ML19309B556
| ML19309B556 | |
| Person / Time | |
|---|---|
| Issue date: | 01/09/1980 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | Advisory Committee on Reactor Safeguards |
| Shared Package | |
| ML19309B552 | List: |
| References | |
| ACRS-1680, NUDOCS 8004040275 | |
| Download: ML19309B556 (23) | |
Text
..s (Y
@\\ @ DC.]. --- i 'f.fc.)
DATE ISSUED: 1/9/80
! '". 8.gi if [{ l} h ', ] h N
- 7.1 y-
.t g
3.1 ni MINtTfES OF ' ale r.,
.i j
.,.i... n
.V d a a d2 e;. !Z2
- f.1 4
~74RS SUBCCMMITTEE MEETING GN
///6/$
CAERGENCY CCRE CCCLING SYSTDd.S OCTCBER 17-18, 1979
'4.SHINGTON, D.C.
W e ACRS Subcommittee on ECCS held a meeting on October 17-18, 1979, in Room 1046, 1717 H St., N.W., Washington, D.C.
H e purpose of the meeting was to review the methods and results of small-break ECCS calculations for W, CE and B&W reactors and the effects of running the reactor coolant pumps on ECCS performance. 'Ihe notice of the meeting appeared in the Federal Register on Tuesday, October 2,1979. A copy of the notice is included as Attachment A.
A list of meeting attendees and a meeting schedule are in-cluded as Attachments B and C.
No written statements or requests for time to make oral statements were received from members of the public.
Dr. Plesset, Subcommittee Chairman, opened the meeting at 8:30 a.m. and indicated that it was being conducted in accordance with the Federal Advisory Committee Act and the Government in the Sunshine Act. Dr. A. Bates was the Designated Federal Employee for the meeting.
Meeting with the NRC Staff Dr. S. Fabic, NRC Research, reviewed a number of topics related to small-break calculations. Critical flow through small-breaks is dependent upon break geo-metry and phase separation at the break may have an influence. Fedeling these effects is necessary to calculate adequately reactor system depressurization I
rates and mass and enthalpy fluxes at the break. Choked flow through small-breaks is dependent upon L/D ratios of the break.
For short tubes (L<D) entrance effects (sharp vs smooth edges), thermal non-equilibrium, and multidimensional effects are important. Various flow coefficients must be used to account for these effects een using Henry-Fauske, Burnell, or HD4 discharge models in the subcooled, saturated, and vapor flow regions.
For short pipes (2 < L/D < 40) lumped parameter analysis can give acceptable results; however, upsteam condi-tions are often not adequately calculated which results in a variety of discharge coefficients that have been chosen to obtain agreement with test data.
nnta s*YE'MtFPT(W th) 5 f0
s ECCS Mtg October 17-18, 1979 Flow through orifices will depend upon whether the downsteam region is confined or not and on the upsteam conditions of the fluid (saturated, subcooled or superheated).
Relief valves need to be tested to observe the type of flow under various conditions. W e effects of non-condensible gases on 2-phase flow have not been studied.
A possible way to approximate critical flow through cracks in pipes may be to use an array of thick wall tubes that have the same area and perimeter as the crack of interest. Bis would allow equivalent wall fraction and heat transfer. te L/D ratio for each of the tubes in the array would also be larger then for an equivalent single hole.
Verification of code calculational ability will make use of tests to be conducted in a' number of expected facilities. Rese include Moby Dick and Super Moby Dick, BNL and IRB nozzle tests, and Marviken tests.
Another area of interest related to small-breaks in the effect of phase separation at flow branches. Tests at RPI in Y's and T's have shown that the vapor phase will turn the corner at the junction while the ~ liquid phase tends to continue in the same direction as it was originally traveling.
nis phase separation may produce enhanced gas phase flow in the Pot legs and could tend to increase the gas phase flow at the break location. Eis effect may cause over-prediction of mass fluxes at the break location. Addi-tional tests are to be done a RPI to characterize and describe the phenomena better. Methods to calculate the fluid behavior are under development.
Steam generator performance will also effect the behavior of the primary system under small-break conditons. U-tube and once through steam generator tests have been or will be conducted. We FLECHT-SEASET program includes a series of steam generator tests. Various secondary side conditions will be simulated and the heat transfer to and from the primary side under varying conditions will be evaluated. Temperature and heat transfer coefficients will be measured alorg with study of the effects of countercurrent flow in the U-tubes and possible instabilities in flow. Tests with non-condensable gases will also be run. MIT has also proposed tests with a transparent system in order to get good' flow visualization.
ECCS Mtg October 17-18, 1979 Meetina With Westinchouse on Small-3reak ECCS Models Westinghouse presented tne results of the studies they performed on small reactor coolant system breaks follewi.g the TMI-2 accident. Se studies were performed in resconse to the NRC I&E bulletins 79-06,79-06A, and 79-06C. Se majority of the work performed was submitted to the 50 in WCAP-9600 with additional calculations on RCP behavior in WCAP-9584 and some steam generator behavior calculations in WCAP-9586.
One of the first results of mI-2 accident was an instruction to the W owners to eliminate the necessity of low pressurizer level in order to get safety injection with low pressure. Calculations showed that for pressurizer vapor space breaks (3 PORV's) no core uncovery would occur with minimum safeguards:
operators had 30 minutes to initiate ECC flow; the system response was sensitive to auxiliary feedwater flow; and the pressurizer level would increase for PCRV " breaks".
I&E bulletin 79-06A directed that one or more RCP's be left running during the transient. W procedures, as had been previous evaluated, specified that the pumps should be tripped if the pressure decreased to 1250 PSIA (plus instrument uncertainties). Additional analysis were conducted and meetings were held with the NRC. At the end of May, the W owner group was formed. About 40 different analyses were performed by W.
We results indicated that the W design is safe for small break LOCA's. W e effects of non-condensable gases coming out of solution in the-primary water was examined and it was decided that the volume of gas.would not be a problem with respect to continued core cooling. On UHI plants the UHI water would be injected on breaks of about 2" in size or larger. We passive accumulator tanks start to inject at about 600 PSIA and nitrogen would not be injected until the system reaches about 200 PSIA.
A six inch break would be required to get the system pressure dcwn to pressures where accumulator nitrogen would inject.
'Jr. Catton suggested that a heat flux map of the reactor system be provided to show heat inputs and outputs with time for snall break events.
l Westinghouse personnel indicated that they have always indicated to their l
operating plant that the reactor coolant pumps should be tripped following l
s, ECCS Mtg October 17-18, 1979 Safety Injection initiation or when the RCS pressure drops below 1250-1300 PSIA.
"he Westinghouse study on small breaks showed that with continued pump Operation there were some cold leg breaks in the range of 2" - 4" in diameter that led to core temperatures abcVe 2200 F.
We continued p=p operation leads to increased mass fluxes at the break and increased core tncovery. We period of high temperatures will occur only if the pumps are tripped within a time window, earlier or later tripping of the pumps is acceptable. The W calculations are reported in NCAP-9584. All of the W calculations were done using E4 methods and assumptions.
In response to questions, W indicated that 1/3 scale pump tests in France had bwn operated under 2-phase and under all steam conditions without damage, however full scale pumps had not operated or been tested under 2-phase flow conditions. Two-phase slug flow is an area of concern with regard to the structural capabilities of the pumps, slug flow was not tested in the 1/3 scale tests. 7 tie W models are based upr.. the experimental pump data from France for the pumps and a combination of homogeneous and heterogenous control volumes that are selected according to the volume location and the postulated flow. Countercurrent flow is not allcwed in the hot legs or steam generators and non-condensable gases are not modeled.
The break flow uses a homogenous model. Some recent work, reported in i
. WCAP-9586, has been done on countercurrent ficw in steam generators; it shows signifient draining of condensed fluid on the hot leg side of the i
steam generators. %ese calculations have not been performed for cases with the pumps running, additional work in this area is continuing.
1 De Subcommittee members and consultants raised a number of g2estions regarding the break ficw models and assumptions in calculations of the upsteam fluid conditions. Results of calculations with the pumps running versus not running are highly dependent upon the degree of fluid separation at the break entrance as well as any phase separation effects of the break. Ccncern was expressed by Subcommittee consultants that overly conservative calculations of break flow may lead to inappropriate guidelines and requirements on plant operation during a small break.
f
ECCS Mtg October 17-18, 1979 l
W indicated that since pump operation cannot be guaranteed indefinitely, in order to avoid having the pumps fail during the time period when high core temperature could result, operator action within the first 10 minutes of the transient must be taken to trip the RCP's.
Bis is adequate to pre-U vent tcm*:eratures greater then 2200 F for all break si::es if the HPI system is operating and A?W is available. W stated that they did not in-clude pump heat in their calculational models.
W reviewed the effect of tripping the RCP's on non-LOCA events. A number of non-LOCA transients produce RCS pressure decreases and actuations of safety injection. Rese events include steamline breaks, feedwater breaks, and excessive feedwater events. W indicated that with their recommendation of RCP trip at 1250 PSIA (plus instrument uncertainty) none of the non-LOCA events would produce unacceptable consequences although control of the plant cooldown is more difficult.
W reviewed their revised operator emergency operating guidelines for their plants. Se W philosophy as to provide a basis for continuing diagnosis of the incident utilizing all available information; automatic systems should stabilize the plant prior to operator action; required operator actions and decisions should be minimized, and h.he procedure uniformity should be maximized. We W guidelines are provided to the individual utilities which then produce the procedures for their operators. Se W guidelines are based upon a two-tiered system where the E-O guideline concentriates on verification of plant automatic responses, immediate operator action and diagnosis of the event. After diagnosis the operator moves to one of three courses of action on the second tier where he takes control of the plant and brings it to a cold shutdown condition. W advises the utilties on the make-up of their energency procedures but they do not need to approve them. Each of the procedures is designed to lead the operator through a series of steps based upon the system behavior (much like a computer flow sheet) which guide the operator to bring the plant back to a stable condition. W recommends that the RCP's be stopped when the pressure drops to 1250 PSIA and that the HPI be terminated when the pressure is 2000 PSIA and risirg, when the pressurizer level is e
w m
w
=-
w
ECCS Mtg October 17-18, 1979 greater than 20% and at least one steam generator has water level in the narrow range span.
If the pressure does not drop bels the SI point following HPI termination the operator can then bring the plant to cold shutdown.
.N indicated that they were not rec; mending that there be a requirement on RCS subcooling for termination of HPI, this as in contrast to NRC reccamendations of at least 50 F of subcooling.
Meeting With Combustion Engineering Combustion Engineering reviewed the features of their small break model, the special features they added for reactor coolant pump operation, the results of small break analysis with the RCP's operating, the behavior of non-LOCA events with the pumps tripped and their guidelines for RCP operation during small LOCA's. B e bottom line of the CE studies of small breaks was the recommendation that two pmps be tripped and two be left operating. CE calculations show that the limiting break for CE plants is~in the hot leg. Eis is due to the loss of fluid at the bottom of the hot leg which has been condensed in the steam generator and is running back down toward the vessel. CE also finds that they do not have a window of pump trip items like W; once a certain point is reached pump tripping produces high core temperature and long delays in pump trip do not help the situation. CE could also keep all of their pmps running if both trains of the ECCS system operated rather than only one.
W e CE small break model was reviewed. We reactor vessel uses a drift flux formulation which has an axial void profile in the core based upon a detailed energy balance and includes a phase separation model to indicate which region is liquid or 2-phase and which region is vapor. An empirical correlation is used for the drift velocity.
Both core decay (proper axial profile). and stored vessel metal heat is considered in the fluid heat transfer.
We hot leg model uses a separated two @ase flow model, with a slip ratio derived from a empirical correlation. Countercurrent flow is allowed. A flow regime map allowing separated flow, slug flow, and distributed flow regimes are allowed.
w
t I.
ECCS Mtg October 17-18, 1979 ne steam generator model is based upon a drift flux model for co-current two phase flow with phase separation at the top of the steam generator and countercurrent flew during reflux boiling. 2e conditions for counter-current flow are based upon a Wallis flooding curve.
l Me cold leg model uses drif t flux in the vertical cceponents, separated ficw in horizontal pipes, and has a dynamic reactor coolant pump model based on a single phase pump homologous curve.
Me CE break flow model consists of three parts, a modified Henry-Fauske j
model for subcooled water, the Moody model for t'.e-phase flow, and a
]
modified Mardock-Bauman correlation for superheated steam flow. A number of different discharge coefficients have been used to determine the sensitivity of.the calculations to the break flow, no significant dif-ferences were observed in the core mixture level with the various valves of C. CE found that the name plate discharge capacity of PORV's was D
generally developed by test at low pressure and extrapolated to high pressure and various sizes by analysis. Mien CE did their calculations on a stuck open PORV they reviewd the process in order to determine the flow area to which they could apply their own correlations for break flow over a range of pressures. CE indicated that they also used a spectrum of flow area for their PORV LOCA cases in order to cover any uncertainties in flow area or quantity.
We various heat transfer regimes and correlations used in the CE models for the primary and secondary side of the steam geneator were described for both forward (primary to secondary) and backward heat flow. Se effects of non-condensables on reducing heat transfer were also evaluated as well as the effects upon natural circulation. Calen1= tion of the maximum amount of non-condensables show that blockage of natural circulation would not be i
exoected. CE believes that the non-condensables will reduce the heat transfer by about 3%.
Dr. Catton indicated that the experimental data used for that l
conclusion was obtained in horizontal tubes with flow throix3h and might be appropriate for use in a reflux situation with vertical tubes or U-tuue steam generators.
In this case gases may collect in the tubes. CE indicated that a 50% decrease in heat transfer in the S.G. would result in primary side pressures that were about 100 PSI higher.
w=
" ECCS Mtg
-o-vetob:r 17-10, 1979 CE indicated that with the PCRV open or with a pressurizer vapor space break the steam flow velocities throtzgh the surge line were high enough to pre-vent countercurrent flow, thus the pressurizer level would stay high and there would be no draining as long as the PORV renained open.
Se calculation for the KCS systen with pumps running was reviewed. A number of changes to the EECS analysis were made to account for the flow behavior with the pumps running (see Attachment D). A ptznp model was used based upon CE/EPRI data which accounts for degradation in head with increasing void fraction and then a period of head recovery at pure vapor flow conditions. Se RCP's have the effect of redistributing the coolant water from the cold leg to the reactor vessel, it maintains a 2-phase flow at high velocity until the cold legs are voided, it pressurizes the downcomer and with all four pumps running steam is pumped into the core region of the reactor vessel. CE indicated that there was no guarantee that the RCP could operate under high void fraction conditions. Heat addition by the pumps is considered. As the void fraction increases, the pump hP decreases from about 80 PSI. to about 4 PSI for steam. Operation of the planps maintains a two-phase mixture in the bottom of the hot leg for a longer period of time and produces a greater system fluid inventory loss; thus the core is uncovered to a deeper level for a longer period of time, which gives temperatures above 2200 F.
Without the pumps running, cold leg break are worse since one must assume that all of the ECC water in that leg is lost. De range of break sizes.for which operating RCP pose a 2
2 problem is 0.02 to 0.1 ft.
For 0.1 ft breaks the pianps need to be tripped in 6 minutes. With best estimate analysis this is extended to 10 minutes.
If 2 pumps are stopped prior to 5 minutes the other two can operate without exceading Appendix K limits.
CE reviewed the impact of RCP trip on non-LOCA events. We impacts fall upon the fuel design limits and accident consequences. CE plants trip the reactor coincident with safety injection. If the ptznps are tripped at the same time the fuel heat rate is still high enough to exceed the specified acceptable fuel design limits on ENBR.
Keeping the peps running until the control rods are fully inserted prevents the reduction
!rCS Mtg October 17-18, 1979 in ENBR. Tripping the peps during a steamline break also reduces the margin to fuel failure, however, no fuel failure is predicted. CE be-lieves that it is preferable to keep une pump in each loop running in order to avoid these challenges to the fuel; at a minimum there should be a 5 second delay before the RCP's are tripped following the reactor trip and a safety i
injection signal.
In the case of a steam generator tube rupture without
.the RCP's operating, cooldown will last longer and it may be lorger before the safety valves on the secondary side close, thus increasing the amount of radioactive material that could be released through the secondary side.
Dose limits would not be exceeded.
Meetine With B&W B&W reviewed their snail break LOCA models, analysis methods, the results of tneir small break calculations' and their conclusions regarding reactor punp operations. B&W uses 3 main codes to do their ECCS analysis, CRAFT calculates the system hydrodynamics, ECAM calculates the core mixture level in a more detailed manner than CRAPF, and TIETA predicts the clad heat up and temperature behavior. Se majority of the calculations are done with CRAFT, only if core uncovery is indicated are the other codes used to obtain the FCT values.
S e CRAFT code consists of a number of control volumes, each characterized by a bottom elevation, a fluid mixture level, a height, and an area (volume).
Each control volume has equations for conservatism of mass and energy and a equation of state. Each volme also has a bubble or vapor balance to account for vapor generation and condensation. Flow paths between volmes are also modeled such that the flow could be vapor, 2-phase, or liquid depending upon the height of the path, the fluid level in the volme and the fluid conditions at the flow inlet.
If the mixture level is below the entrance than only vapor will flow.
Flow is determined by a momentum equation d11ch is homogenous and one-dimensional. Imak flow is calculated using an orifice equation for low quality flow and the Moody model for high quality flow with a linear extrapo-lation in between. Se reactor core heat transfer depends upon the mode of boiling (Attachment E).
Se pJmpS are represented by a Change in fluid head l
within that coltrol volume. Se head is based upon a pmp degradation multi-plier applied to single piase homologous curves. Se head mitiplier is
EECS Mtg October 17-18, 1979 1
based upon Semiscale pump data, and Bigham 1/3 scale pump data. We energy addition at the pump is not accounted for. me CFAPr Steam Generator model consists of two volumes in the primary side and one volume on the secondary side. Be heat transfer can be modeled as a function of time, primary level or direction of heat flow.
'ha secondary. side can model feedwater coastdown, auxiliary feedwater and relief valve actuation.
B&W reviewed the use of the CRAFT code for calculating the effects of small breaks with the RCP's tripped and running. We key' parameters are the break location, feedwater to the stem generator, and the operation of the RCP's.
We steam generator is important for a group of small breaks that cannot remove enough heat at the breek location.
Small breaks that are large enough to remove the decay heat do not need the steam generator.
In the noding used with the CRAFT code B&W is careful to make sure that the volumes at the bottom of the steam generators, in the loop seals, and below the vent valves is properly assigned in the lowered loop arrangement.
Rese regions are very important in locating the fluid in the system as it boils down. B&W can have separated countercurrent flow in the fut legs.
B&W has added an additional node at the top of the hot leg U bends to help model natural circulation, interruption of natural circulation, and reflux boiling. We pressurizer is generally modeled with one node but 4 nodes have also been used and the agreement has generally been satisfactory. We four node arrangment does help give detail to the non-equilibritzn effects during insurges and outsurges.
We Wilson bubble rise model is used in all control volumes to calculate phase separation. Within the pressure vessel multipliers are used to' account for the bubble gradients.
The worst break location occurs in the cold leg pump discharge because of the loss of HPI in the broken loop.
ECCS Mtg October 17-18, 1979 Natural circulation during a small break goes through several phases.
The first is a water solP.etural circulation with heat removed at the break and in the steam,..arator; as the system depressurizes natural circulation is interrupted by steam bubbles at the tcp of the hot legs.
Crowt.h of the steam bubble eventually allows stean to contact the icwer regions of the steam generator tubes where heat can again be transfered to the secondary side. At this point a boiling - condensation heat re-
-oval process begins.
Dr. Catton questioned the use of a single heat transfer coefficient in the steam generator for both forced circulation and natural circulation.
B&W indicated that it did not make much difference because changes in the heat transfer coefficient were adequately ecrnpensated for by small changes in the fluid elevation in the tubes and corresponding large surface area changes for the heat transfer.
In one case B&W reduced the coefficient by 50% and there was only a 20 to 30 PSI difference in the calculated system pressure. B&W indicated that they muld continue to look at their steam generator model and would examine possible changes to it to make it more realistic.
2 B&W indicated that for a.01 ft break it took about 10 minutes for the steam generator to drain down far enough to establish a condensing surface.
B&W indicated that they examined the effect of non-condensable gases on natural circulation.
If there is no metal water reaction there are not encugh non-condensables to block the hot leg U bends. 'Ihe effect of the non-condensables on the heat transfer was also looked at.
B&W estimated 2
that the system pressure would be about 25 PSI higher for a.04 ft break 2
and 40 PSI higher for a.01 ft break. Drs. Catton and Plesset indicated that they were surprized that the non-condensable had that small of an effect on the system response. Collection of non-condensables at the condensing surface will give an effect which is different then the experimental data of Colburn and Hougen which as based on a continously flowing fluid.
l i
)
i
ECCS Mtg wcober 17-18, 1979 In response to a question from Dr. Michelson, B&W indicated that heat losses from the pressurizer were not accounted for in the calculations.
Cooling of the water in the pressurizer could prevent its draining into the reactor vessel and lead to less vessel inventory.
B&W reviewed the effects of RCP trip on non-LOCA events. Se major events of concern include those which cause increased heat removal from the primary system (i.e. steamline breaks, and excessive feedwater transients). Tripping the BCP's affects the pressurizer refill rate, steam void formation and distribution in the BCS, and the adequacy of natural circulation core cooling. Prior to March 1979 B&W plants had experienced 8 events that caused engineered safety features actuation.
The majority of these events were feedwater overcooling.
B&W reviewed the Rancho Seco " light bulb" event and indicated that the new pmp trip and HPI criteria were acceptable and adequate given this event.
Two other cases looked at include one designed to maximize the HPI refill rate and one designed to maximize the RCS cooldown. S e first event was an excess steaming rate which causes the plant to trip on high flux and a subsequent pump trip on low pressure. AEW was delayed to enhance the overfill of the HPI system. W e second event was a double ended steamline break with blowdown of both stean generators. B&W indicated that both events have acceptable results. te second event does result in voiding 3
of about 366 ft of one of the hot legs causing interruption of natural circulation; however, the duration is about 8-10 minutes and as the system is repressurized and refilled natural circulation is reestablished. Steam was only produced in the hot leg connected to the pressurizer, the other hot leg maintained natural circulation. Se Subcommittee raised a number of questions about the results of several initial calculations where B&W made calculational errors which showed that the stean volumes were much less. Se Subc:xnmittee explored the question of when B&W determined that they had done sufficient checking and study of a problen to determine that
)
\\
ECCS Mtg October 17-18, 1979 they had correct answers. B&W indicated that they explored their results until they were confident they understood the behaviour of their sytem and the phenomenia involved.
Meeting With Se 'GC ?GR Staff te NRC Staff reviewed the results of their audit calculations on small breaks done by the !G&G staff. Se audit was done to provide a reasonable assurance that the vendors calculations were accurate. S e calculations were done with REIAP-4/ MOD-7.
Calculations were performed for W, CE and B&W plants for a range of break sizes and with pumps tripped and running.
EGEG found that some of the information included in the W calculation was not accurate and additional calculations will be made with corrected data.
In general the. EG&G calculations follow the trends of the vendor calcula-tions and indicate that for a range of breaks the planps need to be tripped.
W e Staff reviewed their assessment of the results of the present information available on mall breaks. Se differences in the methods of analysis and results of the different vendors was pointed out (Attachment F). Se one common agreeement was that use of the BCP's makes the results of some small breaks worse. Se staff has thus concluded that the RCP's must be tripped following a snall break and that the pump tripping will have to be automatic.
Se staff will be meeting with the vendors and utilities in order to attempt to find agreement on the basis for the signals to actuate the pump trip requirenents.
Dr. Resztoczy indicated that additional work would be needed in the area of small break LOCA calculations. Se staff will be reccr. unending that the present methods used for small break analysis should be revised, documented, and sutznitted to the NRC for review within six months, the NRC should establish a position on required conservatisms in small breaks and issue it by 6/30/80, and that plant specific calculations using NRC approved methods should be provided by all licensees prior to 12/31/80. mis w>rk would be done in connection with a accelerated review of Appendix K for small breaks.
ECCS Mtg October 17-18, 1979 i
The second aspect of the additional work muld be the necessity for experimental verification of the small break calculational results.
Bis could be accmplished through the upcming small break tests scheduled for LCFT, Semiscale and TLTA.
Tests muld also be necessary for PORV's and safety valves.
Plant simulators muld have to be up graded to allow operator training on a range of small break LOCA events.
We staff also recommend that the various modes of two-phase flow natural circulation should be demonstrated experimentally by 12/31/80, anu that instrumentation should be provided to verify that natural circulation has been achieved. Challenges to the PCRV's should also be reduced by altering the reactor trip setpoints and the PCRV setpoints.
B e meeting was adjourned at 3:30 p.m.
A complete transcript of the meeting is available in the NRC Public Doctznent Room, at 1717 H St., N.W., Washington, D.C.
A copy of all the slides used in the meeting is included with the record copy !7 the ACRS office.
redes =8 Esthese / vat. es. m. ser 1 meesday o+r 2. nere / Neeiam NUCLEAR REGULATORY
,/
dose three asemene to psesset Cotet0488000t V
Preynmaary information (5118C.
552b(c)(46 Advisory Committee on Reactor Sofeguards Subcommittee on Fwtherinfarmation regardag topio to be rbrw=ed. whether the meetmg Emergency Core Cooling Systems has been marwa=d er se=4=iulsd. the (ECCS); hiseting Chairman's ruimgon reqorsts for the ne ACRS Subcommittee on opportimsty a pmeest mal statements Emergency Core Cooling Systems wt!!
and the sirme alloted therefor can be hold a meeting on October 17-18.1979 in obtamed h a pmpeid t@ csH to Room 1167(October 17) and Room 1046 the Designated Federal Employee far (October 16) at 1717 H St NW.,
the seenas. k Andrew I. Baus.
Washington. DC 20555 to review polspb 202/679stween m material Submitted by Westinghouse, sa mE Combustion Engineering. and Babcock lismas to be canadered at this meeting al reak E c culat on nd the can be foundin documeents on fue and Three Mile Island. Umt 2 Accident svadabue for poWe inspection et the implications regarding the small break NRC Public Doassent Room.1717 H models. Proper operator action to be StnetR. Weslungton. DC 20555 and followed after a small break will be at the Government Pubbcations Section, reviewed. Notice of this meeting was State Lrbrarycf Pennsylvania, published September 20.1979 (44 FR hi"2 MM
~~
"*U M 54559).
Walnut Street. Hamsburg, PA 17128.
In accordance with the procedures Dated september 25. ters.
outlined in ihe Federal Register on
' John C. Hoyle.
October 1.1979. oral or wntten statements may be presented by Adrisory Cbamittee Manegemarrt Oficer.
members of the public, recordings will pm, ame ra.e so-i-re au ami be permitted only during those portions museo caos mi.e of the meeting when a transcript is being kept. and questions may be asked only by members of the Subcommittee. its consultants and Staff. Persons desiring to make oral statements should notify the Designated Federal Employee as far in advance as practicable so that apreopriate arrangements can be made to allow the necessary time during the meeting for such statements.
{
The agenda for subject meeting shall l
be as follows:
Wednesdayand Thursday. October 17and 1A 1979. &30 am. untilthe conclusion ofbusiness each day.
i ne Subcommittee may meet in Executive Session. with any ofits consultants who may be present. to explore and exchange their preliminary opinions regarding matters which should be considered during the meeting and to formulate a repor* and recommendations to the full Committee.
At the conclusion of the Executive Session, the Subcommittee will hear presentations by and hold discussions with representatives of the NRC Staff.
Westinghouse. Combustion Engineering.
l and Babcock and Wilcox and their consultants, pertinent to this review.
'In addition. it may be necessary for I
the Submittee to hold one or more closed sessions for the purpose of exploring matters involving proprietary information.1 have determined, in
)
accordance with Subsection 10(d) of Pub.1.92-463 thet. should such sessions be required. It is necessasy to ATTACHMENT A l
o ACRS SUBCOMMITTEE MEETING ON EMERGENCY CORE COOLING SYSTEMS OCTOBER 17-18, 1979 WASHINGTON, D.C.
ATTENDEES LIST ACRS NRC M. Plesset, Chaiman J. Guttmann H. Etherington, Member S. Fabic J. Ebersole, Meter T. Murley L. S. Yao, Consultant L. S. Tong I. Catton, Consultant Z. R. Rosztoczy C. Michelson, Consultant E. D. Throm Z. Zudans, Consultant B. W. Sheron F. Zaloudek, Consultant R. F. Audette T. Theofanous, Consultant N. H. Wagner A. Bates, Staff
- D. E. Bessette, Fellow COMBUSTION ENGINEERING G. G. Young, Fellow J. Longo
- Designed Federal Employee G. Menzel J. M. Westhoven EG&G - ICAHO R. P. O'Neill C. L. Lling C. Davis F. L. Carpentino T. C. Kessler WESTINGHOUSE J. H. Holderness J. A. Blaisdell R. Skwarek B. Leyse DUKE POWER K. Kesavan W. J. Johnson G. Swindlewhurst B. Steitler R. J. Contratto YANKEE ATOMIC ELECTRIC S. Kellman R. D. Kelly A. Husain R. A. Muench B&W CONSOLIDATED EDIS0N C0 0F NY H. Bailey D. M. Speyer -
GAI NORTHEAST UTILITIES P. L. Bunker P. L'Heureux KEPC0 OMAHA PUBLIC POWER DISTRICT K. Ota J. K. Gasper ATTACHMENT B
,s i
ITEMS FOR DISCUSSICN AT DIE ACRS ECCS SUBCOMMITTEEE MEETING O M ER 17-18, 1979 Approximate Time 45 min.
1.
Small Break aiodel (a) fluid models (b) break flow models (c) steam generator model (d) reactor coolant pump model. (response to 2-phase flow)
(e) sensitivity to nodalization 45 min.
II.
Results of Calculations performed in response to I&E Bulletins 79-05 A,- B, C and 79-06 A, B, C.
(Evaluation model calculations, any best estimate calculations performed and the differences between the two.)
30 min.
III. Effect of tripping or not tripping the reactor coolant pumps and how long they can be allowed to operate.
30 min.
IV.
Effects of tripping the pumps on non-LOCA events.
30 min.
V.
New Operator Procedures following HPI Actuation.
ATTACHMENT C
e.
TENTATIVE SCHEDULE FDR ACRS ECCS SUBCOMMITTEE MEETI!C OCTOBER 17-18, 1979 4
Wednesday, October 17, 1979 8:15 am Executive Session - Opening Comments - M. Plesset 8:30 - 9:00 am NRC Research Work on Break Flow 9:00 - 12:30 pm Review of Westinghouse Small Break Calculations 12:30 - 1:30 pm
- Lunch -
1:30 - 5:00 pm Review of Combustion Engineering Small Break Calculations Thursday, October 18, 1979 8:30 - 12:00 pn Review ot Babcock and Wilcox Small Break Calculations 12:00 - 1:00 pm
- Lunch -
1:00 - 3:00 pm Status of NRC Review of Submittals by B&W, Westinghouse, and Combustion Engineering, Ebture NRC Action to be taken
c 3
MODIFICATIONS TO THE C-E SMALL BREAK LOCA
{
EVALUATION MODEL FOR CONTINUED RCP OPERATION EVALUATION LICENSING RCP BEST-ESTIMATE i
FEATURE '
MODEL MODEL MODEL
- 1. FLUID MODEL DRIFT FLUX DRIFT FLUX FOR UPWARD FLOW OR DOWNWARC FLOW-AT LOW VELOCITY, HOMOGENEOUS FOR DOWNWARD FLOW AT HIGH VELOCITY 2,' PUMP HEAD' NOT CONSIDERED CALCULATED AS A FUNCTION OF VOID FRACTION DEGRADATION
' BASED UPON ANC DATA
?
- 3. DOWNCOMER TO NOT CONSIDERED EXPLICITLY MODELED U.P. BYPASS
- 4. SECONDARY SIDE PASSIVE ONLY TURBINE BYPASS AND ATMOSPHERIC DUMP VALVES PRESSURE CONTROL (SAFETY VALVES)
OPERATIONAL i
- 5. TWO-PHASE LEAK MOODY HOM0G. EQUIL.
I FLOW
- 6. INNER VESSEL LICENSING MODEL P
IMPROVED MODEL VOID DISTRIBUTION O
Reactor core heat transfer model L
CORRELATION MODE DITTUS BOELTER 1.
SUBC00 LED FORCED CONVECTION THOM 2.
NUCLEATE BOILING McDONOUGH, MILICH AND KING 3.
TRANSITION BOILING DOUGALL-ROHSENOW 4.
FILM BOILING GROENEVELD 5.
FILM BOILING MORGAN
- 6.
POOL BOILING DITTUS BOELTER 7.
StPERHEATED FORCED CONVECTION
- LOWER BOUND SET FOR FILM BOILING REGIME WHEN 2
5 LB/FT -HR-MASS FLUX s 2 x 10 OTHERWISE BOUNDED BY INPUT VALUE CRITICAL HEAT FLUX MODEL (PRESSURE DEPENDENT)
P 2 1500 PSIA B&W-2 INTERPCLATION-B&W-2 AND BARNETT 1500 > P > 1300 BARNETT 1300 2 P 2 1000 INTERPOLATION-BARNETT AND MOD-BNT 1000 > P >
725 MODIFIED BARNETT 725 2 P m
nn-4 4
I MAXIMUM AVAILABLE li of EffECT OF TK ?? Ilk. '.
ilHr FOR PUMP INUOUS ONE PtMP IN I ACH l
BREAK LOCATION BREAK SIZE.
.I IL..P 0PERATION
. LOOP i
'l' B&W RESULTS NOT LIMITING BREAK m 3 MINUTES ACCEPTA8LE CORE NO EVALUATION SENSITIVE DUE SIZE ABOUT (BASED ON PREllMINARY COOLING 2
TO HOMOGENEOUS 0,02 0,2 FT CALCULATIONS)
MODELING ASSUMPTION 1
s1 m
2 CE
- FOIMO HOT LEG LIMITING BREAK 6 MINUTES AFTER O'.1 F1 BREAK ACCEPTA8tE CORE BREAKS LIMITING /
SIZE ABOUT TRIP + SIAS FOR IN HOT LEG LEADS COOLING FOR BE 2
SOME. COLD LEG 02
,1 FT EM ANALYSIS 10 TO PCT's> 2200* P ANALYSIS PROVIDED BREAKS COULD MINU(ESAFTER TWO PUMPS TRIPPED EXCEED 2200 *P TRIP iSIAS FOR WITHIN S MINUTES BE ANALYSIS AFTER BREAK
'0 Westinghouse COLD LEG BREAKS LIMITING BREAK 10 MINIJTES FOR ALL ACCEPTABLE CORE NO EVALUATION LIMITING, N0 HOT Sil[ '.02
.05 PLANTTYPES(2,3 CmLING LEG BREAKS FT 4 LOOPS)
ANALYZED RESULTED IN PCT'S?2200* F e
'* CE ANALYSES PERFORMLJ FOR PLANTS WITH 200 PSI SIT'S',
1200 pst HPSI PUMPS, ANALYSES CONSIDERED CONSERVATIVE WRT PLAHis WITH 600 PSI SITS AND/0R HIGHER llEAD HPSI PIMPS.
i i
?
4 Y
s
<b.,
,,a MOOR DIFFERENCES DURENG 58LOCA WITH PUMPS RUNNING p
i i
noca l
Item W
CE BW RELAP/P.00-7 Cold Leg Stratified Flow Homogeneous Homogeneous Heterogeneous Pump Discharge flow flow flow Pipe
- Downcomer Heterogeneous Model switches Homogeneous Heterogertous model from homogeneous flow flow to heterogeneous model when drift velocity criteria met.
Core Heterogenecas Heterogeneous Homogeneous Heterogeneous flow flow flow flow Hot Leg Pipe' Homogeneous flod Homogeneous Heterogeneous
.Heteroceneous flow flow. No counter-
. for CL breaks' flow with pro-current flow Heterogeneous
. vision for control volumes drainin9 allowed for hot leo
' breaks. No CC
[
flow allowed
(
for either case 5:eam Homogeneous Drift flux Homogeneous Heterogeneeus fic.
Generator flow model - allows flow no vertical sli;/
Hot Side liquid fallback fluid runback to Tubes to het leg if hot leg possible Steam Homogeneous Homogeneous Homogeneous Heterogeneous f1:-
Generator flow flow flow no vertical slip Cold Side Tubes C:'.d Leg Homogeneous Homogeneous Homogeneous Homogeneous Loop Seal flow flow flow flaw (suction pipe)
e-f-3 Model/ Method W
CE B&W EE&G Tdaho ECC Injection No injection No spillage
- 30%
Consistent with assumed in assuned for -
spillage of vender broken loop hot leg water assumptions for cold leg breaks -
injected in breaks no injection broken loep assumed in broken for cold lors for cold leg breaks les breaks ECC Injection Downcomer/ lower Downcomer Cold Leg h[ - upper downcc-(cold le (coldlegby CE - cold leg design)gby design)
B&W - cold leg Locatien pienum node (cold le design) g by Quench No carryover No carryover No carryover No carryover Behavior during accounted for accounted for accounted for accounted for Recovery Steer. Super.
Sucerheating axial coolant No superheat 3 axial coolant Fett Calcu-censidf ed nodes in core, calculated due nodes in core.
la: ion (description Superheating to single Superheating proertetary) of each node control volume of each node allowed model of core. allowed.
All core heat added to liquid phase.
Separate heat-up model cal-culates super-heat but uses CRAFT mixture level.
Cc e fluid Themodynamic Thermodynamic Themedynamic Themedynamic quality equilibrium equilibrium equilibrium equilibrium assumed -
assumed -
assumed -
assumed -
actual actual quality actual actual quality quality not not calculated quality not not calculated calculated calculated "
l W
O*
N*
- 9 6
6 *
,