ML19320A531
| ML19320A531 | |
| Person / Time | |
|---|---|
| Issue date: | 04/22/1980 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-1722, NUDOCS 8006250351 | |
| Download: ML19320A531 (28) | |
Text
9..
A N-/Ye20' 11 s IS$UE DATE 4/
80
,'.3'q
!gjf f g MEETING DATE: 3/ /80 l'i.
4 4
.. i ;. 2 tilNUTES OF Tr!E ACRS AD HOC TMI-2 ACCIDENT IMPLICATIONS SUBCOMMITT REGARDING NUCLEAR POWER PLANT DESIGN WASHINGTON, DC MARCH 5, 1980 1
The ACRS Ad Hoc Subcomittee on the Three Mile Island 2 Accident Implications Regarding Nuclear Power Plant Design held an open meeting on March 5,1980 in Room 1046, 1717 H St.,,'NW, Washington, D.C.
The purpose of this meeting was to discuss the implications of the March 28, 1979 accident at the Three Mile Island, Unit 2 station and to discuss recent studies on additional engineered safety features at Indian Point Units 2 and 3 and Zion, Units 1 and 2.
Notice of this meeting was published in the Federal Register on A
February 19, 1980. A copy of this notice is included as Attachment A.
list of attendees for this meeting is included as Attachment B, and a schedule for this meeti.ng is included as Attachment C.
Selected portions of the meet-ing handouts are included as Attachment D.
A complete set of handouts has been included in the ACAS Files. There were no written statements or requests The for time to make oral statements received from members of the public.
Designated Federal Employee for this meeting was Mr. R. Major.
EXECUTIVE SESSION Dr. Okrent, Subcommittee Chairman, opened the meeting by stating the purpose of the meeting, which was to discuss recent studies on additional engineered The safety features at Indian Point, Units 2 and 3, and Zion, Units 1 and 2.
Subcommittee, in joint session with the ACRS Ad Hoc TitI-2 Accident Action Plans Subcommittee, heard a briefing on the February 26, 1980 transient that took place at the Crystal River-3 Nuclear Station.
INTRODUCTION - ZION /INDI AN POINT TASK FORCE - J. 01shinski, NRC Staff Mr. Olshinski said that as 1t result of TMI-2 follow-up actions, the Staff has looked in mora detail at emergency planning and evacuation in general.
Because of the high population densities surrounding the Zion and Indian Point Plant's, the Staff is undertaking a special rev N of these plants.
The Task Force is attempting to address the question
. whether or not these plants, because of the high population densities, should add addi-Severe tional accident mitigation features not required at other plants.
8C*06260 Y
j/
-c-accident mitigation features being considered include filtered-vented containments, core retention devices, and hydrogen control methods.
Mr. 01shinski mentioned a Union of Concerned Scientist petition regarding Indian Point (IP) that requested decomissioning IP 1 and shutting down The Staff held meetings in early February with the Commissioners.
of IP 2&3.
A showcause order was issued in relation to decommissioning IP-1; the However, confirmatory orders were issued petition was denied regarding IP 2&3.
to IP 2&3 and then on Zion 1&2 regarding a number of interim operational actions to be taken at these plants because of the high population densi-ties involved. The Commissioners have issued a solicitation for public coment on the NRR Director's decision concerning the petition regarding the IP piants and concerning the orders issued to these plants.
The Staff concern centers on assuming the PWR plant design of WASH-1400 was moved to the Zion or IP sites.
If the design is assumed essentially the same as in WASH-1400 together with the higher population density, there would be a significant increase in risk.
Certain interim operational actions, for the Zion and IP plants, will be re-quired by the Staff while reviewing and evaluating severe accident mitigation In addition, the Staff is pursuing, under an accelerated schedule, features.
The in-any outstanding plant specific or generic actions at these plants.
terim operational actions include certain staffing and training requirements, improving testing and maintenance, augmentation of the onsite technical staff, certain operational requirements, certain analyses that are being conducted, and certain actions and reviews by the NRC Staff. Slides 1-5 list these items and indicate the number of days after the order is issued for imple-mentation of each item. Some example items included: aditional SRO manning, containment leak test, vendor representation onsite, restrict plant to base load operation, and analyze control room habitability.
The Staff noted that the long-tenn design changes for severe accident miti-gation features will require evaluation, development, and consideration as
to whether or not they are feasible and what benefits they will provide.
In the meantime, a number of interim operational actions (dealing with staffing and training, test and maintenance, augmented onsite technical staff, etc.) will be required. The Staff feels these items taken collec-tively are of value and should be done as the long-term effort proceeds.
OPERATING REACTORS AT HIGH POPULATION DENSITY SITES _ - J. Meyer, NRC S Mr. Meyer noted that the IP and Zion sites are believed to present a disproportionately high contribution to the total societal risk from re-The cumulative population around these sites is greater actor accidents.
The Staff than that suggested in Regulatory guide 4.7 and the average site.
has asked the IP and Zion licensees to determine what additional measures and/or design changes can and should be implemented that will further re-duce the probability of a severe reactor accident and will reduce the consequences o,f such an accident by either reducing the amount of radio-active releases and/or by delaying any radioactive releases which would provide additional time for evacuation near the sites.
Three key points Mr. Meyer made were:
Based on population density and assuming similar accident proba-bilities, reduction of accident consequences by a factor of 10 would bring IP and Zion, down to the same consequences as those of an average site (assuming similar meteorology and evacuation routes).
Based on a number of studies (Sandia, BCL, UCLA)'a number of mitigation features (such as a filtered-vented containment system for example) have been identified that have the potential to re-duce the risks from severe accidents by at least and order of magnitude.
NRC has a progrant underway, in parallel to the utilities program, to identify that package of viable systems that is sufficient to I
produce at least an order of magnitude in risk reduction and to require that these systems be installed in a timely manner (~2 yrs.)
in these plants.
4 The purpose of NRC's severe accident mitigat1cn reaturva study is to di-temine how immediate and practical technical fixes can be implemented in the' Zion and IP units that assure a real and significant reduction in societal and individual risk due to severe accidents, including core melt.
The general approach is to pursue actively those design features that contribute favorably toward the mitigation of the consequences of a severe Mr. Meyer noted that it is not the Staff's inte' tion to design n
accident.
mitigating features,as this would be the responsibility of the utilities.
The NRC does plan to work through conceptual designs considering all practical alternatives..
There are three severe accident mitigation features that are being addressed These are filtered-vented containment systems, in the Staff's, program.
In addition, the Staff core retention devices, and hydrogen control methods.
is proceeding with steam explosion studies and accident risk evaluations.
Mr. Meyer continued with an overview of the basic components of the NRC Zion /
A key input into the Staff's evaluation will be the Zion /IP spe-IP program.
cific integrated reliability evaluation program reviews. This program will be used-as input to determine significant risk contributors and accident sequences Part two of the program will be to detemine the evolution for Zion and IP.
The MARCH / CORRAL system of codes will be used to de-of accident scenarios.
velop a pressure, temperature and radiological source term history in the This in turn will be input to a consequence evaluation for Zion containment.
and IP (CRAC analysis). In addition to determining the consequences speci-fically for the Zion and IP plants as they are currently built; this program can also consider the change in consequences when nitigating features such as filtered-vented containments are added to the system Mr. Meyer noted that in terms of mitigating features, presently the most u
important in terms of effort is the filtered-vented containment system.
The goal is to determine the feasibility of a filtered-vented containment system, assuming the accident-history determined by the MARCH / CORRAL system of codes. Effectiveness and reliability of conceptual designs will be studied.
l l
f
The Staff has a program in piaca 1.u cuurm 6 = yue=6 vn of hydrogen control. The goal is to cetennine the behavior of hydrogen in contain-ment and practical methods of controling it through burning or burn suppression. A third mitigation feature being studied is a core re-The goal of this study is to determine the utility of tention system.
these devices vis-a-vis their contribution to lowering risk from both atmospheric and liquid pathways.
Dr. Okrent asked the Staff if there were recent studies on liquid pathways which were site specific. Dr. T. Speis of the NRC Staff oelieved a recent Sandia study on the liq'uid pathway had site specific aspects and agreed to send Dr. Okrent a copy of the draft report.
Mr. Meyer returned to comconent five of the Staff's program, which deals with structural response, and structural failure mode analysis. The goal of this effort is to determine realistic failure pressures and failure modes (Dynamic / Static)from Zion /IP containments and reactor vessels.
(NOTE: At this point, the TMI-2 Accident Action Plan Subcommittee joined the TMI-2 Accident Implications Subcommittee and together they heard a 26, 1980 Crystal. River-3 Transient. This presentation on the February portion of the meeting is recorded in the minutes of the March 5,1980 TMI-2 Accident Action Plan Meeting.)
The The sixth component of the Zion /IP program is e study of steam explosions.
purpose of this effort is to assess the potential and magnitude of a steam explosion (based on "BEST" current analytical and experimental information) and the impact of realistic " steam explosion" events on vessel and contain-ment failure. The final element in the program is to establish current thinking for coment and review on mitigative system design criteria.
fir. Meyer explained that the containment failure modes under consideration The are essentially the same as the modes considered in the WASH-1400 study.
six failure modes, using the WASH-1400 nomenclature are:
)
a mode - containment rupture due to steam explosion a mode - containment failure resulting from inadequate isolation y mode - containment failure due to overpressure (hydrogen burning) 6 mode - containment failure due to overpressure (non-condensibles and steam) c mode - containment failure due to melt-through
Y vent - low pressure injection system check valve fails resulting in direct path frcm primary system to atmosphere The filtered-vented containment system (FVCS) program is divided into three components. The input data that is necessary in order to do conceptual designs is the first component. The second component is the exploring of the conceptual designs themselves and the third component is the performing of consequence analysis using the CRAC Code.
Some of the key input parameters taken into account in considering filtered-vented containment systems included pressure, temperature, aerosol, and radiological source term histories from the MARCH / CORRAL and independent analyses. Variation of histories due to the presence of core retention devices, hydrogen control, accumulator water control, and r estoration of AC power are also considered.
Various histories are calculated assuming certain prominent accident se-There are three basic categories taken from WASH-1400. The first quences.
is TMLB' which assumes a loss of offsite and onsite AC power for an indefi-nite period.
In this sequence of events the primary system maintains its integrity until the core melt fails the lower vessel head. A second acci-dent scenario that is considered in generating the input for the ionceptual designs is a large LOCA with hydrogen burn. A third scenario is a small LOCA with felure of ECC inspection, again, anticipating a hydrogen burn.
The Staff is looking at conceptual designs for filtered-vented containment systems (FVCS). They are taking into consideration practical layouts, presence or lack of AC power; decontamination factors achievable; practical design flows; activation levels; operator / automatic controls; venting to atmosphere vs to special building; and environmental requirements (seismic, tornado,etc.).
In response to questions, Mr. Meyer explained that based on preliminary Sandia work a four foot diameter penetration is appropriate and can accom-date a certa.a spectrum of the pressure histories studied. However, in certain transients depending on how fast the pressure spike rises (for example the accumulator dumping its inventory on a molten core in the re-actor cavity) a 20 foot diar:eter vent would be required. The feasibility of such a vent is questionable from a design standpoint. Also being
-M
?
7-considered is the practicality of early venting to accomodate pressure spikes.
In response to a question from Dr. Okrent, Dr. T. Speis (NRC Staff) noted that in some recent studies, involving a core melt scenario, 70% of the cases involving containment failure occurred befcre molten core-concrete basemat interaction.
In 30% of the cases, interaction of the cSre melt and concrete baskmat, leads to the generation of additional noncendensibles which contribute to containment overpressure. Dr. Okrent observed that this might be a departure from the analysis contained in WASH-1400.
One concept being explored is venting to another building. This building would act as a heat sink enabling cool / dry air to be returned to the containment.
Consequence analysis will be done with the CRAC code.
It will analyze the impact of mitigation features with and without: noble-gas attenuation, and AC power to FVCS.
Another mitigating feature under consideration is a hydrogen control system.
Practical methods of hydrogen control include controlled burning and/or providing an environment which suppresses ignition. Also under study are A third hydrogen gas dynamics; its diffusion and mixing characteristics.
area of study is the actual hydrogen burning / explosion dynamics and the pressure time space evolution of a hydrogen burn or a hydrogen explosion.
A fourth important area of study is hydrogen detection and operator inter-vention capabilities. ' Another area of study invotes the potential for control in and damage to a FVCS from' hydrogen.
An investigation of the proposed benefits of a core retention device is a third area of study into,, accident mitigating features. The benefit of a core retention device in regards to the liquid pathway is being explored j
as a delaying or complete stopping device of a molten core penetrating the basemat. A core rehntion device also effects the atmospheric pathway by providing a reduction in: containment pressure, aerosol generation from molten core concrete interactions, and hydrogen gas generation (and other
combustibles) from molten metal constituents (steel) in contact with For. Zion and Indian Point in particular an assessment must concrete.
be made of the practicality / feasibility of installing a retention de-vice, including the negative aspects of installing such a device in an already built and operated reactor.
Considering the various mitigating features, the Staff will perform re-actor consequence evaluations. The purpose will be to evaluate and assess the consequence reduction achievable by various mitigation features.
The Staff is currently performing case calculatioqs for Zion and IP based on the Surry design, but taking into account site specific meteorology, and population distribution, plant specific power levels, release fractions of WASH-1400 PWR release category probabilities. The calculations will then be repeated postulating the addition of a FVCS. Several options for the treatment of noble gases, plus other potential mitigation features (e.g., hydrogen control) will be considered.
A third consequence evaluaticn will be made assuming containment basemat melt-through, and using migration times, estimate consequences via liquid or groundwater pathways with and without mitigation features (e.g. core ladle).
Mr. Meyer discussed design and quality requirements for Class 9 accident mitigation systems. This effort involves the setting of design and quality requirements for the mitigative system (s) whether they involve a filtered-vented system, a core retention system or some combination of these other systems. Staff requirements are preliminary and subject to change pending further licensee / Staff interactions and evolution of system designs. Con-servative design criteria applied to design of ESFs will be avoided, con-sidering the low probability of the events considered.
In general, the design approach should be reasonable and evaluated on a realistic basis
- where possible.
Group C quality standards as defined in Reg. Guide 1.26 The should be applied to mechanical and fluid systems where appropriate.
systems shall be designed and analyzed to remain functional for all of the operating basis environmental conditions, including the loads imposed
.by an operating basis earthquake.
For more severe conditions the plant should be shut down for inspection of the mitigative systems.
In adoition,
the mitigative systems should be evaluated for design basis loads, in-cluding the safe shutdown earthquake, in order to insure that there will be no gross failure which might impart or impair the functioning of safety class components / systems needed for design basis events.
Dr. Okrent observed that he was at loss to judge the adequacy of the preliminary requirements given without more explaination of t'v rationale used in deciding on the criteria. Mr. Meyer noted dat these criteria were presented only as a starting point for discussion and to Mr. Marchese of the Staff felt before final criteria gene: ate comments.
were set, cost-benefit studies should be done.
Mr. Meyer presented a tentative schedule for the Staff's task force work.
By the first week of April, a preliminary NRC FVLS design study to set pre-liminary design criteria should be complete.
In mid-July the NRC wi 1 update FVCS design criteria. By the first week in December, the Staff intends to complete its FVCS design study.
In parallel with these efforts, the Staff, together witl the utilities, is planning technical meetings on
" key" issues / areas relating to a FVCS and other technical areas related to severe accident mitigation.
The Currently, there are five meetings planned to address key issues.
objective of these meetings will be to obtain relevant information and expert opinion on a number of technical areas pertaining to designing, selecting, and evaluating the effectiveness of severe-accident mitigating features. The subject areas of the five meetings planned are:
1.
Accident Scenarios And Related Phenomenology-Evolving to Core Meltdown.
2.
Material Interactions.
3.
Hydrogen Dynamics,and Hydrogen Control Measures.
4.
Mitigating Features Filtered-Vented Containment and Core Re-tention Systems.
- 5. _ Structural Response to Dynamic / Static Loading.
l Answers A number of questions which highlight tha program were given.
are sought for the following questions:
Are the present analysis methods and their experimental /thec-retical basis sufficiently adequate to be used as a basis for the design of severe accident mitigation features?
What will be the role of probabilistic evaluations in choosing the set of accident scenarios that will form the design bases
~
for the severe accident mitigation features?
For that subset of accident scenarios that are amenable to immediate technical fixes, is there reasonable assurance that a significant reduction in risk can be achieved?
Finally, practical problems / considerations in designing and implementing a severe accident mitigation feature; design criteria, potential system interactions (i.
e., creation of new accident paths) must be addressed.
In conclusion, Dr. Speis offered the philosophy that the Staff believes they have reached a point of diminishing returns in significantly re-ducirg the probability of events outside of the current design basis.
If a general improvement in safety beyond that level is required, then the area of accident consequence mitigation seems to have the potential for the largest risk reduction.
PRESENTATION BY LICENSEES _- L. Peoples, Director of Nuclear Licensing for Commonwealth Edison Mr. Peoples presented the results of the Indian Point / Zion Near Site Study along with the program results to date. He noted that the owners of IP and Zion feel very strongly that these plants do not pose any greater risk This conclusion to the public than the average plant in this country.
takes into account the specific design features that were built into these plants, when they were first licensed, as well as their site location and meteorology.
Mr. Peoples noted that the Staff requested tne owners of IP 2&3 and Zion 1&2 (Consolidated Edison, the Power Authority of the State of New York, and
, Commonwealth Edison) initiate studiet in three areas:
1.
Means to mitigate the effects of a core melt.
- 2. -Means to reduce the probability of a core melt.
3.
Potential interim actions.
The owners were given 60 This meeting took place on December 5,1979.
The owners were told the general days to complete the st,udy program.
objective was to improve the time available for public evacuation given The results of the 60-day program were presented to the a core melt.
.NRC Staff on February 20, 1980.
The interim actions have been addressed via NRC confirmatory orders to Indian Point and Zion Stations. Mitigation of core melt consequences, received the bulk of the attention from the NRC Staff during the early meetings. Substantial emphasis was placed on the filtered vented contain-ment concept by the Staff. The owners were directed to consider this concept and others regardless of the probability of a core melt. The owners were also directed to investigate means to reduce the probability of such a core melt.
Mr. Peoples explained that Zion and IP were viewed by Mr. Denton as pre-senting substantially more risk than the WASH-1400 plant at its composite site. Mr. Denton informed the Comissioners that it was his goal to re-duce the risk posed by IP and Zion by a factor of about ten so as to bring these plants in line with an average plant. The owners contend, based on more detailed evaluations, that Zion and IP, as built, are lesser con-tributors to risk than the WASH-1400 average plant.
As part of the 60-day Neap Site Study, the owners conducted a mini-WASH-1400 study of IP and Zion taking into account major design differences between these plants and the reference plant. Actual site data for Zion and IP were used. Consequence models were used to conduct a very preliminary evaluation of ideas for the mitigation of risk (filtered vented containment) and reduc-tion in probability of a severe accident (training, testing of components, etc.)
s
-4
- l j
The owners of IP and Zion are actively engaged in a more detailed, longer This work will con-term, probabilistic risk assessment of these plants.
firm that the accident sequences selected represent the doninant contrib-utors and may identify additional areas where the probability of severe The firm of Pichard, Lowe, and Garrick has accidents could be reduced.
been hired as consultants to aid in this effort.
i The starting point for"the present Zion /IP work was the table of dominant PWR accident sequences from WASH-1400, the initial assumption being that the accident sequences which dominated risk in WASH-1400 would also domi-nate risk for IP ar.d Zion. During examination of the Zion and IP systems cause was found to both add and delete from this set of accident sequences.
Risk characteristics for the reactors at the Zion and IP sites were cal-culated in a preliminary fashion using the probabilities of each release code developed for the Reactor Safety Study (RSS).
category and the CPAr The core inventories t,f fission products used in the RSS were adjusted proportionately for the respective power outputs of the. Zion and IP re-Actual demographic data and site meteorology for the Zion and actors.
The evacuation model employed in the RSS was used.
IP sites were used.
A rore detailed quantitative WASH-1400 type evaluation of the Zion and IP This detailed study will plan is part of the longer term follow-on studies.
accident sequences and indicate develop refined risk estimates for dominar.
whether sequences with substantial contributions to risk have been dmitted However, the short-term study is regarded as an in the short-term study.
adequate basis for study of major alternatives.
, Program Results To Date Base Plant Probabilities and Risks One of the conclusions reached in this study was for a core melt accident.
Containment failure by melt through may be postulated to occur if the con-Given that water is tainment has not failed earlier by some other mode.
likely to be present underneath the reactor vessel at melt through and that reflux cf water to this region after vessel melt-through will con-Basemat melt-through is tinue, basemat melt-thr'bugh is highly unlikely.
assigned the residual probability remaining after other failure mode probabilities are subtracted from 1 for TMLB'.
Assignment to Release Categories _
For each combination of accident sequence and containment failure mode, there i
n will result a particular quantity of fission products released to the env ro -
The total spectrum of containment releases was divided into seven ment.
discrete categories for core melt events, this followed the pattern from Types of sequences assigned each category can be summarized WASH-1400.
steam explosion failures without containment cafety features as follows:
functioning were assigned to release Category 1 and steam explosion failures Containment with safety features functioning were assigned to Category 3.
overpressure failures without safety features functioning were assigned to Category 2 and overpressure failures with safety features functioning were Containment failure via basemat melt-through assigned to Category 5.
wero assigned to Category 6 if containment safety features for fission product removal were not functioning and to Category 7 if they were Release Category 4 was not utilized in this study because functioning.
none of the sequences evaluate for Zion and IP resulted in release within Category 4.
4 o-Accident ~ sequence and containment failure mode combinations were assigned to appropriate release categories and total probabilities for each release The result was that accident sequences category were obtained by summing.
in fission product release Categories 2 and 5 exhibited the highest prob-abilities of occurrence.
Mr. Peoples argued that, Zion and IP represent risks consistent with the This is in part he said the result of care taken in the initial industry.
A few of the specific differences.noted from the design of these plants.
at Zion reference plant include:
the inclusion of containment fan coolers and IP; the third diesel-driven containment spray pump and Zion; additional diesel generators at Zion and IP; more extensive use of power-operated I
This valves rather, than manual valves on Zion and IP safety systems.
also includes the use of confirmatory ESF signals to appropriate valves.
The use of a third series check valve at Zion on the RHR cold leg in-jection lines; and the use, at Zion and IP, of three types of ECCS pumps These in each train as opposed to two types in the reference plant.
examples were given as the kind of differences which affect the risk curves.
Mr. Peoples again stressed the fact that the important conclusion from the risk calculations is that major contributors to risk are Category 2 and Category 5 releases. The principal failure modes for these release Risk through categories are the containment overpressure failures.
these failure modes are potentially amenable to reduction by containment failures which reduce the probability of or eliminate these failure 1
modes.
For overpressure due to noncondensible gases or excessive steam generation, containment vent systems with filter capabilities for fission product re-moval or retention may be one of several potentially useful concepts.
l
. Steam Explosion A mechanism Mr. Peoples described steam explosions within the vessel.
for vessel failure was described as a steam explosion in the lower plenum of the reactor vessel which accelerated a continuous overlying liquid It layer in a piston-like manner until it impacted on the vessel head.
is then postulated that the vessel head fails and is propelled against the containment wall with enough energy to cause failure of the contain-He then noted that available experiments show that a pressure of ment.
150 psi is sufficient to eliminate explosions. Consequent'ly, when the primary systerr pressure is above this value, steam explosions will not occur.
Mr. Peoples noted that for some postulated accident sequences, such as a large break LOCA, the system pressure can be less than 150 psi at the time core melt conditions are hypothesized.
In this case, the molten core f
- However, material could fall into the water in the lower vessel plenum.
the fuel melt and fragmentation process itself is responsible for ensuring that a continuous overlying liquid layer, which is required to fail the containment via missile generation does not exist.
It was also noted that the behavior of a steam explosion within the vessel at low system pressures would resemble that of a shallow unde'rwater ex-If it is postulated that a continuous overlying layer exists, plosion.
even though it cannot, the radius of a subsurface explosion would quickly approach the depth of the mixture and the bubble would break through in its The phenomena to be observed would be a hollow splash.
first expansion.
This would definitely not lead to the slug type of impact used for contain-ment failure analysis.
. For conditions in which molten fuel is assumed to melt through and be discharged from the reactor vessel, the core material would come into contact with water in the reactor cavity at pressures where explosions The major issue to address in the case of an ex-vessel could occur.
steam explosion is the shock from the explosion itself.
Typical maximum interaction pressures from steam explosions were given to several hundred psi, but even using a conservative value of half the (1600 psi) at the cavity radius, the expansion from critical pressure the reactor cavity to the containment walls reduces the shock wave over-presture to about I atm. Such a shock wave does not pose a threat to the containment integrity.
Considerations of the spectrum of conditions representing possible core meltdown scenarios show that:
(1) For the in-vessel case, vapor explosions are eliminated when the system pressure exceeds 150 psia.
(2) For the low-pressure in-vessel case, the continuous overlying liquid layer required to fail the containment via missile generation is precluded for all reason-able fragmentation levels.
(Fragmentation levels would have to be more than 2 meters in dianeter in order to nullify this conclusion.)
(3) For the ex-vessel case, the shock waves resulting from the s+.eam explosion are below the containment design pressures when they reach the containment walls.
Core Coolability The coolability of the core in an unconfigured geometry is investigated by finding the minimum amounts of water needed to maintain coolability.
In order for a core to become badly damaged, water must not only be lost from the primary systa but it must also be kept out of the core for an In this accident scenario, fuel heatup and melting extended period of time.
would begin near the top of the core, eventually leading to the collapse of A system core material upon melting and refreezing at lower core elevations.
such as this could become completely blocked at the bottom but would still Thermal conduction analysis of a sphere have water accessible from the top.
with uniform internal heat. generation at decay heat equal.to 1". full power
. and the outer surface in contact with water show that particles of up to 10 inches in diameter are permanently coolable. Such characteristic sizes would present a coolable configuration if water could be supplied to the core region.
For in-vessel coolability, Mr. Peoples discussed a damaged core configuration which is completely bigcked at the bottom. This is the most conservative case since any leakage path through the bottom would greatly enhance the coolability of the core and would permit water from e cold leg injection to be available for heat removal.
In the event of a complete blockage, the damaged core must be cooled by water supplied from above the core such as hot leg injection or leakage around the outlet nezzles from the downcomer.
Two phenomena must be concidered in evaluating the coolability of a damaged j
first, the ability of the water to contact the top of the core and core; The second, the penetration of the water down through the distorted core.
particular plants in question have hot leg 17jection, which allows water to be added directly to the upper internal region. In addition, a reflux heat removal path can be established and can be accomplished with a small
- fraction of the heat tranfer area in one steam generator which is operable in the presence of noncondensible gas such as hydrc;en. In summary, con-siderations of a badiy damaged core, which is assumed to be completely blocked at the inlet, show that the core is coolable if water can be sup-plied to the upper surface.
To begin the ex-vessel cooling scenario, the complete absence of primary cooling water must be assumed. The core material would averheat due to the decay power and failure of the reactor vessel could not be ruled out.
If such a failure occurred, and its coolability in this region is determined Conclusions reached by the availability of water to the reactor cavity.
were that the cooling water that must be lost from the primary system to resul't in core damage is sufficient to establish a heat transport path th Considerations of the ex-vessel debris bed show the remove thc decay power.
bed will indeed be coolable. The onwers of IP and Zion have, as a result, concluded that a core ladie is not required and have dropped this concept from furtner consideration. The Staff noted that they had heard the same
'"F y
m-,
. However, the Staff has not presentation from the owners of IP and Zion.
reached any conclusions regarding their study.
Speis of the Staff noted that he though it was premature to reach a Dr.
conclusion that a core ladie was not required. Dr. Speis noteo that there were questions remaining in the field of coolability of debris beds such as how far the core material spreads and what size the fragments will (The owners claim that a 10-inch size or smaller fragment ultimately take.
represents a coolable configuration for a debris bed.
Dr. Okrent suggested the seismic design basis for these plants might be reviewed to determined the probability per year with some uncertainty that is defensable, that a larger earthquake might occur which could change preliminary conclusions. A more quantitative evaluation of the seismic contribution to risk for IP and Zion might prove such considerations to be a significant contributor to risk. Such a conclusions might affect decisions that have been drawn regarding the possible usefulness of some Dr. Okrent noted a concern about abandoning kind of core retention device.
additional engineered safety features at this stage. He though such a conclusion before more information is developed to be premature.
Transient Containment Analyses Containment transients were evaluated for core melt sequences similar to Those sequences those found limiting in the short term WASH-1400 study.
ECCS, and were a large break with no ECCS, a small break LOCA with no The mass and energy releases a loss of all AC power with loss of heat sink.
This data to the containment were obtained for each of these sequences.
was used in scoping calculations for the containment response.
Each plant with its individual heat sinks, containment volume, and con-tainment safeguards systems were modelled. Among the phenomena studied hydrogen burning, containment venting, and continuous core melt were:
These parametric studies indicated that the possibility of cooling.
combusti-I. gas burning was one of the key parameters which affected containment response.
l
. Preliminary results from the studies showed:
(1) LOCA melt sequences with minimum containment safeguards and no hydrogers burn are acceptable in terms of containment capability.
In fart they remain within contain-ment design pressure. (2) LOCA melt sequences with minimum containment burn go slightly above containment design safeguards and continuous H2 (3) LOCA melt se-pressure but do not exceed containment capability.
burn when satisfying the quences with containment safeguards and H 2
" flame temperature criteria" (if the calculated flame temperature of the containment mixture exceeds this criteria, then the containment atmosphere was flammable and could burn if a spark or ignition source The 710 C criteria corresponds to 81/2". hydrogen in 0
were present.
room temperature dry air at which spherical flame propagation commences),
(4) TMLB' sequences can exceed containment capability in some cases.
(loss of all AC power with loss of heat sink) and either continuous burn when the " flame temperature criteria" is exceeded H burn or H2 2will exceed containment capability. Calculations with a continuous steam may not burn. However, generation in containment indicate that the H2 the containment design pressure can be exceeded from the steam generation alone.
The vented containment study performed by Battelle (NUREG-CR-0138) has been reviewed and used by the owners as a guide to examine containment f
Calculations performed by the owners used as a reference, two 12-venting.
The calcu-inch vents with a realistic back gressure and line resistance.
lations indicate the vent would help some sequences, particularly T!!LB' and the LOCA sequences with minimum safeguards.5ich had continuous steaming.
The scoping calculations indicated that if a very rapid and large energy additions occurred in the containment, this size vent could not keep the Mr. Peoples stressed, however, pressure below the containment capability.
that these were preliminary results and are subject to confirmation or changes as investigations continue.
. 20 -
Directions of Conceptual Design The effects associated with steam explosion phenomena as analyzed earlier, were of minimal significance and readily accommodated by presently available structures and components in the plant.
~
The investigation of combustible gases and their potential effects on con-tainments suggested consideration of measures which include the contain-Several other approaches have also been studied, each ment vent concept.
of which deals with either consuming the gases in a controlled fashion of adjust the vapor mix constituents in containment such that combustion C3ntinued work in this area will be conducted to insure is prchibited.
that a complete program evaluation is performed.
The potential' for core debris melt-through and resulting containment failure was dismissed based on the technology review and analysis dis-cussed previously.
The first Two methods for a filtered containment vent were discussed.
candidate uses a sparger condenser and water scrubber. This system provides the means to vent the containment in a controlled manner and prevent containment failure by overpressurization. The system consists of two pressure operated isolation valves, a vent line with an orifice, and a sparger tank. Gases enter the vent system when the containment An pressure reaches 60 psia and the containment isolation valves open.
When orifice limits initial flow through the valves and the vent line.
containment pressure drops, the containment isolation valves close.
Tne vent line discharges into a manifold which is located in the bottom of
. Gases exit the manifold and pass up through the water the sparger tank.
The sparger tank has the capacity to absorb a total where heat is removed.
Water saturated gas is vented from the sparger tank to of 2 billion BTU.
0 0
The process is the atmosphere at temperatures ranging from 50 F to 180 F.
estimated, on a preliminary basis, to provide a decontamination factor s
of 100 for particulates and a decontamination factur of 10 for meiecular iodine.
. A tank with a usable volume of 1,875,000 gallons is required. Such a tank would be approximately 80 feet in diameter by 50 feet high.
The second candidate is a filtered containment vent with a submerged The gravel filter.
It is similar in concept to the first candidate.
size of the submerged gravel scrubber is 130 feet square by 25 feet To fill the gravel bed, almost 25,000 tons of gravel is required.
deep.
Mr. Peoples noted that the vent system examples described,.do not offer It is important a clear, well-defined reduction in risk at this time.
to take.into account the as yet unquantified reduction in overall safety Such reductions in safety could which may be caused by such systems.
result from the possibility of failures, inadvertent activation during The lesser transients or accidents and possible system interactions.
work of the owners in evaluating these concerns and balancing them against possible gains in safety is far from complete.
Future Action Mr. Peoples noted that continued action in this program consists of main-The owners plan to complete taining two parallel, closely coupled efforts.
a long tern probabilistic risk assessment program currently underway with consultants, Pickard, Lowe, and Garrick. Secondly, it is planned to proceed with further investigations of the core melt related technology and of potential mitigation concepts.
The owners program with Pickard, Lowe, and Garrick should be complete All event and fault trees will be initially completed f
during late summer.
This program uses plant specific design information to develop by spring.
The study will also use plant specific reliability data for these tools.
Site sequence evaluation where such data is available and is meaningful.
~
specific data on demography and meteorology is also used for both sites along with an improved CRAC program for modeling consequences.
- 22 The A second phase of the continuing action concerns mitigation concepts.
owners plan limited further work in the areas of containment transient analyses and containment ultimate strength, and in the area of combustible gas evolution, behavior and combustion. This work combined with the results of the probabilistic assessment program serves as the basis for further discussions with NRC StaYf.
The second portion of the continuing action on the part of the owners con-cerns concepts. Limited further work in the areas of containment tran-sient analyses and containment ultimate strength, and in the area of com-bustiole gas evolution and behavior is planned. That work, combined with the results of the probabilistic risk assessment program will serve as the basis for further discussions with the NRC Staff.
Following these studies, if necessary, the owners will begin to prioritize alternative design concepts for mitigation features.
M.. Peoples stressed, however, that the owners still need a clear unambiguous safety goal for these plants.
Such a goal wauld have to come from the Staff.
The owners believe that the definition of uniform, quantitative criteria and methodology for evaluating Class 9 events and evaluating potential features to reduce the probability of occurrence or mitigate the conse-The owners quences of such events is necessary for further progress.
believe that the risk assessment methodology, applied reasonably, is the proper tool. They believe that a criteria based on WASH-1400 average f
risk is appropriate to use for decision making at this time. NRC concur-1 rence with both of these approaches has been requested.
In concluding, Mr. Peoples noted that the results of their current assessment f
~
of the relative risks from Zion and IP show that these plants do not con-tribute excessively to public risk. These results coupled with the Interim Actions ordered by the NRC and coupled with the actions taken in response to NUREG-0578 give the owners a great deal of assurance that the mutual
]
interest in excellence has been and is still being satisfied at these plants.
j The owners stressed there are still areas where substantial work remains to be dora before any final selections of mitigating features are possible.
I
. Mr. Peoples suggested that it was possible that studies would be completed in June of 1981. This is the present target date of the owners of IP and Zion.
CYRSTAL RIVER IREP PRELIMINARY INSITES ON CRYSTAL RIVER 3 - J. Murphy, PAS Staff Mr. Murphy reported on the present progress on the Crystal River IREP study.
Currently, the Staff has.,gone through the event trees, fault trees on all the systems the Staff deemed to be important. The Staff is trying to conduct the IREP. study in the same way WASH-1400 was done -- with the minimal use of computers. The amount of hand calculations are increasing the amount of time necessary for this project. The Staff noted that they were finding quite a bit of coupling between supporting systems for this particular plant.
In performing the IREP study on Crystal River, the PAS Staff assumed that the things Crystal River has committed to have been performed and have been performed correctly. The Staff took the plant as they found it during s
the first week of December plus what the plant and committed to change at that time. That was the basis of the PAS analysis.
Mr. Murphy stressed that the IREP study was never intended to be complete and the PAS Staff is making no completeness argument. What was hoped to be accomplished in the IREP study was to find obvious problems and go This means through Crystal River and other plants in a rapid fashion.
a thorough job would be impossible. To put it in perspective, Mr. Murphy noted that approximately 5 man-years of effort was focussed on Crystal River.
Probably less than that amount will be expended on the rest of the IREP plants, maybe by a factor of 5.
By contrast, he noted the Reactor Safety Study cons'umed 60 man-years of effort.
Mr. Murphy noted the preliminary nature of the IREP results. There are a whole class of accidents which have not been considered yet, the best example of those are-the ICS related initiators like the February 26th event at Crystal River. This transient itself is on the event trees (the loss of feedwater with an opened PORV). However, the Staff has not gone into the depth of detail necessary to analyze this event.
i
\\.
Mr. Murphy indicated that he recognize that there are a whole family of These other accidents accidents that are not included in this study.
He did include those related to non-nuclear instrumentation and the ICS.
note that the PAS would take a quick look at those, lir. Murphy noted that the IREP is still being organized with regard to what plant would He noted that the Staff thinks it will take about be studied first.
26 weeks to study a particular plant. Currently, the Staff believes that three plants will be done locally in Bethesda and three plants At this stage, PAS is still trying to develop will be contracted out.
It is hoped to be able to standardized techniques for the program.
put together relatively short manuals that describe the type of analysis In this manner, all six or seven groups working PAS is looking for.
They on the various plants will have generally the same basic format.
would draw their fault trees in the same basic style. Mr. Murphy concluded by saying that the main value he sees coming out of the IREP The final output is a qualitative one rather than a quantitative one.
should be a listing of what effectively says "if you want to buy the greatest risk reduction for your buck, this is what you should attack."
For additional details, a complete transcript of the meeting is NOTE:
available.in the NRC Public Document Room,1717 H St., NW, Washington, DC 20555 or from International Verbatim Reporters, Inc., 422 South 20002,(202)484-3550.
Capitol Street, SW, Suite 107, Washington, DC
Foeral Regist:r / V 1.'45; W. 34 / Tuesday, Mbru,ry 19, 1980 1 noncms mo.,
8 e
vee cs fai Advisory Comndttee onineoctor
)
(:)!)ghting fixtume on the side of the the Design:tedFederalEmy #ct in advance as practicabl;r so u Safeguarda Ad Hoc Subcommittee est equipment would " blind" the opersfor appropriate arrangements can be made Three Mae ladend. Unit 2 Accident and ne:rby miners or requi:w conetant to allow the necessary thee daring the Action Plan; Meeting adjustment to changes is !!!umination:
meeting for such statements. ' shall he hm hGe Island.UndtI Amident De ACRS Ad Hoc hhr==ittee en Ratures would be sheared off ce broken he agenda for subject t
Incro.cing the likelihood of snore serious be as follows: Wednesday.
5.
Action Plan willhold a meeting se
- equipm:nt failure, wedging. janumns or 19e0 a:30 a.m. until the conclusion of M 1,19e0 in Roosa 1187.1717 H St.
upset. Also, as lighting fixtures on the business.
NW. Washington.DCansts tm enemider side or top are sheared off. roof bolts, The Subcommittee may meetin Draft 3 of the NRC"Acham Plane far...
seoss beams and strape wu! be sheand Executive Session.with any ofits Implemntag Recousmandattens of she.
eE thereby damaging er destroying roof consultants who mey bepresent. to..
President's f>=i===rm and Otbar....
exploremad exchange their prelsamary i
oP~ intone regarding1 natters winch should StadaestdtheDreeWelsiempelfJ p,,
suhport.)instauntion of stadonary Ughting
.n._.: ;.:w w Anddent.r,:..;.M~ he procedures
. '~e'quipraent would similarly impair the be considend durmg the meeting.... n In accordan:.a with t 3 eperst:rs' and nearby miners' vision. It -
. At the conclusion of the Executive entlined in theFederalRagineeran.
? would c!so create additionalbestin the'
' meAnIregly small aress in which the -;
Senalos. the tab===dttee wu! hear
, October 1.1979 (44 FR 5eens). oral or i,
. Presentations by and hold esenssions writtan statcments may be,. _
- by~
N
.~ -
' min:rs mustwork.
with representatives of the NRC Stag members of the public. recordings wit!
7 r* q.For these reasons,the petitioner f
' York. the Consolidated E& san Co. of be permitted only dunas those portions.
3 the Power Anthority of the State of New e dests a modi 5 cation of the ~-m~
4 of the meeting whw4 transcript la being
"~tpplication of the standard toit.i.adoe.-
Now York. lac b Commonwealth.
kept, and questions may be asked only -
7.
n for h=====8=
Edison Co their consultants, and other
.by members of the Subcomsdttes,its
_. c
- Raquest Persons interested in this petithm,ney intenstedpersons.
consultants, and Staff. Persons dseiring In ad& tion.It maybe necessary for to make oral statements should motify Surnish written comments on or before the Subcommittee to holdone or more the Designated Federal Employee as far March 20.1980. Comments must be filed closed sessions for the purpose of in advance as practicable so that with the Office of Standards.
exploring matters involving proprietary appropnate arrangements canbemsde Regulations and Varjances. Mine Safety Information.I have determined. In
'and Health Administration. Room 627 accordance with Subsection 20(d) of the to allow the necessary1hne dareg the "J.615 Wilson Boulmed. Arlington. ~-
Federal Addsory Committee Act (Pub.
meeting for such statements.
Virginia 22203. Copies of the petition are The agenda for subject meeting shah cv:ilable for inspection at that address.
- 1.92-483), that, should such sessions be Be as follows: Wednesday. March S.
required. It is necessary to close these 1980. 8:30 km. antil the concluaica of
. Dated February it teso. -
sessions to protect proprietary
'- Q A, w yg.,
information. See 5 U.S.C.552b(c)(4).
business.
Further information regarding toples The Subcommittee may meet le
' g g g -_. J.
..\\'.
lo be escussed. whether the meetms Executive Seasmwith any ofits
- og y,,,
,,, b
., J.,,, has been canceHed or renEhadalad. the canaultants whoany be prenant.to Chairman a ruling on requests for b explom and - r hange their prehminary
.; auses oceas m opportunity to present oral statements opinions regardag mattars wiuch should and thd time auctted therefor can be be considered dunng the===hreg leuct. EAR REGULATORY
- obtatned by a prepaid telephone cab to At the "haf the Eascative COMESSION the cognizant DesignatedFederal Session, the Shdttee will hear Employee.Mr. Richard K. Major presentations by and held discussions Adytaory Committee on Reactor (telephone 202/534-1414) between 8:15 with representatives of the NRC Staff.
Saf;guarcis, Subcommittee on Three a.m. and 5:00 p.tc. EST.
the nuclear industry,various utilities.
kile latand, Urdt 2 Accid 6nt Background informat on concerning and hir cxmenitanta.and o&ar dmplications; Meeting hams to k &,scussed at this mting laterested persons.
The ACRS Subcommittee on the Three *** be found in documents on Sie and in addition.It may be necessary for
- Mil) Island. Unit 2 Accident pu e ns a at the Subcommittee to hoidaneer name..
Impileations wdl hold a meeting on -
cD qn closed sessions forthepurposeof.
March 5,1980 in Room 1046.Tn7 H St. - Stmet. N.W.Wa:W n. E 2c555and axploring mat'._rs involving propsietary information.Ihave determined.in
.NW Washlegion.DC 20555 to consider at &&aznment Publications Secuan b potentialinstallation of moltan core State 1 Aran of Pennsylvania * ~
accordmee with Subsectiomdbfee.
/ seucibles mder the Indien Point 2 & 3 -
EducationBuilding Commonwealth and Federal Advisory Comrnittee Act(Pub.
- 1. 82
&at.s on such mions be
'.and tM Zion 1 & 2 Reectors. Notice of this meeting was published January 22.
h*8 d
fis required. it is necessary to close these.
~~
White 8
Plains public1ibra*y.100 Maritime J sessions to protect proprietary
- 1980, In accorcance with the procedures A nnue.
te Plains.New York 10001*
information. See 3 U.S.C.552b(c)(4).
i Further information regarding topics outlined in the Federalltesister on I '
October 1.1979. (44 FR 56406). oral or B
ry.
Em to be discussed. whether the meeting wntten statements e ay be presented by As unne. Zion.11. 60099 (regarding Zion).
het been cancelled or rescheduled. the Cha> man's ruling on requests for the msmbers of the pubhc. recordings wt!!
Deted. February 13.1seo-opper; unity to present oral stataments be permitted only during those portions 1
of the meeting when a transcript is being John C.He3 e.
and the time allotted therefor can be h"t. and questions may be asked only Murcry Committre Monopreens Offser. -
obtained by a prepaid telephone cal! to l
by members of the Subcommittee,its in o= is.m r m -is se t the cognizant Designated Federal consultants and Staff. Persons destrm; se p c coot rss a Employee Mr.lohnC.McKinley i
13 make oral staternents shouhl notify kkbEd n
l
]D p ]d D " " 'D
- b...,9,,u,,u g.
._. & W
ACRS AD HOC SUBCOMMITTEE MEETING ON TMI-2 ACCIDENT IMPLICATIJNS WASHINGTON, DC MARCH 5,1980 ATTEDEE LIST E
NRC STAFF ACRS R. DISalvo D. Okrent, Chairman M.'Medeiros, Jr.
M. Carbon J. Long S. Lawroski S. Acharya W. Mathis L. Olshan J. Ray E. Reeves R. Major, Designated Federal Employee A. Schwencer A. Marchese L. Soffer CONSOLIDATED EDISON T. Speis J. Meyer W. Bennett J. Olshinski M. Scott G. Zech POWER AUTHORITY OF THE STATE OF NEW YCRK COMMONWEALTH EDISON CO._
J. Davis G. Klopp J. Schmieder W. Naughton W. Sayed J. Deress C. Pratt D. Peoples J. Bayne J. Marianyi R. Goyette WESTINGHOUSE ELECTRIC CORP.
EPRI - NSAC R. Slember R. Leyse D. Goeser-E. Zebroski L. Hochreiter N. Liparulo D. Paddleford MITSUBISHI S. Jacobs J. O'Cilka K. Okabe
'R. Marchese A. Hoizumi j
H. Keller W. Kortier NEW YORK STATE ENERGY OFFICE BURNS & ROE J. Dunkleberger J. Rubin t
e.
TMI-2 Accident-Implications 3/5/80 -
ATTENDEE LIST (CONT'D):
SARGENT & LUNDY DEPARTMENT OF ENERGY N. Weber.
J. Yerick J. LaVallee TENNESSEE VALLEY AUTHORITY TAEC T. Price
- 0. Akalin SNUPPS KMC F. Schwoerer R. Boyd 0FFSHORE POWER SYSTEMS ARGONNE NATIONAL LAB.
R. Walker B. Spencer R. Henry D. Cho BECHTEL CORP.
R. McDermott ATOMIC-INDUSTRIAL FORUM N. Willoughby R. Szalay
,ISHAM LINCOLN & BEALE IVF _
P. Steptoe A. Young-S. Glocker
UNITED STATES f
h NUCLEAR REGULATORY COMMISSION
'/
- j ADVISORY COMMITTEE ON REACTOR SAFEGUARDS l' '
. y.
r W ASHINGTON, D. C. 20555 k
v.
/
s.,o....j February 29, 1980 TENTATIVE DETAILED SCHEDULE - SUBCOMMITTEE ON THREE MIL ACCIDENT IMPLICATIONS, ROOM 1046, 1717 MARCH 5, 1980 Times are appro ximate Executive Session (0 pen) 8:30 a.m.
Review schedule and add or delete topics Meeting With NRC Staff (D.Eisenhut et al.)
8:45 a.m.
Sumary and discussion of additional engineered s Meeting With Utility Representatives (L. Peoples et al.)
10:00 a.m.
1.
Introduction History Summary 2.
Diset : ion Study Objectives Utility Study Program Utility Methodology Utility Program Results Future Plans 3.
Concl usion 12:30 p.m.
LUNCH Meeting With NRC Staff - Discussion of other possibl 1:30 p.m.
caused, e.tc.) such as the Rancho Seco occurrence of Jenuary 5, accident 1979.
5:00 p.m.
Adjourn
/
J. C. McKin y, Chief
.I Project Review 8ranc m e,nL c
M