ML20248L204
| ML20248L204 | |
| Person / Time | |
|---|---|
| Issue date: | 06/08/1998 |
| From: | Scaletti D NRC (Affiliation Not Assigned) |
| To: | Smith E AFFILIATION NOT ASSIGNED |
| References | |
| NUDOCS 9806110008 | |
| Download: ML20248L204 (1) | |
Text
. - _ . - _ _ _ _ _ _ _ _ - _ _ _ _ - _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ -
- o s June 8, 1998 b !
- Eric R. Smith Licensing Information Services l NUS Information Services 2650 McCormick Drive, Suite 300 Clearwater, FL 33759-1049 Enclosed are the doci'ments you requested in your e-mail to Ted Quay on May 4,1998, and our recent telephone conversation; a staff evaluation, dated July 31,1996, and two recently issued evaluations, dated May 18,1998. The three documents have been put in the Nuclear Regulatory l . Commission (NRC) Public Document Room (PDR) and are accessible to the public. Other Ii-
- Simplified Boiling Water Reactor (SBWR) correspondence that you expressed an interest in may be made available in the near future. ' As it becomes availablo, the correspondence will be put into the PDR and a c.opy will be sent to you.
The last NRC staff actions regarding the SBWR are covered in the two enclosed letters dated May 18,1998. Also, I was not able to find non-proprietary versions of NEDE-32177 and NEDE-32178 in the NRC Central Files.
Sincerely,
)
original signed by:
Dino C. Scaletti, Project Manager.
Standardization Project Directorate Division of Reactor Program Management Office of Nuclear Reactor Regulation
Enclosures:
As stated DISTRIBUTION:
Docket File PDST R/F TQuay PUBLIC DScaletti JNWilson ACRS (11) JMoore,0-15 B18 1
/, Q
~
9801y a Y
DOCUMENT NAME: A: SMITH.DS ROWD M Ok &lb PCk T* receive a cop r of this document, ineScate he the bom *C* = Copy without attachment / enclosure *E* = Copy with attachment! enclosure 'N' = No copy OFFICE PM:PDST:DRPM J D:PDST:DRPM l c l l NAME DCScaletti:sg M TRQuay-1W4 DATE- 06/f/98 06/ V /98 OFFICIAL RECORD COPY n
' 9866bOOO8 980608 PDR TOPRP ErfVGENE C PDR
.J
i a cec y- 4 UNITED STATES g
j
's NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 July 31,1996
\ .C7
- .++ /
Mr. James E. Quinn, Projects Manager LHR and SBWR Programs GE Nuclear Energy 175 Curtner Avenue, M/C 165 San Jose, California 95125
SUBJECT:
STAFF REVIEW OF GENERAL ELECTRIC'S LICENSING' TOPICAL REPORT (LTR),
NEDE-32176P, "TRACG MODEL DESCRIPTION " REVISION 1, RELATED TO CONTAINMENT AREA l
Dear Mr. Quinn:
In response to your letter dated March 13, 1996, the staff has prepared the i enclosed report on its evaluation of the GE's TRACG Model description,
! Revision 1, related to containment area.
The staff has performed a review of the acceptability of the revised LTR and identified a few concerns related to containment models as discussed in the enclosed report. Therefore, the staff concludes that in order to accept TRACG containment models, the code will have to go through extensive comparative studies using both experimental data and other containment models.
You are requested to review the enclosed report to determine if it contains any GE proprietary information and provide your response within 30 days of the date of this letter.
If you have any questions regarding this matter, please contact Son Ninh at (301) 415-1125 or Andrzej Drozd at (301) 415-2807.
Sincerely, I 4- M 8 ^7 Theodore R. Quay, Director Standardization Project Directorate Division of Reactor Program Management Office of Nuclear Reactor Regulation
Enclosure:
As stated cc w/ enclosure:
See next page i
c> h -- i
Mr. James E. Quinn Docket No.52-004 i GE Nuclear Energy l
cc: Mr. Robert H. Buchholz Mr. Tom J. Mulford, Manager GE Nuclear Energy SBWR Design Certification 175 Curtner Avenue, MC-781 Electric Power Research Institute San Jose, CA 95125 3412 Hillview Avenue Palo Alto, CA 94304-1395 Mr. Steven A. Hucik GE Nuclear Energy Mr. Rob Wallace 175 Curtner Avenue, MC-780 GE Nuclear Energy San Jose, CA 95125 1299 Pennsylvania Avenue, N.W.
l Suite 1100 Washington, DC 20004 Enclosure to be distributed to the following addressees after the result of the proprietary evaluation is received from Simplified Boiling Water Reactor:
- Mr. Brian McIntyre l Westinghouse Electric Corporation Energy Systems Business Unit Box 355 Pittsburgh, PA 15222 Director, Criteria & Standards Division Office of Radiation Programs U.S. Environmental Protection Agency 401 M Street, S.W.
Washington, DC 20460 Mr. Sterling Franks U.S. Department of Energy NE-42 Washington, DC 20585 1
4 3
STAFF EVALUATION OF THE TRACG MODEL DESCRIPTION FOR CONTAINMENT APPLICATION GENERAL ELECTRIC REPORT NEDE-32176P INTRODUCTION AND BACKGROUND In February 1993, General Electric (GE) submitted a topical report NEDE-32176P, Revision 0, describing the TRACG models. TRACG is GE's thermal-hydraulic code being proposed as a tool for licensing safety analysis of the Simplified Boiling Water Reactor (SBWR).
After reviewing the 1993 report, the staff issued initial comments in December 1994, where the staff identified major roadblocks in the TRACG review process.
The staff concluded that the Revision 0 report did not contain sufficient information regarding' containment models for the staff to make meaningful comments.
This position was reiterated several times by the staff during consequent telephone discussions and meetings with GE. As a result, GE agreed to expand the description of TRACG containment models in the revised model report. As an interim step, GE issued a " road map" describing the use of generic TRACG physical models and correlations in the major containment regions and compo-nents. In a letter to GE (November 3, 1995), the staff acknowledged that the. '
" road map" is a good starting point.in reviewing the containment models employed in the code.
In February 1996, GE submitted Revision 1 of the NEDE-32176P report. The following is the staff's evaluation of Revision 1.
GENERAL REMARKS TRACG containment models utilize the same conservation equations and constitu-tive correlations as applied to the reactor system xdels, i.e., the code, which was initially developed to model primary side of BWRs, is currently being used to model full plant including containment. Similar attempts in the past to apply models developed for small, pipe-like volumes to containment calculations, with its multi-dimensional, multi-phase, and multi-component characteristics, where mostly unsuccessful. Because of that experience, the staff is reluctant to accept " generic" thermal-hydraulic containment models i based solely on possible " theoretical correctness" of TRACG equations. For example, a " correct" flow solution for a pipe break may not be applicable to the flow between containment cells. In order to accept TRACG containment models, the code will have to go through extensive comparative studies using both experimental data and other containment models.
The staff did not perform a detailed evaluation of the derivation of TRACG model as applied to containment. However, the staff is concerned that even preliminary review suggests that some equations (e.g., Eq. 6.6-60) and models (e.g., fogging, hydraulic diameter based on cell size) may be erroneous.
Enclosure
The. following comments address the applicability of the generic conservation equations and the constitutive relationships to containment, as presented in NEDE-32176P.
DISCUSSION
- 1. Conservation Equations TRACG code is based on two-fluid approach with three conservation (mass, {
energy, and momentum). equations for each phase. There are additional mass I balance equations for noncondensible gases. All of the gases are assumed to j be in mechanical and thermal equilibrium, implying that they have same I velocities and temperatures.
The TRACG code has two sets of formulations differing in their approach to the momentum balance. The first one is for one-dimensional flows (PIPE compo-nent) and the second is for three-dimensional flows (VSSL component). The constitutive relationships are the same.
The containment is modeled with three dimensional VSSL component with PIPE component representing connections and PCCS heat exchanger.
The assumptions in the current TRACG formulation can not address the mixing phenomenon of steam and gases or predict the distribution of noncondensibles.
The dominant mechanism of mixing is the momentum transfer between gases. The ;
steam jet from the break into drywell and its effect on surrounding gases can j not be addressed. The current formulation assumes complete mixing between the gases in a cell. Furthermore, the gas mass balance equation has a mixing term which increases the exchange of noncondensibles between the adjacent cells but this process, as implemented in the code, is cell size dependent and can not i l be used. Furthermore, TRACG report (Section 6.7.2) indicated that the mixing l model is not used in the containment, j I
The current formulation also has a limitation that it does not apply to open spaces as it does not account for viscous diffusion terms in the momentum balances. The momentum equations require the presence of walls for any. .
- viscous effects. '
The mixing of steam with noncondensibles is an important phenomena for the initial pressure peak in the containment as the assumption of complete thermai ,
. mixing will lead w rain out and lower peak pressure.
Another important' phenomenon is thermal stratification in the' suppression !
pool.- Various injections of steam will- heat-up the fluid around the injection ,
points. However, it is an unstable situation as cold fluid cannot stay above the hot fluid. There will be thermal conduction and thermal currents (free l' convection) which will lead the pool to reach another stable state with the hot fluid on the top. However, the 1-D version of the code does not have the capability to model mixing or an unstable situation. The 3-D VSSL component will allow some circulation, however, the, timing and extent of the mixing has
l l
not been assessed. .A vigorous mixing may lead to uniform temperature rather than stratified tuperature leading to smaller surface temperature and smaller partial pressure of steam.
There is a~ need for a statement in the manual regarding the applicability and limitations of the current set of conservation equations for diverse contain-ment applications.
- 2. Constitutive Relationships There are interactions between the wall, liquid, and gases and they lead to six sets of. closure laws, i.e., wall momentum transfer and wall heat transfer to the liquid and gases and, mass and momentum transfer between the gas and the liquid. These laws require further information about the shape and the size of interface between the phases which is obtained from flow regime maps.
2.1 Flow Regime Maps The flow regime maps provide the critical information about the interfacial area density and the shape for two-fluid formulation.
The SBWR containment consists of many sections where two-phase flow conditions exist. These sections vary in size and orientation. The drywell and suppres-sion chamber consists of large volumes which may have film on walls and droplets in the gas phase. The suppression pool will get an inflow of a jet of mixture of noncondensible' gas and steam which will break up in bubbles.
There are also liquid pools with free surfaces. The horizontal vents during early blowdown undergo vent clearing and two-phase flow. The heat exchangers of the' PCCS have small diameter (0.05m) tubes with downward flow of film on the wall.
TRACG contains a simple flow regime map (Figure 5.1-1) for vertical pipe flows based on vapor flux and void fraction. Generally, this type of flow regime map is based on' mixture mass flux and not on vapor mass flux. No basis is provided_for this map.
The report correctly mentions that in most cases the flow regimes were developed from visual observations, although in some tests pressure drop measurements were used to discern the regimes. However, there is large uncertainty in these regimes. The report also states that the applicability of the code should be based on the assessment with the tests which simulate interfacial transfer as they include flow regime map. This is one weakness of a two-fluid approach where the formulation calls for detailed information at the phasic level which are not yet available.
The transition between annular flow and dispersed flow regimes is given by entrainment inception. However, no information about entrainment inception is provided. . The entrainment rate correlation described in the report, is based on pipe data with the diameter less than 0.032 m and, therefore, the entrainment correlation does not apply to any part of the containment except to PCCS tubes. A liquid film is expected on the heat structures and liquid droplets in the drywell atmosphere. However, the droplet field can not be
I 9
predicted by the entrainment criteria in the code as the mechanism is fogging and not shear at the interface. Therefore, the flow regime map does not apply to drywell and suppression chamber.
There is also a question about the applicability of pipe flow regime map to the suppression pool and to the downward flow in PCCS tubes. The Tables 6.1-1 and 6.2-1 summarize GE's assessment of flow regime maps for different contain-ment sections. The indirect assessment through interfacial shear and mass transfer data base covers the pressure, void fraction and mass flux range, but the diameter range is not covered for drywell and suppression chamber.
2.2 Wall Friction Wall friction or momentum transfer is important in PCCS tubes and horizontal vents. The friction on containment walls is also computed in the code. The ,
single phase friction factors are calculated from the curve fit to Moody's diagram which is valid for pipe flows. The data base covers a very large Reynolds number range. The uncertainty for square channels is about 10 percent. However, the applicability to drywell geometry and large diameter channels is questionable. This model was assessed with the data base, limited to diameter less than 0.33m, which is close to PCCS tubes (dia 0.05m) but too small for horizontal vent (dia 0.7m). Furthermore, the two-phase multipliers were based on the data with steam qualities of less than 50 percent while in the drywell as well as in the horizontal vents, the quality could be close to 100%. Furthermore, it is not clear if the two-phase multiplier is valid for down flow as expected in the PCCS tubes and in the horizontal vents.
There is another uncertainty in the implementation of the friction factors in the 3-D component used for containment. It is not clear how the friction factor in the transverse direction are estimated from the Moody's curve which was developed from vertical tube flows.
The additional uncertainty is in the partitioning of the wall friction contribution between two phases. The correlations for single-phase flow along with two-phase multiplier are for mixture models and are being used for two-fluid formulations. The report does not indicate the method of dividing wall friction between two phases.
It is recognized that the wall friction is not very important in the drywell or gas space of the suppression chamber.
2.3 Wall Heat Transfer The wall heat transfer occurs in every component in the containment. The important areas are heat transfer to vertical and horizontal structures and inside and outside of the PCCS tubes.
The single-phase heat transfer is based on Dittus-Boelter for forced flow and McAdams correlation for free convection on vertical walls. However, applica-bility of these correlations for large open spaces has not been shown.
I Dittus-Boelter correlation was developed from pipe data and requires hydraulic diameter for Reynolds number calculation. Similarly, McAdam's correlation
also requires hydraulic diameter for computing Grashof number. These correla-tions have been implemented with hydraulic diameter based on cell size which is not a good practice. However, for open spaces, if the cell hydraulic diameter is computed with only wetted perimeter in the denominator, the diameter will come out to be correct. It will be more appropriate to use correlations for flat wall which are based on wall length.
The correlations used in modeling heat transfer require an estimate of Reynolds number, but it is not shown how it is estimated. For the 3-D formulation, there are three components of velocity and the code documentation does not indicate which component of the velocity is used to estimate Reynolds number. The other uncertainty is in the use of cell edge velocity. As the calls are large, the velocity is averaged over a large area and the effect of a no slip condition at the wa'.. is negligible. The correlation was developed from pipe flow data where the average velocity is affected much more by a no slip conditions at the wall.
Furthermore, the wall heat transfer is partitioned between two phases but it is not explained how this partitioning is performed.
For the horizontal surfaces, TRACG uses the same heat transfer correlation as for the vertical walls. The assessment provided indicates that for GrxPr >
1.0 Ell the heat transfer coefficient is over predicted by 550 percent. There is a need to provide an assessment of the effect of this discrepancy on long term pressure. The heat transfer from the floor will be different than from the ceiling. This is not distinguished in the code.
The other area of importance is heat transfer due to condensation on cold surfaces. With the accumulation of noncondensibles the condensation rate will degrade. The TRACG model employs this heat transfer with Nusselt's correla-tion for condensation and degradation due to noncondensible calculated as a minimum of Kuhn-Schrock-Peterson correlations, derived from vertical pipe data, and Uchida' correlation, which is based on experiments in a large volume. In principle, the staff accepts such an approach. However, the applicability of this model to the containment analysis should be discussed in more details given the fact that the nodalization may affect the non-condensible concentration near the interface and therefore, the heat transfer degradation. The data base for these correlations covers pressure up to 4.5 bars which is appropriate for containment application.
The PCCS consists of heat exchangers submerged in a pool. The condensation oc. curs inside the tube and most likely nucleate boiling on the outside. In ti3 TRACG model, the PCCS pools are represented by two rings each with many vertical levels. The two rings will allow for convection in the pool. The heat transfer on the pool side of the tube is correctly modeled with Forster-Zuber correlation for pool boiling.
The two-phase flow in the tubes is modeled with conventional approach of film flow regime. The critical aspect of this component is the heat transfer inside the tubes. TRACG used Dittus-Boelter model for single-phase flow and Kuhn-Schrock-Peterson correlation for condensation heat transfer. The correlation is appropriate as it was developed from 50.8 mm 0.D. tube data.
i which is the same diameter as the PCCS tube, and for pressures up to 5 bars.
However, implementation as described in Section 6.6.11.1 has an error in Eq 6.6-60. The average heat transfer coefficient ir a function of the length over which averaging was done and a derivative with respect to [z] should
-account for this dependency. This model should be revisited.
2.4 Interfacial Momentum Transfer j The interfacial momentum transfer occurs at the interfaces and affects the distribution of the liquid and vapor phases and therefore, eoid fraction. It i is important to predict the void fraction accurately as it has an effect on '
heat transfer and the two-phase multipliers for wall friction and local pressure loss coefficients. The containment has many sections where interfa-cial momentum transfer needs to be modeled, such as the film on the wall or the spillover from the vessel in the drywell, droplet phase, PCCS tube film flow, flow in the horizontal vents and the flows over liquid levels in GDCS tank, horizontal vents and suppression pool. i The general approach in TRACG is to use mixture information or drift flux correlation and to partition it into interfacial shear for different regimes.
The description lacks an assessment about applicability to the containment.
The areas where the models may not be applicable are drywell, horizontal vents, and suppression pool. In the drywell area, the liquid will be in the form of film on structures and of fog in the atmosphere. The flow regime maps will not predict a film flow and therefore, the code may select dispersed flow regime. Furthermore, the fogging in the bulk due to cooling of steam will lead to droplet regime. However, the size of drops should not be determined from Weber number equal to 12, as this critical Weber number represents the largest drop size while fog will consists of much smaller drops. The fogging phenomenon will produce a spectrum of drop sizes which cannot be represented by a drop size calculated from the critical Weber number, and thus resulting in a different behavior of the droplets.
The other area where applicability is not certain is in the horizontal vents as interfacial shear was derived from Vg j and Co for vertical flows and it will not apply to horizontal vents. No assessment has been presented for application to horizontal flows.
The suppression pool will get injection of a mixture of steam and noncondensibles from different sources. The steam condensation will depend upon the residence time of the bubble and the interfacial area. The report :
does recognize the difficulty of the modeling the pools (see Eq. 6.1-33 in the i report). If void fraction is overpredicted, then interfacial shear is l underpredicted and bubbles will have larger residence time and larger interfa- l cial area leading to more condensation. l 2.5 Interfacial Heat and Mass Transfer The heat and mass transfer at the interface are related and predictions of one will provide an estimate of the other. The model consists of predicting the i
I
3.- -
- I a
i flow regime, interfacial area density and heat transfer coefficients at the interface. The condensation phenomenon in the presence of cold walls has been
' covered in the wall heat transfer section.
The interfacial heat transfer occurs in the suppression pool when there is
- flow through horizontal vents, or venting from PCCS or flow from SRVs. In all cases, there will be a flow of a mixture of steam and/or noncondensibles into the pool. The liquid side interfacial heat transfer coefficient is obtained from the correlation developed for heat transfer over evaporating drops. The vapor side interfacial heat transfer coefficient is obtained from the conduc-tion heat transfer solution for the solid sphere with a correction for internal convection. However, the degradation due to noncondensibles was not accounted for. The current model will condense all steam while only partial condensation is expected in the presence of noncondensibles. The complete condensation will raise the pool water temperature which will increase the partial pressure of steam in the suppression chamber (SC) space, but under predict the pressure of the gas space due to the smaller amount of gases going =
to this space. It is difficult to estimate the net effect as it depends upon the concentration of noncondensible in the flow.
Another major area is the condensation on the free surfaces. It is an important heat sink and the primary mechanism of heat transfer is by condensa-tion. However, the heat transfer is impaired due the accumulation of noncondensible gases near the interface and the heat-up of the liquid layer near the interface. It is important that these effects be considered as these two phenomena will degrade the heat transfer. In the absence of these phenomena, the condensation process will continue and will create low
. pressure at the interface which will bring in more steam. TRACG manual does not describe this phenomenon.
- 3.
SUMMARY
AND CONCLUSIONS The balance equations do not allow for different velocities or temperature for the gases. Therefore, the . code will predict good mixing, fogging, and lower pressure.
There is a mixing term in the mass balance equations which is nodalization dependent but it is not used in containment input.
The stratification in the suppression pool is important as it controls the partial pressure of the steam. The TRACG code may not be able to predict this stratification when steam condensation is taking place at different locations.
~The mixing process will be nodalization dependent. No assessment is provided for accuracy and the timing of achieving stable stratification. l
!. Flow regime maps were developed for vertical upflow and are applied to large j volumes and down flow in the PCCS tubes. The document does not provide a direct assessment for applicability. The entrainment rate determines the dispersed phase but in the drywell the dispersed phase is fogging and the mechanism is not entrainment.
o .
There are four concerns with the wall friction: Moody's correlation is based on tube data and will lead to large uncertainty in containment application with 3-D VSSL; the two-phase multiplier was developed from data with flow oualities below 50 percent while higher quality is expected in containment; two-phase flow multipl,ier was developed for an upflow, and there will be downflow in PCCS tubes; and horizontal flow in the vents, and partitioning of wall friction between the phases.
There are two concerns with the wall heat transfer; the wall heat transfer partitioning between the phases and degradation of condensation on vertical walls. The TRACG model employs heat transfer with Nusselt's correlation for condensation and degradation due to noncondensibles calculated as a minimum of Kuhn-Schrock-Peterson correlations, derived from vertical pipe data, and Uchida's correlation, derived for large volumes. The applicability of this model to the containment analysis should be discussed in more details.
There are two concerns with the interfacial momentum transfer; interfacial I shear is developed from drift velocity for vertical pipe, while there is horizontal flow in the vent and, the fog in the drywell consist of drops whose size will be over predicted by the Weber number criterion and therefore, drops may rain out instead of flowirig with the gases.
There are two areas of concern for interfacial heat and mass transfer; the bubbles of a mixture of steam and noncondensibles in the suppression pool will have complete steam condensation and no degradation due to noncondensibles, and the lack of adequate discussion of the condensation heat transfer and its degradation due to noncondensibles at the free surfaces in the containment.
I
\
l _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ - - - - - _ - - - -_ A
g y, u.,\ UNITED STATES
. =-
!' E NUCLEAR REGULATORY COMMISSION WASHINGTON D.C. 30006 4 001 l
k * . ., .
- May 18,1998 i
1 Dr. Craig D. Sawyer, Manager Advanced Reactor Programs GE Nuclear Energy 175 Curtner Avenue, M/C 754 San Jose, CA 95125
SUBJECT:
STAFF REVIEW OF GE REPORT" GIRAFFE SBWR SYSTEMS INTERACTIONS TEST REPORT" (NEDC-32540P, REVISION-0)
Dear Dr. Sawyer; l
When the design certification review of the Simplified Boiling Water Reactor (SBWR) was terminated in 1996, the Office of Nuclear Reactor Regulation agreed to provide General Electric Nuclear Energy (GE) with a preliminary review of Simplified Boiling Water Reactor (SBWR) testing and analysis program reports. This letter documents the staff's preliminary review of the GIRAFFE SBWR Systems interactions Test (GIRAFFE / SIT) report.
The Staff agrees in general with the conclusions reported by GE in Section 7 of NEDC 32540P, Revision-0 that: (a) the test results are consistent with and generally indicate expected performance of the simulated SBWR passive safety systems; (b) there were no notable adverse systems interactions; (c) the test strategy was shown to be appropriate; and (d) that an acceptable database has been obtained to aid in the validation of the capability of the TRACG
' computer code to calculate SBWR behavior during the phases of design basis accidents represented in the GIRAFFE / SIT program.
' The following additional comments are being provided for consideration by GE in the event that the SBWR design certification review (including the test and analysis program) is restarted at some future time.
The staff recommends that further data analysis be performed to reduce, if possible, the noted errors and uncertainties in some of the data. This includes: (a) correction of the reactor pressure vessel (RPV) level for dynamic effects (discussed on p. 6 2); and (b) correction of DP/ level measurements by using actual fluid temperatures rather than 277K (discussed on P. 6 1).
The emphasis in the report is on demonstrating that there are no adverse systems interactions l
and on test-to-test comparisons of key parameters. The staff agrees that these are important i objectives; however, further consideration of actual phenomenology during the tests is also valuable, and would be necessary in a more comprehensive review to demonstrate TRACG's ability to represent SBWR phenomena consistent with the SBWR Phenomena identification and Ranking Tables. For example, in Fig. 6.2-10, between about 200 and 1000 seconds (the end of the plot), small amplitude oscillations occur in the collapsed liquid level in the chimney during Test GS2. Since the suppression chamber pressure is greater than the drywell pressure during this period (per Fig. 6.2 2), the staff believes that these oscillations may occur as a result of repeated actuations of the wetwell-to-drywell vacuum breakers. (This would be consistent with results from the Office of Nuclear Regulatory Commission's confirmatory tests in the PUMA facility.) While this does not appear to cause adverse interactions, it nevertheless demonstrates A5VJKVX%hiGr2
Dr. Craig D. Sawyer May 18, 1998 the very close coupling that the staff believes exists between the SBWR containment and RPV during accident conditions, and could be an issue in a future review of TRACG modeling capabilities.
A more comprehensive and quantitative orror and uncertainty analysis would be expected in a report that was part of an active certification review. N GE does reopen SBWR certification, the report should be revised to include auch an analysis.'
With this letter the staff concludes its review of the GIRAFFE / SIT. N you have any questions regarding this matter, please contact Dino Scaletti at 301-415-1105.
Sincerely, original signed by:
Theodore R. Quay, Director Standardization Project Directorate Division of Reactor Program Management Office of Nuclear Reactor Regulation Docket No.52-004 a
l b.. _
b' o
Dr. Craig D. Sawyer Docket No.52-004 GE Nuclear Energy cc: Mr. Rob Wallace Mr. Brian McIntyre GE Nuclear Energy Westinghouse Electric Corporation i 1299 Pennsylvania Avenue, N.W. Energy Systems Business Unit Suite 1100 Box 355 Washington, DO 20004 Pittsburgh, PA 15222 Director, Criteria & Standards Division Office of Radiation Programs U.S. Environmental Protection Agency 401 M Street, S.W.
Washington, DC 20460 Mr. Sterling Franks U.S. Depadment of Energy NE-42 Washington, DC 20585 l
l Mr. Robert H. Buchholz
! GE Nuclear Energy 175 Curtner Avenue, MC-781 San Jose, CA 95125 Mr. Steven A. Hucik GE Nuclear Energy 175 Curtner Avenue, MC-780 San Jose, CA 95125 Mr. Tom J. Mulford, Manager SBWR Design Certification '
Electric Power Research Institute 3412 Hillview Avenue Palo Alto, CA 94304-1395 I
l
. _ . _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _A
r May 18, 1998 t
Dr. Craig D. Sawyer, Manager Advanced Reactor Programs GE Nuclear Energy 175 Curtner Avenue, M/C 754 San Jose, CA 95125
SUBJECT:
STAFF REVIEW OF GENERAL ELECTRIC'S PANTHERS - IC TEST REPORT (SIET 00458RP95)
Dear Dr. Sawyer:
l . When the design mortification review of the Simplified Boiiing Water Reactor (SBWR) was terminated in 1996, the Office of Nuclear Reactor Regulation agreed to provide General Electric Nuclear Energy (GE) with a preliminary review of SBWR testing and analysis program reports.
The enclosu o documents the staffs preliminary review of the PANTHERS-lC test report. The j l
subject report was issued by SIET Laboratories and was reviewed by GE before submit'11 to the '
Nuclear Regulatory Commission.
The staff agrees in general with the conclusions reported by SIET (for GE) regarding the thermal-hydraulic testing phase of the PANTHERS-IC program. The staff has a few comments on this phase of testing as shown in the enclosure.
Additiona'ly, the staff has comments regarding the stmetural phase of testing, which was not completed when testing was suspended in November 1995. The staff would require that this test ,
program be completed in the event that the SBWR design certification review is restarted at some future time.
l This letter closes the staffs review of the PANTHERS-IC test program. If you have any questions regarding this matter, please contact Dino Scaletti at 301-415-1104.
Sincerely,
- TNohoS'k. bay',Ndtor l Standardization Project Directorate Division of Reactor Program Management Office of Nuclear Reactor Regulation Docket No. S2-004
Enclosure:
As stated oc w/ encl: See next page I
i l I i
l \
- 9 R D d 2n(~ N W Y _
t Dr. Craig D. Sawyer Docket No.52-004 GE Nuclear Energy cc: Mr. Rob Wallace Mr. Brian McIntyre GE Nuclear Energy Westinghouse Electric Corporation 1299 Pennsylvania Avenue, N.W. Energy Systems Business Unit Suite 1100 Box 355 Washington, DC 20004 Phtsburgh, PA 15222 Director, Criteria & Standards Division Office of Radiation Programs U.S. Environmental Protection Agency '
401 M Street, S.W. {
Washington, DC 20460 Mr. Sterling Franks U.S. Department of Energy NE-42 Washington, DC 20585 Mr. Robert H. Buchholz GE Nuclear Energy 175 Curtner Avenue, MC-781 San Jose, CA 95125 Mr. Steven A. Hucik GE Nuclear Energy 175 Curtner Avenue, MC-780 San Jose, CA 95125 Mr. Tom J. Mulford, Manager SBWR Design Certificate n Electric Power Research Institute 3412 Hillview Avenue Palo Alto, CA 94304-1395 l
l j
i,- i. .
t STAFF REVIEW OF GE PANTHERS-IC TEST REPORT (SIET 00458RPg5)
The staff agrees in general with the conclusions reported by SIET (for GE) regarding the thermal-hydraulic testing phase of the PANTHERS-lC program. dowever, the staff has a few comments on this phase of testing, and has additional comments regarding the structural phase of testing, which was not completed when testing was suspended in November 1995.
The staff agrees that substantial data was acquired on the thermal-hydraulic performance of the prototype IC host exchanger, and fudher, that the data support a conclusion that the component, as tested, is capable of meeting its design performance requirements. The quality of the data appears to be adequate for use as a validation database for the TRACG computer code. The data cover a range of conditions for normal IC operation that appears to be sufficient to bound this mode of IC performance. The staff believes that the non-condensible gas injection / purge tests and the testing with varying water level serve to add breadth to the database, and further confirm the operational capabilities of the prototype heat exchanger. While considerable testing was accomplished over a range of accident conditions, the staff is unable to draw conclusions about the component performance or the adequacy of the database due to the failure to complete the full test matrix.
The staff does not fully agree with the conclusion that there is no water hammer in the tested IC
( operational modes. As the report states, initial tests did indicate water hammer, which was subsequently eliminated by modifying facility operation. It is not clear that the range of conditions that could occur in an SBWR can be anticipated at this time, and the possibility of water hammer cannot be ruled out. The staff agrees with the apparent conclusion that tne suspension of the testing program makes it premature to attempt to infer information about water hammer during
! accident conditions. The staff expects that, if this program were resumed in the future, the test matrix would be completed, and a more extensive evaluation of the likelihood of potential water hammer loads would be made. The final design of the IC system components should take these potentiallouis into account.
Although the structural testing performed in the PANTHERS-lC does not bear directly on the results of tine thermal-hydraulic testing that was performed prior to the suspension of the test program, the staff is, nevertheless, concemed about the occurrence of leakage during this program. Originally, both the PANTHERS-PCCS and PANTHERS-IC programs were to include substantial structural performance (thermal cycling) testing. In both cases, leaks occurred in the heat exchanger headers relatively early in the testing. Since the SBWR design places the heat exchanger pools outside the containment, any leakage of primary effluent from these heat exchangers represents a potential containment bypass. GE recognized this fact in establishing a "zero leakage" performance requirement for the heat exchangers, which has yet to be demonstrated. The staff encourages GE to continue to evaluate the heat exchangers to determira the reason for the leakage in both the PCCS and IC tests. If significant modifications to the heat exchanger designs are necessary to eliminate the problem, it may be necessary for the staff to evaluate the need for further testing of new prototype components. In any event, the staff would require that the original IC test program be completed.
1 Enclosure