sprays.
ment size. Some h1 ARCH calculations were carried The burn time corresponded to the flame speed used out using a burn time corresponding to the flame n the CLASIX-3 analysis,6 ft/s. The ignition limit speed considered in the CLASIX-3 study,6 ft/s. hiost used was 0.08 hydrogen mole fraction. The extinction of the h1 ARCH calculations were carried out using limit for combustion was 0.012, corresponding to an burn times corresponding to higher flame speeds, in 85G burn completion. Figure 3.5 seems to indicate a agreement with some experimental work on hydrogen minimum hydrogen mole fraction higher than 0.012.
combustion at SNL." In agreement with HECTR and CLASIX-3, we have used values from 0.08 to 0.10 for This is probably due to the introduction of hydrogen during the burn, which is not consumed. The amount the hydrogen mole fraction required for ignition. The of hydrogen generated is shown in Figure 3.2 and in lower limit of hydrogen mole fraction for hydrogen burns (the extinguishment limit) has been set at either Appendix C. h1 ARCH predicted three burns. The highest pressure was 60 psia (4.1 atm). This value is 15% of the ignition limit or zero. In each of the high, but below the estimated containment failure following sections e table is presented with the case descriptions and results. pressure of 71 psia (4.8 atm). After the last burn, h1 ARCH predicted the containment was inerted due to oxygen depletion. The oxygen mole fraction varied somewhat as steam condensed, but h1 ARCH predict-3.3 Analysis of MARCH ed it to be about 0.045, below the inerting limit or 0.050. For case A-1, HECTR predicted four burns, a Calculations for Configuration A .
peak pressure or 60 psia (4.1 atm), and a final oxygen In configuration A, the containment is divided mole fraction of 0.040. Both codes agree on the most into two compartments: the drywell, and a wetwell important result, a peak pressure of 60 psia (4.1 atm).
consisting of all the free volume between the suppres- The data in Figures 3.3 and 3.4 show that the sion pool and the top of the containment dome. As temperature and pressure peaks associated with the stated in section 3.2, the drywell is isolated in the burns have almost fully decayed by the time subse-model used in this section and hence ignored. The quent burns occur. Hence the initial conditions for the effect of the drywell will be considered in section 3.8. subsequent burns are not sensitive to the rate of heat Since here we consider only one compartment, all transfer from the previous burns for this case. We hydrogen burns are global burns. The use of configu- investigated the effect of heat transfer,in the form of ration A causes h1 ARCH to instantly mix any hydro- different burn times, on the peak pressure of a hydro-gen or steam entering the wetwell uniformly through- gen burn for configuration A in Appendix C. We found out the volume. Configuration A then represents the that without sprays activated h1 ARCH and HECTR limiting case of very rapid mixing. If the time for predicted nearly the same small reduction in peak mixing is shorter than the time between burns and pressure due to heat transfer for reasonable burn shorter than the characteristic time for release of times. This was surprising, since h1 ARCH does not I hydrogen'io the wetwell, the atmosphere will tend to include radiation heat transfer from the hot steam.
be nearly homogeneous, and configuration A will tend {
Evidently, h1 ARCH compensates by having a higher j to more accurately model the situation than the more convective heat transfer loss. With sprays activated, l complex multicompartment models. There is some the two computer codes give different results, as will l ividence that mixing due to natural convection can be he discussed in the following cases that have sprays relatively rapid (order of minutes), and this mixing activated.
101
Case A-2 differs from case A-1 only in that the containment had oxygen mole fractions near the inert-burn time is reduced, corresponding to our estimates ing limit after the last burn, and that hence there was of faster flame speeds, 8 m/s. The results were as the possibility of one further burn.
expected (Figure 3.8). h1 ARCH predicted three burns, Case A-5 models an accident involving a break in with a maximum pressure of 65 psia (4.4 atm), slightly the reactor coolant system inside the drywell. This is higher than the 60 psia (4.1 atm) of case A-1. The modelled by assuming the air in the drywell is pushed reduced burn time reduces the heat transfer during out into the wetwell, raising the pressure to 17 psia.
the burn and gives results closer to the adiabatic, The adequacy of this assumption is addressed in isochoric combustion pressure. HECTR also predict- section 3.8 which discribes the situation where the ed three burns and a peak pressure of 65 psia (4.4 break flow is allowed to be directed into the drywell atm). Note that the predicted peak pressure is close to and the drywell is interconnected with the wetwell.
the estimated containment failure pressure of 71 psia Aside from the use of the higher initial pressure, the ,
(4.8 atm). MARCH predicted that the containment input for case A-5 is identical to case A-4. MARCH I was inerted by oxygen depletion after the third burn, predicted four burns. The peak pressure of 76 psia (5.2 while HECTR predicted the final oxygen mole frac- atm) occurred during the first burn, before the sprays tion was 0.07, which is above the inerting limit, were turned on. This is above the estimated contain-Case A-3 differs from case A-2 only in that sprays ment failure pressure. The containment was predicted are turned on early in the accident (long before the to be nearly depleted of oxygen after the last burn.
first burn) and stay on. The results shown in Figure HECTR predicted only three burns, but with suffi-3.9 indicate MARCH predicted four burns. Each of cient oxygen remaining to have a fourth. The peak the burns has a lower peak than those of cases A-1 and pressure predicted by HECTR was 66 psia (4.5 atm),
A-2, and the pressure peak decays much more rapidly with nearly equal values for the first and second due to the cooling of the gases by the containment burns. Again notice that MARCH predicts greater sprays. The peak pressure was 48 psia (3.3 atm) on the peak pressure reduction when sprays are activated first burn. After the last burn, the containment was than does HECTR.
inerted by oxygen depletion. HECTR predicted three Cases A-1 through A-5 used an ignition criterion burns, a peak pressure of 59 psia (4.0 atm), and of hydrogen mole fraction equal to 0.08. Cases A-6 to sufficient oxygen remaining after the last burn for a A.9 used a more conservative criterion of 0.10. For fourth burn. Case A-3 is the first case discussed with comparable situations one expects this to give higher sprays. It shows that both MARCH and HECTR peak pressures, and this was so.
predict reductions in peak pressure due to the pres. Case A-6 can be compared to case A-2 since nei-ence of sprays, but MARCH predicts a greater reduc- ther has sprays in operation. For case A.6, MARCH tion. This point is discussed in Appendix C. predicted only two burns, but with the oxygen mole Case A-4 differs from case A-3 only in that the fraction after the second burn just below the inerting sprays are turned on after the first burn. As a result, limit; HECTR predicted three burns and very little the pressure peak associated with the first burn,62 oxygen left after the last burn. The peak pressure psia (4.2 atm), is far larger than the highest pressure predicted by MARCH was 80 psia (5.4 atm), and that associated with the next two burns,44 psia (3.0 atm). predicted by HECTR was 78 psia (5.3 atm). Both HECTR also predicted three burns, but with a peak values are higher than the estimated containment pressure of 57 psia (3.9 atm). The slight increase in failure pressure A bomogeneous 10% hydrogen burn peak pressure because of the presence of sprays
- without spre a day fail containment.
before the first burn in HECTR is discussed in section in ca M w in case A-3) sprays are in operation 2.2.3 and Appendix B. For all the cases considered, earl: J e e ident. Both MARCH and HECTR MARCH predicted a benefit in turning the sprays on pruin e u . hurns with subsequent inerting by before the first burn. Both codes predicted that the oxyp a depw 63 The two codes differ in that MARCH predicted a peak pressure of 56 psia (3.8
- atm) and HECTR predicted 70 psia (4.8 atm). Again, MARCH predicts a greater reduction in peak pressure 4
due to spray cooling of the containment gases than
- The spray temperature in HECTR is specified to bc 135*F does HECTR. Both programs indicate an increase of while the initial air temperature is specified to be 80* F. This about 10 psia (0.7 atm) in peak pressuf e over case A-3.
mismatch artificially causes the preburn air pressure to be higher when sprays are initiated early. MARCH considers In case A-8 (as in case A-4) the sprays come on the energy additions to the containment sprays and hence after the first burn. As expected, the peak pressure has a change in spray temperature. caused by the first burn is the highest. MARCH 102
predicted three burns for case A-8 and a peak pressure Case A.9. like case A-5, starts with a pressure of 17 of 74 psia (5.0 atm). HECTR predicted three burns psia (1.16 atm) in the containment, and sprays are and a peak pressure of 68 psia (4.6 atm). Both codes turned on after the first burn. For case A-9 MAllCH l predicted inerting due to oxygen depletion after the predicted three burns followed by oxygen depletion, third burn. These pressures are near the containment while HECTR predicted two burns, with sufficient failure pressure. Again, the resulta of cases A-6 A-7, oxygen for a third. The peak pressure predicted by and A.8 indicate the value of turning on the sprays MARCH was 78 psia (5.3 atm) for the first burn, and early in the accident. 78 psia (5.3 atm) was also predicted by HECTR.
Table 3.1. Configuration A - Case Descriptions and Results i I I I I I i l Initial i Hydrogen Mole I Hydrogen Mole 1 Containment i Burn Time ! No. of l Peak Pressure Case ! Pressure i Fraction For i Fraction For l Sprays l Seconde l Burns l Psia No. I psia l Ignition ! Extinguishment l l l l (atra) 3 1 l l l l l l j l l I I I I I A-1 l 14.7 i O.08 i O.012 I Off 1 38 1 3 1 60 t I i i i I I (4.1) i I I I I I I A-2 l 14.7 1 0.08 1 0.00 l off I 8 1 3 1 65 l t 1 i ! I i (4.4) i I I I I I i A-3 l 14.7 I O.08 1 0.00 I On 1 8 1 4 1 48 I l i I 1 I l (3.3) l I I I I I I A-4 1 14.7 1 0.08 I O.00 i Auto l 8 1 3 1 62 l ! l l l l I (4.2) i I i i i l l A-5 1 17.0 1 0.03 I O.00 l Auto 1 8 1 4 1 76 I i i i l l l (5.2) i I I I I I I A-6 l 14.7 1 0.10 1 0.00 1 Off I 8 l 2 ! 80 t i l I l i ! (5.4)
I i 1 I I I I A-7 l 14.7 1 0.10 1 0.00 t On 1 8 I 3 1 56 l I I I I l 1 (3.8) "~'
I I I l l l 1 A-8 l 14.7 1 0.10 1 0.00 l Auto l 8 l 3 1 74 l l I
I I I I I (5.0) 3 I I I I I A-9 l 17.0 1 0.10 1 0.00 l Auto l 8 1 3 1 78 I I l l 1 l l (5.3) l l
1 103
i i
] 1400 Ignition Fraction - 0.08 H, 2 Extinguishment Fraction - 0.012 Ha
$ 130.0 - Burn T!me - 38 sec D Sprays - OFF 100.0 -
O t-Z 80.0 -
C.2 2
- < M0 -
! Q.
3 o
O 40.0-a __
h 20 0 - --
0.0 , , , , , , , ,
~
OD 20 0 40.0 00 0 80.0 100.0 120.0 140b 150 0 180.0 200 0 220.0 TIME - (MINUTE)
Figure 3.3. Case A-1, Preesure in Containment, psia 2500 0 Ignition Fraction - 0.08 Ha Extinguishment Fraction - 0.012 Ha w Burn Time - 38 sec E 2000D- Sprays - OFT p
H<
' Q:
t.J h
m 1500.0 -
E~
E-=
5 y 1000.0 -
b Q. f o 500 0 - t U '
l 0.0 , , , , , , , , , , ,
OD 20 0 40.0 80.0 80.0 100.0 120.0 140.0 160 0 180.0 200.0 230.0 TIME - (MINUTE)
Figure 3.4. Case A-1. Temperature in Containment, 'F 104 i
10 Ignition Fraction - 0.08 H Extinguishm1nt Fraction - 0.012 H, 7 ,
Burn Time - 38 sec Ca3 Sprays - OFF 8
=
Q Z 04-
'. Z O
b 04-l
! Ca3 i' 3 O
y 02- -
~
0.0 , , ,
OD 20 0 40.0 00.0 80D 100.0 120 0 140.0 1600 18 0.0 EOS E304 TIME - (MINUTE)
Figure 3.5. Case A-1, Hydrogen Mole Fraction in Containment 1.0 Ignition Fraction - 0.08 H.
Extinguishment Fraction - 0.012 H, Burn Time - 38 sec 00- Sprays - OFF z
ls3 0
02-Z O
t<
Q: 0.4-l ta3
( 3 o
3 og- .
OD , , , , , , , , , , !
OD 300 400 000 000 100.0 120.0 140.0 100 0 100.0 200.0 220.0 TIME - (MINUTE)
Figure 3.6. Case A-1, Oxygen Mole Fraction in Containment l
l 105 i
T , . _ _ _ . _ _ . , .,
1.0 Ignition Fraction - 0.08 H, Extinguishment Fraction - 0.012 H, Burn Time - 38 sec 0.8 - Sprays - OFF y
Ca3 b
Z 08-O t
M Es 0.4 -
Ca.1 1
O N_
2 02- ,,
OD , , , , , , , , , ,
OD 20 0 40.0 000 80.0 1000 180 0 140 0 100 0 100 0 3100 W4 TIME - (MINUTE)
Figure 3.7. Case A-1, Steam Mole Fraction in Containment 140 0 Ignition Fraction - 0.08 H, Extinguishment Fraction - 0.00 H, 2 1300- Burn Time - 8 sec Sprays - OFF 100.0 -
l Q.
b l Z 80.0 -
tal 2
b
< 80 0 - $
Q.
3 O
V 4JD -
1 h
O g 200 -
0.0 , , , , , , , , , ,
OD 20 0 400 00D 80.0 10E0 120.0 14E0 1000 1510 20110 220D TIME - (MINUTE)
Figure 3.8. Case A-2, Pressure in Containment, psia 106
i F <.1 O O 5 n, < b 2 I.aZH Q.
My F .
t p <. O1 o 2 O. 4 b 2 a. Z H O. h gi gi ar u
1 1 1 r 1 e 2 4 8 8 0 2 4 e 1 0 4 3 O D
0 D
0 D
0 0
0 D
0 0
0 D
0 0 3 0 2
0 4
0 0
0 8
0 0 2 0 1
9 4 0 D D D 0 8 0 0 O
- - - - - ~
D C O C D I
S B E g a I a p u x n s S B E g s
e r r t ti e _
pr ru xt n a n in A i i t ya n n y
A 2 0, s Tg oi ,3 2, a Tgu oi
- imsuih Frn
,4 0 0 0
P P
r - imsi nF r
e Ae U - ma e c
e s
s Oe N - ma h r e c s t s 4, T 8 n io u 4, t u 0 O t r
e 8 nt i o r A s n 0 e e Fr - i n
0 s n ni c a e Fr -
c 0 C c a C t .
o c 0 o 0, i o 08 n 0 t .
n 0 n
t a 0, i0 o 8 t 4 D n ai i
- aH n n
m - aH m
e 8,
~ 0 0
e n - 0 n 0 ,t 8, .
0
,t TD 0 p 0
p I S p s T0 0 s
M r ,H i a
I i
a a M ,H E10, y s E1 0, l
- 00 O n
- a0
(
M I 1
(
M 1
- k 2 I N n, 2 N a, D
U U 0 T T E
)1 E 4, )1 4, a
0 a
0 1 1 6, 6, 0 0 0 0 .
t 1 e, 8, n
o a 0
2 0,
2 0,
0 1 0
) 0 2 2 2 2 0 0 D D 1
0 7
1l llll
i P
i 140D 1 Ignition Fraction - 0.08 He Extinguishment Fraction - 0.00 H, 1a0D -
Z Burn Time - 8 sec D
r.n Sprays - AUTO 100.0 -
3 Q:
1 F
, Z 80D -
i t.2 2
4 $< 80D - Sprays On O.
3 O l V 40D -
l a
F 20D - -
J OD , , , , , , , , , , j OD 20 0 40D 80.0 80D 100.0 120.0 14a0 100 0 18a0 2DED 220.0 1
- TIME - (MINUTE)
Figure 3.11. Case A-5, Pressure in Containmem, usia j
i l 1400 1 Ignition Fraction - 0.10 H, g Extinguishment Fraction - 0.00 H, c: 1204 -
- Burn Time - 8 sec i D i y) Sprays - OFF 100D -
4 c.,
F
! Z B00-ta3 1 2 b
4 000-O.
., 3 O
l U 40D -
3 O 20 0 - - - - -
00 , , , , , , , , , ,
OD 200 40.0 80.0 80D 100.0 120.0 140.0 100 0 18 0.0 2003 220D TlME - (M1NUTE)
Figure 3.12. Case A-6, Pressure in Containment, psia 108
140D Ignition Fraction - 0.10 H.
Extinguishment Fraction - 0.00 H, y la04 - Burn Time - 8 sec
@ Sprays - ON 100.0 -
L E-.
Z 80D -
W 2
b
< t10D -
o O 40D -
J 200 - b _
OD , , , , , , , , , ,
OD 20 0 40.0 80.0 80D 100 0 12a0 IMO 1000 18E0 20110 230D TIME - (MINUTE)
Figure 3.13. Case A-7, Pressure in Containment, psia 140 0 Ignition Fraction - 0.10 H Extinguishment Fraction - 0.00 H, y 1200 - Burn Time - 8 sec
$ Sprays - AUTO 100D -
Q.
E-=
Z 80D -
W 2
< GOD-O.
3 o
O 404 -
3 h
O '
20.0 - _
OD , , , , , , , , , ,
OD 200 40D 00D 80D lonO 12110 1440 1000 18E0 2Dt10 230.0 TIME - (MINUTE)
Figure 3.14. Case A-8, Pressure in Containment, psia 109
_ _ _ _ - _ _ _ \
140D Ignition Fraction - 0.10 H 2 Extinguishment Fraction - 0.00 H 2 130D - Burn Time - 8 sec h Sprays - AUTO 100D -
L E-Z 80D -
W Sprays On 2
N
< .0D -
L 3
o O 40D -
J O 20D - -
E-=
t OD , , , , , , , , , ,
i DD 260 40.0 00D 80D 10n0 1260 140.0 100 0 18E0 2DEO 220.0 I
TIME - (MINUTE)
{ Figure 3.15. Case A-9, Pressure in Containment, psia 1 3.4 Analysis of MARCH large enough to fail containment from simple defla-grations. Because of inerting of the w Cdculations for Configuration B oxygen depletion, very high mole ions offract;etwell hydro- due i in configurat. ion B, the containment ,is divided gen can accumulate in the wetwell.. The potential for into three volumes: the drywell, a small wetweil (the local detonations exists if this hydrogen mixes with I
volume between the suppression rool and the 135 ft sufficient oxygen. Detonations are discussed in Chap-elevation), and a very large upper containment vol- ter 5.0.
ume.This configuration is the same as that used in the 'I he typical sequence of events for configuration B 3
majority of the CLASIX-3 calculations. As discussed predicted by MARCII is as follows:
- in Appendix C, this appears to be the first time 2
MAllCH has been used in BWR calculations with 1. Numerous wetwell burns models having more than two compartments. Uncer- 2. Inerting of the wetwell by oxygen derdetion tainties exist on how MARCH treats certain phenom- 3. Buildup of the hydrogen mole fraction te high ena in such multicompartment BWR models, particu- values in the wetwell and to combustible larly the containment sprays. As mentioned amounts m, the upper contamment previously, for these MARCH calculations there is no 4. Burns in the upper containment, wetwell, or connection from the wetwell or upper containment to both the drywell; hence the drywell is isolated. We will 5. Wetwell is again inerted by oxygen depletion; consider only the behavior of the wetwell and the the upper containment is either oxygen deplet-upper containment volume. The effect of the drywell ed or has insufficient hydrogen for ignition is considered in section 3.8. For all the cases considered using configuration B, Wetwell burns involve the combustion of such MARCH and HECTR predicted one or more upper small masses of hydrogen that the resultant total containment burns, and hence significant pressure pressure rise in containment is small. Combustion in peaks. For several of the comparable cases, CLASIX-3 the upper containment, because of the large volume did not predict any upper contMnment burns, and involved, is nearly a global burn. Consequently, only hence it predicted only small pressure peaks. The upper containment burns can cause pressure peaks number of wetwell burns, the occurrence and duration i10
of wetwell inerting, and the number of upper contain- Figures 3.17 and 3.20 show simultaneous spikes of ment burns differ somewhat among AIARCH, temperature and steam in the wetwell late in the HECTR, and CLASIX-3. Part of the reason for the accident. These spikes are not hydrogen burns. We differences is that the configuration B model is "ill- have investigated the possibility that they are due to conditioned;" i.e., small changes in the input values or the release of either hot hydrogen or steam from the differences in the assumptions used in the codes can reactor coolant system. We have not been able to give rise to considerably different results. explain them. R. O. Wooton of Battelle Columbus Our configuration B runs were initially carried out Laboratory implied that MARCH may be having com-using the normal MARCH value of 0.065 oxygen mole putational difficulties at this point since the reactor fraction for inerting. This was later altered to 0.050 to coolant pressure is close to the containment pressure.
agree with HECTR and CLASIX-3. In the following HECTR predicted 22 wetwell burns and, late in discussion of configuration B cases we will first de- the accident,1 large upper containment burn with a scribe results obtained using the 0.050 value. We will peak pressure of 82 psia (5.6 atm). CLASIX-3 predict-then compare those results to the ones obtained using ed no upper containment burns, and hence a low peak the 0.065 value. Th>: results are often quite different, pressure. In later work with CLASIX-3, an upper giving additional confirmation of our statement that containment burn was " forced" when the mole frac-the configuration B model is ill-conditioned. tion of hydrogen ended up just below the ignition The case descriptions and a partial summary of limit. Of course these results are closer to those of the results of the MARCH configuration B calcula- MARCH than the earlier CLASIX-3 results. Case B-1 tions (for both the 0.050 and 0.065 oxygen inerting is typical of all the B configuration cases. MARCH limits) are given iri Table 3.2. The results of case B-1 and HECTR predict upper containment burns and are shown in Figures 3.16 through 3.24 using the 0.050 high peak pressures, while CLASIX-3 generally does limit. The pressure history for case B-1 with the 0.065 not predict upper containment burns. The peak pres-oxygen inerting limit is shown in Figure 3.25. Pressure sure predicted for case B-1 is close to or above the histories for cases B-2 to B-7 with the 0.050 limit are predicted failure pressure for the containment,71 psia shown in Figures 3.26 through 3.31. (4.8 atm).
Case B-1 modelled one of the cases considered in Case B-2 is similar to case B-1 except the initial the CLASIX-3 analysis. The bvak into the drywell pressure was 14.7 psia (1.0 atm). The pressure history and expulsion of air into the wetoell was simulated by predicted by MARCH is shown in Figure 3.26.
using a wetwell/ containment pressure of 17 psia (1.16 MARCH predicted 13 wetwell burns, a period of atm). 3 prays came on after the first burn. Combustion tvetwell inerting, two more wetwell burns, an upper occurred only in compartments that met the ignition containment burn and a final wetwell burn, for a total criterion, hydrogen mole fraction equal to 0.10. The of 16 wetwell burns and I upper containment burn.
results are shown in Figures 3.16 through 3.24. Figure The peak pressure was 52 psia (3.5 atm). After the last :
3.25 shows the pressure for case B-1 with the 0.065 burn the oxygen mole fractions in the upper contain-oxygen inerting criteria. ment and wetwell were just below the inerting limit of For case B-1, with the 0.05 oxygen inerting criteri- 0.050. When case B-2 was run using the oxygen inert-on MARCH p:edicted 15 wetwell burns, a period of ing limit of 0.065, MARCH predicted seven wetwell wetwell inerting due to oxygen depletion, another burns, a period of wetwell inerting, an upper contain-wetwell burn, a further period of wetwellinerting, and ment burn, a wetwell burn, and a final upper contain-then a burn in both the upper containment and ment burn. The peak pressure was 63 psia (4.3 atm).
wetwell. The peak pressure due to the last burn was 65 For case B-2, HECTR predicted 23 wetwell burns, a psia (4.4 atm). After the last burn, the containment period of wetwell inerting, and a final burn starting in was inert due to oxygen depletion, the mole fraction of the upper containment and propagating to the oxygen being 0.045 in the upper containment and wetwell. The peak pressure predicted was 71 psia (4.8 0.043 in the wetv. ell. atm). HECTR and the two MARCH runs for case B-2 By comparison, when the oxygen inerting limit of show that the number of burns in the wetwell is very 0.065 was used for case B-1, MARCH predicted nine sensitive to the assumptions made, but all three calcu-wetwell burns, a period of wetwell inerting, and then lations predict one or two upper compartment burns an upper containment burn followed in seconds by a with peak pressures in the range 52 to 71 psia (3.5 to final wetwell burn. The peak pressure was 70 psia (4.8 4.8 atm). By comparison, CLASIX-3 predicted 43 atm). After the last burn the cor.tainment was inert wetwell burns and no upper containment burns. Con-due to oxygen depletion, but the oxygen mole fraction sequently, the peak pressure predicted by CLASIX-3 was above 0.05. was low,22.5 psia (1.5 atm).
l i
l 11I l
l l
l l
l Case B-3 differs from case B-2 only in that the codes predict one or more upper containment burns hydrogen mole fraction was reduced to 0.08 for igni- with peak pressures between 56 and 68 psia (3.8 and l tion and was raised to 0.012 for extinguishment. 4.6 atm). These pressures are high, but they are below h1 ARCH predicted 14 wetwell burns, a period of the predicted failure pressure.
wetwell inerting, a rapid sequence of wetwell/ upper- Case B-5 is the first of the h1 ARCH cases in which containment /wetwell burns, and finally two more up- we allowed Game propagation from one compartment per containment burns. The peak pressure was 42 psia to all others which contain more than the minimum (2.9 atm). After the last burn the containment was amount of hydrogen required for name extinguish-
! oxygen-depleted.The corresponding h!ARCil run us- ment. Since these h1 ARCH runs are to try to duplicate ing the 0.065 oxygen inerting limit predicted 11 those llECTR runs having complete combustion, we wetwell burns, a period of oxygen inerting, a wetwell took the hydrogen mole fraction for extinguishment to burn, a nearly simultaneous upper containment and be sufficiently low,0.002, as to be essentially zero. By wetwell burn, and a final upper containment burn. using a small positive number, we force A1 ARCH to The peak pressure predicted was 44 psia (3.0 atm). avoid burning negligible amounts of hydrogen in a HECTR predicted 31 wetwell burns, a period of compartment and hence help clarify the output with l wetwell inerting, and an upper containment burn respect to which compartments burned significant
! propagating into the wetwell. The peak pressure pre- amounts of hydrogen. The hydrogen mole fraction for dicted was 56 psia (3.8 atm). As in case B-2, ignition was taken as 0.08 and the name speed as 8 Cl,ASIX-3 predicted many wetwell burns (58) and no m/s, and the sprays came on after the first burn.
l upper containment burns. The pattern is clear. While For case B-5, h1 ARCH predicted 11 burns mainly l H ECTR generally has at least one upper containment in the wetwell, a period of wetwell inerting due to burn, and Cl.ASIX-3 predicts no upper containment oxygen depletion, a combined wetwell and upper com-burns, h1 ARCH most easily inerts the wetwell and has partment burn, a wetwell burn, a second period of the most upper containment burns. The fact that the wetwell inerting, and finally a burn mainly in the h1 ARCH runs include more hydrogen generated may upper containment. The peak pressure predicted was explain this pattern. 60 psia (4.1 atm). After the last burn both the wetwell Case B-4 is similar to case B-3 except that there and upper containment were inert due to oxygen were no sprays and the combustion completeness was depletion. When the 0.065 oxygen mole fraction limit
, assumed to be unity. Without sprays, h1 ARCH pre- for inerting was used, h1 ARCH predicted nine burns dicted 24 wetwell burns in a 20-min time period. The mainly in the wetwell, a period of wetwell inerting due pressure rise for each burn was about 4 psi (0.3 atm) to oxygen depletion, a combined wetwell and upper and the pressure did not have time to fall back com- containment burn, and finally a burn that was mainly pletely to the initial pressure before the next burn. in the upper containment. The peak pressure for the The peak pressure for the series of wetwell burns was last two burns was nearly identiwl,56 psia (3.8 atm).
27 psia (1.8 atm). After the 24 wetwell burns, the For case B-5, HECTR predicted 15 wetwell burns, wetwell was inert due to oxygen depletion. The hydro- a wetwell burn propagating into the upper contain-gen concentration increased in both the wetwell and ment,2 more wetwell burns, and then an upper con-upper containment. h1 ARCH then predicted two up- tainment burn propagating down into the wetwell. If per containment burns with a peak pressure of 56 psia case B-5 had been carried out further in time, HECTR (3.8 atm). After the last burn, the containment was might have predicted another large burn. The peak inert due to oxygen depletion. When case B-4 was run pressure predicted was 72 psia (4.9 atm).
using the 0.065 oxygen inerting limit, the results were A comparison of the various results for case B-5 altered. h1 ARCH predicted 14 wetwell burns with a shows that the exact number of burns varies greatly peak pressure of 28 psia (1.9 atm), and 3 upper con- with the input and code. All the runs indicate upper tainment burns with a peak pressure of 65 psia (4.4 containment burns and high pressures. h1 ARCH pre-atm). After the last burn the oxygen mole fraction in dicted lower peak pressures than HECTR,60 and 56 the containment was very low. psia vs 72 psia. This is believed to be primarily due to For case B-4, HECTR predicted 30 wetwell burns h1 ARCH *s over-estimate of the cooling effect of con-and I upper containment burn, with a peak pressure tainment sprays.
of 68 psia (4.6 atm). After the last burn, there was still Case B-6 differs from case B-5 only in that the enough oxygen in the upper containment for another hydrogen mole fraction for ignition was raised to 0.10 hurn. Comparing the results of the two h1 ARCH runs and the sprays were turned on early in the accident.
with the HECTR run for case B-4,it is clear that the h1 ARCH predicted 13 burns mainly in the wetwell number of burns v9 ries greatly. However, both of the and I burn in the upper containment. The peak i12 1
1 pressure predicted was 63 psia (4.3 atm). After the last h1AllCII predicted 12 burns mainly in the wetwell, a burn, the oxygen mole fraction in both upper contain- period of wetwell inerting due to oxygen depletion, ment and the wetwell was 0.04.'When the 0.065 oxygen and then 2 burns in Imth the upper containment and mole fraction limit for inerting was used, MAllCll wetwell.The peak pressure predicted was 62 psia (4.2 predicted eight burns mainly in the wetwell, a period atm) during the first of the two global burns. After the of wetwell inerting due to oxygen depletion, a large last burn there was very little oxygen left in contain-burn in both compartments, a second period of ment. When the 0.065 oxygen mole fraction inerting wetwell inerting, and a final upper containment burn. limit was used, MAllCil predicted eight burns mainly The peak pressure predicted was 66 psia (4.5 atm). in the wetwell, a period of wetwell inerting, and two For case 11-6', llECTil predicted 17 wetwell upper containment /wetwell burns. The peak pressure ;
burns, I wetwell burn propagating into the upper was predicted during the last burn,61 psia (4.2 atm).
containment, and I upper containment burn propa. For case 11-7, llECTil predicted 18 wetwell burns, gating into the wetwell. The peak pressure predicted a period of wetwell inerting, a wetwell burn propagat-was 84 psia (5.7 atm). Just as in case 11-5, the number ing into the upper containment, another wetwell burn, of burns predicted by the two MAllCII calculations and a final upper containment burn propagating into and the llECTil calculation differ, all calculations the wetwell. The peak pressure predicted was 81 psia
- predict upper containment burns, and the peak pres. (5.5 at m). One expects case 11-7 to give nearly identical
' sures predicted by IIECTil are almve those predicted results to case 116. The results are very similar, but by MAllCII. All the peak pressures predicted are high the exact number of burns differ. The peak pressure and near the estimated failure pressure of contain- predicted by IIECTil is significantly different, but ment MAllCil predicted nearly the same pressures in both Case 11-7 differs from case 116 only in that the 116 and 11-7.
l sprays are turned on after the first wetwell burn.
i i
! Table 3.2. Configuration B Case Description and Results l I I I I I I I Number or i l i l l l I Hydrogen I i i ISignificant l l Oxygen Motel l 1 Initial i nydrogen 1 Mole Fraction l Burn i l l Upper i Peak l Fraction l lCaeol Preneure IMole Fraction l for l Time. 1 Flame l l Containment IPressurel Required to!
} l No.1 (peia) ifor Ignition '
xatirquishment lSecondeL Propagation lSprayel Burne 1 (pela) l Inert i i i l I I I I i i i i i l R-Il 17 0.10 i i 1 1 0.0 l 4.8 l No l Auto 1 65 .05
} l l l l l l 1 70 .065 1 I I I I I I I B+2l 14.7 1 0.10 1 0.0 4.8 I No l Auto 1 52 .05
) i i l l 2 63 .065 l l I I I l B-31 14.7 i O.08 1 0.012 1 4.4 i No 1 Auto l 2 42 .05
!.. l l l l l t i I 2 44 .065 #
l 7 I I I I i i l l B-4l 14.7 I O.08 i O.002 l 4.8 l No l Of f 2 56 .05 j l l l l l I l 3 65 .065 J l i i I i i I
- 1 B-51 14.7 1 0.00 1 0.002 l 2.0 l Yes l Auto 2 60 .05
! l I l I I l l 2 56 .065 I I I I I I I i B-6 14.7 0.10 0.002 1 2.0 I Yes I th 1 .05
._ _. 6 6 l l 1 2 66 .065 i i l I l B-71 14.7 1 0.10 1 0.002 1 2.0 l Yes l Auto l 2 l 62 l .05 l l l t i 1 l t i 2 I 61 . 0M I I i
i b
i 1
l i
f
. ll3 i
s
~.-n . . . , , - . , -, ,-- , , , . , - v- --- - e,.-- , , - . , , a e-,---wa. - - . , . - , - - . - r - - , , . . . , , - -.,n.,- -n,.-.---..-n-.. , , ,
- 140.0 j Ignition Fraction - 0.10 Ha l I
I Flame Propagation - NO g 180.0 - Extinguishment Fraction - 0.00 H D Burn Time - 4.8 sec Sprays - AUTO 100.0 -
- E O ]
b Z 80.0 -
r.2 3
$< 80.0 -
O.
3 o
- O 40.0 -
J j p 20.0 - M a- w . ... . ._
i 4
0.0 , , , , , , , , , ,
! 0.0 20 0 40.0 00.0 80.0 100.0 120.0 1400 160 0 180 0 200.0 220.0 i
TIME - (MINUTE) ;
- Figt're 3.16. Case 11-1, Pressure in Containment, psia i I i 2600 0 Ignition Fraction - 0.10 H, Flame Propagation - NO Ca3 Exti.1guishment Fraction - 0.0 H, M O-p Burn Time - 4.8 sec Q Sprays - AUTO
] %
w fl.
t y 15000 -
i [:3
.1 g f-j b
y 1000.0 -
4 g
i O.
2 3
- o 500 0 -
U k s,sL, , k ,1. Ihem J ,
! 0.0 , , , , , , , , , ,
00 20 0 40.0 80.0 80.0 100.0 120 0 140.0 100 0 180.0 200.0 220.0 TIME - (MINUTE)
Figure 3.17. Case 11 1, Temperature in Wetwell, 'F f
- 114
1.0 Ignition Fraction - 0.10 H, Flame Propagation - NO
( 0.8 -
Extinguishment Fraction - 0.0 H, Burn Time - 4.8 sec 8
x Sprays - AUTO Q
- I
- 0.8 -
Z O,
b< 0.4 -
+
E k
ta 3 /
h 02-0.0 0.0 2b0 4b.0
)h>
0'0.0 80.0 id0.0 INO0 140.0 ido0 INO0 2$0.0 220.0 TIME - (MINUTE)
Figure 3.10. Case IL1, Hydrogen Mole Fraction in Wetwell 1.0 Ignition Fraction - 0.10 H, Flame Propagation - NO Extinguishment Fraction - 0.0 H, 0.8 - Burn Time - 4.8 sec z
Sprays - AUTO 08-Z
,O,,,
b m 0.4 -
k l
Ed 3
0 2 02-
~h 0.0 , , , , , , , , ,
OD 20 0 40.0 00.0 80 0 100.0 120.0 140 0 100 0 180.0 200.0 220 0 TIME - (MINUTE)
Figure 3.19. Case 11-1, Oxygen Mole Fraction in Wetwell 115
i ID Ignition Fraction - 0.10 Ha l
Flame Propagation - NO Extinguishment Fraction - 0.0 H O. - Surn Time - 4.8 sec y
y Sprays - AUTO b
z 0-9 b
d Es. 0.4 -
cs2 0
02-d
) I h Dfl . .
OD . . . . . . .
OD 20 0 40.0 00.0 80D 100.0 120.0 14a0 160 0 10 0.0 20a0 220.0 TIME - (MINUTE)
Figure 3.20. Case B.1, Steam Mole Fraction in Wetwell 25000 Ignition Fraction - 0.10 H, Flame Propagation - NO W Extinguishment Fraction - 0.0 He 2 20000-O Burn T.ime - 4.8 sec kD: Sprays - AUTO W
h W
15000 -
P l b d
y 1000.0 -
b Q.
X 5000 -
o l U
L \ .,_ _
OD 50 40.0 0'0.0 ND INO.0 INO.O 14O.0 160 0 INO.0 250.0 220.0 TIME - (MINUTE) 1 l
Figure 3.21. Case B 1, Temperature in Upper Containment, 'F l
l 116 l
l
l 1.0 Ignition Fraction - 0.10 Ha Flame Propagation - NO Extinguishment Fraction - 0.0 H 6 08- Burn Time - 4.8 sec h
=
Sprays - AUTO Q
lI; 06-Z O
t
< 0.4 -
m ta.
ca 3
O y 02-
~
0.0 , , , , , , , , , ,
OD 20 0 40.0 00.0 80.0 100 0 120 0 14 0.0 160 0 180 0 200.0 220.0 TIME - (MINUTE) j Figure 3.22. Case B.1, Hydrogen Mole Fraction in Upper Containment f
10 Ignition Fraction - 0.10 H, Flame Propagation - NO Extinguishment Fraction - 0.0 H, 08- Burn Time - 4.8 sec 7
y Sprays - AUTO X
08-Z O
t*C m 0.4 -
Es Ca3 J
O 7 02-r 0.0 , , , , , , , , , ,
00 20 0 40.0 00.0 80.0 100.0 120 0 140.0 100 0 ISO 0 200.0 220.0 TIME - (MINUTE)
Figure 3.23. Case H.1, Oxygen Mole Fraction in Upper Containment i17
i i
18 f
Igriition Fraction - 0.10 Ha i Flame Propagation - NO
- Extinguishment Fraction - 0.0 H
0A - Burn Time - 4.8 sec l y y Spreys - AUTO M
z OD-O b
M Es. 0.4 -
f ca
.3
? o 2
02 -
04 , , , , , , , , , ,
OD 200 40.0 004 8DD 100,0 taa0 14a0 1000 ISEO 2DEO 220D TIME - (MINUTE)
, Figure 3.24. Case B-1, Steam Mole Fraction in Upper Containment 140D Ignition Fraction - 0.10 H.
Flame Propagation - NO 120.0 Extinguishment Fraction - 0.0 He i p Burn Time - 4.8 sec Spreys - AUTO 100.0 -
Q E-Z 80.0-Em1 2
$ 80.0-4 O !
V 40.0 - )
e.-l l l
N O 20.0 - _
b" ' h - - -x .a- - - -
l OD n , , , , , e s , s OD 20.0 40.0 00.0 80.0 100.0 12 0.0 140.0 10 0.0 100.0 200.0 220.0 ,
l TIME - (MINUTE)
Figure 3.25. Case B-1 (6.5% Oxygen Inerting Limit), Pressure in Containment, psia 118
140 0 Ignition Fraction - 0.10 H Flame Propagation - NO
$ M8 ~ Extinguishment Fraction - 0.0 Ha h Burn Time - 4.8 sec d
0:
gg . Sprays - AUTO Q.
E-=
Z 80D -
c.3 2
4 80D-O.
2 o
O 40D -
J O, 20D - p- _a ->
E- .__--
1 OD , , . . . . 1 OD 20 0 40D 004 80D 10R0 1220 14 & O IdOO 1$0.0 2d(10 220D TIME - (MINUTE)
Figure 3.26. Case B.2, Pressure in Containment, psia 140.0 Ignition Fraction - 0.08 H Fisme Propagation - NO y 120D -
Extinguishment Fraction - 0.012 He I
$ Burn Time - 4.8 sec d
0:
100D . Sprays - AUTO C.
E-=
Z 80D -
La3 2
b
< 80D -
Q.
o U 40D -
3 h
E. 20D - g _.
( (, , , , , , ,
OD , . . . . . . . . .
OD 200 40D SOD 80D 10(10 1200 14a0 1000 18(10 200.0 2304 TIME - (MINUTE)
Figure 3.27. Case B-3, Pressure in Containment, psia i19
4 l 140j o l Ignition Fraction - 0.08 H, g Flame Propagation - NO
. t2: IED ~ Extinguishment Fraction - 0.002 H, D
i Burn Time - 4.8 sec 100D - Sprays - OFF g
l A 3 E-*
i Z 80D -
4 Ca3
- 2 j 4 800-1 C.
O O 40D -
- .3 C 20 0 -
- OD , , , , , , , , , ,
j OD 2I10 40D 80D 80 0 10(10 1211 0 1440 1800 18(10 20110 220D
! TIME - (MINUTE)
) Figure 3.28. Case Il-4, Pressure in Containment, psia i
i e
! 140 0 ,
j Ignition Fraction - 0.08 H, I Flame Propagation - YES 0:
124- Extinguishment Fraction - 0.002 H, j
a D cn Burn Time - 7.0 sec l b Q:
100.0 - Sprays .- AUTO l i I Q.
I E-*
I Z SOD -
i C.3
) 2
! M 800-i O.
3 O
j U 400 -
- a f
h< 20D - % C ( ,_,_,,,
I OD , , , , , , .
Ort 20 0 40.0 OOD 80D 100.0 12110 140 0 1800 18110 200.0 220D TIME - (MINUTE)
Figure 3.29. Cane 115, Pressure in Containment, psia 120
1400 Ignition Fraction - 0.10 H, Flame Propagation - YES a: la0D - Extinguishment Fraction - 0.002
$ Burn Time - 2.0 sec b
0:
100 0 -
Sprays - ON Q.
I E-~
Z 80D -
W 2
$4 800-O.
3 o
O 400-3 h
h 20 0 - .. ,, , _ -- -
OD , , , , . . i OD 20 0 40D SOD 80D 100.0 1200 14a0 180 0 IN110 250.0 220.0 TIME - (MINUTE)
Figure 3.30. Case 11-6. Pressure in Containment, psia 1400 Ignition Fraction - 0.10 H, Flame Propagation - YES laDD - Extinguishment Fraction - 0.002 H, c:
$ Burn Time - 2.0 sec i
b 0:
100 0 -
Sprays - AUTO Q.
E~
Z 80.0 -
M 3
$ 800-O 3
o V 40.0 -
A 20 0 -
l 00 , , . . . . . . . .
0.0 200 400 80 0 80.0 100.0 120.0 140 0 180 0 1800 20E0 220.0 TIME - (MINUTE)
Figure 3.31. Case 11-7, Pressure in Containment, paia 121
3.5 Analysis of MARCH For case C-1 IIECTil predicted six wetwell burns, a seventh burn starting in the wetwell and propagat.
Calculations for Configurat. ion C ing into the uppe, containment, ana then two more Cor. figuration C consista of three volumes: the wetwell burns. The peak pressure was 40 psia (2.7 drywell, a large wetwell (the volume from the suppres- atm). After the last burn the upper compartment sion pool to the 209 ft elevation), and the still larger contained sufficient oxygen for a second burn, but upper containment volume. The ratio of wetwell to insufficient hydrogen for ignition. The wetwell con-upper containment volume is 0.600 compared to 0.129 tained a high mole fraction of hydrogen, but was for configuration 11. For configuration C, wetwell inerted due to oxygen depletion. A comparison of the burns give rise to moderate pressure peaks (larger MAllCil and ilECTil results for case C-1 shows than wetwell burns in configuration 11) and upper- many similarities. The predicted peak pressures are containment burns give rise to somewhat higher peak reasonably close, llowever, the number of wetwell pressures (smaller than upper containment burns in burns differs.
configuration B). For the cases considered in this In case C-2 the hydrogen ignition limit was raised section, there was no connection from the wetwell to 0.10 and the sprays were turned on early in the back to the drywell, as stated in section 3.2. Conse- accident. MAltCil predicted five wetwell burns, all quently, we ignore the drywell in this section. The with peak pressures of about 28 psia (1.9 atm). After effect of the drywell is considered in section 3.8. One the five wetwell burns, the wetwell was inerted due to upper containment burn occurred in all of the cases oxygen depletion, and late in the accident MAllCil considered in this section, but none in the case consid- predicted a burn mainly in the upper containment ered in section 3.8. with a peak pressure of 48 psia (3.3 atm). After the last The case descriptions and a partial summary of burn, the entire containment was oxygen depleted.
results for the configuration C calculations are in llECTil predicted three wetwell burns, one Table 3.3. Graphical results fo. case C-1 are shown in wetwell burn propagating into the upper containment, Figures 3.32 through 3.40. The pressure histories of and then three more wetwell burns. At the end of the cases C-2 to C-4 are shown in Figures 3.41 through IIECTIt calculation the upper compartment had an 3.43. oxygen mole fraction of 0.10 but a hydrogen mole The qualitative behavior of the four cases consid- fraction of only 0.023. The two codes are in reasonable cred for configuration C is as follows: agreement about the peak pressure (both have one
- 1. Several wetwoll burns with moderate pressure burn in the upper containment), but disagree on the peaks number of wetwell burns. The peak pressure is high
- 2. Wetwell inerting due to oxygen depletion but below the estimated failure pressure of contain-
- 3. A buildup of hydrogen mole fracti<m to high ment.
values in the wetwell and combustible values in in case C-3, the sprays came on after the first the upper containment burn, but otherwise the case is identical to C-2. In this
- 4. One upper containment or combined wetwell/ case MARCil predicted five wetwell burns, a long upper-containment burn period in which the wetwell was oxygen depleted, and
- 5. Inerting of the wetwell by oxygen depletion. then a combined wetwell and upper containment Insufficient hydren in the upper contain_ burn. The peak pressure of the first wetwell burn, ment for a second burn, but sufficient oxygen before the sprays were operated, was 38 psia (2.6 atm),
and the peak pressure after the last burn was 47 paia For case C-1, MAllCII predicted six wetwell (3.2 atm). After the last burn, the wetwell and upper burns, a period of wetwell inerting, and then a com- containment were oxygen depleted.
bined wetwell/ upper-containment burn. Smce the For case C-3, IIECTil predicted three wetwell sprays came on after the first burn, MAllCII predict- burns, a wetwell burn propagating into the upper ed that the first burn generated a higher peak pressure containment, and then three mere wetwell burns. The than the next five wetwell burns,33 psia (2.2 atm). peak pressure predicted was 42 psia (2.0 atm). After The peak pressure due to the combined wetwell/up- the last burn, the upper containment was not predict-per-cor.tainment burn was 45 psia (3.1 atm). After the ed to be oxygen depleted.
last burn, the wncentration of hydrogen and oxygen In case C-4, the sprays were off. As a result, the in the upper ecntainment approached that required to peak pressures and temperatures were higher and ignite another burn. The wetwell contained a high their decay after the burns was much slower. MAllCil mole fraction of hydrogen, but was inerted due to predicted four wetwell burns, with gradually increas-oxygen depletion. ing peak pressures up to 50 psia (3.4 atm), and a final i
122 l
l
upper containment and wetwell burn with a peak In cases C-1, C-2, and C-3, MAllCH predicted pressure of 70 psia (4.8 atm). This is about equal to the somewhat higher pressures than HECTR. However, in estimated failure pressure for the containment. After case C-4, without sprays, the MARCH prediction of 70 the last burn, the upper containment was not oxygen psia (4.8 atm) is much higher than the HECTR pre-depleted. diction due to the nature of the propagating burn in For case C-4 HECTR predicted six wetwell burns, the HECTR calculation. The MARCH case C-4 peak a wetwell burn propagating into the upper contain. pressure prediction is the only one that approaches ment, and then two more wetwell burns. The peak the estimated containment failure pressure of 71 paia pressure predicted was 49 psia (3.3 atm). (4.8 atm).
Table 3.3. Configuration C Case Descriptions and Results I 1 I i 3 i i l I i l Hydrogen l Hydrogen I i l i I No. of l No. of Casel Initial IMole Fraction! Mole Fractionl Flame l Comp. l l Peak (Significantl Significant No.! Pressure,l for i for IPropaga-l Burn Time,lSpraysl Pressure,l Upper l Wetwell l psia l Ignition IExtinguishmentl tion i Sec. l I paia lContainmentl Burns l l l l 1 l l atm i Burns i I I I I I I I I I C-1 1 14.7 0.08 1 0.002 1 Yes l 4 l Auto l 45 l 1 l 7 1 1 1 1 I (3.1) l l I i i i I i i l C-2 l 14.7 0.10 1 0.002 Yes 1 4 l On l 48 l 1 1 5 l l l l l l l l (3.3) l l 1 I I I I I i I l C-3 1 14.7 1 0.10 1 0.002 l Yes I 4 1 Aato l 47 1 1 6 I I I I l l l l (3.2) l
! 3 I I I I I I I i -4 1 14.7 1 0.10 1 0.002 I Yes 1 4 i off l 70 1 1 I 5 i l l l l 1 l (4.8) I l 140D Ignition Fraction - 0.08 H Flame Propagation - YES U
180D - Extinguishment Fraction - 0.002 Burn Time - 4 sec 1000 - Sprays - AUTO D:
O.
f-Z 80D -
r.:
2 s
Ct:
4 00D -
O.
3 o
U 40D -
r d
h f 3DD -
OD . . . . . . . . . .
OD 200 40D 00.0 80.0 1710 18tLO 140.0 1600 ' 18110 200.0 220 0 TIME - (MINUTE) l Figure 3.32. Case C-1, Pressure in Containment, psia '
l 123 l
j ,
! i l
I i
4 1
2500.0 Ignition Fraction - 0.08 H:
j Flame Propagation - YES
{ [a3 Extinguishment Fraction - 0.002 H, j E 2000.0 -
a o Burn Time - 4 sec 1
k% Sprays - AUTO C3 h
ca 1500.0 -
i 6 i- n--
Z y 1000.0 -
i f O.
5 i o 500.0 - .
I U 4
uut L __
l 0.0 , , , , , , , , , ,
f OD 20 0 40.0 00.0 80.0 100.0 120.0 140.0 100 0 180.0 200.0 220.0 l TIME - (MINUTE)
Figure 3.33. Case C-1, Ternperature in Wetwell, "F t
i 1 1.0
! Ignition Fraction - 0.08 H, Flame Propagation - YES i
Extinguishment Fraction - 0.002 Ha d 02 - Burn Time - 4 sec 8
Sprays - AUTO C
x 02-Z O
t 0.4 - -
co 2
O y 02-0.0 '
00 b0 40.0 00D 80.0 ld0.0 l$0.0 140.0 10'0 0 lb.0 250.0 220.0 TIME - (MINUTE)
Figure 3.34. Case C-1, liydrogen Mole Fraction in Wetwell 124
lD Ignition Fraction - 0.08 H, Flame Propagation - YES OA -
Extinguishment Fraction - 0.002 H, y Burn Time - 4 sec y Sprays - AUTO M
04-Z
.O b
E 0.4 -
[s.
[a3 N
2 02-m OD 1 OD NO N.0 NO ND l$0.0 t$0.0 140.0 l$00 l$0.0 2d0.0 220.0 TIME - (MINUTE)
Figure 3.35. Case C-1. Oxyge , Mole Fraction in Wetwell 1.0 Ignition Fraction - 0.08 H, Flame Propagation - YES -
Extinguishment Fraction - 0.002 H, 02-3 Burn Time - 4 sec
$ Sprays - AUTO 15 Z OA -
O Es. 0.4 -
Ca2 J
O 2
02 -
s i4 L 1
CD OD NO ND N.0 ND l$0.0 t$0.0 14Q0 l$00 l$0.0 2$n0 220.0 TIME - (MINUTE)
Figure 3.36. Case C-1, Steam Mole Fraction in Wetwell 1
125 l l
l
2110 0 0 Ignition Fraction - 0.08 H, Flame Propagation - YES Ca3 Extinguishment Fraction - 0.002 H, 0: 20000 -
- -) Burn Time - 4 sec k Sprays - AUTO 2
y 15000 -
ta3 P
b Z
IOOOD -
m 4:
O.
3 O S00 0 -
U I%
0.0 , , , , , , , , , ,
OD 20 0 400 00 0 800 100.0 1200 140.0 100 0 180.0 2000 220 0 TIME - (MINUTE) l l
Figure 3.37. Case C-1 Temperature in Upper Containment, 'F 10 Ignition Fraction - 0.08 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, 08-Ca3 Burn Time - 4 sec b
0:
Sprays - AUTO O l l
>=
lI: 08-Z O
I
-t; 04- l 1
Q:
La La3 3 .
O y 02- l l
s _
i 00- , , ,
-, - , , , , , , , 1 00 200 400 00 0 800 1000 120 0 140.0 100 0 180.0 200.0 220 0 TIME - (MINUTE)
Figure 3.38. Cane C-1,Ilydrogen Mole Fraction in Upper Containment 126 u _ _ ._ .. _
1.0 Ignition Fraction - 0.08 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H 2 88~ Burn Time - 4 sec Sprays - AUTO M
03-Z O,,
b Q: 0.4 -
Es.
Ca3 b
02-L OD . . . . . . . . . .
OD 20 0 40.0 00D 80D 10110 12(10 140.0 1000 IEE10 200.0 220.0 TIME - (MINUTE)
Figure 3.39. Case C-1, Oxygen Mole Fraction in Upper Containment 1.0 Ignition Fraction - 0.02 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, OA -
2 Burn Time - 4 sec
$ Sprays - AUTO b
Z OA -
O
- E O
00 Es. 0.4 -
CaJ 3
O 2
02 -
1 0.0 '
OD 20 0 40.0 0O.0 80.0 ido.O l$0.0 I4'O.0 IMD0 l$t10 2NO.0 220.0 TIME - (MINUTE)
Figure 3.40. Case C-1, Steam Mole Fraction in Upper Containment 127
1400 Ignition Fraction - 0.10 H.
Flame Propagation - YES 1808 ~ Extinguishment Fraction - 0.002 H ct:
Burn Time - 4 sec 100D -
Sprays - ON
% f Q.
E--
Z 800 -
c.3 2
b4 000-Q.
3 o
U 40D -
3 h
12 20 0 -
ld L L OD , , , , , , , , ,
OD 20 0 40.0 80.0 800 100.0 120.0 140.0 1800 180.0 2D(LO 220D TlME - (MINUTE)
Figure 3.41. Case C-2, Pressure in Containment, psia 140 0 Ignition Fraction - 0.10 Ha g Flame Propagation - YES 1 0-0: Extinguishment Fraction - 0.002 H.
D en Burn Time - 4 sec b
D:
t00 0 - Sprays - AUTO Q.
E-Z 80.0 -
c.3 2
b< 800-Q.
3 O
O 40.0 -
3 O 20 0 - -
0.0 , , , , , , , , , ,
OD 20 0 40.0 80D 80.0 100.0 120 0 140.0 180 0 180.0 200.0 220A TIME - (MINUTE)
Figure 3.42. Case C-3, Pressure in Containment, psia 128
I l
l 140D Ignition Fraction - 0.10 He l Flame Propagation - YES l 1sDD - Extinguishment Fraction - 0.002 H Burn Time - 4 sec MIDD . Sprays - OFF Q:
Q.
H Z 80D -
ra3 2
b
< MO-O.
3 o
U 40D -
a O h 20 0 -
CD , , , , , , , , , ,
OD 20 0 40.0 00.0 80D 100.0 12 0.0 1400 100 0 1800 200.0 220.0 TIME - (MINUTE)
Fsgure 3.43. Case C-4, Pressure in Containment, psia 3.6 Analysis of MARCH Case D-i used a hydrogen mole fraction of 0.08 for Calculations for Configuration D ignition and had sprays after the first burn. Of our cases,it most closely matches the one configurat,on i D Configuration D consists of a drywell and three case calculated with CLASIX-3. MARCH predicted compartn:ents: a wetwell volume up to the 135 ft seven burns mainly in the wetwell, a period of wetwell elevation, an intermediate annular volume up to the inerting due to oxygen depletion, four burns mainly in 209 ft elevation, and an upper containment volume. In the intermediate annular region, and finally a burn this section the drywell was not interconnected to the mainly in the upper containment with a peak pressure rest of the containment; hence it was isolated and of 61 psia (4.2 atm). For case D-1, HECTR predicted ignored. The effect of the drywell is considered in 17 wetwell burns, a period of wetwell inerting,7 burns section 3.8. in the intermediate annular region, and a final burn The case descriptions and a partial summary of starting in the intermediate annular region and propa-results for configuration D are in Table 3.4. Graphical gating into the upper containment. The predicted
- results for case D-1 are shown in Figures 3.44 through peak pressure was 39 psia (2.7 atm). Although the 3.56. Pressure histories for cases D-2 and D-3 are general pattern of burns is similar in HECTR and shown in Figures 3.57 and 3.58. MARCH, the number of burns and the peak pressure The general pattern of results can be summarized differ greatly. CLASIX-3 predicted no upper contain-as follows: ment burn and consequently low peak pressure (27
- 1. Numerous wetwell burns psia,1.8 atm).
- 2. Inerting of the wetwell due to oxygen depletion Case D-2 differs from case D-1 only in that the
- 3. Buildup of hydrogen mole fractions in all com- hydrogen mole fraction for ignition was increased to partments 0.10. MARCH predicted eight burns mainly in the
- 4. Burns in the intermediate annular volume wetwell, a period of wetwell inerting, three burns
- 5. Possible inerting of the intermediate annular mainly in the intermediate annular region, and a final volume burn in all three compartments with a peak pressure
- 6. One burn mainly in the upper containment of 46 psia (3.1 atm). This result is surprising, since one 129
expects a larger peak pressure in case D-2 than D 1, upper containment burn. The peak pressure,41 psia but the reverse was found. The mole fraction for the (2.8 atm), was generated by the first of the three 1:st burn of hydrogen in the upper containment was intermediate annular region burns.
lower for case D-2 than for D-1,and therefore the peak For case D 3,llECTil predicted 19 wetwell burns, pressure was lower. a period of wetwell inerting,7 burns in the intermedi.
For case D 2,llECTit predicted 12 wetwell burns, ate annular region, and a final bsua in the intermedi-a wetwell burn propagating into the intermediate ate annular region propagating into the upper con-cnnular region,3 burns in the intermediate annular tainment. The peak pressure predicted was 48 psia region, a burn starting in the intermediate annular (3.3 atm).
region and propagating into the upper containment, The results obtained for the configuration D cases and a final burn in the intermediate annular region. show a variation in peak pressure, number of burns, The peak pressure predicted was 41 pain (2.8 atm). etc, that indicate that the compartment model is Case D-3 involved a hydrogen ignition mole frac- ill-conditioned. Small variations in the input give tion of 0.08 and no sprays. MAltCil predicted nine considerably different results. Using MAltCil, case wetwell burns, wetwell inerting, three burns mainly in D-1, which one would expect to generate the lowest the intermediate annular region, and inerting of the peak pressure, generates the highest peak pressure; intermediate annular region. At the end of the calcula- case D-3, which one would expect to generate the tion, the oxygen mole traction in the upper contain- highest peak pressure, generates the lowest. The ment volume was far above the inerting limit, and the llECTIt. predicted peak pressures do agree #ith the hydrogen mole fraction,0.062, was approaching the expectation of having D-1 the lowest and D 3 the ignition limit. Ilowever, MAltCil did not predict an highest.
Tcble 3.4. Configuration D - Case Descriptions and Results l I I I I I I I i i l I case linitial l Hydrogen l Hydrogen 1 l Flame ISprayal Peak l Number or l Number of I
! No. lPressurelMole FractionlMolg Fraction 1 Durn IPropagationt l Pressure,lSignificantlrigniricantl l l psia i ror l ror i Time, l l l pela 1 Upper l Annular I i l l Ignition lEstinguishmenti see l l l (ata) IContainmentl Region i I I I l l l l l l Burne l Burna l l I I I I I i l l I I I i i i i 1 ~l 1 D-1 l 14.7 i O.08 1 0.002 1 2 1 Yes Auto j 61 1 1 1 4 I I I I I I I I i l i l i 1 1 1 I I I (4.2) i I I I I I I I I I i i 1 'e l l D-2 l 14.7 1 0.10 O. 002 l 2 I Yes Auto l 46 l 1 1 4 1 1 I i 1 1 'l I I I I I i I (3.1) i I i 1 1 I I l l i I I i 1 1 D-3 1 14.7 0.00 0.002 2 1 Yes i No 1 41 1 O I 3 l l i i l i I (2.8) 1 I I I 1 1 I l I I i 130
I 140 0 Ignition Fraction - 0.08 H Flame Propagation - YES g 130 0 - Extinguishment Fraction - 0.002 H D Burn Time - 2 sec i
Sprays - AUTO a
Q.
l H i i Z 80.0 -'
w l 2 b
< 800-CL.
3 o
O 40 0 -
l 4 A
20 0 - . - . ,,,,,,,_,____r_,_ _ , _ ,_ _
l t
00 , , , , , , ,, , , ,
0.0 20 0 40.0 00 0 80.0 100.0 120 0 140.0 1000 h30.0 200.0 220.0 TIME - (MINUTE)
Figure 3.44. Case D 1, Pressure in Containment, psia 25000 Ignition Fraction - 0.08 H.
Flame Propagation - YES W Extinguishment Fraction - 0.002 H.
D: 2000 0 -
p Burn Time - 2 sec Q Sprays - AUTO a: ,
! w I Q. l y 1500 0 -
w l-* 1 E-=
5 y 1000 0 -
i 5
Q.
i 3
O S00 0 -
U
_ N A h J L dim n i m JLLh jo ,
00 , , , , , , , , , ,
j 00 200 400 000 800 1G40 120 0 140.0 1800 18E0 20n0 ETO
! TIME - (MINUTE)
Figure 3.45. Case D-1, Temperature in Wetwell, *F l
I 131 l
. -- , - - - - - _ + - . .- -% - . - - .-, - , , - .
4 ,
l 10 Ignition Fraction - 0.08 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, 8~ Even Time - 2 see b
M Sprays - AUTO
> 08- 4 Z
Z e I D
< 0.4 - If m I. /
y 02-400 2500 0 80.0 100.0 120 0 140.0 100 0 180 0 200.0 220.0 0.0 20 0 TIME - (MINUTE)
Figure 3.46. Case D.1, Ilydrogen Mole Fraction in Wety cil 10 Ignition Fraction - 0.08 H Flame Propagation - YES 5xtinguishment Fraction - 0.002 Ha 08- Burn Time - 2 sec z
Sprays - AUTO M
06-Z O,,
b<
M 0.4 -
[.s.
Ca3 3
m:
-w s'
00 20 0 4C.J 80 0 80 0 10'0 0 120.0 140.0 100 0 1t$0.0 250.0 220.0 TIME - (MINUTE)
Figure 3.47. Cane D.1, Oxygen Mole Fraction in Wet ~ ell 4 l32
l 1.0 Ignition Fraction - 0.08 H Flame Propagation - YES Extinguishmerit Fraction - 0.002 H 04-3 Burn Time - 2 sec j Sprays - AUTO b
Z 08-O b
4:
CY Ca. 0.4 -
02- _
b OD 20 0 40.0 'S 0 8'0.0 WTMWr$renW ido.0 ido.O 14O.0 1d00 IND.0 2dE0 220.0 TIME - (MINUTE)
Figura 3.48. Case D 1, Steam Mole Fraction in Wetwell 25000 Ignition Fraction - 0.08 H, Flame Propagation - YES y 0- Extinguishment Fraction - 0.002 Ha g Bern Time - 2 sec y Sprays - AUTO W
h W
15000 -
b b
Z 1000.0 -
b Q.
3 o 300 0 -
U N
00 . . . . . . . . . .
00 20 0 400 00 0 80 0 100 0 1200 14a0 100 0 180.0 200.0 220 0 TIME - (MINUTE)
Figure 3.49. Case D-1. Temperature in Intermediate Annular Rep,.on, 'F i 133
_ _ _ _ _ _ _ _ _ _ _ l
to ignition Fraction - 0.08 H, Flame Propagation - YES l Extinguishment Fraction - 0.002 H, 7 Os-c.2 Burn Time - 2 sec b
0:
Sprays - AUTO O
Z 0A-Z O
b
< 0.4 -
0:
k Cal 3
h 02 -
^=W i
I m
.. , , , . , , , , . 1 OD 20 0 400 00 0 802 It.C0 120 0 1400 100 0 180.0 200.0 220 0 TIME - l MINUTE)
Figure 3.50. Case D 1,Ilydrogen Mole Fraction in Intermediate Aanular llegion t.0 ignition Fraction - 0.08 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, 08- Burn Time - 2 sec z
$ Sprays - AUTO 08-
. Z j O, b
00 0.4 -
6 Cal 3
O E .g- _
00 , , , ,
h ,
OD 20 0 400 80 0 80.0 100.0 120 0 140.0 100 0 180.0 200.0 220.0 TIME - (MINUTE)
Figure 3.51. Ca e D 1, Oxygen Mole Fraction in Intermediate Annular llegion 134 l
l
l.0 Ignition Fraction - 0.08 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, OA- Burn Time - 2 sec 3
y Sprays - AUTO b
2 04-O m
ca. 0.4 -
ta)
J O (
2 1(
02-
^
w___ _
i 00 , , , , , , , , :., ,
OD 20 0 40.0 00.0 80.0 100.C 120 0 140.0 1000 180.0 Eu0 220.0 TIME - (MINUTE)
Figura 3.52. Case D-1, Steam Mole Fraction in Intermediate / nnular Region 2500 0 Ignition Fraction - 0.08 H, Flame Propagation - YES ta2 Extinguishment Fraction - 0.002 H, g 0,- Burn Time - 2 sec Q Sprays - AUTO E
t.2 h
ta3 15000 -
6 E-=
Z h 1000D -
t ti.
3 O S00.0 -
U
- w. L_
00 , , , , , , , , , ,
0.0 20 0 -40.0 00.0 80.0 1000 12 0.0 14G0 1000 180.0 2000 220.0 TlME - (MINUTE)
Figure 3.53. Case D.1, Te.aperature in Upper Containment, *F 135 t
10 Ignition Fraction - 0.08 H, Flame Propagation - YES Z Extinguishment Frartion - 0.002 H, 0A-c.2 Burn Time - 2 sec b
m Sprays - AUTO o
Z 08-Z O.
b
<: 0.4 -
k Ca3 3
O y 02-
~
-M i i e i s a s s OO 200 400 00 0 80.0 100.0 120 0 140.0 100 0 180.0 200.0 220 0 TIME - (MINUTE)
Figura 3.54. Case D 1, liydrogen Atole Fraction in Upper Containment to Ignition Fract'on - 0.08 H, Flame Propagation - YES 08- Extinguishment Fraction - 0.002 H, h Burn Time - 2 sec Q Sprays - AUTO x
08-Z O
b=5
% 0.4 -
W Cd 3
O '
2 92
..~.
00 , , . . . . . . . .
00 20 0 400 00 0 80 0 100.0 1200 140.0 100 0 180.0 20110 220.0 TIME - (MINUTE)
Figure 3.55. Case D 1, Oxyge; hiole Fraction in Upper Containment 136
8.0 Ignition Fraction - 0.08 H Flame Propagation - YES Extinguishment Fraction - 0.002 H, 08- Burn Time - 2 sec y
j Sprays - AUTO b
Z 08-
.O b
Q:
f.s 0.4 -
C.d 3
O 2
02-0.0 , , , , , , , , , ,
00 20 0 40.0 00.0 80.0 100.0 120.0 140.0 160 0 1800 200.0 220.0 TIME - (MINUTE)
Figure 3.56. Case D-1, Steam Mole Fraction in Upper Containment 1400 Ignition Fraction - 0.10 H, Flame Propagation - YES
$ 120.0 -- Extinguishment Fraction - 0.002 Ha
$ Burn Time - 2 sec d
0:
100 0 -
Sprays - AUTO C.
E-=
Z 800-c.2 2
,c 000-O.
o V 400 -
J
- i EC-h 20 0 -
gobQ ,,,,,,,,
00 . , , , , . , , , ,
00 20 3 400 00 0 80.0 100.0 120 0 140.0 100G 100.0 200.0 C00 TIME - (MINUTE)
Figure 3.57. Case D 2, Pressure in Containn,ent, psis 137 l
l
140 0 Ignition Fraction - 0.08 Ha g Flame Propagation - YES I ~
2: Extinguishment Fraction - 0.002 H.
D Burn Time - 2 sec 100 0 - Sprays - None g
C.
E--
Z B00-W 2
- C 000-Q.
3 O
O 400 -
3
- C 20 0 -
0.0 , , , , , , , , , ,
0.0 20 0 40.0 00.0 80.0 100.0 120 0 140.0 100 0 150.0 200.0 220.0 TIME - (MINUTE)
Figure 3.58. Case D-3, Pressure in Containment, psia 3.7 Analysis of MARCH calculations. consequently, we do not have a HECTR result to compare to the result of this section. As in the Calculations for Configurat,on i E' previous sections, there was no flow path from the The configuration E used in the HECTR analysis wetwell to the drywell. The drywell is therefore isolat-involved parallel flow paths from the bottom of the ed and we will not discuss it.
wetwell to the upper containment. Since MARCH In order to handle multiple compartments, we cannot treat parallel how paths: we have modified the found it necessary to make changes in the MARCH compartment configuration frma E to E'. Configura- code. For example, the plot subroutine had to be tion E' consists of five compartments: a drywell, a modified to give us graphs. In carrying out the follow-wetwell, a lower annular region (from the 135 ft eleva- ing calculations, the MARCH code stopped the run tion to 165 ft elevation), an upper annular region about 165 min into the accident, instead of going out ;
(from the 165 ft elevation to the 209 ft elevation), and the full 220 min as in all other cases, an upper containtnent volume. Examination of the The behavior expected for configuration E' was HECTR configuration E results shows that the pie- multiple wetwell burns, inerting of the wetwell, burns shaped sector volume, eliminated in going from con- in the lower annular region, inerting of that region, figuration E to E', is of great importance to the hurns in the upper annular region, inerting of that l
138
region, and finally burning in the upper contiinment. region, which was then inerted due to oxygen deple-This is roughly the behavior that was found, but tion. The final three burns took place in the upper MARCli did not predict an upper containment burn annular region, which was then inerted due to oxygen during the first 165 min of the accident. depletion. At 165 min into the accident the lower three in case E'-l the hydrogen mole fraction for igni- containment compartments were inerted due to oxy-tion was 0.10 and sprays came on after the first two gen depletion, but the upper containment still had a wetwell barns. The initial pressure was 14.7 psia and considerable amount of oxygen. The wetwell and low-initial temperature was 80'F. The mole fraction for er annular region had very high mole fractions of extinguishment was 0.002 hydrogen, burn time was 2 s hydrogen, the upper annular region had a mole frac-and flame propagation was allowed into all compart- tion of hydrogen approaching 0.20, but the upper ments with more than 0.00211. 2 Graphical results are containment still had very little hydrogen. It is not shown in Figures 3.59 through 3.75. MARCII predict- clear if there would have eventually been an upper ed seven burns mainly in the wetwell, after which the containment burn if the MARCH calculation had wetwell stayed inerted due to oxygen depletion. The continued. The peak pressure predicted without an next four burns took place mainly in the lower annular upper containment burn was low,35 psia (2.4 atm).
I40.0 Ignition Fraction - 0.10 H, ca ,
Flame Propagation - YES
]
._ Extinguishment Fraction - 0.002 H, CD Burn Time - 2 sec 100 0 - Sprays - AUTO Q.
H Z 80 0 -
i4 2
b
< 800 -
O.
l O U 40.0 -
l 1 J p 20 0 -
[ ,
00 , , , , , , , , , ,
OD 200 400 000 80 0 100.0 12a0 1400 100 0 18a0 20a0 220 0 TIME - (MINUTE)
Figure 3.59. Case E', Pressure in Containment, psia l
l 1
139
2500 0 ','
Ignition Fraction - 0.10 H, ,
Flame Propagation - YES cd Extinguishment Fraction - 0.002 H, g 200a0 - Burn Time - 2 sec Sprays - AUTO Q
Q:
c.2 1500D -
m E-=
b Z
1000D -
1 12 Q.
3 o 500 0 -
U I L d _i II _1 -
OD , , , , , , , , , ,
OD 20 0 400 00 0 80.0 100.0 120.0 140.0 1600 18a0 2000 220.0 TIME - (MINUTE)
Figure 3.60. Case E', Temperature in Wetwell, *F to Ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, y 08 - Burn Time - 2 sec Sprays - AUTO Q
lI: 08-M Z
O N< 0.4 -
h fl.
Q: r Ca.
Cd 3
O y 02-00 20 0 40.0 00 0 80 0 10a0 120 0 140.0 100 0 18a0 200 0 220.0 TIME - (MINUTE)
Figure 3.61. Case E', Ilydrogen Mole Fraction in Wetwell 140
1.0 ,
Ignition Fraction - 0.10 Ha Flame Propagation - YES Extin2uishment Fraction - 0.002 H OA -
Z Burn Time - 2 sec Ed c Sprays - AUTO
> i
\
X OA-Z 9
b
-C Q: 0.4 -
Ca.
Ed 3
O 2
~1
/
0.0 , , , ,
OD 20 0 40.0 00 0 800 100.0 120.0 140.0 180 0 180.0 20a0 220.0 TIME - (MINUTE)
Figure 3.62. Case E', Oxygen Mole Fraction in Wetwell to ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 Ha OA- Burn Time - 2 sec 7
-C Sprays - AUTO
[4 b
Z 0A-9 b
-c D: n Is. 0.4 - 0 i .
ca
.J O
2 02 -
/
W . . . . . . . . . .
OD 20 0 40.0 80.0 80.0 100.0 120 0 140.0 180 0 18a0 200.0 220.0 TIME - (MINUTE)
Figure 3.63. Case E', Steam Mole Fraction in Wetwell 141
2600 0 Ignition Fraction - 0.10 H, Flame Propagation - YES Ca3 Extinguishment Fraction - 0.002 H.
M 20000 - Burn Time - 2 sec g
y Sprays - AUTO M
Ln ca2 1500D -
H E--
Z 1000 0 -
t Q.
O 500.0 -
U l Y I -
OD , , , , , , , . . .
OD 20 0 40.0 00.0 800 100.0 12GO 140 0 100 0 18 0.0 200.0 220.0 TIME - (MINUTE)
Figure 3.64. Case E', Temperature in I,ower Annular Region,'F ID #
Ignition Fraction - 0.10 H.
Flame Propagation - YES Z Extinguishment Fraction - 0.002 H, 0A-Ca3 Burn Time - 2 sec b
2 Sprays - AUTO C
>=
Z 0A -
b< 0.4 -
Q*
is.
Ca3 I
h 02-TM OD y
200 I
40 0 5
00 0 J 80.0 3
I IMO I
12a0 E
14a0 B
1000 I
1800 I
200.0 220.0 TIME - (MINUTE)
Figure 3.65. Case E*, Hydrogen Mole Fraction in lower Annular Region 142
to j Ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, 0A -
Z Burn Time - 2 sec ta o Sprays - AUTO 0A-Z O i ct: 04-Ca.
Ca3 3
O G2 - -
l ff OD , , , ,
h , ,
CD 20 0 40.0 SOD 80D 10a0 12 0.0 140.0 100 0 10a0 20a0 223.0 TIME - (MINUTE) rigure 3.66. Case E', Oxygen Mole Fraction in Lower Annular Itegion 10 Ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, 2 Burn Time - 2 sec
$ Sprays - AUTO M
Z 04-9
[3 (E
ta. 04-Ca2 2
(
O 2
02-
~
~t 0 Q 00 20 0 400 00 0 80 0 1200 100.0 14 0.0 160 0 180.0 20a0 220 0 TIME - (MINUTE)
Figure 3.67. Case F/, Steam Mole Fraction in Lower Annular llegiori l
l 143
2500.0 l Ignition Fraction - 0.10 H Flame Propagation - YES Ca3 Extinguishment Fraction - 0.002 He O~
b Burn Time - 2 sec M
Sprays - AUTO ca3 C1.
y 15G30 -
ta3 E-*
E-.
Z 1000D -
t Cl.
3 O 500D -
Y l y Lbk 08 - . . . , , , , , ,
OD 20 0 40D 00D 000 10(10 120.0 IMO 100 0 18(10 20(10 220.0 TIME - (MINUTE)
Figure 3.68. Case F/, Temperature in Upper Annular Region. *F ID Ignition Fraction - 0.10 H Flame Propagation - YES z ,
Extinguishment Fraction - 0.002 H C'3 Burn Time - 2 sec Sprays - AUTO o
Z QA -
Z O
b< 0.4 -
M Ca.
Ca3 3
I O
y 02-
" 3 i , s , , , , ,
OD 20 0 40.0 00D 000 120 12(1 0 IMO 100 0 18110 20110 220.0 TlME (MINUTE)
Figure 3.69. Case F/. Ilydrogen Atole Fraction in Upper Annular Region 144
ID Ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H,
{ Burn Time - 2 sec o Sprays - AUTO 0A-Z 9
l C<
M 0.4 -
Ca.
ta3 2
0 W-C OD OD 20 0 404 00D 800 100.0 120 0 14Q0 100 0 180.0 20a0 220.0 TIME - (MINUTE)
Figure 3.70. Case E', Oxygen Mole Fraction in Upper Annular Region 10 Ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 Ha 0A- Burn Time - 2 sec i
y y Sprays - AUTO b
z QA -
9 b
M ta. 04-Ca3 3
0 2
02-l 0
Y . . . . . . . . . ,
OD 20 0 40D 80 0 800 10E0 1230 1400 100 0 180 0 20C0 220.0 TIME - (MINUTE)
Figure 3.71. Case E', Steam Mole Fraction in Upper Annular Itegion i
i 145 1
1
2600 0 Ignition Fracti on - 0.10 H, Flame Propagation - YES ta3 Extinguishment Fraction - 0.002 H, g 20000 - Burn Time - 2 sec Q Sprays - AUTO N
w hW 15000 -
E-=
E--
Z 1000.0 -
la !
l O.
- E o f000-U L M 00 , , , , , , , , , ,
00 20 0 400 80.0 80 0 100.0 120.0 140.0 180 0 180 0 200.0 2200 TIME - (M1NUTE)
Figure 3.72. Case E', Temperature in Upper Containment to ignition Fraction - 0.10 H.
Flarne Propagation - YES Extinguishment Fractica - 0.002 Ha os- Burn Time - 2 sec b
x Sprays - AUTO Q
>=
lI: OA -
Z O.
b
.d; 0.4 -
2 Ca.
Ca3
.1 O
llE 02 -
-48 . , . .'" ,' . . . . .
OD 200 40 0 00D 80D 100.0 1200 140.0 180 0 180.0 200.0 220.0 TIME - (MINUTE)
Figure 3.73. Case E', Hydrogen Mole Fraction in Upper Contaimnent 146
t0 Ignition Fraction - 0.10 H.
Flame Propagation - YES g, ,
Extinguishment fraction - 0.002 Ha d Burn Time - 2 sec '
O Sprays - AUTO O
0A -
Z
.O b
% 0.4 -
k
[a3 3
O 2 __
v4 -
0.0 , , , , , _,__
OD 20 0 40.0 80.0 800 100.0 1$0.0 14O.0 l$00 l$00 2d0.0 220.0 TIME - (MINUTE)
Figure 3.74. Case E', Oxygen Mole Fraction in Upper Containment ID Ignition Fraction - 0.10 H.
Flame Propagation - YES Extinguishment Fract on - 0.002 H 8
0A- Burn Time - 2 sec y
y Sprays - AUTO b
Z QA -
O 10 0
4 0.4 -
Ca3 3
I O
2 02 -
l 5 -
I
, CD j OD $0 ND da da Idtto gfglo g4QO '
l$00 i$(10 2d(10 220.0 TIME - (MINUTE)
Figure 3.75. Case E', Steam Mole Fraction in Upper Containment i
147
antien used thia hydrogen was burned, but in no case 3.8 The Effect of Connections was the burn mole fraction of hydrogen high enough to the Drywell from the for a burn to be caHed significant. In case A 6 without a connection to the drywell, the final state of the Wetwell/ Containment wetwell was close to igniting a third burn. The resulta in the M AltCil calculations discussed in sections .
shown m, 'Iable 3.5 imh,cate that the effect of the 3.3 to 3.7, a connection was used from the drywell h) d is N lower the peak pressure (for the TPE the wetwell, but no connectm, n was used back to the accident using configuration A), but by less than 10 drywell. Our mmlels isolated the drywell so that com- .
psia (0.7 atm).
pari ons could be made with the results of the ,
Cases A4 and A.9 mmlel 1,0CAs with a break 11EUI 11 cmle. In th,is section, we mehule the effects of inside the drywell. Graphica! results for case A4 are the drywell by adding a zero flow resistance connec- shown in Figures 3.78 through 3.86, and the pressure tion from either the wetwell or a region correslumdmg history for case A-9 is shown in Figure 3.87. The to the entrance of the vacuum breakers. In Appendix results for the two cases with and without a connec-( , we mentum that flow back to the drywell may occur g g .g g ; ,ru m F dm when the suppression pool is forced d<>wn beh>w the wo cases, air is driven from the drywell early in the level of the drywell vents, and that th,s i flaw may he .
accident. The drywell is therefore m.erted due to oxy-much more important than that through the vacuum Mi 'rk proce< hire of assuming an initial breakers. CI.ASIX-3 uses a fimte-flow resistance con- .
pressure of 17 ps.ia in the wetwell and neglect.mg the nect. ion to mmlel the vacuum breaker flow,. and ap-Irywell is reasonable for the two cases shown in Table parently alsa considers How batk to the drywell 3.6. ,I,he peak pressures with the connected drywell are f.hrough the suppression pool. a little lower than without the drywell.
For the TPE accidents, all that was changed in ,
For configuration H, one has a choice of connect-this section from the corresponding cases in the previ- ing the wetwell ba+ to the drywell, s.imulatm.g back-ous section was the addition of a connection to the now throudi the suppression pool, or connectmg the drywell. For the accidents mmlelling breaks from the uppu conta,mment back to the dryweU, simulatmg reactor coolant sy% tem into the drywell, additional now through the vacumn breakers. %e have carried changes were made in the cmle to send the steam and hydrogen from the reactor system into the drywell out runs wa pow oph,ons Wuns M8, W ayo Figum 126).1 he msults for case H-2 amshown in rather than the suppression pool and to set the initial ,
pressure in all compartments to 11.7 psia (1.0 atm). ,I,able 11 It is ckar dujt for configurat,um H the numbu of wetwell burns is as highly sensitive to the The TPE accidents that are reconsidered in this sec- ,
tion are caws A4, A 7,11-2, H 6, H 7, C-2, and D-2; '.nsumpdons mah almut didywd connections as it
.a to oWu input assumpt,mns. ,The variation m the -
the drvwell i,reak accidents that are reconsidered are ,
A4, [4 and H-1. Fa psurpmtwnn th caws is sigmficant. The use
" " """"U"" fron dg uppu contamment to the For configurations A and C, there is no distinction drywell (the cho. ice used in Cf,ASIX-3) gave the low-hctween a connection through the vacuum breakers est peak pressure and avoide<1 a drywell burn.
and one through the suppression ptml For configura- \V th the flame propavn,on option used in case tion H, the vacuum breaker entrance is in the upper containment. For configuration D, the vacuum break- U4 '"", burns utend diro@ aH connutp cempas ments he msuha for ca7 R6 am simwn m yaw 3.8 er entrance is in the intermediate annular region. ,
(Figures 3.90,3.91; also I igure 3.30). W,e in heate the Case A-7 considered an ignition criterion of 0.10 numbn of burns with significant combustion of hy-hydroger; mo'c fraction and no containment sprays; drogen m, the compartment. I hese resuits indicate no case A.9 is identical except containment sprays are on.
Consistent pattern of peak pressure vs drywell connec-The results are shown in Table 33 and Figures 3.76 n.on urept that the model with the upper contmn- {
and 3.77 (also in Figures 3.12 and 3.13). For hoth cases ment to drywell connection prmluced the lowest peak in which there was a connectior. to the dr3well, some pmssums, and the wetw ell to aywell connectmn led to hydrogen did enter. Because of the flame propagation a drywell burn.
Case H 7 differs from H4 only in that the sprays were turned on after the first burn insteml of early in
'We lwlieve that the CI.ASIX 3 calculation of flow through We accident. The results for this case are shown in the vacuum breakers (1.09 ft5 total flow area) may be too Table 3.9 and Figures 3.92 and 3.93 (also Figure 3.31).
large. Details of that calculation are not available at present. The results agree in pattern with those of case B4.
14F
Case B-1 also considered a 1,0CA break inside the Using configuration D, one has two choices:
drywell, as in cases A-5 and A-9, but employed the . h1aking a connection from the wetwell to the configuration B compartment model. The results are drywell, simulating flow *hrough the suppres shown in Table 3.10 and Figure 3.94 (also Figures 3.16 sion pool through 3.24) '.n this case, the addition of a drywell . hiaking a connection from the intermediate an-gave a significant reduction in peak pressure, contrary nular volume to the drywell, simulating flow to the cases considered with a break in the drywell for through the vacuum breakers.
j configuration A.
In configuration C, the vacuum breakers are in the We have carried out a case D : calculation with a
- large wetwell volume, so for h1 ARCH it does not connection from the wetwell to the drywell and then l matter if one assumes flow through the suppression from the intermediate annular region to the drywell.
pool or the vacuum breakers. The results for case C-2 The pressure histories are shown in Figures 3.108 and with a wetwell.to-drywell connection are unique. 3.109. The results are shown in Table 3.12. The addi-Graphical results are shown in Figures 3.95 through tion of a connection from the wetwell to the drywell I
3.107. This h1AllCil calculation ir. the only one in resulted in a somewhat lower peak pressure and one which there was no uppersontainment burn (except drywell burn. The addition of a connection from the possibly for case F/-1). h1 ARCH predicted six wetwell intermediate annulcr region to the drywell led to a burns. At the en(1 of the run, the upper containment slight increase in peak pressure and also one drywell had mole fractions of 0.082 for hydrogen and 0.084 for burn.
oxygen. The mo?e fractions in the wetwell were 0.487 for hydrogen and 0.029 for oxygen. Clearly, if the 3.9 References atmosphere werc allowed t<> mix, or a bit more hydr * M. Herman, ed, Light Water #cactor Safety Re-gen were produced, or the ,gmtson s hmnt for hydrogen search Program &miannual Report, April-September mole fraction were reduced, there would be an upper- 1981, NUREG/CR.2481, SAND 82-0006, February 1982.
containment burn. Without an upper-containment arR. O. Wooton and H. I. Avci, AfARCH (Afaltdoten burn, the peak pressure was low,27 psia (1.8 atm).The Accident Response Characteristics Code Description and results are compared in Table 3.11 to the case without User's Afanual, Battelle Columbus Laboratories, a drywell connection (Figures 3.95 through 3.107; also NUREG/CR-1711, HMI.2064 R3 October 1980.
Figure 3.41). 38,J. IL Rivard et al, Interim Technical Asses mer t of
, the AfARCH Code, Sandia National 1.aboratories, 1
NUREG/CR-2285, SAND 81-1672 R3, November 1981.
Table 3.5. Effect of Connections to the Drywell on Cases A-6 and A-7 ,
Case A-6 Case A-7 No No Connection Connection Connection Connection l l'eak I'ressure (psia) 80 71 56 51 (atm) 5.4 4.8 3# 3.5 No. of Significant Burns in 2 3 3 3 Wetwell No. of Significant llurns in 0 0 0 0 Drywel!
i 149 l 2
j l
140.0 Ignition Fraction - 0.10 H, g Extinguishment Fraction - 0.002 H,
~
g Burn Time - 8 sec Sprays - OFF g 100D -
L F
Z 80 0 -
W i 2
< 800-O.
2 o
O 40D -
.4 Q ADD -
OD , , , , , , , , , ,
OD 200 40D 00.0 80 0 10GO 1200 1400 100 0 18(13 20a0 220 0 TIME - (MINUTE)
Figure 3.76. Case A-6 (With Containment to Drywell Connection), Pressure in Containment, psia 140D Ignition Fraction - 0.10 H, Extinguishment Fraction - 0.G0 H, y laDD ' Burn Time - 8 sec U Sprays - ON 100D -
g Q.
H Z 30D -
W 2
< 80D-O.
2 o l U 4on -
l J
20D - >
OD , , , , , , , , , ,
OD 20 0 40.0 00 0 80D 10(L0 1200 14a0 1000 18a0 20n0 220D TIME - (MINUTE)
Figure 3.77. Car,e A-7 (With Containment to Drywell Pennection), Pressure in Containment, psia 150
140D Ignition Fraction - 0.08 Ha c,3 ,
Extinguishment Fraction - 0.00 H,
$ Burn Time - 8 sec Spreys - AUTO ID0D -
Q.
E~
Z SOD -
W 2
4 80D -
O.
2 o
O 404 -
J
^
j mD-E~ e-i OD , , , , , , , , , ,
OD 300 40D 004 80D 10(LO 12 0.0 1440 1800 18Q0 20Q0 EBILO TIME - (MINUTE)
Figure 3.78. Case A.5, Pressure in Containment, psia 2500D Ignition Fraction - 0.08 H, Extinguishment Fraction - 0.00 H, ta3 Burn Time - 8 sec ME -
$ Sprays - AUTO a:
C.2 O.
y 15060 -
Es3 E~
fa b
y IC%D -
b i
Q.
2 O SCOD -
U l '
I
% w_
OD , , , ,
04 300 40D N ED 10110 laQO 1440 1800 1810 EltLO EBOD TIME - (MINUTE)
Figure 3.79. Case A.5, Temperature in Containment, 'F 151 l
t
LD Ignition Fraction - 0.08 H Extinguishment Fraction - 0.00 H Burn Time - 8 sec h 0A - Sprays - AUTO 8
x Q
>=
T, QA-Z O
w: 0.4 -
M Ca.
Ca3 A
O y 02-
)
CD , , , , , , ,
OD 200 40D 00.0 SDD 10(LO 12E10 IMO 1000 18(10 20(L0 230D TlME - (MINUTE)
Figure 3.80. Case A-5, liydrogen Mole Fraction in Containment to ignition Fraction - 0.08 H Extinguishment Fraction - 0.00 H, Burn Time - 8 sec OA-Z Sprays - AUTO c.3 0
O g,,
Z O
b
<C M Q4-Ca.
Cn3 A (
O 02 L
OD , . , , , , , , , ,
OD 200 40D GOD SDD 10GO 12(10 IMO leQO lELO 20110 2204 TIME - (MINUTE)
Figure 3.81. Case A 5, Oxygen Mole Fraction in Containment 152
LO Ignition Fraction - 0.08 H, Extinguishment Fraction - 0.00 H, Burn Time - 8 sec y os- Sprays - AUTO ta2 b
Z 0A -
9
[3 M
Ca. 0.4 -
CaJ 3
o CD , , , , , , , , , ,
OD 20 0 40D SOD 80 0 10E0 1200 14a0 100 0 18G0 20a0 220 0 l
TIME - (MINUTE) l Figure 3.82. Case A-5, Steam Mole Fraction in Containment 2500 0 Ignition Fraction - 0.08 H, Extinguishment Fraction - 0.00 H, taJ Burn Time - 8 sec h
~
Sprays - AUTO 4
x t.2 150Q0 -
Ca3 i
b \
E-=
b y 1000D -
Q.
3 O SOOD -
U f
On , , , , , , , , , ,
OD 20 0 40D 80 0 80D 10E0 !!nD 14a0 100 0 18WLO 20a0 220.0 TIME - (MINUTE)
Figure 3.83. Caw A.5, Temperature in Drywell. *F 153
j to Ignition Fraction - 0.08 Ha 4 Extinguishment Fraction - 0.00 H Burn Time - 8 sec 0A -
i Sprays - AUTO 8
x C
> 1 x GA -
5 N
b
< 0.4 -
N k
Ca2 3
0 3 02 - ,
l 08 . . , , , , , , , ,
OD 200 40.0 00.0 SDD 1040 12 0.0 140 0 1000 18 0.0 20a0 220.0 TIME - (MINUTE)
Figure 3.84. Case A.5, Ilydrogen Mole Fraction in Drywell 10 Ignition Fraction - 0.08 H, Extinguishment Fraction - 0.00 H, Burn Time - 8 sec 0A-Z Sprays - AUTO Cn3 0
x 0
0A-Z 9
b
% 0.4 -
k Ca3 A
o 02 -
' ^ l ^
I I I E I , 5 3 , g I OD 200 40D 00D 80D 1000 lano 1400 100 0 18a0 2Da0 330.0 TIME - (MINUTE)
Figure 3.85. Case A.5, Oxygen Mole Fraction in Drywell 154
tD hnition Fraction - 0.08 H, Extinguishment Fraction - 0.00 H,
, 8 urn Time - 8 sec 2 Sprays - AUTO
[a3 b
2 DA-
.9-b 0:
Ca. 0.4 -
EaJ J
O 2
02 -
OD -
OD 20 0 40D 00D 80D 1000 12 0.0 1400 180 0 1800 2000 220.0 TIME - (MINUTE)
Figure 3.86. Case A.S. Steam Mole Fraction in Drywell 1400 Ignition Fraction - 0.10 H, Extinguishment Fraction - 0.00 H, WO -
Burn Time - 8 sec Sprays - AUTO
$ 100 0 -
O.
H Z 80D -
E.J 2
< 800-O.
2 o
O 40-3 800-OD , , , , s , , , , ,
00 200 40D 00D 80D Kd0 120D 1440 180 0 1840 3100 230D ;
TIME - (MINUTE) '
Figure 3.87. Cane A.9, Pressure in Containment, psia 155
Table 3.6.- Effect of Connections to the Drywell on Cases A-5 and A 9 Case A-6 Case A-7 No No Connection Connection Connection Connection Peak Pressure (psia) 76 69 78 75 5.2 4.7 5.3 5.1 (atm)
No. of Significant llurns in 4 3 3 3 Wetwell No. of Significant llurns in 0 0 0 0 Drywell Table 3.7. Effect of Drywell Connections on the Results of Case B-2 No Connection Wetwell to Upper Containment to to Drywell Drywell Connection Drywell Connection 1
No. of Upper Containment 1 1 Iturns Peak Pressure (psia) 52 44 40 3.5 3.0 2.7 (atm) 16 20 21 No. of Wetwell Ilurns No. of Drywell Iturns 0 1 0 Table 3.8. Effect of Drywell Connections on the Results of Case B 6 No Connection Wetwell to Upper Containment to to Drywell Drywell Connection Drywell Connection No. of Uppe[ Containment 1 1 2 ,
Ilurns Peak Pressure (psia) 63 66 60 4.3 4.5 4.1 (atm) _
No. of Significant llorns in 0 1 0 0 1 0 Dr>well No. of Significant llurns in 13 15 18 Wetwell 156
140 0 Ignition Fraction - 0.10 H, Flame Propagation - NO 120 0 -
Extinguishment Fraction - 0.00 H, h Burn Time - 4.8 sec b 100 0 - Spreys - AUTO 0:
Q.
H Z 800-i W
l E
< MO-0- 1 5 '
O O 40 0 -
3 h
R 20 0 -
w A......... .
00 , , , , , , , , , ,
00 20 0 40.0 00 0 80 0 100 0 120 0 1400 160 0 180 0 200 0 220.0 TIME - (MINUTE)
Figure 3.88. Case 11-2 (With Containment to Drywell Connection), Pressure in Upper Containment, psia 1400 Ignition Fraction - 0.10 H, g Flame Propagation - NO 130.0 -
g Extinguishment Fraction - 0.00 H, f4 Burn Time - 4.8 sec 100 0 - Sprays - AUTO '
0:
O.
H Z 80.0 -
w 2
< 800-O.
3 o
O 40D -
3 h
h} 20D - g. _ -- - d,,,_,__
i l
OD , , , , , , , , , ,
00 20 0 40D 00D 80 0 10n0 1200 1400 100 0 18110 2E110 220D TIME - (MINUTE)
Figure 3.8g. Case 112 (With Wetwell to Drywell Connection), Pressure in Upper Containment, psia 157
1400 Ignition Fraction - 0.10 H, y Flame Propagation - YES Extinguishment Fraction - 0.002 H,
$ Burn Time - 2.0 sec 100 0 - Sprays - ON g
Q.
E-=
Z 80D-W 2
b< ikl 0 -
O.
5 o
O 400-3 h '
- (__. . . _
OD , , , , , , , , , ,
OD 20 0 400 00 0 80 0 100.0 120 0 140.0 180 0 180.0 2000 220D TIME - (MINUTE)
Figure 3.90. Case 114 (With Containment to Drywell Connection), Pressure in Upper Containment, psia 140 0 Ignition Fraction - 0.10 H, m Flame Propagation - YES 0-Extinguishment Fraction - 0.002 H, Burn Time - 2 sec g 100 0 - Sprays - ON Q.
E-=
Z 80 0 -
W 2
s-.
M 800-O.
2 o
O 40.0 - I 3
h
[2 20D - . _.
OD . . . . . . , , . .
OD 20 0 40D 80D 80D 10E0 12nD 1440 1800 181LO 201L0 220D TIME - (MINUTE)
Figure 3.91. Case B-6 (With Wetwell to Drywell Connection), Pressure in Upper Containment, psia 1
158
Table 3.9. Effect of Drywell Connections on the Results of Case B-7 Connection Wetwell to Upper Containment to Drywell Drywell Connection to Drywell Connection No. of Significant Upper 2 1 2 Containment Ilurns Peak Pressure (psia) 62 71 60 (atm) 4.2 48 41 No. of Significant Drywell 0 1 0 Iturns No of dignificant Wetwell 14 15 16 Iturns 1400 Ignition Fraction - 0.10 H.
Flame Propagation - YES IEE~
, Extinguishment Fraction - 0.002 H, g Burn Time - 2 sec Q goog _ Sprays - AUTO I Q.
H Z SOD -
c.2 2
h d 80D-Q.
\
I O O 400 -
J E-12 2n-
. Mt. . , ,
OD , -, , . . . . . . .
OD 20 0 40.0 00D 800 1000 1200 1400 100 0 18 0.0 2000 220.0 TIME - (MINUTE)
Figure 3.92. Case 117 (With Containment to Drywell Connection), Pressure in Upper Containment, psia j
1 159
1400 Ignition Fraction - 0.10 H Flame Propagation - YES IMO- Extinguishment Fraction - 0.002 H, Burn Time - 2 sec 1000 - Sprays - AUTO Q:
Q.
F=
Z 80D -
ta3 2
b4 000-Q.
3 o
V 40D -
A 20D - _.
CD i i i i i OD 20 0 40D 80D 800 10(10 120.0 14Q0 l$00 INEO 2da0 220D TIME - (MINUTE)
Figure 3.93. Case 11-7 (With Wetwell to Drywell Connection), Pressure in Upper Containment, psia Table 3.10. Effect of Wetwell to Drywell Connection on Case B-1 No Connection Wetwell to Drywell to Drywell Connection
~
1'eak Pressures (psia) 65 52 (atm) 44 3.5 No. of Wetwell Ilurns 15 14 g Ilefore First Wetwell Inerting No. Upper Containment 1 1 ;
llurns No. Drywell Iturns 0 0 Total No. Of Wetwell 16 19 Ilurns 160
i 140 0 Ignition Fraction - 0.10 Ha g Flame Propagation - NO 120D -
Q: Extinguishment Fraction - 0.00 Ha D
Burn Time - 4.8 sec 100 0 SPrays - AUTO 0:
C.
H Z 80D -
La.1 2
b
< SOD -
C o
O 40.0 -
1 h
12 =0- _.4 - -
(..._ __ .
OD- , , , , , , , , , ,
OD 20 0 40.0 00D 80D 100 0 1200 14a0 160.0 18a0 20Q0 220.0 TIME - (MINUTE)
Figure 3.94. Case 11-1, Pressure in Upper Containment, psia Table 3.11. Effect of Wetwell to Drywell Connection on Case C-2 No Connection Wetwell to to Drywell Drywell Connection No. of Significant Upper 1 o Containment Ilurns Peak Pressure (psia; 48 27 (atm) 3.3 1.8 No. o.' Significant Wetwell 8 6 Ilurns No. of Significant Drywell 0 e llurns l
161
140 0 Ignition Fraction - 0.08 H, g Flame Propagation - YES
~
% Extinguishment Fraction - 0.002 H, D
Burn Time - 4 sec 1000 - SPrays - AUTO g
Q.
H Z 80D -
W 2
b< 80D-Q.
2 o
O 40D -
3 20A ~
OD , , , , , , , , , ,
OD 20 0 40.0 00D 80D 10Q0 120 0 140.0 100 0 18a0 20a0 220.0 TIME - (MINUTE)
Figure 3.95. Case C-2 (With Wetwell to Drywell Connection), Pressure in Upper Cor.tainment, psia 2500 0 Ignition Fraction - 0.10 H, Flame Propagation - YES W Extinguishment Fraction - 0.002 H, Q: 2000 0 -
D Burn Time - 4 sec kQ: Sprays - ON W
Q.
y 1500.0 -
W E-*
b Z
10000 -
f m
Q.
3 l o 500.0 - 1 0
L ( L L. O OD , , , , , , , , , ,
OD 20 0 40D 80.0 80D 100.0 120.0 146.0 100 0 18a0 2Da0 220.0 TIME - (MINUTE)
Figure 3.96. Case C.2 (With Wetwe'.. to Drywell Connection), Temperature in Wetwell, *F 162
to ignition Fraction - 0.10 H, Flame Propagation - YES 7 ,
Extinguishment Fraction - 0.002 H,
[^3 Burn Time - 4 sec b
x Sprays - ON Q
Z OA-Z O
b-c 0.4 -
=
k Es3 A
O i
y 02-OD , , ;
/ W/ -
OD 20 0 40D 004 80D 100.0 12 0.0 1440 160 0 18a0 20GO 220.0 TIME - (MINUTE) l Figure 3.97. Case C-2 (With Wetwell to Drywell Connection), Hydrogen Mole Fraction in Wetwell l
ID Ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, 2, 0A - Burn Time - 4 sec
$ Sprays - ON 08-Z
, O
~
L 0* 44-k l
Es3 A
O 02- -- _
Ir OD 5 5 5 5 5 5 g 3 , ,
OD 20 0 400 000 80D 1000 tano 1420 10 0.0 lea 0 2000 220.0 TIME - (MINUTE)
Figure 3.98. Case C-2 (With Wetwell to Drywell Connection), Oxygen Mole Fraction in Wetwell f
163 l
t0 Ignition Fraction - 0.10 H.
Flame Proparp:lon - YES Extinguishment Fraction - 0.002 H 08-y Burn Time - 4 sec
$ Sprays - ON b
Z 08-O b
Cr:
Ca. 0.4 -
W J
O 2
02-a L LLL 0.0 , , , , , , , , , ,
OD 20 0 40D 00.0 80 0 100.0 12a0 1400 160 0 18a0 20a0 220.0 TIME - (MINUTE)
Figure 3.99. Case C-2 (With Wetwell to Drywell Connection), Steam Mole Fraction in Wetwell 25000 Ignition Fraction - 0.10 H.
Flame Propagation - YES W Extinguishment Fraction - 0.002 Ha
% 20000 -
p Burn Time - 4 sec
$ Sprays - ON W
Q.
y 1500 0 -
W t--
f--
Z 1000.0 -
b<
Q.
3 )
o 500 0 -
U i
0.0 , , , , , , , , , ,
OD 20 0 40.0 00.0 80.0 10a0 120.0 14a0 100 0 18a0 200.0 2 .0 TIME - (MINUTE)
Figure 3.100 Case C-2 (With Wetwell to Drywell Connection), Temperature in Upper Containment, 'F 164
l 10 Ignition Fraction - 0.10 ha Flame Propagation - YES Extinguishment Fraction - 0.002 Ha OA-
[4 Burn Time - 4 sec 0
o Sprays - ON 2
Q OR -
Z
.O b
< 04-k C4 1
0 3 02 -
- f OD , , , , , , , , , ,
OD 20 0 40.0 00.0 80D 100.0 12 0.0 140.0 100 0 lino 20a0 220D TIME - (MINUTE)
Figure 3.101. Case C-2 (With Wetwell to Drywell Connection), Hydrogen Mole Fraction in Upper Containment 1.0 Ignition Fraction - 0.10 H Flame Propagation - YES Extinguishment Fraction - 0.002 H, OA- Burn Time - 4 sec Z
Sprays - ON X
02 -
Z 9
b
% 0.4 -
W C4 1
O I o2- -
0.0 , , , , , , , , , ,
OD 20 0 40D 00.0 80D 10E0 1200 140.0 100 0 18a0 20a0 2204 TIME - (MINUTE)
Figure 3.102. Case C-2 (With Wetwell to Drywell Connection), Oxygen Mole Fraction in Upper Containment 165
ID Ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H OA- Burn Time - 4 sec y
$ Sprays - ON b
Z 0s-O,,
ti<
Ca 04-ca2 3
0 2
02 -
(
OD , , , , , , , , , ,
OD 20 0 40.0 00 0 80D 100.0 120.0 140.0 100 0 18a0 20a0 220.0 TIME - (MINUTE)
Figura 3.103. Case C-2 (With Wetwell to Drywell Connection), Steam Mole Fraction in Upper Containment i
2500 0 Ignition Fraction - 0.10 H, Flame Propagation - YES
[83 Extinguishment Fraction - 0.002 Ha
% 2000D -
p Burn Time - 4 sec
< Sprays - ON c.2 hc.2 1500D -
E-*
E-.
Z 1000.0 -
t Q.
'E J O 300D' \
U I i
OD , , , , , , , , , ,
OD 20 0 404 e04 OOD 10E0 12n0 140.0 1800 120 200D 220D TIME - (MINUTE)
I Figura 3.104. Case C-2 (With Wetwell to Drywell Connection), Temperature in Drywell, 'F 1 166 l
1 l
(D Ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 Ha h 0A- Burn Time - 4 sec 8
=
Sprays - OPi c
lI: OA-Z O
D
< 0.4 2
Cs3 N
y 02-CD -
> < - "~. . . , , , ,
OD 200 40D 00D 80D 10a0 120.0 l4no 100D 18 0.0 200.0 220.0 TlME - (MINUTE)
Figure 3.105. Case C-2 (With Wetwell to Drywell Connection), Hydrogen Mole Fraction in Drywell 1.0 Ignition Fraction - 0.10 H Flame Propagation - YES Extinguishment Fraction - 0.002 H, OA-Z Burn Time - 4 sec ta3 i o Sprays - ON 02 -
Z O
D<
Cd 0.4 -
Es.
L Ca3 A
O 02-CD , , , , , , , , , ,
OD 20 0 40.0 00.0 80.0 100.0 120.0 14R0 1600 18a0 200.0 220D TIME - (MINUTE)
Figure 3.106. Case C-2 (With Wetwell to Drywell Connection), Oxygen Mole Fraction in Drywell 167 l
I0 Ignition Fraction - 0.10 H, Flame Propagation - YES Extinguishment Fraction - 0.002 H, 08- Burn Time - 4 see y
y Sprays - ON b
Z 0A-3 D
M L 0.4 -
D.2 2
0 2
02-00 , , , , , , , , , ,
OD 20 0 400 60.0 80 0 100 0 120.0 1400 160 0 180.0 200.0 220.0 TIME - (MINUTE)
Figure 3.107. Case C.2 (With Wetwell to Drywell Connection), Steam Mole Fraction in Drywell teos Ignition Fraction - 0.10 H, Flame Propagation - YES W ISDA - Extinguishment Fraction - 0.002 H, g
D Burn Time - 2 sec Sprays - AUTO NinD -
O.
H Z OOD -
W 3
4 ND-D.
3 o
O 40D - j 3
8 =n- m t C L , . ,_ . _ .
OS . . . . . . . . . .
04 WD 40D 000 50D 10lLO 1300 1440 1000 18110 SOELO 23110 TlME - (MINUTE)
Figure 3.108. Case D.2 (Wetwell to Drywell Connection), Containment Pressure, psia 168
140 0 Ignition Fraction - 0.10 H, g Flame Propagation - YES
% IND-Extinguishment Fraction - 0.002 H, D
Burn Time - 2 sec 100 0 - Sprays - AUTO O= I I
H Z 80D - !
w <
2 j 12
< 80.0 -
O.
3 O i V 400 -
A h
8 20 0 -
%O L L . , . . . . _. C .. .
On , , . , , . , , , ,
OD 20 0 40.0 80.0 80D 100.0 I200 140.0 180 0 1800 200.0 220.0 TIME - (MINUTE)
Figure 3.109. Case D-2 (Intermediate Annular Itegion Connection to Drywell), Containment l'ressure, psia Table 3.12. Effect of Wetwell to Drywell Connection on Case D-2 Intermediate No Connection Wetwell to Drywell Annular Region to to Drywell Connection Drywell Connection l'eak I'ressure (psia) 46 41 50 (atm) 3.1 2.8 3.4 No. of Significant llurns 1 1 1 in Upper Containment No. of Significant llurns 3 4 4 in Intermedidate Annular Volume No. of Signifier.nt llurns 9 12 11 in Wetwell i No. of Significant Burns 0 1 1 l in Drywell 169 170
i i
4.0 RALOC Hydrogen Transport Calculations for the Grand Gulf Nuclear Station 4.1 Summary liydrogen transport calculations made with the RALOC nodalization. Once the bottom compartment computer code RALOC are discussed in this chap- reached 8?;, an additional 5 min were typically re-ter. The objective of these calculations was to provide quired to bring the upper compartment to the same an estimate of the volumetric hydrogen concentration value. As expected, the hydrogen transport was domi-in the upper compartments of the Grand Gulf con. nated by convection.
tainment building when the bottom compartment For initially isothermal conditions, the hydrogen achieves a reliable combustible hydrogen level at the concentration becomes uniform (independent of verti-igniters (assumed to be approximately 8?e to 10?L). A cal and azimuthal location) within 15 to 20 min after sensitivity study, rather than a best-estimate tech- the end of injection. Inverting the initial temperature nique, was used to attain this objective. This was done distribution in the containment produced nonhomo-for two reasons: First, there are large uncertainties geneous hydrogen concentrations as late as 10 000 s associated with the hydrogen release models following (2.8 h) into the calculation (there was a lower hydro-a LOCA; second, the RALOC code itself has had only gen concentration at the top than at the bottom).
limited assessment to date.** It must be noted that all Homogenization of the hydrogen concentration was RALOC calculations involve precombustion condi- inhibited by a thermal barrier produced by the tem-tions-no mixing during or after combustion was perature distribution.
considered. The concentration of hydrogen achieved at the Two groups of parameters were varied in this end of the calculation (10 000 s) is clearly a function of study. The first group contains parameters which test the total hydrogen injected. Injecting 500 lb of hydro-the reliability of the code. These are the maximum gen produced volumetric concentrations on the order time step allowed during the implicit integration of of 30?e. Injecting less hydrogen produced lower final the differential equations, and the number of zones concentrations, as expected.
and connections used to model the Grand Gulf con-tainment (nodalization). The second group consists of physical parameters which are either uncertain in the 4.2 Introduct. ion hydrogen source description or are known to affect the Predicting the vohimetric concentrations of hy-transport of hydrogen in a containment. They are the drogen in individual compartments of a nuclear power amount and rate of hydrogen injection, the hydrogen plant following a degraded. core accident is extremely injection temperature, and the initial temperature important for reactor-safety calculations. Knowledge distribution present in the containment building. of these concentrations could be crucial to the design Although the quantitative results of our sensitiv- of possible hydrogen mitigation schemes.
ity studies varied a little, the qualitative 'results were Sandia has recently obtained the computer code the same. During the injection period, nonhomogenei- "RALOC-MODI with 1980 updates"in order to per-ties existed in the hydrogen volumetric concentra- form hydrogen transport calculations for severe acci-tions. Increasing the rate of hydrogen injection, or dents. The code was obtained from GRS (Gesell-inverting the temperature distribution (higher tem- schaft fur Reaktorsicherheit) under an NRC-BMI perature at the top of the containment than at the (Bundesministerium des Innern) agreement consis-bottom) produced larger asymmetries. (Temperature tent with the existing NRC-BMFT (Bundesminister-inversions appear to be possible if containment sprays ium fur Forschung und Technologie) agreement.
cre inoperable, especially if burns have occurred pre. Using RALOC, calculations were performed to viously.) When the bottom compartment reached the analyze the transport of hydrogen in the Grand Gulf essumed lower combustion limit of 896, the upper Nuclear Station following a hypothetical degraded-containment levels had a hydrogen concentration be- core accident. In particular, we wanted to determine tween 3?h and 6"c, depending on the rate of hydrogen the upper-containment volumetric hydrogen concen-injection and the degree of resolution used in the trations when the bottom compartment concentration 171
was near some lower combustion limit for hydrogen (hydrogen, helium-steam injection into a multiroom igniters. The calculations were performed using a geometry).*
- sensitivity approach in which parameters relevant to the calculation were varied over a wide range to deter mine their overall relative importance. Qualitative 4.4 RALOC Grand Gulf Models conclusions can be drawn from this study, but no Figure 4.1 shows a diagram of the Grand Gulf Nuclear Station containment building. Figures 4.2 precise quantitative results should be deduced.
The following sections will discuss the RALOC and 4.3 show coarse- and fine-zone ItALOC models for the Grand Gulf containment building. Zone numbers computer code, the ItALOC models for Grand Gulf, are shown in large type and connection numbers in hydrogen source models used for the transport calcu.
lations, two ItALOC code-reliability calculations, and small type, and the arrows indicate the initial gas now directions assigned (negative flows are in a direction the sensitivity studies performed for the physical pa.
opposite to the initial assigned direction). Altitudes of rameters.
the zone centers are given in meters to the left of each diagram. Volumes and flow areas for these nodaliza-4.3 RALOC tions were based on values calculated by itALOC is a computer code developed by 11. Jahn M. P. Sherman." Cells 8 and 17 were used as hydro-of GitS under the sponsorship of IlMI. IIALOC was gen sources in the coarde- and fine-nodalization mod-designed to calculate time histories of local gas con- els, respectively. These sources simulate injection of centrations in subdivided containments. In particular, hydrogen into the annular region cf the containment ItALOC was developed to determine the distribution building between 110 and 135 ft, just above the sup-of radiolytically-produced hydrogen in a nuclear plant pression pooi. Zones 1 through 3 of Figure 4.2 repre- l following a LOCA. Hydrogen concentrations calculat- sent the dome and regions above an elevation of 209 ft. l ed with ItALOC represent mole percentages. For ideal Zone 5 represents a pie-shaped shaft at azimuth 225*
l extending from 135 to 208 ft. Zones 4 and 6 model the j gases, the mole percentage is equivalent to the volu-metric percentage. annular regions between 135 and 209 ft, and zone 8 l
in solving the hydrogen transport problem, models the entire annular region from 110 to 135 ft IIALOC uses lumped-parameter control volumes con- elevation. Zone volumes are given in Tables 4.1 and 4.2 nected by Dow junctions. Within a given zone, up to for the coarse- and fine-zone models, respectively.
five constituents can be treated assuming thermody-namic equilibrium end saturated steam. Possible con-stituents are liquid water, steam, hydrogen, air, nitro-I gen, and helium. .lunction gas flow is driven by both i n 8 C convection and diffusion and is opposed by resistance I between the zones. Differential equation integration is "" N explicit during the problem startup and fully implicit courainment.
after the calculation has stabilized. I Other code options include fans /recombiners, op. ,
, u,, y .,,.
tional heat transfer to constant-temperature heat y q sinks, and optional beat transfer to and from variable- "'**"
temperature heat sinks with one heat slab allowed per ontrol volume. In addition, the user can specify one ME su n -o-zone as a hydrogen : air source and can specify a a,5,^, c,T,0,a _ __[_
coolant path for transport of recirculation coolant
' y ~
l j
_ u ,g ,,,,,. I water with dissolved gases.
- "*"- - ""8*
To date, llALOC has had limited assessment. - I DRYWEn - l l l Qualitatively, good agreement has been obtained with ,
ru n -c l the experimental data of Ilattelle-Frankfurt (multi- Protstat- --
I r l suppRresion - .-
l room geometries with various initial temperature dis- %
l tributions and hydrogen injection rates)." ** Addi-g g "'" #
tional verification of IIALOC will be performed as more experimental data become available.
Currently, ItALOC is being modified by SNL t Figure 4.1 Simplified Diagram of the Grand Gulf Nuclear simulate the EPill/HEDL tests now being performed Station 172 t
l l
l 1 1 2 a t.4u - g s-~ ,
T s. s u ----
7s.1 = - i * .
I - t 2 s 3 s 5 e 6 ,
6 7.e ---- 4=s 67.a - -
- -7
. . n d T T
~~~~~~~
59.9------------ $
58.2 - - - - - - - - - - -
9d- h,'0 is i* +-- + "
- 6 i*
- i. 12 II 52.7 - . - - - - - - - - . , g,, 5 2.s - - - - - - - - - - -- -.---ie Sa.a --------
5 t --
- 14 7 it is 4 5. 2 -- - - - - - - - - + ie 4 5.2 ---- - - - - - - -- IS .----se y q w
" 8si
3 9.3 - - -- -- - -- - - - - 15 ~~ ss 16 3 7. 3 - - - - - -- - - -- 1
{,
3 5.4 - - - - - - - - - - - - 17 --se s* 18 Figure 4.2 IIAI.OC Coarse-Zone Model for the Grand Gulf Figure 4.3 RAI.OC Fine-Zone Model for the Grand Gulf
('ontainment fluilding Containment lluilding Table 4.2. Volumes for RALOC Fine-Zone Grand Gulf Model Zone Number Volume (m')
Table 4.1. Volumes for RALOC Coarse-Zone Grand Gulf Model 1 4.335 x 10 3 2 2.735 x 108 3 4.335 x 10 8 Zone Number 4 2.735 x 108 Volume (m')
5 5.876 x 10' 1 1.414 x 10' 6 3.701 x 10' 2 5.876 x 10' 7 6.489 x 102 3 3.701 x 10' 8 7.543 x 102 4 2.808 x 10 8 9 7.543 x 102 5 1.133 x 10' 10 7.543 x 102 j 6 3.987 x 10 8 11 5.408 x 102 7 3.728 x 10' 12 3.987 x 10' 8 4.341 x 10' 13 5.852 x 102 14 3.728 x IC' 15 5.726 x 102 16 1.506 x 108 17 3.792 x 10' 18 1.225 x 108 173
4.5 Grand Gulf Hydrogen Source 4.6 Results of a Typical RALOC Models Calculation Due to the uncertainties in the release of hydrogen IIALOC results obtained with the coarse nodaliza-following a core melt, four different time. dependent tion shown in Figure 4.2 and the model B hydrogen sources were used in this study. These are shown in injection surce shown in Figure 4.4 are given in Figure 4.4. Sources A,11, and C were used to investi- Figures 4.5 through 4.8. Isothermal initial conditions gate the effects of the rate and temperature of the were assumed in this calculation (60*C). Figure 4.5 hydrogen injection. Source A injects hydrogen twicc as shows the time histories of the aydrogen volumetric fast as B and four times as fast as C. Approximately concentrations for all zones in the calculation. Ilydro-1500 lb of hydrogen (6)0 m') and 400 m' of air were gen injection began at 200 s and ended 1800 s later. A injected with each of these sources. Curve D of Figure aniform concentration of hydrogen was predicted ap-4.4 shows a hydrogen source for Grand Gulf based on proximately 1000 s after the end ofinjection (~ 14.67; calculations performed with the computer code in all zones). When the bottom compartment (zone 8)
MAltCII." The release of hydrogen in curve D is reached a hydrogen concentration of 8?h, zones 1 and assumed to be slow for the first 45 min after the 2 (upper containment) had concentrations of approxi-accident, and then rapid for the next 15 min. The mately 3?i. Discontinuities in the slope of the hydro-release rate for curve D resulted in a total injection of gen time histories were the result of the edit frequency 504) Ib of hydrogen (total injection - 22 000 m 811, used in plotting the results (400 s),
and 11000 m8 air). This model was used primarily to Mass flows for connections 1 and 2 are shown in investigate effects caused by variations in the initial Figures 4.6 and 4.7. The mass flow in Figure 4.6 is containment temperature distribution. Injection of negative; negative flows correspond to flows in a direc-t he hydrogen took place in the annular region immedi. tion opposite to the initial direction assigned (arrow of ately above the suppression pool (llALOC does not Figure 4.2). The mass flow curves of Figures 4.6 and have a suppression pool model). 4.7 are approximate mirror images of each other be-cause most of the gas convects through zone 1. After hydrogen injection stopped (2000 s), the mass flows
- i. coa , , , , , , , ,
decreased and nearly reached zero flow at 10 000 s into the calculation.
Two convection loops were calculated for the Grand Gulf containment building by IIALOC. These loops are shown in Figure 4.8. An upper convective O c.roo -
a e c = _ h>op consists of zones 1 through 4, while zones 4 through 8 participate in a lower h>op. Zone 4 is a E
i member of both hmps and transfers hydrogen from i ,,,, _ _
the lower elevations to the upper containment.
a ga4oo
!! ..soo - -
4.7 Code Reliability Calculations Two llALOC code reliability calculations were performed in this study.These calculations were done e,oo because the code had been recently acquired and we had not had much experience in using it, and because i the code had only limited previous assessment. The
" o ooo o$oo oloo oloo o$oo t. coo first calculation varied the max, imum allowable time m mounsi step during the implicit integration in order to test the stability of the solution. The second calculation tested Fture 4.4 Ilydr igen Source Models Used in the llALOC Grznd Gulf Transport Calculations. Total injected hydro-the ability of the code to asymptotically approach a gen was approximately 1500 lb for cases A,11, and C. Case D converged solution when the nodalization was made injected 5000 lb. finer.
174
l 4.7.1 Effect of Maximum Allowable increased. To accomplish this objective, the coarse Time Step nodalization of Figure 4.2 was modified to the fine Figure 4.9 shows a comparison between two calcu-nodalization shown in Figure 4.3. Zones which were lations differing only in the maximum allowable time tall (zone 5 of Figure 4.2) were split vertically into two step used during the implicit portion of the calcula zones to reduce any gravitational effects (properties in RALOC are evaluated at cell centers). Also, zones tion (50 s vs 200 s). Decreasing the time step did not significantly affect the calculation. which had a high degree of activity or importance (zones 1,4, and 8 of Figure 4.2) were subdivided to provide smoother convective transport. To reduce I computational time, the remaining coarse zones were 4.7.2 Effect of Nodalization not changed (zones 2,3,6, and 7 of Figure 4.2).
One of the attributes of a good computer code is Comparisons between the fine-zone and coarse-that the calculated results approach an asyraptotic zone calculations using the isothermal source model D limit as the degree of resolution is increased. A calcu- of Figure 4.4 (60*C) are shown in Figures 4.10 and lation was performed to determine if the RALOC 4.11. As hoped, the RALOC calculations appear to be solution would change appreciably if the numher of farly insensitive to the degree of discretization used in zones and connections in the Grand Gulf model were the nodalization.
40.0 , , , , , , , , ,
~
~
0-PERCENT HYDROGEN M ZONE 1 35.0 -
o-PERCENT HYDaoGEN N ZONE 2 _
A-PERCENT HYDROGEN W ZONE 3 32.5 -
0-PERCENT HYDROGEN M ZONE 4 -
- -PERCENT HYDROGEN 81 ZONE 8 30.0 -
4-PERCENT HYDROGEN M ZONE S ~
g 27.5 - $-PERCENT HYDROGEN M ZONE r
- 9. PERCENT HYDROGEN M ZONE 8
~
a:
25.0 -
o 22.5 -
Z 20.0 -
$ 17.5 -
15.0 -
- ; y -
' 12.5 -
10.0 -
7.5 -
5.0 -
2.5 -
0.0 i ' ' ' ' ' ' ' ' '
O 2000 4000 6000 8000 10000 l
TIME (SEC)
Figure 4.5 IIAI,0C Grand Gulf Hydrogen Transport Pred:ction for Source Model 11 and Coarse Nodalization 175
25.0 , , , , , , , , ,
0.00 g -
-25.0 - -
-50.0 - -
E o -75.0 - -
g -100.0 - -
@ -125.0 - - .
E - 1 s0.0 - -
- -175.0 - -
o w
-200.0 - -
3 -225.0 - -
o
.J is.
- 250.0 - -
en -275.0 - -
m 4 -300.0 - -
3
-325.0 - -
-350.0 - -
-375.0 ' ' ' ' ' ' ' ' '
O 2000 4000 8000 8000 10000 TIME (SEC)
Figure 4.6 ItAI,0C Mass Flow Prediction for Connection 1 (source modelIl and coarse nodalization) 375.0 , , , , , , , , ,
350.0 -
325.0 -
N g 300.0 - -
h 275.0 -
N 250.0 - -
3
- 225.0 - -
E 200.0 - -
E s 175.0 - -
o 5 150.0 - -
R 125.0 - -
o
(;8 100.0 - -
$ 75.0 - -
l 50.0 - -
25.0 -
0.0 -
-25.0 ' ' ' ' ' ' ' ' '
O 2000 4000 6000 8000 10000 TIME (SEC)
FigJre 4.7 ItAI,0C Mass Flow Prediction for Connection 2 (source model 11 and coarse nodalization) 176
1 1
18 s h -----
i . d T t 2 iU 3 8
-~
er.a---- <.
. . , I T
4.> T.
s o . o - - -- ----- -----
i d .
L. T, u 6 ss.1------~~----- +,e 4a.s------------. 5 t
___ 7 l 4 5. 2 ---- --
+"*
d i i n 8 is U s7.3------------
Figure 4.8 ItAI.OC Calculated Convection I, oops for Grand (iulf Containment. Small arrows indicate flow direction.
1.arge arrows represent the initial gas direction assigned.
20.0 , , , , , , , , ,
18.0 -
)
16.0 - -
wt :sc .w &--
14.0 -
Z w
o 12.0 -
f -
E O
$ 10.0 -
>=
2 y 8.0 -
E W
& 6.0 -
o-PaRCawT HvoRocam n Zone t ,,g,
_ O-PfRCENT MTOROCEN m ZONE S 2.0 - 4-PERCENT HVOROOFN m ZONE ax 1l T MCMOS-0-PeRCrwT uvoRoorn m Zone 8j 0.0 % ' ' ' ' ' ' ' ' '
o 2000 4000 6000 800c 10000 TIME (SECONDS)
Figure 4.9 Effect of Maximum Allowable Time Step on the RALOC Calculation (source model H and coarse nodalization) 177
40.0 , i i i i i a i 37.5 - -
35.0 - -
32.5 - -
30.0 -
~4- '4 W-Z 27.5 - -
@ 25.0 - -
22.5 - -
Z 20.0 - -
>=
Z 17.5 - -
W y 15.0 - -
W a 12.5 -
- o. pencent nvonocen in rowe O. pEnCENT MvonooEN IN rout S p,mg -
0.0 -
A. Panceur uvonoosuiu rome a rows 7.5 - O'""C'"'""O*'"'"8'"'* -
5.0 - -
- e. Pancsur HvonoosE in rows i C@e_
0.0 O 2000 4000 6000 8000 10000 TIME (SECONDS)
Figure 4.10 Comparison of IIALOC Fine and Coarse-7ene Itesults for the Top of the Grand Gulf Containment
.0 , , , , , , , , ,
37.5 - -
35.0 - -
32.5 - -
30.0 - '- O T T-Z 27.5 - -
E 2 5.0 - -
o 22.5 - -
I 20.0 - -
1 >"
Z 17.5 - -
N 15.0 - -
W 10.5 -
- o. Pencent uvonoosm au rows is -
10.0 -
"".e"taCGNT A HYo#ooGN Wer'o"se't roset- er 7.5 . o.pencm um mamrm u -
5.0 -
,.S _
..p m To.oo..mro .l u,,r:
0.0 ' ' ' ' ' ' ' ' '
O 2000 4000 6000 8000 10000 TIME (SECONOS)
Figure 4.11 Comparison of It ALOC Fine- and Coarse-7one Results for the flottom of the Grand Gulf Containment 178
l 4.8 Physical Parameter 4.8.3 Initial Containment Temperature l Sensitivity Studies Distribution The previous section demonstrated that the Ilattelle-Frankfurt experimental data demon-ItALOC calculations are numerically well bc haved. strate thet the mitial temperature distribution m a The solutions obtained are reasonably independent of amta,mment can significa>;aly affect the hydrogen vol-the maximum time step specified by the user and of umetric concentrations. 'Io mvestigate this effect for the degne of resolution used in the model nedaliza- the Grand Gulf calculglic ns, concrete slabs were add-tion. This section discusses the effects of physical '.d to the fine nodahzation model of Figure 4.3 as parameters on the ItALOC calculations. These pa- time-dependent heat sources. Two imtial temperature rameters are associated with the time-dependent re- distributions were assigned to the heat slabs and the lease of hydrogen following an accident, and the intial gas imtially present in the containment. Both distri-conditions present at the start of hydrogen injection butions were hnear with elevatmn and started at 60*C.
into the containment. The first temperature distribution assumed that the temperature in the coatamment decreased linearly 4.8.1 Hydrogen injection Rates with elevaiion (20*C hotter at the bottom than at the top of the containment), while the second profile Figure 4.12 shows a comparison between RALOC assumed an inverted distribution; i.e., the tempera-calculations using source models A and H of Figure ture was assumed to increase linearly with elevation 4.1. Increasing the hydrogen injection rate increases (20*C cooler at the bottom than at the top of the the spatial nonhomogeneity of the hydrogen. Faster containment).
rates of injection allow less time to mix. Decreasing Figure 4.16 shows a comparison betvecan an i>o-the injection rate decreases the hydrogen asymmetry thermal containment and the decreasing temperature by extending the mixing time (Figure 4.13 comparison model. The results are based on source model D of of smirce models B and C). Differences in the final Figure 4.4. Differences in the hydrogen concentration concentration values of Figures 4.12 and 4.13 were during the injection period were insignificant. Final caused by small differences in the total amount of hydiogen concentrations were somewhat higher for hydrogen injet:cd (source A injected 85 lb more hy- the decreasing temperature model. This occurred be-drogen than source H, while source C injected 25 lb cause the convecting gases cooled the containment lessh and condensed some additional steam.
Figure 4.14 compares tha hydrogen concentra- Figure 4.17 shows a comparison made between the tions at various altitudes when the bottom compart- isothermal base case and the inverted temperature ment reaches 85. Larger hydrogen injection rates model. During the injection period, a significant dif-produce greater volumetric differences in the local ference is seen between the hydrogen concentrations hydrogen concentrations.
in the top compartment. The higher temperatures present in the top of the containment produced a 4.8.2 Hydrogen injection Temperature thermal barrier which inhibited the convection of in the event of a degraded. core accident in Grand hydrogen. Final concentrations for the inverted tem-Gulf, the temperature of the hydrogen injected into perature model were lower than for the isothermal the suppression pool will be hotter than the water case. Convecting gases warmed the containment and temperature. M. L. Corradini has indicated that the created more steam. Increasing the temperature dif-temperature of the hydrogen exiting the suppression ference between the top and bottom of the contain-pool will be nearly the same as the water temperature ment from 20' to 50*C results in even more pro-(or perhaps somewhat higher)."To test the effect of a nounced hydrogen asymmetries (Figures 4.18 and higher hydrogen temperature, the injection tempera- 4.19). Again, the higher overall containment tempera-ture was increased from 60* to 300*C and a compari- ture produced lower end-of-calculation hydrogen con-son was made with the base-case calculation (source centrations due to an increased production of steam.
model H,60*C isothermal, and coarse nodalization). Figure 4.20 compares the hydrogen concentra-This comparison is shown in Figure 4.15. The effect of tions at various elevati'ms when the bottom compart-a higher hydrogen injection temperature on the ment reaches 8G . A decreasing temperature distribu-IIALOC calculation is negligible. tion has little effect on the hydrogen volumetric 179
concentrations. Inverting the temperature distribu- In the absence of containment sprays,our RALOC tion significantly affects the concentrations, especially calculations clearly demonstrate the importance of an in the uppermost compartments. Middle elevations inverted temperature distribution. Asymmetries (less than 209 ft) were not as affected by the inverted caused by the temperature inversion are greatest dur-temperature pwfile and actually showed a slight in- ing the injection period. These asymmetries can sig-crease in mixing. nificantly affect the design of possible mitigation schemes.
20.0 , , , , , , , , ,
18.0 - -
16.0 -
\ T T $--
z 14 0 ~ ~
su o
o 12.0 - -
e z 10.0 - -
z j g 8.0 - -
i Fa
- 6.0 -
,,,c,,,
,,,,,,,, , ,,,, , ll EL* I -
o ,
o- Pencent avonoorm m zone e 4.0 - -
A- PencENT o- nacent HYDaootN wronoorn m rome m aZONEj 1l'*"SounCE 2.0 _ _
1 0.0 4$d O 2000 4000 6000 8000 10000 l TIME (SECONDS)
Figure 4.12 Effect of Hydrogen injection llate on the itALOC Calculations. Source A versus source B.
1 1
I i
l l
l I
f 180
-. .. .. -- -~ _
i 20.0 , , , , , , , , i 18.o -
16.0 -
, 14.o -
w o
o 12.o -
a:
O r 10.0 -
z w ,o o- nacant wronoaan m zone i I sounce y _
o- Pancant wronocen se zone a j mooEL e -
S' 6.0 -
A- Prncant wronoorn se zone s ) sounce -
O- Pencrut wronoorn se zona e j woont c 4.0 -
2.0 -
0.0 ' ' ' ' ' ' ' ' '
O 2000 4000 6000 8000 10000 TIME (SECONDS)
Figure 4.13 Effect of flydrogen Injection Rate on the RALOC I Calculations. Source C versus H.
a.ooo n , , , , ,, , , , , , , , , 1 7.500 -
te-sorromi 7.000 _ -
6.500 -
6.000 -
, 5.500 -
o 5.000 -
O
$ 4.500 '- -
Z 4.000 -
6 3.500 -
u 4 g 3.000 -
a 2.500 -
2.000 -
is. :::
1.ooo -
1.500 -
0.500 -
o.ooo ' ' ' ' ' ' ' ' ' ' ' ' '
o.54 s' 2.cas INJECTION RATE Figure 4.14 Percent of flydrogen in a Zone When Zone 8 has 89, as a Function of the Rate of liydrogen injection. Numbers in parentheses are zone number and elevation of the zone in feet.
l l
181
20.0 , , , , , ,
18.0 - -
16.0 - .
E I
+
I
+
W-z 14.0 -
W O
@ 12.0 - -
O I 10.0 - -
k I
u 8.0 -
5 6.0 - ,
O* PenCeNT MYonooeN W Zone ti _
O- PenceNT MTDacoeN m Zone s'i '
4.0 -
a- penceur wronoosu m zone Tel -socc -
0- pencent uronocen a rome el 2.0 0.0 O 2000 4000 6000 8000 10000 TIME (SECONOS)
Figure 4.15 Effect of Hydrogen Injection Temperature on the ilALOC Calculation 40.0 , , ,. , , , , , ,
j 37.5 - -
35.0 -
32.5 -
30.0 -
8 _$ g-6 27.5 -
o 25.0 -
oc o 22.5 -
Z 20.0 -
)
I 2 17.5 -
E 15.0 -
E 12.5 -
10.0 -
""* j o-pencentuvonocen in zone t 7.5 -
o-pencenT uvonocem m zone it 'T"e[gknY -
5.0 -
6-PenCeNTMYonooeNIN Zone 1 lssoTuenMAL 2.5 -
o.pencenT uvenooen m zone erj moce' - )
0.0 '
9 2000 4000 6000 8000 10000
. TIME (SECONOS)
Figure 4.16 Comparison of Isothermal Base Case With a Decreasing Temperature Model. Results are shown for the top and bottom of the Grand Gulf containment.
I 182
l l
i 40.0 , , , , , , , , ,
37.5 - -
I 35.0 - -
l 32.5 -
.x ,.
n 30.0 - -
4 0 -D -
6 27.5 - -
o o
E 25.0 - -
$ 22.5 - -
Z p 20.0 - -
17.5 - -
g 15.0 - -
a 12.5 -
o- nncewvHvonocem m zona i necerasmo-10.0 - ' -
7.5 -
- a- nacent Hvonceau m zone t l esornanasat-5.0 -
0- necewT wronocen m romair) asomet -
2.5 - -
0.01 -
O 2000 4000 6000 8000 10000 TIME (SECONOS)
Figure 4.17 Comparison of Isothermal Base Case With an Inverted Temperature Model. Results are shown for the top and bottom of the Grand Gulf containment.
35.0 , , , , , , , , ,
32.5 - -
30.0 -
- 3-~
27.5 - -
E 25.0 -
0 O : -.
W 8 22.5 - -
e
$ 20.0 - -
Z p 17.5 - -
E y 15.0 - -
a y 12.5 - -
l 10.0 - -
l l 7.5 -
m-Ar= 0.0*c -
l O - a7= 3o.o *c
! 5.0 -
e - AT= so.o *c -
t l 2.5 - -
l 0.0 r_ ' ' ' ' ' ' '
l 0 2000 4000 6000 8000 10000 TIME (SECONOS) l l
Figure 4.18 RALOC Comparison for Hydrogen Concentrations in I the Top of the Grand Gulf Containment (Zone 1) as a Function of Time and Magnitude of the Initial Temperature Inversion (0.0*C, 20.0*C, and 50.0*C) 183
4 i s i i a i i i 27.5 - -
., +
25.0 - - - r : -
22.5 - -
z g 20.0 - -
O E 17.5 - -
O Z 15.0 - -
6-E 12.5 - -
o 9c w 10.0 - -
EL 5 = PERCENT NTDROGEN SM ZONE 1 D- PEACENT NYDROGEN 6N 200EE 10, 5.0 -
2.5 -
0.0 : . _
O 2000 4000 6000 8000 10000 TIME 03ECONDS)
Figure 4.19 ItAI.OC Comparison of the flydrogen Concentration lletween the Top and the llottom of the Grand Gulf Containment for the Initial Temperature Inversion of 50*C (t?-SOTTOut
-m (10-173) mm
~
(F-30s) -
)
6.500 F __ __ n. ::
6.000 - -
5.500 - -'
6 5.000 - -
O g 4.500 - -
0 Z
4.000 - -
is-sefi g 3.500 - -
U sc 3.000 - -
E 2.500 - -
2.000 - -
1.500 - -
1.000 - -
0.500 -
0.000 ' ' ' ' ' ' ' ' ' ' ' ' ' ' '
-20.0 -15.0 -10.0 - 5.0 0.0 5.0 10.0 15.0 20.0 MAXIMUM TEMPERATURE DIFFERENCE (K)
BETWEEN TOP ANO 80TTOM OF CONTAINMENT Figure 4.20 Percent of Ilydrogen in a Zone When Zone 17 has 87, as a Function of the Initial Temperature Distribution. Numbers in parentheses are the zone number and the elevation of the zone in feet. (Fine nodalization).
1 I84 ,
i
4.9 Conclusions 7. The upper containment compartment reaches 8'T hydrogen concentration approximately 5
,I'he following conclusions are based on the sensi-tivity studies that we performed with RAI OC. How- min after the bottom has reached 8'76.
ever, until additional assessment calculations have been completed, t hese calculations should be regarded 4.10 References as tentative. 11. Jahn, "IIALOC-MODI with 1980 Updates," un-documented revision of proprietary itALOC-MODI, RA-
- 1. The concentrat. ion of hydrogen .in the upper toc. MODI - A Computer Program for the Determination compartments of the Grand Gulf containment of Local Gas Concentrations in Subdivided Vessels (Parti-are probably combustible (~G'b) when the culary: H, Distribution in PWR Full Pressure Contain-bottom compartment concentration reaches ments Follouing a Loss-of-Coolant Accident, NIIC Trans-lation W, GRS-A-263 (January 1979), Contract No. 82260, 8'7 for isothermal initial conditi'ms. Gesellschaft for Iteaktorsicherheit, Federal Itepubhc of Increasing the hydrogen mjection rate m-
,2.
Germany.
creases the degree of nonhomogeneity in the
'fl.Jahn, Status Report on the Hydrogen Distribution hydrogen concentrations due to reduced mix- yoffyyjng a cy,.lant loss Accident, NItC Translation 796, ing time. GIts.A-333 (August 1979), Gesellschaft fur Iteaktorsicher-
- 3. An initial temperature distribution which is heit, Federal Itepublic of Germany, hotter at the top of containment than at the "G. Langer, It. Jenior, ana 11. G. Wentlandt, L'xperi-hottom increases hydrogen asymmetry during mentalinvestigation of the Hydrogen Distribution in the injection and long-term stratified concentra. Containment of a Light Water Reactor Following a Coolant tions (10 000 s) by establishing a thermal barri- Loss Anident, NitC Translation 801,IlF-F-63.363-3 (Octo-
. ber 1980), llattelle Institute e.v. I renkfurt, Federal Itepub-er to convection. lie of Germany.
- 1. A temperature profile which is colder at the top "G. It. Illoom, L. D. Muhlestein, and A. K. Postma, than at the bottom produces nearly the same -Ilydrogen Mixing in Containment Atmosphere," Propri-results as isothermal conditions. Mixing is not etary paper presented at the EPiti flydrogen Project Ite-significantly enhanced. view and Workshop (October 5,1981), Palo Alto, CA.
- 5. Hydrogen injection temperatures greater than "M. P. Sherman, Sandia National Laboratories, Pri-the temperature of the suppression pool do not vate communication (September 1981).
significantly affect the mixing calculations. "II. W. Ilurnham, Sandia National Laboratories, Pri-
- 6. Nearly steady-state solutions are achieved ap- vate communication (September 1981).
proximately 20 min after the end of hydrogen 4'n L unds Wm@ of Wisconsin, Private injection. communication (September 1981).
185-186
5.0 Dynamic Combustions and impulsive Loading 5.1 Introduction in this section we will discuss qualitatively the extensively in laboratory-scale devices. A more de-like.lihood of a detonation occurring in a mixture of tailed discussion of C4 detonations and the impulsive hydrogen and air at the Grand Gulf Nuclear Station. pressures that they produce is given elsewhere.*'
We begin by listing the conditions which seem to be Quasi-detonations and accelerated flames can also necessary, but not always sufficient, to cause a detona- produce impulsive loads that significantly exceed the tion in hydrogen : air mixtures. Then we discuss the calculated dedagration pressures. This happens forms of combustion that can produce impulsive whenever the Dame-front velocity is comparable to or loads: detonations, quasi-detonations, traruitions to greater than the speed of sound in the unburned gas detonation, and accelerated names. Finally, we exam- ahead of the front so that strong shock waves are ine the Grand Gulf plant and various accident scenari- produced.
os in order to make a qualitative estimate of the Quasi-detonations are defined here to be super-likelihood of detonations. sonic combustion waves that, for one reason or anoth-er, have not developed into full CJ detonations. An 5.2 Necessary Conditions for example would be a combustion that is initiated in an Detonation bstacle. filled tube using a detonable mixture, or a fully developed detonation that enters an obstacle-For a gaseous mixture containing hydrogen, oxy.
f lied tube. Figure 5.1 illustrates this behavior. Re.
gen, nitrogen, and steam, a detonation can propagate if the component gas concentrations fall within a searchers at McGill University *' used a 5-cm-dia by given range (recently it has been shown that a concen- ll.m.long detonation tube to produce the results tration of 13.8th hydrogen in air is detonable). The shown in the figure. The first 3 m of the tube con-tained orifice-plate obstacles (area blockage ratio 0.G; fact that a mixture is detonable does not mean that it will detonate. In general, it is also necessary that at distance between orifice plates ~5 cm). The ignition least one of the following initial conditions be satis- source was a glow wire (" weak" igniter). Flame speeds fied: were measured as a function of distance from the igniter for hydrogen : air mixtures containing 209h to
. A " strong" ignition source exista 3076 hydrogen (by volume).
. A " weak" ignition source followed by either a The data shown in Figure 5.1 illustrate four phe-long run-up distance, or an obstacle-filled run, nomena of interest to this discussion. For all mixtures, exists the Game accelerates to supersonic speeds by the time Geometries (boundary conditions) also play a very it has travelled 0.5 m from the igniter. The 3096 important role in the initiation and propagation of hydrogc.n case is apparently unstable (with the 0.6 detonations. These conditions are given in a qualita- ama blockage orifice plates) because it can behave as tive manner because quantitative criteria are not yet either a quasi-detonation (Game speed ~1300 m/s) or generally available, accepted, or applicable. If the an accelerated name (name speed ~850 m/s) while it mixture is detonable, the boundary conditions are travels through the obstacles. The quasi-detonation right, and at least one of the above initial conditions is becomen a C-J detonation (~2000 m/s) shortly after it satisfied, detonation is possible but not a certainty. emerges fr m the obstacles, but the accelerated Games actually decelerate for 0.5 to 1.5 m before making a 5.3 Detonations, Quasi. deilagration.to-detonation transition (ooT).
Detonations, Transitions to The net effects of name acceleration can range from miid to strong depending on the degree or accel.
Detonation, and Accelerated cration. For name velocities, v, that are small com-Flames pared to the sound speed (v < 400 m/s), the major effect or an increase in velocity is to decrease the heat Classical or Chapman 4ouguet (C-J) detonations kisses; consequently, the combustion pressure more in hydrogen : air minores have beesa studied closely approaches the adiabatic value. When the 187
'/.
a -_ _ _ _ _ _ _ _ _ _
Dame speed becomes comparable to or greater than the materials in a given problem. Either Cartesian or the sound speed, significant shock waves are pnxluced cylindrical coordinates may be chosen; the latter op-
! by the flame; consequently, impulsive pressure loads tion was used in the calculation described here. The greatly exceeding the adiabatic, equilibrium pressure numerical method used in the program includes an can occur. The damage potential of these dynamic artificial viscous pressure, which smootha strong com-
. loads depends on the specific structures that are ex. pression waves so that results agree well with steady posed to them. shock wave solutions. This smoothing, and the finite spatial mesh size, result in peak and reflected pres-nures for a steady detonatiun wave (C-J wave) that are
- ' ' ' ' ' ' ' ' lower than the theoretical valuer The " conservative" i
l effect of the assumption of axial symmetry is thus
' counterbalanced, although the amount of this com-pensation is unknown. For the hydrogen : air detona-2000 -
~
_- tion modelling, a method was used which converts unburned matenal to burned material at a constant 3
/ ,,,, rate whenever a threshold pressure is exceeded; the i
rate and threshold values are input cbnstants. The containment building boundaries were treated as rigid 4
3
~
/ ~
and smooth, and pressure histories were maintained at various points on the boundaries.
3 y in our CSQ calculation for the Grand Gulf con-j tainment we only modelled the containment dome g
region and the wetwell above the suppression pool 3 1000 -f ,
water level.The NitC estimated static failure pressure l (0.48 MPa) would be exceeded if a mixture well below f
\ ,,,,,,
t y g He- 8' 18", hydrogen (a previously quoted detonation limit) filled containment, and underwent an adiabatic, iso-
--o- a r s 8 800 ~
ass H,- choric deflagration. Therefore, only detonations of
~ 2 0' He' localized hydrogen: air mixtures need to be consid-9""7' ~
- l , , , , , ered-the remainder of the volume can be modelled as 0 1 2 3 4 5 6 7 8 containing only air. A 20"a hydrogen mixture in dry TUBE LENGTH - meters air was assumed to occupy the bottom of the wetwell to a height of 7.7 m,and was detonated at the approxi-Figure 5.1 Flame speed as a function,of distance from the mate center of its interface with the air (Figure 5.2).
igmter for various 112 : air mixturen. The first 3m of the detonation tube contained orifice-plate obstacles with an The wetwell was given irregular boundan.es in an area blockage ratio equal to 04. effort to approximate the presence of equipment and l now obstacles. For the initial conditions chosen for the mixture (275 K,0.12 MPa), the C-J pressure is about i.e MPa. necause of the factors mentioned 5.4 Local Detonation Calculation previously, and because the calculation requires a fcr Grand Gulf propagation distance of ~20 m to establish a steady Detonation of hydrogen : air mixtures will result wave, the maximum detonation pressure attained was in spatially varying dynamic loads on the containment only about one. half the theoretical C-J value.
structure. In order to obtain ertimates of the magni- Pressure histories in the wetwell region exhibited tinie and nature of such loads, the two. dimensional high, very narrow peaks early in the calculation and w ave propagation code, CSQ," was used for a calcula- many complicated interactions, as shown in Figure tion modelling a local detonation in the Grand Gulf 5.3. Peak pressures caused by the shock propagating containment. This computational tool has previously through the air remained fairly high as far up as the been applied to detonations in large, dry contain- wetwell flow path entrance to the containment dome ments and ice condenser plants." " (Figures 5.4 and 5.5). As this shock wave emerges into CSQ is a well. tested computer program which the containment dome, the toroidal expansion of the solves finite. difference analogs to the differential wave results in lower loads on the boundary (Figure equations describing conservation of mass, momen- 5.6). Because of the uncertainties in modelling the tum, and energy, together with equations of state for boundaries of the wetwell region, we are unable to
' I 188 I
I
assena the threat to containment posed by the calcu- but the period is probably in the range of 40 to 60 ms.
lated pressures, either by direct wall loading or by With this ertimate, and a pressure of 0.48 h1Pa, the missile generation. criterion gives acceptable specific impulsea of 6.2 to The load histories produced on the dome bound- 9.3 kPa.s; according to h1 ark, these values shouhl be aries had peaks of much longer duration than those in conservativt for a reinforced concrete structure. Ile-the wetwell. The 0.48 h!Pa value was exceeded for cause our computed loads are not spatially uniform, several milliseconds on the central re ;; ion of the dome we computed specific impulses for all 10. and 15-ms to a radius of about 4.6 rn (Figures 5.7 and 5.8). intervals, rather thsn just the first. The results are Without dynarnic structural analyses, we can make no shown in Figuret 5.9 and 5.10 for the top center of the realistic estimate of the threat to the dome region dome. It may be seen from the figures that for this from the predicted loads. Ilowever, a very rough idea calculation and the estimated periods chosen, the of the possibility of structural damage may be gained predicted impulsen are below the acceptable valuca.
by considering an impulsive failure criterion proposed flowever, because the calculated loads are not spatial.
by hlark. The criterion considers spatially uniform ly uniform, we are uncertain as to the real meaning of loading of Ihe st ructure, and deterrnines un acceptable this comparison. In addition, variations in the amount impulse value which is proportional to the product of and concentration of the hydrogen : air mixture, in the mrsimum acceptable pressure and one. quarter of detonation h> cation, and in the modelling of the flow the first fundeu.sental period of the structure. We path between the wetwell and the upper compartment know of no formal estimates for the lat4r quantity, ma; all be expected to affect the numerical results.
+25 i i i .
E
- +10 -
E .
o ,
E ~
us 2 -5 -
a *100%
< _ j AIR .
$-20 w
P rd -35 ' IGNITION 80% AIR E SOURCE +
LOCATION ,
20% H2 '
-50 '
-37.5 -22.5 -7.5 +7.5 +2 2.5 + 37.5 RELATIVE RADIAL POSITION (m)
Figure 5.2 Grand Gulf Containment Model for CSQ Calculation (axisym-met ric) 189
l 1.4 ,
1.3 - -
1 1.2 -
1.1 -
1.0 -
e I
I0.9 w
. 0.8 -
'I o,7 .
l l
1 l 0.6 0.5 -
} O.4 -
1
- 0.3 - * -
! 0.2 -
0.1
)
0.00 0.04 0.08 0.12 0.16 0.20 i
i TIME (sec) f Figure 5.3 Pressure IIistory on Outer Wall. Ilelow Detonation Point (CSQ)
]
- l
! 1.2 ,,,,,,,,,,,,iii I 1.1 - -
a I
1,0 - -
0.9 - -
l 7 j
53
- n. 0 - -
4 I -
W 0.7 -
cc 3
Wtt) 0.6 - -
w
( 0.5 -
y i
0.4 - -
0.3 - -
! 0.2 -
0.1 "' ''''
O.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18
! TIME (SEC)
- Figure 5.4 Pressure History at Outer Edge of the Wetwell Flow Path into the Containment Dome (CSQ) 190
850 , , , , , , , , , , , , , , ,
800 -
750 -
700 -
650 -
T 600 c.
-) 1 1
6 550 -
$ 5M -
a m 450 -
M f' E400 -
J -
c.
350 -
300 -
250 -
200 -
150 -
100 ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' '
O.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 TIME (SEC)
)
Figure 5.5 Pressure IIistory at Inner Inige of the Wetwell Flow Path into the Containment Dome (CSQ) 375 , , , , , , , , , , , , , , ,
350 -
325 -
300 -
g 275 -
- c. I 5 250 -
Lu j
CC j
3 m
225 -
m E 2M -
c.
175 -
150 -
125 -g -
100 ' ' ' ' ' ' ' ' ' ' ' ' ' ' '
O.03 0.05 0.07 0.09 0.11 0.13 0.15 0.17 0.19 TIME (SEC)
Figure 5.6 Pressure Ilistory at Lowest Point of the Containment Dome Wall (CSQ) 191
l l ' i '
l '
I l l l -
8o0 -
700 -
-e 600 -
- o. .
M.
m e
500 -
a .
us E
m 400 -
Q. .
300 -
200 100 O.06 0.08 0.10 0.12 0.14 0.16 0.18 TIME (sec)
Figure 5.7 Pressure llistory at Dome Center (CSQ) 550 i , , , , , , , , , ,
500 -
450 -
3 M -
g _
[ 350 -
e _
3 y300 g _
Q- 250 -
200 -
150 -
J I I I i t i 1 I I i 1 g i 0.06 0.08 0.10 0.12 0.14 0.16 0.18 TIME (SEC)
Figure 5.8 Pressure Ilistory on Dome 4.7 m From Center (CSQ) 192
7.0 .
i .
6.0 - -
g .
2 o.
4 5.0 - -
g . -
E a
4.0 - -
o .
Q.
E 3.0 - -
o g
8 a.
2.0 - -
m -
m E 1.0 - -
d 0.0 ' ' ' '
O.06 0.08 0.10 0.12 0.14 0.16 0.18 TIME (sec)
Figure 5.9 Specific impulse at Dome Center for 10 ms intervals i -
i i 8.0 - -
3e 7.0 - -
a . .
e i 6.0 .
m 3
5.0 - -
a.
E 4.0 - -
o . .
$3.0 w
a.
"m 2.0 - -
E, 72 1 . 0 0.0 ' ' ' ' ' ' '
O.06 0.08 0.10 0.12 0.14 0.16 0.18 TIME (sec)
Figure 5.10 Specific Impulse at Dome Center for 15 ms Intervals l
l 193
l 5.5 Likelihood of Detonation at from the formation of detonable mixtures, is very small. The probability of a DDT occurring is also tho Grand Gulf Nuclear Station yoite smaii iioweyer, we feel that some degree of It appears to us that several conditions could lead flame acceleration upward through the wetwell obsta-to the formation of detonable clouds of significant cles is almost a certainty. The degree of acceleration extent. In general, the igniters would have to be off for will depend significantly on the effectiveness of the some time and other ignition sources (hot surfaces, igniters in burning very lean mixtures and on the sv. itching sparks, motor sparks, etc) would also have spatial concentration gradients. Measurements in ob-to be temporarily off or ineffective. A sudden large stacle-filled tubes at McGill" indicate that flame increase in the hydrogen release rate or the movement acceleration mechanisms function with as little as 6ch of a large air m.ss into an oxygen-depleted, hydrogen- hydrogen present in air, and that high flame speeds rich compartment (e.g., the drywell) might also lead to can be achieved (>20 m/s for 8"6112 and >150 m/s l local regions of detonable proportions The most likely for 10"r 11,3 in the recent experiments).
( regions for localized de'onetions appear to be the drywell and the bottom of the wetwell. liigher wetwell regions could become mvohed if steam mertmg 5.6 References and/or oxygen depletion occurred due to previous ht. P. Sherman et al, The Behaeior of Hydrogen I combustion events. During Accidents in Light Water Reactors, Sandia Nati ,n-l Several complicating factors make it difficult to al 1 aboratories, SAND 80-1495, NUREG/CR-1495, August 1980.
discuss the l}elihood of forming a detonable mixture.
Ilydrogen, oxygen, and steam concentration gradients "Himonthly Progress I,etter (October and November, occur it; the vertical (z), radial (r), and azimuthal (6) 1981), hl. Herman (SNI.) to J. T. I,arkins (NRC), December
( 23,1981.
directions. This would be true without combustion.
With numerous combustion events occurring '% L Thompson, CSQH - An Eulerian Finite Differ-throughout the wetwell/ containment region e* various ence Progrnm for Two-Dimensional Afaterial Response -
Pad L AfWrial Secdons, SAND 77-1339, Sandia National times, the picture is extremely complex. For each
' 9"#'9" ' """*'Y igniter we can envision a sequence of" pulsed" barns in which some hydrogen is locally combusted. A period 8'W. B. Afurfin, Report of the Zion / Indian Point Study: V lume I, SAND 80-0617/1 (NUREG/CH-1410),
of mixing follows each burn. In the region between the *n adona Laboratones, Albuquerque, Nht, August top of the suppression pool and the first wetwell f93fa igniters (z = 135 it), the contentration of iiydrogen u
- n. K. Byers, CSQ Calculations of H2 Detonations in
! may reach detonable proportions if the hydrogen re- the Zion and Sequoyah NucIcar Plants, SAND 8122t6.
l lease rate is high enough. A pulsed burn from z = 135 (NUREG/CR-2385), Sandia National Laboratories, Albu-
! ft down to the top of the pool would remove all querque, NN1, to be published.
hydrogen (or oxygen) from the region and permit high a sJ. C. Stark, hiemorandum for ACRS hiembers, Notes hydrogen concentrations to form before the new mix- on Hydrogen Burn with Igniters, December 4,1980, l
ture rose to the igniters. The spatial extent of such u nB LiW Water Reactor Safety Research detonable mixtures woull probably be h,mited. Program Semiannual Report, April-September 1981, San-In our judgment, given current information, the dia National Laboratories, SAND 82-0000, NUREG/CH-likelihood of a detonation occurring, as distinguished 2481, February 1982.
l l
194
l l
l 1
l 1
6.0 Assessment Results and Recommendations In this chapter we discuss the results of our assess- and 19 are ocated near elevation 204 ft. In the con-ment of the adequacy of the IIIS to meet the threat tainment dome there are 8 igniters at elevation 262 ft, posed by hydrogen combustion at the Grand Gulf 8 at elevation 285 ft 10 in., and 2 at elevation 295 ft.
Nuclear Station. Chapters 2.0 and 3.0 discussed our In our judgment the lilS design could be im-calculations of the containment atmosphere pressure proved in two ways: an increased number of circuit l and temperature response to hydrogen combustion. breakers gewer igniters per breaker) and either a Comparisons to similar CLASIX-3 calculations are modified igniter support design or an increased num-madc in those chapters. Containment atmosphere her of igniters in the wetwell/ containment regios We mixing mechanisms are discussed in Chapters 2.0 believe that if a short circuit at one igniter will deacti-through 4.0, with special emphasis in Chapter 4.0. The vate 22 other igniters, the circuit breaker design is likelihood and consequences of detonations are dis- lnadequate. In our opinion, no more than two igniters cussed in Chapter 5.0 of this report, at any elevation should be protected by the scme The following sections of this chapter present our circuit breaker. This means that a total of S or 8 assessment of the adequacy of the filS design, actua- igniters can be protected by one breaker if those tion criteria, and testing. The final section of this igniters are apatially distributed throughout tha vari-chapter discusses the adequacy of the accident spec- ous regions. We strongly recommend that the drywell trum considered in the Mi>&L evaluation, igniters be protected by their own circuit breakers, since our analysis of suppression pool motion during 6.1 Adequacy of the HIS Design c nt inment burns indicates that significant pool surge may occur inside the drywell.
In our opinion the design of the ills is basically sound (the deliberate igmtion concept, however, i' The present design for igniter housing and sup-port structure is such that it tends to limit the effec-judged to be marginally adequate to meet the threat posed by hydrogen combustion). We do have a few tiveness of combustion ignition. In our opinion, the system will reliably ignite hydrogen : air mixtures in suggestions that we feel would improve the Grand Gelf system in terms of reliability.The present design the ~10?a hydrogen
- range, but not necessarily in the 6"e to 8?e hydrogen
- range. Since it is desirable to calls for a total of about 90 igniters distributed throughout the drywell and wetwell/conta, burn the hydrogen in containment at the lowest possi-m ment. Ile ble mole fraction, the igniters should be positioned so dundancy is factored into the design by using tw separate power distribution panels (Divisions I and that they can burn out large volumes with upward propagation of flames. Ilydrogen is unique in having a 11). Each panel uses two circuit breakers (approxi-large difference between the upward and downward mately 23 igniters per breaker) to protect its igniters.
The power system is attached to the station diesel flame propagation limits, the upward propagation limit of 4.lfe being much lower than the downward generator in case of loss of offsite pcwer.
The igniter distribution and location design is limit of 9?e or tl e horizontal limit of 61 such that the igmters are no more than ~30 ft apart if The design of the igniter housing incorporates a spray shield (horizontal plate with vertical sides) a few all are operable (except vertically between elevations inches above the igniter. This shield tends to locally 209 and 262 ft in the containment dome, and near the bottom of the drywell); the ,gniters i are no more than inhibit flame propagation. A much more serious prob'-
lem, for all but those igniters located on the contain-
~60 ft apart if only one Division is operable; all enclosed regions are served by at least one igniter from ment dome wall, is that most igniters are located within 2 ft of ceiling structures. Some of these ceilings each Division. The drywell will be served by 18 ignit-ers,6 each at elevations 146,161, and 179 ft. The are solid concrete, while others are metal gratings. The wetwell di3tribution of igniters is such that 12 igniters are located near elevation 135 ft,15 are located near
- Average concentration in the region served by w.e igniter elevation 160 ft,8 are located near elevation 183 ft, (~30 ft x 30 ft x 20 ft - 18 000 ft 3).
i 1
195 t
l l
_ _ _ _ _ _ _ _ . _ . _ - - - - i
i solid ceilings clearly inhibit the effectiveness of the We suggest ? hat MP&l, consider incorporating l igniters, and a similar inhibition of igniter effective- hydrogen detector information into their procedures ness will occur if the grating area above igniters is for containment spray activation and for lilS deacti-l covered with equipment. The main point here is not vation. We also recommend that the containment 4
that the igniters will not work as proposed, but that sprays be activated at the same time the 111S is reliable ignition of lean mixtures is questionable for activated or, at the latest, upon indication from the the volume served by a siagle igniter. hydrogen detectors of perhaps 2G hydrogen concen-We feel that the proper design of a support for an tration in containment. Our analyses indicate that the igniter in a specific locasion should incorporate con- sprays are very important to the plant safety and sideration of the likely Hame path, in general, it is should be maintained in operation throughout the desirable to have the igniter away from large solid accident.
1 surfaces, especially tho< e above. If such an approach is not feasible at this time, we would recommend in- 6.3 Adequacy of the HIS Testing l creasing the number of igmters, especially m the In our judgment the operat.ional m.spect. ion and wetwell region (from elevation 135 to 209 ft). We also recommend that the igniters in the containment dome testing program specified by MP&L for the lilS sys-tem *' is reasonable and adequate (however, the fre-at elevation 285 ft to in. be symmetrically staggered in azimuthal position with respect to those at elevation quency f the periodic testing was never defined). We assume that the station's diesel generators are also 262 ft, so as to take advantage of the likely Hame subjected to periodic testing. As discussed in section paths.
6.1, we feel that the design of the 1118 7 juld be changed in several ways. The impact of t!%e design changes to the operational testing program is benefi-cial. If fewer igniters are powered by one breaker, a change in the current drawn by that breaker is more 6.2 Adequacy of the Actuation easily detected and finding the cause of that change is Crj{grig simpler.
Testing of the system design with respect to ignit-
,l'he stated actuation entena,, for the 111S .is an er, transformer, or wiring failure under accident con-mdication that the pnmary-vessel water level has ditions is very important since one failure (short cir-fallen to the top of active fuel. Our judgment is that cuit) can shut off a number of igniters. The planned actuatmn of the filS at this pomt m an accident would submergence test is a step towards addressing this provide sufficient time for the surface temperature of issue. We feel that it is desirable to test the igniter the gmters to reach an acceptable value for reliable assembly under accident conditions for extended peri-igmtmn. It should be noted that we have not examined ods beveral hours). These conditions would include the ava,ilability or reliability, under accident condi- hydrogen burn environments, high steam pressures, tions, of those systems involved m the primary-vessel and water spray impingement.*
[
water level measurement. We are somewhat concerned ,
that the lilS actuation is done manually-apparently without an automatic backup system. Such an auto- e Previous igniter tests at Fenwal may have addressed these l
matic activation system might be a desirable addition. issues partially.
1 s
1%
l
l l 6.4 Adequacy of the Spectrum of There are severai kinda or accidenta not consid-ered by MP&L or SNL that present a more severe Accidents Considered in the challenge to the containment integrity. The possibili-Applicant's (MP&L) Evaluation ty of significant hydrogen releases directly into the We consider the degraded. core accident spectrum wetwell/ containment without first traversing the sup-considered by MP&L in their evaluation to be ade- pression poolis apparently extremely unlikely. Acci-quate for small-break LOCA and TPF,. liasically, the dents involving loss of all offsite and emergency power hydrogen threat to a HWit Mark Ill plant manifests are of course much more threatening but less likely.
itself in one of two ways. The release path is either The inclusion of accidenta involving the loss of the through a break in the drywell and then through the Ilesidual Heat Itemoval (111111) system may not be suppression pool, or it is directly into the suppression warranted because containment failure apparently oc-pool (from the pressure vessel safety relief valves). curs due to static over-pressurization before core deg radation rather than due to hydrogen combustion. ,-
Due to the uncertainties in hydrogen generation and release rates,* we consider the MP&L approach for this parameter to be a reasonable approach, especially 6.5 References since the effect ofincreasing the rate was examined in s' Letter and attachments from L. F. Dale, Mp& L, to H.
their sensitivity study. II. Denton, USNitC (December 21,1981).
CLASIX-3 Containment Response Sensitivity Anal-ysis for the Mississippi Pou er & Light Grand Gulf Nuclear Station, Off. Shore power System Rept No. 37A15 (Decem-
'As discussed in a number of sections in this report, the I hydrogen generation and release ratea are uncertain. The '8 M. Ilerman, Light Water Reactor Safety Research release rate is such a dominant input parameter to our codes Program Semiannual Report, April-September 19M, San-that we feel a separate study should address the issue for dia National Laboratories, SAND 82Mei, NUllEG/ Cit-IlWit plants. 2481, February 1982.
197 198
APPENDIX A Excerpt From Task 1 Report October 1,1981 Request for Additional Information We recommend that the applicant modify the wetwell volume and the neglect of natural convection, compartmentalization model used in CLASIX-3. As a the hydrogen is largely confined to the wetwell and result of studying the applicant's final report (June rapidly reaches the assumed concentration for igni-19, 1981), touring the Grand Gulf Nuclear Station, tion.The pressure rise associated with the combustion and studying the plant drawings at the end of this of this quantity of hydrogen is small because of the appendix, we feel that the wetwell/ containment com- transport of gas from the wetwell to the much larger partmentalization used in CLASIX-3 is inadequate. containment. Subsequent inerting in the wetwell, due The pressure rise caused by a hydrogen deflagra- to the depletion of oxygen or the generation of steam, tion is mainly dependent on the initial pressure before is computationally prevented because fully mixed the burn and on the number of moles of hydrogen containment air is transported back into the wetwell consumed. If the hydrogen that enters the wetwell is (at an artificially high rate) when the post-combustion mainly confined to the region between the top of the gases cool. Consequently, CLASIX-3 predicts a large suppression pool and elevation 135 ft, if there is number of burns confined to the wetwell region with sufficient oxygen present, and if no steam inerting associated small pressure increases. We believe that occurs, the ignition of even 109 hydrogen in that the net result of using the CLASIX-3 code with the limited volume leads to a very small pressure rise (q.v., present compartment model is an unrealistic calcula-the CLASIX-3 results). On the other hand, if the tion of the mixing and combustion processes.
mixing of hydrogen is rapid, so that the concentration In order to reduce the artificial mixing caused by of hydrogen is nearly uniform throughout the wetwell/ the use of just two volumes for the wetwell/contain-containment region, or if the lower wetwell becomes ment, we suggest developing a model with that region inerted after several burns due to inadequate mixing divided into five subvolumes. The five subvolumes are downward of oxygen and/or the presence of too much
- 1. Annular region from top of suppression pool to steam, then an ig iition can lead to the consumption of a large mass of hydrogen and a resultant large pres- elevation 135 ft (150 000 ft').*
- 2. Partial annular region from elevations 135 to sure increase.
161 ft (130 000 ft8). This excludes the volume In the CLASIX-3 code, mass is transported only occupied by the steam tunnel and the section because of pressure differences between compart-considered to be part of volume 5.
ments. Mass entering a compartment is assumed to be
- 3. Free region between elevations 161 and 209 ft instantly mixed with the entire contents of that com-(170 000 ft'). This complex region excludes the partment. Mass transport due to natural convection volume of the various pools, internal concrete and diffusion is not considered by the CLASIX-3 walls, and the sector considered as volume 5.
code. The two-compartment model of the wetwell/
containment used by the applicant in the CLASIX-3
- 4. Region above elevation 209 ft (840 000 ft').
This region is open and free of obstructions calculations consisted of a very small wetwell volume except for the polar crane.
(152 000 ft8) and a very large containment volume 8
(1 250 000 ft ). In view of the limitatione of the CLASIX-3 code, we f.elieve that the results obtained with this two. compartment model may underestimate th2 pressure rise due to hydrogen combustion.
Hydrogen will enter the wetwell/ containment
.The volumes indicated are based on our calculations and from the suppression pool. Because of the small are accurate to no more than two significant figures.
199
- 5. Pie-shaped sector from elevations 135 to 203 ft 3. liydrogen concentration criteria for combus-(60 000 ft'). This sector, near azimuth 225*, is tion ignit:on and combustion propagatien set free from obstructions since it is used in the at different values. We recommend an ignition
- movement of equipment from the equipment criterion of 976 - 10% hydrogen and a propaga-hatch. tion criterion of 4?; - 576 hydrogen. A variation A more detailed model would break the containment in assumed combustion completeness From 1.0 into even more volumes, separating the annuli azi- at 976 to perlaps 0.1 at 4 ?l) should accompany muthally into additional sectors. The cruciform geom- these changes. More sophisticated propagation etry between elevations 161 and 209 ft suggests this and combustion-completeness criteria would course, include factors to account for propagation di-In addition to modifying the compartment model rection (upward, horizontal, and/or down-used in CLASIX-3, we recommend that the applicant ward).
perform additional computations with variations in 4. Burn propagation delay time (from one com-certain key parameters. We suggest that the sensitiv. partment to another) decreased to O sec. Since ity of the CLASIX-3 results to the following parame- the compartments are intimately connected we ter variations be determined. feel that a zero delay time is realistic.
- 5. Containment spray carry-over fractions from
- 1. flydrogen source term increased and decreased the dome region into the wetwell (i.e., from by a factor of 3. The present value represents a almve 209 ft to below 209 ft) in the range:
high production rate but a lower rate might 0.0 s droplets s 0.20 decreasing as eleva-actually pose a greater threat, especially if the tion decreases hydrogen concentration for ignition and propa- 0.33 s sheet flow s 0.53 increasing as eleva-gation are not made equal (see item 3 below). tion decreases An increased source term could model a scenar- These values are based on our estimates of io such as that at TMI-2, where the production blocked areas at the 209 ft elevation.
rate and release rate are decoupled (i.e., pro-duction can be relatively slow, but the hydro- A number of specific items need to be addressed gen is trapped in the primary system for some before we can complete our evaluation of the Grand time and then released rapidly). Gulf HIS. These items are listed below (following the
- 2. Flame speed increased to 10 and 100 times the topic of evaluation).
present value (1.8 m/s). In recent experiments
- 1. Testm, g of the lilS at McGill University flame velocities in excess of 150 m/s were measured for a 1076 concentra- a. How will the system be tested? Specifically, tion of hydrogen in air.* These experiments what indicates that a particular igniter is or also indicated that the accelerating flame had is not functioning properly?
not attained a maximum value of speed. For the b. Ilow often will the system he tested once the computer model it may be desirable to make plant is operating normally?
the flame speed a function of hydrogen concen- c. Are hydrogen detectors to be used as part of tration and compartment geometry (i.e., the the ills? If so, please specify the types of flame would be expected to have a higher speed detectors, number, location of sampling at higher concentration and in more cluttered ports, system response time, and testing compartments). format and frequency.
- 2. Location and distribution of igniters Pleasa provide construction drawings for sever- I al " typical" igniter mounts in the wetwell and 1 containment regions. Also provide a list of the ;
approximate elevation coordinates for each ig.
- June / July Himonthly Report to John Larkins (NRC) on " n ese r ns an C m spondng the flydrogen Program. M. Herman, ed, published August 'I Vat.ion coordinates of the nearest floor or T
1981, celh.ng.
1 1
200
- 3. Calculations to determine the containment at. h. What fraction of the spray carryover (from mo*phere pressure and temperature response: containment to the wetwell) is assumed to remain as a spray and what fraction is as-
- a. Please check the list of CLASIX-3 input sumed to be in liquid layers?
parameters given in the June 19,1981 re- i. What is the nominal elevation of the top of port and notify us if any errors exist. the suppression pool during an accident
- b. Justify the values of the following with and without sprays activated? What is parameters and/or coefficients used in the the level of the upper pools under similar CLASIX-3 calculations: circumstances?
- j. How and under what conditions are the (1) beam lengths upper pools drained into the suppression (2) gas emissivity, absorptivity, heat capac-pg9p sty, heat conductivity, and viscosity k. What is meant by " satisfactory" conver-gence of the CLASIX-3 solutions?
(3) convective heat transfer rates and coef-ficients 1. Where have the sensitivity studies for
. . .. CLASIX-3, mentioned in section C.3 of the (4) wall thermal properties (emissivity, ab-sorptivity, heat capacity, conductivity) June 19 final report, been published (list (5) How coefficients (define and j,ustify val- references)? Please provide copies.
"*8 m. Please provide an estimate of free volume (6) spray droplet fall time.
and surface areas (with appropriate thermal (7) heat smk surface areas (identify loca- preperty estimates) as a function of eleva-tmn) tion for the annular wetwell region for the (8) upper pool surface thermal propert.ies following discrete elevation ranges: 110 -
(9) pam, ted surface thermal properties 135 ft,135 - 161 ft,161 - 209 ft.
- n. During normal operation, does the equip-
- c. What is the form of the momentum equa- ment " hatchway" (at azimuth angle 225*,
tion used to link the control volumes? radius 42 - 62 ft) have gratings in it? If so, at what elevations?
- d. What model is used for heat transfer to and evaporation of the spray droplets? 4. Containment atmosphere mixing mechanisms:
- e. Why was the number-mean diameter a. Describe the now rates of the ventilation (230 um) used for the spray droplets instead system in the contain nent/wetwell regions.
of the mass.mean diameter (370 - 400 pm)? h. What are the elevations and radial positions
- f. How are the various Door gratings treated of the spray rings?
thermally? Are they assumed to be a lump- c. Which spray rings operate when a single ed mass? Are they assumed to be thermally RHR loop is operating and what is the now ,
isolated from the rest of the Door and/or rate under such conditions? Does the spray walls? water contain additives? :
- g. Are the presence of liquid layers on walls d. Describe any sprays, fans, or other systems and condensation heat transfer treated con- that could move air in the annalar wetwell sistently with radiative heat transfer to the .!
region and estimate the veh> cities in the
walls? region due to these sources.
{
e i '
l 1
'l l
l 201 i
-- -~ ..-,v- ,,~~. . - . .~. , ,- - e. , - - - . ,, - , - ,. -,, - , ..w.c -
i
- 5. Actuation criteria: steam directly into containment without having passed through the suppresion pool?
- a. Under what conditions are the sprays acti-vated?
- b. Briefly explain the workings of the "drywell
! b. How long after the sprays are activated does purge system" including " purge compres-the spray system attain full flow rate? sors" and " vacuum breakers." Estimate flow-l
! c. When during an emergency situation would rates from this sytem during an accident.
l the HIS be activated? c. Briefly explain the workings of the "back up
- d. Wl..it role would hydrogen detectors play in containment purge system."
actuating the HIS? d. Justify the zirconium inventory in the reac-
- e. What role, if any, would the hydrogen re- tor used to estimate hydrogen inventory combiners play with respect to the HIS? from the metal-water reaction.
- 6. Sprectrum of accidents considered in the appli- e. Are there any accident sequences that might cant s evaluation: involve a localized injection of hydrogen into
- a. Are there any accident sequences that might the suppression pool (e.g., a single stuck-lead to the introduction of hydrogen and open safety relief valve)?
I I
(
202
+=m. wee = m ess-m- e m*
w w b
R 8
llE A y 1J -
- il I'l s)l I i'
ll!
, : y -
I ll I E l 3
! I b!
! lf-h 1
{ l i
i kh 7md s
g j4 i
, 4.a;iu A I w' s L
e- [ 4 e @ (s =
Y
\ 4 =ii',
W ! f i =
g _
L -
Lirf -
.li..
i C
't g=-
4 4
1;l II;I'f.i h hl - m :... . ! 0# L '_d I :!!-
Li Hl 1 l@. < il E gu =mm em - - '
i,d~hdllb k U- bF [EIClh _.'3 g
}R.1 Ar4A l h
+f gJ f 3 i i 2i l ll!
r .....i..:.wagg T0 god
- !p#p,qj i rn g- ..
!I n, ( l!!!
ii i n-11 1/ N!,9 15 !
~ j 1
1
- In : jg p; N !,l. L i
Y --
ig-I kj ! '
'l! e l .J,
- hi s-t
.a 1
l
@!Neg@!!Ih1lg'@
e se e e
-_ s. ..
'.,r '
e ..
E
. $' r. i i ij e
I i n !i! !I! ! N I, j ,ii'li 'i !JIII!h ..
e 'N l il[G}M,gcc fBid ,4dmrne..,y#H J 1 i 4
j .. . . Dpavheg%r) b
' e g .p.o t s di,:
ij g.. .
e
.. . L. i klu/.. dGMN e, , ;. 1 i k j i "
4 .j. .!,
m, i i i .. LS. l l:
i,
, n. ,
d I ' "
lb.:
sll!ll
! l hj }l
!i!f (*
e l [i ~ b,'Mfi 3 '
- i. ;,ingu m.m
. li I!
l lij,. '%d 'J1f[
. 3 .
- l. . .i y.
- i, L f: 'JJ/ p :-4
'11- !
d 1
"=
]sllllH'A'pdlsT a spe r . i. . 1
- N5,. U il4
~
i
!ff \1 J .:, c jjlr A~* 'j d . bi EC "I i
! l LO ItI$ll ,
l ; 1' N j h h [ hi M .(I{I i 4
f
= N U
O I I ut sm
,.n ***"
,,,g.o mQ.--- 3 , .,
T@l-4 w{ & ir a- - -- Im . . .
- ._c ...y , , , , , .
_ ~_ ,c l n . . C e >l .m J_ .
igll . -
.o=-
~
N ** h ***" (*c3 I h* -
o
[Ei'.com t =, l' d;g Am.c.R -
i O'D=l' q.
g
'=--
y i
",Tr.'. 'r",,,,, .oo. ... %uag p uu
. . .o.o. .
a .= -.
, ,a " _
- r . - ,ons '" . . ,
'"N m * * * " - " /
noce a.r4 , , , , , , l coeu.
9 . _-- , : ,:- ,
,. 'y,m ..
9 EoD..
cm q" 4 \ n ia =-~$l o '" ~
27o 2-
- ' .>. j 5 - - ' "
! 90 1-
{ '; s 4..,..
s./s.
c'o' acma"
- L a ,. soo. .'"c'*,", ',7 8,.:" ,.,,
ii;"m '
- c.c . .
% O," "."
YNE -
. Ja."Ja., 7 r
,y g 3
- x. q x
=rg m maa O
= _ m.,, . .._
=h ,
,,o . ,,.,
. -- a ..
~ 3.- . . ..
% = .oo. ,:::: r=='.
WM re -[ :,
)
,,,n,,,, <
- s. g3 ogo . l y,, g
.- v.a q . .. ,
.m,,,,,,
-. me o...
.,o..,.. g
'" " '*fo" '**"* .\ . _
= = <
m...- ..-. - l l' T T Tg n i? '
i -oi
- u. ioo, . s =. ..c l Riefg/Centelament Elevation 9 3
- 0 .100 *9
205
+,m-
, y, .. , , , , ,, .
i
.. -i
.LA1. - '-
f ** ;,, a /"*' m r 2 -. -
p "c4 ~ , , , .
. ,<=.
,,,, )\//\
. !l v.- J \ l - - ,o., ., L
- 1. -. . ~~ ~~ ~~\. \ \ . . s, ,
q=,h 7'.
h[ !
p r
] .
Q-..Mm, d g
,l]
..i 1[ ->
j t
a-y . y.g. .= m e --
u, -, w.
l' . -%. m .w 7 -:, Q p : L!
f N
- '. ~
2 * " " "
.o.ain , C. $ > " ' * *- A. w l
E p u- w ->
a' ps =Q' .cc i..
.3
.. o.- <
K '= , ' . r ' c -.
a 2g I f p, %'TR <--
,., j 7g" ""
s ';,.,
.< k ..m..,, C '-
_MC 3 SE .00 . *. 4 51 cCpot.a. gg
- 8'""" *
-7 (l* . ... .. . I ein
- Of 3 ,,
- / _: x g a ,
. s 4.- -.
..i .... ~~ c N.
a.
wi o. axi. -. , 3 l ',
4, s
ame.a
- mm c-- l' ge./ .
- g ,-( '-
e LOG
".4 .'. * # .'
N -
Co 0L .
% L 6 810 se.Et gg,g,g, .. T -
"l."<
.". ,"'.t
.,- s .
'"=
1 N'
, 5 i C.IB G5. . 'g
- E.
. , , m
,,. . ./ e ,
~
e 6 ,.,'
mie a. es so. liguigur-*-- . 7,i .'. g
=,=,,o,,',,=, j inouma ' * -.r.,,, -
j' h. 5 .oc. .
Nim N 8
/
ao A ,, ,,
j gE c c -.
. .o.o.n. ,,,,,".,"",,,
iig .
x _.a'ca._
Z
==g .
,p tec wu ice _i i .
tec m.. ut 'gai l
Il I
_B) _
! (Il
~__
1
- s. am&
A
_ - . .. i
" M '" LAJ I AU -*a < m.-
"'* LAJ \L. .. m . .m coaus -i se
.". - ' ,7
< e y "gfi , .-
,s . .
.o.o.. * '
c_\ _
6 2:. s 6 '-
%. 'i-w
.u _
g
=c me.
,, g-
.a.'az. :s U
.<c s= '" "' "-"
E E
uc~m. uG=' mis Jr. oi g c3
. i __._ . . . _ _ . _ _ . . - - ., -
- *V
_ __. . .. '* ..,,.n.. - -
Aux Bids /Centelament. Elevetiens 114' 6 ". 119' 0 ". 120* 10 "
206
=%
r G .U .=
C' S rm,,
~( -
-._ . - - < ,c e :. x- ,'1
-y
-. N'
',, l ii
\
g \ l y A,/-==.
9 / r= =7~:.=
i ~
ll r- -n c w-r-N- . :- .
i, 4 w bv -
- tj p s.
l'l p A ssi. .-_. . -_. r e-I iF- m--
4.. s. , M W
L _d.jT. pr f
l
- % - [u7,.s.-. =.g ^ l .r.-- - .
A y.
g ',f
,' .w s- .
j Lc t ' z.== '-]
,- i
,. i i -
g- c. .
+
+ / i i 1 -
- - M* ~j
- f. -
Pl ./ d i i l: ll3- g% r-R p '
ll =-Mu 2-
~-'_J- _
~ ~ - -
', j l
4
- >_ . - _~
s
- l f.h y .
- , q 3 < .> s < 2
,J,
'$. h E * *
. .e =,i ll
/
- ,-Y.,, fE -
\.a aang
--- = =_6 .
. \ . ggio p =
ggeFB
_A T
%g e555g
\
/ ;.
=e g#
m 'v : p,\ - ,
] s- aggy l
l gs9B \ ,)a g ... -- iz. SiSD55=
s 3 --,
n angg [.
v=.
--= W ' -
, h j .
o* ,
g }~--- .=- ~ s. ,_. ,~ __ Y ...-
%g, ,g_
. ' , -=
__,,~5 j , ?
a ,
__. . / . - --.
l h-e - -
/
g J = N- -Z
'~
M= - M._ _, / _a -;
. ,s ---f ' ~ -
o E i._b-[y 1
.- __ M '~-- D i . = .:: , _ . - ,
q _._
, g 1- ,-
.
-ti .-
a-
. __ v= -
.r= =. s.- p 3,j l m :- =- g -
u --
=%7 -f ._
~ ~---
=---=.
]
v
=.- i-L
.}l .lg -=,~~9 i -
g-t_" .
I!
(5::u. .' y
- - - . . u- -
I l --- r K. s_ i
, n<=..- ._ .i , ;c '
B
-'=='
j E
-,F': ,
^^
3
.-N ./ 4. /
- 1 l E a .c 's r .- .-
k F *--- .- r.. - - - -
Aux Ridg/ Containment 135'4'*, 139.0". 147*1" 207
f
~
- . i ,W. . =. .
. . . ~
" ~*
Q EEE EE~33 -E h ,, . . . . .f ~..=. ,
l X l 1
/
7 g t
s
)
}
Lh 6 '
) ** "..' "a" g h-1 l
1(
,"::"s
.v .
({ F 1
~ . , _ ,
,L,='--"
Jw/ '
x (I')
"" y-- '
w ,
- .w w E W N0 ue^
- p., M
,_ m _f___
i d. - _ -- .= N.~ hl. (a- ._- -
a 0
e_j e**h. s b ..
- * =., a.-
= ::
-3 s .
~=' :o . - ,
~~ * -
.-. ~w- ;*,. z= ..,,.,
] q *
.. . . g ,,,,.
.I' i
=
~ *~ 1. *.2..?d ; +:~~ -
e
- L ;y
- ==-r p 4
%.%%.,1
-.. g yW
~
= ..
h i
I i
.,, :A.
~~ g D4y,. .
~~1.
'll"."
M/'
- . Im j
I j
0 .
_ m ja ._. ..
p - g L" a h [g -
a ->9 i r __ _ _
x _/
C
- q. ;
~
4 r(
e .1
- ..~. - - - - n::/
%s
-- . - s
\\ *; - .
~~~
. 1.
k.
-o e
/ _- - e, - ( l) ,4
_ _ . . i
.: l .-
- N.a.i '
f /
=
...,. b ~~.e I I -
= .
- u.a I " .I.Ea T **
Aun EidglCentainment Elevation . Elevation 161'10** S 166'0**
208
.m s' 7 i "7
?7 , , , , = . -- Q l 'i #=#i= -
n \ A _ _.==~VLa ~
l R
- .;;.s AJ \
<= h o
I ,..: = 1 m#
'=.:. =
), ,,, r ~. ,tf
/ N -
d
.[ Oho OE' ),
w . -
t _
't m . . . i
.%*- - -)
~ -
,,g - .
gam *',,: _.,y \ -....
f<l. -
c -
l.
= K, l 7}-'[ /
on_,.,
. - - ,. s. [g li -l,l
. -- l -
2 '. ;
t g =. .
- . a g 'l f. w=
~
.f ,
U
.:. =. -
"1,~_-
- - hb / N ya f
c ~~r - 7- r -
jg
/=m. . _..:: t y 2qe
-]
,j
[- :. ( .i L. /
.:=,
i..F
. i i /- - --
z.
E 'y
__/ 4a c ( .? Q l- - ,,,,,,,,,,
=- \ q.W E e a q. 1 u _. m
== e og-so
> h"--
=a- -
W 1.
-g1 i' "a""J r a ,4J:".ypllllll . .
j.
{ ] :
- -? H- . l e
- t o.- .. =.
.=
~ "
- - ,p
~ -..
- r-, .'
3 fl l-i.
- =
p f.
. ! }
- v. , . . .g, ,; , , ..
Bldg /Centelament. Elevation 184*6** & 185'0'
,+__ -., __ _
r - -:
a
- - ~l n .;.._
=- 7. ,1.- .
lI l
l< . :::=~~
(. ====
'My .
M:
. :. l'
= } < s ., e : - .~%
" , = = -= .Y l 4l j,.._ s. . . ,
w p..:;,; ~7; f,y ,
- ,.. y =-- .
==
g ) - =: a r '-
's s
~'
N s__..
/
,- -- l[
==}
o i > == 1
\>& :; o r -
.m_
,e B -.iIll'tr ;r= . .
i.
j T j A, ,
. / 'N h ;- ] _g.
.. st..< lk r-,
.;__.A>,,%..
.. ., u y.
d' , ,
. . .@t. M. [ ql. t - =) Ei , i% =-
_ fs _A w_..: t . .
.a sy . i..
l(e,=_=_'/\ -g,,, '-'u,
-K+ ; J. ., - S. =.u ,
6.7hv u.
\ -
/
,o.p ,
- ! x* e._
'i{ g M= tri
== , ;, %f g F- #.01 f:
S -- 7 _..
=.= >
s -
LTDT- . WE:- N__~_,_ : _ _
c- m-- ~~0: ;n.,_=
-- ,x. :
.u.
d r - ;i t,%
_,. s_=~. .-- .. qe
- - b. h_- < --
, 3 ,, -
.- - m/ .- o o-- - - _ - -
.__l
-- g i m_r _,,. _ sx'ry_= .7 . . . _.:x.
Lr._ --
I c_J -]..s/ _-i = .,y-
.== 4 _..
. =-
i,. s.i jr_e t,..a-t=___
o l
=_ , , ,, .
, r.- . :.
.s _. _._
- .~s
- i.,_. -s s 1 8 u
- l 1 - g L ----- , , , , , , ,
g* t, L. L.
Aux Eldg/ Containment. Elevation 208'-10**
210
l t
1 l
APPENDIX B HECTR Description Each of the various models in HECTil is dis- The j'" species of gas (1 s j s 4) in the i'r cussed in detail belew except the hydrogen burn mod- ompartment (i s i s 10) will have its molar rate of el which is discussed in section 2.1. change described by B.1 Multicompartment Mass and A Energy Equations -dN'! - 3, Aa va at y vi N,3 + y7v 4 v4 N. i The multicompartment model is a system of ordi- (v < o) (v,i 2 o) nary differential equations expressing mass and ener-gy conservation. The dependent variables are the dN,,\ [dN,np}
wnperature and number of moles of hydrogen, nitro- + ([T),i,,,,,,,, + \ at / .,,.twwon w.
gen, oxygen, and water vapor in each compartment.
Gas flow between compartments is caused by pressure differences and gravitational forces. The gases in each + (/dN,3\T),,,, (83) compartment are assumed to be ideal, homogeneous, and completely mixed at all times. Compartment in-terconnections are modelled as orifices with flow resis- where tance.
Consider two compartments, i and k. The effec- N,3 - number of moles of gas tive pressure di!ference between them is ,
speciesj present in com-partment i g
Aa = interconnection area be-AP - Pi - P + 3 (Z -i Z )(p, + pi) i (B1) tween compartments i and k where Vi - volume of compartment Pi - pressure in compartment i I g -acceleration of gravity Z,-elevation of compartment I [dN,q ~
\ at /ch, i p,= gas density in compartment i # ""E" "" C"" ""
The gas flow velocity, va , to compartment i from compartment k is then /dN'"#3
^
l l - rate of change of water
( dt j ,,.uo. a '"
vapor due to evapora-2AP if AP > 0
- tion or condensation en p,K,, -
v- the wall surfaces and (B2) the spray dropht .
if AP < 0 dN,,
vi Ki-rate of addition of hy-at nin '"""
drogen or water vapor Ka is a coefficient defining the now resistance encoun-into compartment i tered going from compartment i to compartment k. from an external source 211
The conservation equation for energy in the i* com- wall emission is negligible, have no radiant heat trans-fer and are ignored. The bands are determined using partment is then an exponential wide-band molecular radiation mo-
/* ST, SN, del."' For each band, a gas emissivity is calculated l {C,3 No ) u,3 - - which depends upon the temperature, pressure, mole k3' ' /7 + [8"' fraction of steam, and beam length. For Qe rest of this
- discussion, the subscript i will refer to the i* spectral band.
T T 'V'"N i oeh+T T Yk
- N,3h,3 The rate of radiative heat transfer from the j*
Fi' _T surface per unit area, Q, will be the sum of the (via < o) (v,i 2 o) 3 quantities calculated for each spectral band where (BQ) (dQ)
(dQ) atJ,au. \ dt / a, \ dt /.,,.n q _1-9 (B, - J,) , (BS)
+ (dNwp di / .,. .. a. u.,
- P where
- 9- the emissivity of the j* surface (assumed to be
+
/dNo \ h,(T,_) (s4) constant; hence the surfaces are considered to 1-' Q Q be gray bodies)
U - the Planck blackbody flux emitted from the j*
A where surface per unit area T,- temperature in compartment i J,-the radiosity of the j* surface (the rate of radiant heat flux leaving the surface per unit (K) area)
C,3- molar specific heat of the j* gas in compartment i (J/ mole-K) The radiosities themselves are determined by solving ug = molarin. ternal energy of the j. a system of N simultaneous equations, where N is the gas in compartment i (J/ mole) total number of surfaces present:
he- molar enthalpy of the j* gas in "
compartment i (J/ mole) 1 Jj - { F _, j ri(k-j)Ju -
h,(T,_)- molar enthalpy of the j,, gas 19 ,_,
evaluated at the source tem-perature (J/ mole) ,'
U+ #
F _iq (k-j)B,(k-j) 3 (as)
= rate of radiative heat transfer IS i.
(dQ-at ,w,.i out of the gas (watts) where
= rate of convective heat transfer (aq-at ""** out of the gas (watts) F3 -k -geometric view factor looking from sur-fact. j to surface k
-rate of heat transfer out of the at f.,,." gas to the spray droplets (watts) 9 (k-j)- gas emissivity as a function of the beam length from surface k to surface j Therefore, for each compartment there are four mass ,
(k-j)- p umini% 9 H conservat,oni equations and one energy conservation B, (k-j)- gas Planck blackbody flux per unit area equation. For N compartments, a total of SN differen.
tial equations are solved simultaneously. Further details regarding this model are presented in Rermnce a2.
B.2 Radiative Heat Transfer
}n this analysis, the radiative spectrury is broken B.3 Convective Heat Transfer up into seven active regions and the emission from As mentioned in section 2.1, convective models are steam is considered in each spectral region. Regions in included for both vertical slab and horizontal pool which the gas has no radiative bands, and in which 212
I surfaces. The model for heat transfer on a vertical where plane consists of a Nusselt laminar flow solution for wall liquid layers and either a Couette-type forced TGI, TW, THULK -liquid. gas interface, wall sur-turbulent flow model for the air : water-vapor bound-face, and bulk gas tempera-ary layer or a natural convection turbulent boundary tures (K) layer correlation."'The temperature of the liquid gas interface, TGI,is varied until the sum of the convec- Pr Ps'Pb -liquid, saturated steam, and tive plus radiative heat transfer rates through the gas bulk gas densities (kg/m8) boundary layer is equal to the heat transfer rate h,,- heat of vaporization (J/kg) through the liquid.
K,- liquid thermal conductivity (W/m-K)
Maximum of (qc, qNc) + qRAD " 9F (0 71 K,-gas thermal conductivity The subscripts C, NC, RAD, and F indicate forced (W/m-K) convection, natural convection, radiation, and liquid. g,-liquid viscosity (kg/m s)
The heat fluxes, q, are the rates of heat transfer per u - gas viscosity (kg/m-s) unit area. If there is no liar.id layer, the interface temperature is set equal to the wall temperature, and C,- water vapor specific heat terms related to the transport of mass and its associat. (J/kg-K) ed enthalpy (the first terms on the right-hand side of C,,- gas specific heat (J/kg-K)
Eqs (B9) and (BIO)) are dropped in the gas boundary layer correlations. D-diffusion coefficient of water vapor through gas (m'/s)
The correlations used are presented below in di-mensional form using SI units (q is in W/m 2).p2 # = coefficient of expansion of gas (K-8)
Nusselt Liquid Laver Lc - characteristic vertical length of liquid film (m) q, = 0.943 (TGI - TW)
L = characteristic vertical length of gas film (m) p,(p, - p,)hr,K,'(9.807) "*
- u,Lc (TGI - TW) -
a) U6 - buW Weeskam) gas veMy (m/s)
Forced ( vnvection Gas Boundary Laver Re - Reynolds number based on length L (U 6p3L/u) qc - 0.037p,U,Rr *2 Sc-*
- Pr - Prandtl number (C,,u/K,)
Sc - Schmidt number, g/(p3D)
In(1 + B) (MFVII - MFVI)
MFVB - mass fraction of vapor in bulk
.(C,(THULK - TGI) + hr,}/B + gas MFVI-mass fraction of vapor at lig-0.037 K, He** Pr"* .(THULK - TGI)/L (B9) uid-gas interface B - mass transfer driving force Natural-Conecction Gas Ikgendary Laver (MFVB-MFVI)/(MFVI-1) 0.0188 p,rg"* Gr-Grashoff number based on (INC " length L, 9.807# (TBULK-(pbd)v4 TGI)L8 p,2/g*
Sc" In(1 + B) (MFVB - MFVI) r = characteristic velocity in natu-ral-convection boundary layer,
.{C,(THULK - TGI) + hr,}/B 1.185 (r,/L)Gr"'
2
.(1 + 0.5Pr /8)-U8 (m/s)
+ 0.0246 K, Gr25 Pr"5 r,-ger boundary layer kinematic
.[1 + 0.494 Pr 2
/3j-2^(TBULK - TGI)/L (sto) viscosity (m 2/s) 213
8 = gas boundary layer thickness, instantaneous size and mass. The drops are tracked to 0.565LG r-* 'Pr
- the bottom of the compartment, with their final mass and temperature determining the heat and mass
. [1 + 0.5Pr"]'" (m) transfer rates.
It is assumed that the compartment gas is homo-The liquid properties, pr, p,, K, are computed at a geneous and does not change during the fall time of temperature the drops. Therefore, the solutions are quas,-steady i
state. The correlations presented are valid for both T - TW + 0.31(TGI - TW) (B11) condensation on and evaporation from the drops. The The gas properties g,C,,K,,D, and # in Re, Pr, Sc and differential egations describing drop behavior are shown below.
Gr are evaluated at a temperature T - TGI + (TBULK - TGI)/3 (B12) - - (1 + .25 Re* Sc")2rDgD,1n(1 + B) . (B15) dt The model for convective heat transfer to a pool surface is one for laminar now over a horizontal Dat dT. dm 1 [Cp(T - TBULK) plate "*
The correlation used is dt dt mCri, \ (1 + B)'d" -1 '*}/
dz ~4 (9.807)D (pa - pf
- q, .664K,Re* Pr*(TBULK - TSURF)/L ,(B13) dt ~
3 p Ca ~
where TSURF is the temperature of the pool surface and the other terms have been defined previously. where Mass transfer from pools to the gas is approximated by assuming that the thermal conductivity of water in m - droplet mass (kg) the pools is zero (i.e., all heat is deposited on the t - time (s) surface and results in the evaporation of water). The mass transfer is then He - Reynolds number Sc-Schmidt number
$1 - (q, + qun)/[Cei,(TSURF - TPOOL) + h's D - droplet diameter (m)
F C,v(TBULK - TSURF)] , (B 14) p - gas density (kg/m')
where Cei, is the liquid specific heat, TPOOL is the D,- Diffusion Coefficient (m'/s) bulk pool temperature, and the other terms were B - Mass Transfer driving force (defined in defined previously. Eq (B13) and (B14) are not valid Section B.3) in all regions that may be encountered (e.g., when T 4-droplet temperature (K) condensation is occurring). However, they are reason-ably valid when the gas is superheated, such as during Crt.-droplet specific heat (J/kg - K) a burn. We have chosen to ignore the pools whenever Cp-gas specific heat (J/kg - K) the gas in the compartment is not superheated. Since THULK -bulk gas temperature (K) the gas is saturated only at relatively low tempera-tures, we feel that ignoring the pools then results in Le -I ewis number negligible errors. Future versions of HECTR will treat hr,-heat of vaporization (J/kg) the pools in more detail. z - fall height (m) p4 = droplet density (kg/m')
B.4 Containment Sprays C. - drag coefficient The containment spray model can treat a droplet distribution that includes up to ten drop sizes. The The above equations are solved using a standard drops are assumed to be isothermal, spherical, and Runge-Kutta differential equation solver for each travelling at a terminal velocity corresponding to their drop size.
214
The first version of IIECTil only allowed sprays BT , qA to be present in one compartment (usually the dome).
(B 18) dt mC, flowever, because we felt that there might be signifi-cant errors due to the absence of sprays in the wetwell where region for Grand Gulf, an effort was undertaken to modify IIECTIt to allow spray carryover from one q -incident heat Oux (watts /m 2) compartment to another. One case (112') which re-flects this upgrade is presented in Chapter 2.0. m - mass (kg)
C,- surface material specific heat (,J/kg-K)
A - exposed surface area (m')
B.5 Wall Conduction The surfaces in IIECTit are treated either as slabs or as lumped masses. In either case, all of the surface properties are assumed to be constant. Walla consid-cred as slabs are assumed to have m, sulated rear B.5 References surfaces. A finite-difference formulation is then used D. h. Edwards, " Molecular Gas Hand Itadiation,
to calculate the forward wall surface temperatures Advances m Heat 7rans/cr, ed T. F. Irvine, Jr., and J. I .
11artnett, Vol. 12, 115-194, 1976.
given the heat flux to those surfaces. Two different mesh sizes are used mside the slabs: A fine mesh is 8'M. Herman, Light Water Reactor Safety Research Program semiannuaI Report January- Afarch 1981, Sandia used near the surface receiving the heat flux to better National Laboratories, NUREG/ Cit-2163, SAND 81-1216, handle the neep temperature gradients that typically July 1981.
occur there; a coarser mesh is used for the remainder s'J.11. Welty et al, Fundamentals v/ Afomentum, Heat of the wall where the temperature varies more slowly, and Afass Trans/ce, John Wiley & Sons,1969.
The temperature (T) of a surface considered as a a* Internal Communication from M. It. Haer, April 24, lumped mass is given as a function of time by 1980.
9 215 216 1
APPENDIX C The MARCH Code The MARCil (Meltdown Accident ltesponse TPE accident is initiated by a transient event such as Characteristics) emle was developed by the Battelle the loss of off-site power. The reactor is shut down and Columbus 1.aboratories (IICL) for the US Nuclear the safety relief valves open when the pressure in the llegulatory Commission and first used in the reactor gets too high. Then one or more safety relief WASli Iwo rep <>rt. The code has been widely used in valves fail to reseat .md there is a failure of emergency nuclear reactor safety studies, even though it is known core cooling. The water level in the reactor pressure to contain many limitations and some errors."2 The vessel is gradually reduced until MARCII predicts code is fast running and inexpensive, so that many that, at 33 min into the accident, core uncovery begins.
computer runs can be made in a study at reasonable Tha small.to-medium break LOCA, S to S2 , in-i cost. MAllCil has mmlels for many phenomena, in. side the drywell differs from the TPE accident in that cluding core meltdown, core slump, pressure vessel steam and hydrogen are first injected into the drywell failure, core-concrete interaction, and containraent and n >t directly into the suppres. ion pool. As a result, failure. For our study of Grand Gulf, we are mainly nearly all the air initially in the drywell it forced concerned with the MARCII subroutines llOIL through the suppression pool and into the wetwell (which predicts the behavior of the reactor core and early in the accident, raising the pressure in the coolant system) and MACE (which calet.htes the wetwell to about 17 psia. A failure of emergency core containment response). cooling is again assumed. The history of reactor cool-Both IIECTit'and CLASIX 3 are essentially con- ant pressure and temperature, rate of steam and hy-tainment response cmles. The two codes used hydro- drogen generated, and reactor core behavior for th5 gen production and steam production rates derived accident is very similar to that for the TPE accident.
from MAllCil calculations. The most widely used For both accidents the reactor is at low pressure when reactor. core and coolant-system code that will predict the core begins to melt and substantial clad oxidation hydrogen prmluction is MAltCil. begins, approximately 60 min into the accident. Acci-M ARC!l has. frequently been used for BWR safe-dents in which the reactor coolant pressure is high ty studies using a two-compartment model, a drywell dwing core uncovery, such as the TQUV accident and a wetwell. We 1: ave been in telephone contact with sequence, were not considered in this report. We H. O. Wooton and P. Cybulskis of BCL about using discuss the history and the generation of hydrogen in M ARCil in this study. R. O. Wooton informed us that the next section of this appendix.
they have not checked out MARCil for llWR studies with more than two compartments (our configura-tions 11, C, D, and E). In particular, Wooten was C.2 Production of Hydrogen 7 concerned with the treatment of containment sprays in multicompartment models. With considerable ef- The rate and total amount of hydrogen produced by the BOIL subroutine of the MARCil computer fort we developed multiempartment MARCII input of the Grand Gulf containment that appears to give program can be varied over very wide limits depend-ing on the input parameters selected by the code user.
reasonable results,in fair agreement with results ob-Predictions of hydrogen generation depend on exist-tained from IIECTR and CLASIX ing code models, experimental data, and engineering judgment. Input parameters are adjusted to provide C.1 Accident Scenarlos some " reasonable" hydrogen production rate.
Considered Let us consider the course of the BWR accidents consiaerea. The pressure in the 7-actor yessei a, ops The accident scenarios considered are variations rapidly as shown in Figure C.I. At the time core on two types: the TPE accident scenario, and the uncovering begins, 3 min into the accident a:nall to-medium break LOCA inside the drywell. The (Figure C.2), the pressure is about 525 psia; by the i
217 f
_ _ . _ . - - _ - = _ . _
)
i time the core is completely uncovered, at 50 min, the However, there is no evidence that the zirconium in presaure is down to about 200 psia. There are slight the debris cannot indeed oxidize very rapidly and differences in these figures for different accidents produce a large hydrogen release spike.
because of the coupling of the reactor coolant system MARCH also contains an option which treats the
' to the containment. During core uncovering the steam slumped debris as composed of spherical particles
, gineration rate is very high, as shown in Figure C.4, with three concentric shells. The initial diameter and ,
- i. and the core stays cool. MAEH indicates that at the composition of the shells are input parameters to the
' time the core is completely uncovered, there is fission- code. We had difficulty using this option and did not
-product release (presumably caused by clad rupture) get it to work until very late in the study. Some l but the amount of clad oxidation is negligible, far preliminary calculations with this debris bed model below I G, and hence there is negligible production of resulted in predictions of rapid quenching-the sur-hydrogen. face temperature of the particles was quickly reduced For our MARCH study we have taken the zirconi- to a point where zirconium oxidation ceamd. A model um inventory of the reactor core as 174 700 lb. This which results in quenching the melt will clearly pro-zirconium will be referred to as " clad," even though it vide a lower bound on hydrogen generation subse- j
- includes clad and other zirconium structures in the quent to core slumping.
core. MARCH contains an option in which i ach node is ,
The rate of steam generation drops greatly after assumed to slump into the lower plenum as soon as it the core is completely uncovered. The clad and reactor melts and then to oxidize in a single time step. Use of core begin to heat up rapidly. MARCH prelicts an this option insures that MARCH will give a series of increasing rate of clad oxidation, limited by the short- small hydrogen production spikes instead of one large
, age of steam. The beginning of core melting occurs at spike. This behavior (small sections of molten core
~
60 min into the accident. The amount of clad oxida- falling directly into the lower plenum) is probably tion at that time is less than IFc. Consequently, the unrealistic, since experiments have indicated that the generation of hydrogen caused by clad oxidation is molten core will tend to refreeze in the lower, cooler closely tied to core melting for the low-pressure BWR portions of the core. A pool of molten core might form j accident. such that gradual dripping into the lower plenum in MARCH, the user specifies the fraction of could occur, as in the option considered, but at a later zirconium at each reactor core node that will be al- time. Even if such late time dripping did occur, it Iowed to oxidize, and the fraction of the core that, could not be modelled in MARCH, which assumes a
! when melted, will slump into the lower plenum of the sudden massive drop of the molten core into the lower pressure vessel. If one uses 75% core melt for the plenum.
criterion for slump, then about 20re of the clad will be For our study we delayed core slump in MARCH
- oxidized at core slump. This corresponds to about by requiring complete core melt before slump. As a 150015 of hydrogen praduced. Since we are consider- result, we arbitrarily eliminated the large hydrogen ing accidents with greater amounts of hydrogen pro- production spike. Various hydrogen production pa-duced, we are forced to consider slumped-core behav- rameters are shown in Figures C.5 through C.10. How-l ier ever, core behavior predicted by this model is unrealis-L The model of molten core behavior most frequent- tic. The hottest portion of the core is predicted to be at ly used in MARCH assumes that the core is held up an unphysically high temperature (above the vapor-l until a given fraction melts. At that time the entire ization temperature).
- core is assumed to drop into the lower plenum and all it can be seen from the previous discussion that the remaining rirconium is oxidized in one time step. the hydrogen production rates predicted for the
! This gives rise to a Wge spike in hydrogen production, HECTR, CLASIX-3, and MARCH computations are and usually a very large burn in containment, result- unreliable, especially after about 1500 lb (680 kg) of ing in containment failure. This difficulty was noted hydrogen are produced. A more reliabh prediction of i iii cur earlier work on the Sequoyah ice condenser hydrogen production rate will require a better model -
i plant and in the CLASIX-3 analysis of Grand Gulf. If of core melting and slumping and experimental data
, one believes the hydrogen production spike is not on these phenomena, so that simple programs like l
realistic, one must try to avoid it by some artifact. MARCII can be adjusted to give desired rates of
[' hydrogen production.
! s, 218 .
Y s
it i
i l
l GRAND GULF PLANT MODEL 28000 Cal M 3000D -
D m
IS00D -
2
[4 b
W 1000D -
2: 300D -
OD , . . . . . . . . .
00 ro00 40.0 00D 8DD 10(10 12(1 0 1400 160 0 18t10 20t10 IEDD TIME - (MINUTE)
Figure C.1. Case A.I. Ileactor coolant pressure, psia.
a t.3 GRAND GULF PLANT MODEL MD Ct: 900D - I f 300D -
5 g 400D -
H g 300D -
M 0:
MIDD -
g<
n.
2 ICOD M OD NO 40D g O0D 80 0 Idt10 l$(10 14'4 0 l$0.0 IN110 25t10 2500 TlME - (MINUTE)
Fspre C.2. Case A-1. Ileactor coolant temprature, 'F.
219
GRAND GULF PLANT MODEL d
15.0 -
M GC
- D
[
10D -
2 2
h SD-M l
00 g OD --
9
~
-sD , . , , , . , . . .
DD 200 40D 00.0 00D 100.0 12 0.0 140.0 100.0 100.o 200.0 220D TIME '(MINUTE)
Figure C.3. Case A-1. Mixture level in reactor, ft (Top of active core at 12 ft).
,w GRAND GULF PLANT MODEL M 30D h
y xD-20D -
m o
15.0 - 1 N
t2 l te t g 10.0 - I O
<r.
M l l h
a 6D-OD y
OD 20.0 4DD 000 80D 100.0 120.0 140.0 100.0 100.0 200.0 220D TIME - (MINUTE)
Figure C.4. Case A.I. Net Steam leakage rate from reactor coolant system (Steam Production rate - steam reacted with zirconium),10'lb/ min.
220
4 GRAND GpLF PLANT MODEL i
M
% 4000D-N l h N
3000D-H 8 2000D-N O
4 C4 1000D -
C OD , , , , , , , , , ,
OD 20D 40D 00D 00D 100.0 12 0.0 140.0 160 0 1f10.0 200.0 220D TIME - (MINUTE)
Figure C.5. Cane A.I. Average reactor core temperature, *F.
GRAND GULF PLANT MODEL l 6000 0 S
U 4000D-8 g xm. --
M Z
N O
2000D-l I
1000D -
l OD NO dD O'0D ND IdC.0 1$n0 14h0 1$0.0 INO.0 Edo.O EE04 l TIME - (MINUTE)
Figure C.6. Ilydrogen Release into Containment for Case A-1 without burns. Approximately true for all l MARCII cases.
l 221 l
- s w GRAND: GULF PLANT MODEL M
p 400D in m 350D-4 l2 5300D -
A 200.0 -
x N
y 200D-g 150D -
O 100.0 -
S 50D -
0 n: CD , , ,
l
' " * ^ * " , ' .""
, r,,,, ,"
A 04 20.0 40D 00D 80D 100.0 12 0.0 14GLO 100.0 100.0 200.0 2202
$ TIME '(MINUTE) !
Figure C.7. Cane A.1. liydrogen leakage rate into containment, Ibhain.
GRAND GULF PLANT MODEL D
o 02-ta3
% OS-Q 5
O y 0.4 -
l 6 b M ,
l k 02 l
I OD , , , , , , , , , ,
OD 20 0 40.0 00D WD 10t10 120 1440 N00 1510 E300 Wie TIME - (MINUTE)
Figure C.8. Case A-1. Fraction of zirconium reacted.
222 l
l
1 j
GRAND GULF PLANT MODEL l is ._
03-Ca3 B
06-8 0.4 -
t 4
02 -
OD , , , , , , , , , ,
OD 20 0 40D 80.0 OOD 10E0 lano 1440 1800 IIRO BBS WOD TIME - (MINUTE)
Figure C.9. Case A 1. Fraction of reactor core melted.
GRAND GULF PLANT MODEL 2004 M
m M
150.0 -
b x
G.
3 100D -
13 E
Z w GOD -
0 k
a 0.0 , , , , , , , , , ,
OD 20.0 40D OOD 80D 1000 1200 140.0 16 0.0 100.0 200.0 220D TIME - '(MINUTE)
Figuro C.10. Case A-1. Ilydrogen mass in reactor coolant system, Ib.
29 \
223 i
C.3 Compartment In MARCH, HECTR, and CLASIX-3, one speci-fies the hydrogen mole fraction at which combustion InterCOnneCliOnS is assumed to start (ignition). We have used values of Compartments are interconnected in MARCH by 0.10 and 0.08. The completeness of the burns can be the use of a series of one-way, zero-flow-resistance specified in each code. In MARCH, one specifies the connections. Usually, the user employs two one-way value of hydrogen mole fraction at which combustion connections between two compartments, one in each is assumed to be extinguished. For complete combus-direction. At each time-step MARCH attempts to tion, this value would be zero.
equalize the pressures of all compartments so inter- A compartment is assumed to be inert if the connected. If the difference in pressure between the oxygen mole fraction is too low. The MARCH code compartments is too great, the code requires a few uses an oxygen mole fraction of 0.065 for inerting; time-steps to equalize pressure. When the drywell was HECTR and CLASIX-3 use an oxygen mole fraction connected to the wetwell by a single one-way connec- of 0.050 for inerting. We have altered the MARCH tion to the wetwell, MARCH did not equalize pres- code to use an oxygen mole fraction of 0.050 for sures between the wetwell and drywell. inerting, and carried out most of our runs with this Our study using HECTR shows that the MARCH value. However, our first runs in the B configuration model of equal pressure in all the wetwell/contain- were made using the 0.065 limit. In section 3.4 we ment compartments is reasonable. Pressure differ- compare the results obtained with the two limita.
ences between compartments based on realistic esti- We have inserted into the MARCH code a criteri-mates of flow resistance were found to be small. on for steam inerting: a steam mole fraction of 0.56 or However, the drywell presents a problem. In the more. In none of the cases considered did steam CLASIX-3 analysis, considerable discussion is inerting occur except in the drywell during portions of given to the vacuum breakers. Some simple calcula- the drywell break accident.
tions made at SNL suggest that the resistance of the A major weakness in MARCH is its lack of inore vacuum breakers is too great to permit large flows into realistic flame propagation options. In all three codes, the drywell after a wetwell/ containment burn. Howev- one may require each compartment to meet the hydro-er, a large pressure in the wetwell can drive the gen ignition criteria in order to have a burn. Several of suppression pool downward and uncover the vents to the cases used this criterion. For most of the cases the drywell. This suggests that most of the flow to the calculated by HECTR, the flame was assumed to drywell may be from the bottom of the wetwell after a propagate into adiacent compartments after some large hydrogen burn. time delay if the hydrogen mole fractions in the adja-For most of the calculations discussed in Chapter cent compartments were above the appropriate flam-3.0 we ignored the presence of the drywell volume by mability criterion (fnr upward, downward, or horizon-having a connection from the drywell to the wetwell, tal propagation) and the compartnents were not but no connection back to the drywell. This results in otherwise inerted. In MARCH, one has an option that having a constant drywell temperature, pressure and will allow flame propagation between adjacent com-gas content, effectively isolating the drywell. This was partments. Apparently, when the option is used the done to get a comparison with HECTR computations. only thing that will stop such propagation between HECTR has no drywell model. However, we have a interconnected compartments is a hydrogen mole number of calculations in which either the wetwell or fraction in the second compartment that is below the upper containment was connected to the drywell. limit considered for extinguishment. If complete com-These calculations are considered in section 3.8 of the partment combustion is considered, MARCH will text. cause all the hydrogen in all the interconnected com- (
partments to be burned (if there is sufficient oxygen to combine with the hydrogen). Combustion will begin in C.4 Hydrogen Burn Model an adjacent compartment even if the initial oxygen or In considering the hydrogen burn model, we must hydrogen mole fraction in that compartment is below '
discuss the criterion used for the initiation and burn- the inerting or flammability limit.
ing of hydrogen in a single compartment. In addition, MARCH prints out the number of moles burned for multicompartment models we must also discuss in each compartment and the adiabatic isochoric pres-the criterion used for propagation of combustion from sures corresponding to the burning of the hydrogen in one compartment to an adjacer.t compartment. We each compartment. Given these numbers, one can will first consider the behavior in a single compart- determine whether a burn is mainly in one compart-mint, ment, or is truly a multicompartment burn. In 224
IIECTIt and CLASIX-3, burns are strictly confined to CLASIX-3 uses a flame speed of 6 ft/s (2 m/s),
a single compartment whenever the hydrogen mole with some runs et 12 ft/s (1 m/s). We have used burn fraction in the other compartments is below some times corresponding to 2 m/s for runs intended to specified propagation limit. duplicate the CLASIX-3 runs llowever, experimental In HECTil and CLASIX-3, one can specify for data from testa carried out at Sandia and McGill each compartment a burn time which la equal to the indicate that ihe effective speed of the Game front ratio of a characteristic compartment length to the may be much higher. Possible causes of these higher flame speed. In MAllCII, one specifies a 0, gle burn speeds are turbulance and the folding of flame around time. We have investigated the effect of burn time on obstacles in the path. Because of the long distances for the peak pressure generated by a hydrogen burn for flame propagation inside the containment building, identical initial conditions. Wnhout sprays, heat and the multitude of obstacles in parts of the flame transfer effects cause a small reduction in peak pres- path, the flame speed is expected to be high, possibly sure for increasing burn times. With sprays, the effect higher than the value (8 m/s) used in most of the of longer burn times on reducing the peak pressure is MAllCH calculations.
mu:h greater. This is shown in Figure C.11.
I 1.0 , , , , , , , , , , , , , , , , , i.
1
, .. ,'~~ ____,
\
0.9 -
. ' . \
, - l
'A, -
' 4, s 's, 0.8 -
k s, '< >
)
s .
0.7 -
's -
X
< :$ 's 2 e s s
- a. a. 0.6 -
-O- HECTR, NO SPRAYS 's -
--G-- HECTR, SPRAYS ,
0.5 - ' *
- G-- MARCH, NO SPRAYS 's N' ]
--1?r-- MARCH, SPRAYS -
's '
', I o.3 ' ' ' ' ' ' ' ' 'd '
O.01 0.03 0.10 0.30 1.00 BURN TIME (minutes)
) Heure C.11. Effect of Hydrogen Combustion Burn Time on the Predicted Single Compartment Peak Pressure by llECTil and MAllCli Codes.
i 225
C,5 Natural Convection pressure after a given burn is reduced and the tem-perature and pressure fall much more rapidly than cnd Mixing without sprays. In Figure C.11 is shown the effect on None of the three codes (h1ARCll,11 ECTR, or peak burn pressure caused by convective heat transfer CLASIX.3) contains provision for natural convection. only (no spray case) and then convective heat transfer Considering ths large density inversions found after with sprays. Note that the reduction of peak pressure wetwell burns and the results of the RALOC calcula- predicted by h1ARCil is greater than that predicted tions, a large ameunt of mixing by natural convection by 11 ECTR for the same burn time.
is expected. If the mixing in cor.tainment is rapid enough, the results of configuration A (in which all the wetwell/ containment volume is considered as one C.7 Containment Fa. dure compartment with a homogeneous atmosphere) may In MARCII one can specify a value of pressure at be more realistic than those of the multicompartment which containment is assumed to be breached. De-models. pending on the option used, the code continues the The mixing that does occur in MARCil and the computation past containment failure with the pres-other codes is artificial. It is cansed by forcing flows sure in containment assumed to be atmospheric. We between the compartments and having each inflow used an arbitrary high containment failure pressure.
immediatey mix with the previous contents of the 100 psia (above the NRC estimated value for Grand compartment. This means that the degree and speed Gulf,71 psia), for all of our MARCH calculations. As a of mixing is governed by the compartmentalization result, the code calculated each run as if the contain-model used in th- code. Consequently, we have carried ment had not failed, since 100 psia pressure was not out studies with both MARCll and 11 ECTR for a reached in any of the cases considered.
variety of compartmentalization. The results vary '
greatly with the compartment model used. We dis-cussed this point further in Chapter 3.0. C.8 References c.iR. O. Wooton and II. L Avci, AfARCH (Afcitdown
^"'"' "Ponse Characteristics) Code Description and C*6 SbraYs User's Afanual, llattelle Columbus Laboratory, NUREG/
The MARCil manualC' is not very clear on the CR.1711, BMI-2064 R3, October 1980.
proper use of the spray option in the code. Neverthe- c. L B. Rivard et al, Interim Technical Assessment of less, we believe we have correctly carried out the thc AfARCH Code, Sandia National laboratories, NUREG/
insertion of sprays. When sprays are on, the peak CR-2285, SAND 81-1672113, November 1981.
226
DISTRIBUTION:
US NRC Distribution Contractor (CDSI) (340) Swedish State Power Board (3) 7300 Pearl Street El-Och Vaermeteknik Bethesda, MD 20014 Sweden 315 copies for R3 Attn: Eric Ahlstroem (2) 25 copias for NTIS Wiktor Frid US Nuclear Regulatory Commission (13) Berkeley Nuclear Laboratory Office of Nuclear Regulatory Research Berkeley GL 139PB Washington, DC 20555 Gloucestershire Attn: R. T. Curtis United Kingdom J. T. Larkins (10) Attn: J. E. Antill P. Worthington (2)
Gesellschaft fur Reakforsicherheit (GRS)
US Nuclear Regulaej Commission (3) Postfach 101650 Office of Nuclear Regulatory Research Glockengasse 2 Washington, DC 20555 5000 Koeln 1 Attn: B. S. Burson Federal Republic of Germany M. Siiherberg Attn: Dr. M. V. Banaschik L.S. Tong Battelle Institut E. V.
US Nuclear Regulatory Commission (6) Am Roemerhof 35 Office of Nuclear Reactor Regulation 6000 Frankfurt am Main 90 Washington, DC 20555 Federal Republic of Germany Attn: J. K. Long Attn: Dr. Werner Baukal J. F. Meyer R. Palla UKAEA Safety & Reliability Director (3)
K. I. Parczewski Wigahaw Lane, Culcheth G. Quittschreiber Warrington WA34NE D. D. Yue Cheshire United Kingdom US Nuclear Regulatory Commission (11) Attn: J. G. Collier (2)
Office of Nuclear Reactor Pegulation S. F. Hall Washington, DC 20555 Attn: V. Benaroya British Nuclear Fuels, Ltd.
W. R. Butler Springfield Works G. W. Knighton Sclwick, Preston T.M.Su Lancs Z. Rosztoczy United Kingdom C. G. Tinkler (6) Attn: W. G. Cunliffe, Building 396 US Department of Energy AERE Harwell (3) l Operational Safety Division Didcot Albuquerque Operations Office Oxfordshire OX11 ORA PO Box 5400 United Kingdom Albuquerque, NM 87185 Attn: J. Gittus, AETB (2)
Attn: J. R. Roeder, Director J. R. Matthews, TPD US Bureau of Mines Pittsburgh Research Center PO Box 18070 Pittsburgh, PA 15236 Attn: M. Hertzberg 227
DISTRIBUTION (cont):
Kstnforschungszentrum Karlsruhe (5) CNEN NUCLIT Postfach 3640 Rome, Italy 75 Karlsruhe Attn: A. Morici Federal Republic of Germany Attn: Dr. S. Hagen (3) CEC Dr. J. P. Hosemann Rue De La Loi 200 Dr. M. Reimann 1049 Brussels Belgium Simon Engineering Laboratory Attn: B. Tolley University of Manchester Director of Research, Science & Education M139PL United Kingdom Bechtel Power Corporation Attn: Prof. W. B. Hall 15740 Shady Grove Road Gaithersburg, MD 20877 Krtftwerk Union (3) Attn: D. Ashton Hammerbacher strasse 12 & 14 Postfach 3220 Northwestern University D 8520 Erlangen 2 Chemical Engineering Department Federal Republic of Germany Evanston, IL 60201 Attn: Dr. K. Hassman (2) Attn: S. G. Bankoff l Dr. M. Pechs l Brookhaven National Laboratory (3) l Gesellschaft fur Reaktorsickerheit (GRS mbH) (3) Upton, NY 11973 1 8046 Garching Attn: R. A. Bari (2)
Federal Republic of Germany A. Berlad Attn: E. F. Hicken (2)
H. L. Jahn Westinghouse Hanford Company (5)
PO Box 1970 Technische Universitact Muenchen Richland, WA 99352 D.8046 Garching Attn: G. R. Bloom (3)
Federal Republic of Germany L. Muhlstein Attn: Dr. H. Karwat R. D. Peak McGill University (3) UCLA 315 Querbes Nuclear Energy Laboratory Outremont, Quebec 405 Hilgard Ave.
Canada H2V 3W1 Los Angeles, CA 90024 Attn: John H. S. Lee Attn: I. Catton AEC, Ltd. Argonne National Laboratory i Whiteshell Nuclear Research Establishment (3) 9700 South Casa Ave.
Pinawa, Manitoba Argonne, IL 60439 Ccnada Attn: H. M. Chung Attn: D. Liu (2)
H. Tamm University of Wisconsin :
Nuclear Engineering Department )
National Nuclear Corp. Ltd. 1500 Johnson Dr.
Cambridge Road Madison, WI 53706 Whetestone, Leicester, LE83LH Attn: M. L. Corradini-United Kingdom Attn: R. May 228 .
t-t-* - - - * *-ee
- w --
P- v- -'
W
DISTRIBUTION (cont):
l Los Alamos National Laboratory (7) General Electric Co. (2) l PO Box 1663 175 Curtner Ave.
Los Alamos, NM 87545 MC N IC157 Attn: H. S. Cullingford (4) San Jose, CA 95125 R. Gido Attn: K. W. Holtzclaw J. Carson Mark G. Schott Duke Power Co. (3)
PO Box 33189 Battelle Columbus Laboratory (3) Charlotte, NC 28242 505 King Ave.
Attn: F. G. Hudson (2)
Columbus, OH 43201 A. L. Sudduth Attn: P. Cybulskis (2)
R. Denning S. Levy, Inc. i 1999 S. Bascom Ave., Suite 725 Power Authority State of NY (3) Campbell, CA 95008 10 Columbus Circle l Attn: C. Roy Jores l New York, NY 10019
{
Attn: R. E. Deem (2) Westinghouse Corporation (5) !
S. S. Iyer PO Box 355 Pittsburgh, PA 15230 Massachusetts Inst of Technology Attn: N. Liparuto (3)
Room 1-276 J. 01hoeft Cambridge, MA 02139 V. Srinivas Attn: M. N. Fardis General Physics Corporation (5)
Offshore Power System (3) 1000 Century Plaza 8000 Arlington Expressway Columbia, MD 21044 PO Box P000 Attn: Chester Kupiec Jacksonville, FL 32211 Attn. G. M. Fuls (2) TVA (2)
D. H. Walker 400 Commerce, W9C157-CD Knoxville, TN 37902 Electric Power Research Institute (7) Attm Wang Lau 3412 Hillview Ave.
Palo Alto, CA 94303 EG&G Idaho Attn: J. J. Haugh (4) Willow Creek Building, W-3 K. A. Nilsson PO Box 1625 G. Thomas Idaho Falls, ID 83415 L. B. Thompson Attn: Server Sadik Fauske & Associates NUS Corporation 627 Executive Dr. 4 Research Place Willowbrook, IL 60521 Rockville, MD 20850 Attn: R. llenry Attn: Richard Sherry Mississippi Power & Light (5) University of Michigan PO Box 1640 Dept of Aerospace Engineering Jackson, MS 39205 Ann Arbor, MI 47109 Attn: S. II. Hobbs Attn: Martin Sichel 229
DISTRIBUTION (cont):
Roger Strehlow 8523 B. R. Sanders 505 South Pine St. 9268 D. Tomasko Champaign, IL 61820 9400 A. W. Snydcr 9410 D. J. McCloskey Applied Sciences Association,Inc. 9411 V. L. Behr PO Box 2687 9411 A. S. Benjamin Pelos Verdes Pen., CA 90274 9411 F. E. Haskin Attn: D. Swanson 9412 J. W. Hickman 9412 A. M. Kolaczkowski Acurex Corporation 9412 R. G. Sp'dak 485 Clyde Ave. 9415 D. C. Aldrich Mountain View, CA 94042 9420 J. V. Walker Attn: Ray Torok 9420A J. B. Rivard 9422 D. A. Powers Bechtel Power Corporation 9422 A. R. Taig PO Box 3965 9424 M. J. Clauser San Francisco, CA 94119 9425 W.J. Camp Attn: R. Tosetti 9425 A. J. Wickett 9440 D. A. Dahlgren Factory Mutual Research Corp. 9441 M. Berman (25)
PO Box 68S 9441 A.L. Camp i Norwood, MA 02062 9441 R. K. Cole, Jr. I Attn: R. Zalosh 9441 J. C. Cummings, Jr.
9441 S. E. Dingman Sindia National Laboratories (87) 9441 N. A. Evans PO Box 5800 9441 D. P. Kelly Albuquerque, NM 87185 9441 L. S. Nelson Attn: 1131 B. Morosin 9441 P. G. Prassinos 1131 W. B. Benedick 9441 E. W. Shepherd 1131 J. Fisk 9441 M. P. Sherman 1500 W. Herrmann 9441 S. R. Tieszen 1513 M. R. Baer 9441 M. J. Wester 1513 D. W. Larson 9442 W. A. Von Riesemann 2510 D. H. Anderson 9444 L. D. Buxton 2512 V. M. Loyola 9444 R. K. Byers 2513 J. E. Kennedy 9444 S. L. Thompson 2513 S. F. Roller 9444 G. G. Weigand 2513 J. E. Shepherd 9445 L O. Cropp 2612 B. V. Burnham 9445 W. H. McCulloch 8513 W. J. McClean 8214 S. L. Tyler 8523 W. T. Ashurst 3141 L. J. Erickson (5) 8523 K. D. Marx 3151 W. L. Garner (3) ,
4 230
l
'O U S. NUCLEAR REGULATORY COMMISSION JU G R-25 N # #
BIBLIOGRAPHIC DATA SHEET SAND 82-0218 4 1ITLE AND SUBTSTLE (Acts Volume No, of appropriate) 2. (Leave btwk)
Review of the Grand Gulf Hydrogen Igniter System l 3 RECIPIENT'S ACCESSION NO.
' ^urnOHm J. C. Cummings, A.L. Camp, M. P. Sherman, 5. ATE REPORT COMPLE TED M.J. Wester, D. Tomasko, R.K. Byers, B.W. Burnham gbr l
ry 1983 9 PE HF OHMING OHGAN1/A TION N AME AND M AILING ADDRESS finclude I,p Code) DATE REPORT ISSUED voNTH l YEAR March 1983 Sandia National Laboratories , ,L,,,, ,,,,,,
Albuquerque, NM 87185 8 (Leave blanki 17 LPONSOHING OHG ANI/ A TION N AME AND M AILING ADOHE SS IInclude 2,p Codel p
Division of Systems Integration Office of Nuclear Reactor Regulation ,,. ny no.
U. $ . Nuclear Regulatory Commission Washington, DC 20555 A1308 and A1246 13 T Y PE OF HEPOHT PE RIOD COV E RE D //4clussve dJfes/
Technical 15 SUPPL E ME N TAH Y NO TE S 14 (Leave o/m A f 16 ABSTH AC T (100 words or less/
The Mississippi Power and Light Company has proposed installation of a Hydrogen Igniter Systems (HIS) at the Grand Gulf Nuclear Station (BWR Mark III) ta burn hydrogen generated during accidents more severe than the design-basis accidents. Sandia National Laboratories, under a con-tract with the U. S. ' Nuclear Regulatory Commission, has perfonned a technical evaluation of the adequacy of the proposed HIS to meet the threat posed by hydrogen combustion. Areas considered in this review include HIS design and testing, location and distribution of igniters containment pressure and temperature response calculations, detonatio,ns, containment atmosphere mixing mechanisms, actuation criteria for the HIS, and the spectrum of hydrogen-generating accidents.
KE Y WORDS AND DOCUME NT AN AL YSIS 1 74 DE SC RIP T O HS l
17h IDE N TIFIE RS OPE N EN DE D TE HYS 18 AV AILABILITY ST ATE ME N T 19 CU t Y A ( h.s report / 21 r F PAGE S Unlimited zos e ngyn eso-i n eniCE s
sac.0 u us.n.,,
UNITED STATES soonin et ass mat NUCLEA7 f.E2ULATORY COMMISSION tostoss a re ss er.o WASHfNGTON. O.C. 20565 $f.'o c
,. .,m. w OFFICIAL BUSINESS PENALTY FOR PRrVATE USE $300 0
1 1 N01048 120555078877 US NRC ADM DIV OF TIDC BR-PDR NUR EG POLICY C PUB MG1 20555 l W-501 DC WASHINGTON 9
3