ML20059F903

Requests Addl Info Needed to Support Review Re Sbwr Design Certification.Response Requested within 120 Days of Ltr Date

ML20059F903
Person / Time
Site:	05200004
Issue date:	09/15/1993
From:	Malloy M Office of Nuclear Reactor Regulation
To:	Marriott P GENERAL ELECTRIC CO.
References
NUDOCS 9311050151
Download: ML20059F903 (11)

Text

_ _ _

= . Cerkal All 8

[~% o

~ ,%

UNITED STATES NUCLEAR REGULATORY COMMISSION 7

WASHINGTON, D. C. 20555

{ :i ,

September 15, 1993

%.....,/

Docket No.52-004 Mr. Patrick W. Marriott, Manager Licensing & Consulting Services GE Nuclear Energy 175 Curtner Avenue San Jose, California 95125

SUBJECT:

REQUEST FOR ADDITIONAL INFORMATION (RAI) REGARDING THE SIMPLIFIED BOILING WATER REACTOR (SBWR) DESIGN The staff has determined that it needs additional information to support its review activities related to the SBWR design certification. Some additional information on the GIRAFFE (Q900.2-Q900.26), PANTHERS (Q900.27-Q900.41), PANDA (Q900.42-900.45), and University Condensation (Q900.46-Q900.47) testing programs is needed. Please provide a written response to the enclosed ques-tions within 120 days of the date of this letter. In order to expedite.the l NRC review, however, we would like to meet with your staff and consultants, as appropriate, as soon as mutually convenient to' discuss your proposed responses to the questions.

The questions on the GIRAFFE tests were based on the draft final test report that GE Nuclear Energy (GE) transmitted to the staff on July 1,1993. The PANTHERS test program questions are based on review of the PANTHERS Test Specification, GE document number 23A6999, Revision 1, dated September 29, 1992. The questions on PANDA are general in nature, since the staff has not received sufficient information to date on this test program in order to  ;

perform a detailed review. We propose that the questions on these three programs would form the agenda for the requested meeting. The enclosed  !

! questions related to the University Condensation tests reiterate the staff's i

requests for information during the meeting with GE on June 10, 1993, (GE has not yet provided the requested information). l You have previously requested that portions of the information submitted in the August 1992, application for design certification of the SBWR plant, as  !

supplemented in February 1993, be exempt from mandatory public disclosure.

The staff has not completed its review of your request in accordance-with the requirements of 10 CFR 2.790, therefore, that portion of the submitted information is being withheld from public disclosure pending the staff's final determination. The staff concludes that this RAI information does not contain those portions of the information for which you are seeking exemption.

• The numbers in parentheses designate the tracking numbers assigned to the questions.

9311050151 93091S- 1  !

?" ^"

• • W NRC HLE CalTBt COPY gI

i l

Mr. Patrick W. Marriott September 15, 1993 I'

However, the staff will withhold this letter from public disclosure for 30 calendar days from the date of this letter to allow GE the opportunity to verify the staff's conclusions. If, after that time, you do not request that :

all or portions of the information in the enclosure be withheld from public i disclosure in accordance with 10 CFR 2.790, this letter will be placed in the  ;

NRC's Public Document Room. ,

This RAI affects nine or fewer respondents, and therefore, is not subject to l review by the Office of Management and Budget under P.L.96-511.

If you have any questions regarding this matter, please contact me at (301) 4 504-1178 or Mr. Son Ninh at (301) 504-1125.

T9 'Eand 9v- ,

Melinda Malloy, Project Manager i Standardization Project Directorate Associate Directorate for Advanced Reactors  ;

and License Renewal l Office of Nuclear Reactor Regulation l I

Enclosure.

l RAI on the SBWR Design  :

I cc w/ enclosure: 1 See next page Distribution (w/ enclosure):

• Central File PDST R/F MMalloy

• PDR SNinh DMcPherson, BE2 RElliott, BH2 Alevin, 8E21 RCaruso, 8E1 TLee, NLN353 JHan, NLN353 YChen, NLN353 CTinkler, NLN344 ANotafrancesco, N.6344 MFinkelstein, 15B18 ATGody, Jr., 17G21 JMonninger, 8H2 MSnodderly, 8H2 Shou, 7H15 SAli, 7H15 Slee, 7H15 JKudrick, 8D1 Distribution (w/o enclosure):

DCrutchfield/WTravers RBorchardt JNWilson PShea TMurley/FMiraglia,12G18 WRussell,12G18 AThadani/MVirgilio, BE2 RBarrett, 8H2 RJones, 8E21 BSheron/TKing, NLS007 FEltawila, NLN344 JMoore, 15B18 ACRS (11) GSuh (12), 12E4 GBagchi, 7H15

• To be held for 30 days / m f 0FC LA:PDST PM:PDST - PM:PDS[ SCS k SCMT/

NAME PShea 3 d b -MMibyItz SNid JKud M k JNWilIon DATE 09/9/93 09/\h/93 M//93\ )

094\/93 4

b 09/l /93 DOCUMENT NAME: SBWR9320.MM

j i

Mr. Patrick W. Marriott Docket No.52-004 l General Electric Company l cc: Mr. Laurence S. Gifford GE Nuclear Energy 12300 Twinbrook Parkway Suite 315 Rockville, Maryland 20852 i Director, Criteria & Standards Division Office of Radiation Programs U.S. Environmental Protection Agency 401 M Street, S.W.

Washington, D.C. 20460 -

Mr. Sterling Franks U.S. Department of Energy '

NE-42 Washington, D.C. 20585 ]

Mr. Jeffrey C. Baechler GE Nuclear Energy 175 Curtner Avenue, MC-782 San Jose, California 95125 -

Mr. Frank A. Ross l Program Manager, ALWR Office of LWR Safety & Technology ,

U.S. Department of Energy  :

NE-42 l 19901 Germantown Road 1

. Germantown, Maryland 20874 i l

l l

i 4

-, , ._w, - . ..,. l

l REQUEST FOR ADDITIONAL INFORMATION (RAI) ON THE SIMPLIFIED BOILIC WATER REACTOR (SBWR) DESIGN GIRAFFE Test Program 1

Questions 900-18 through 900-26 below are on the overall test program. The remainder of the following questions are based on review of the GIRAFFE draft final test report submitted by GE Nuclear Energy (GE) to the staff on July 1,1993. j l

900.2 Page 3-6. What is meant in the third paragraph by "...Since the 1 flow rate through the new GDCS-to-drywell line was not scaled, the I water levels in the drywell and GDCS pool prior to start of a test were adjusted to maintain correct total water inventory in the GDCS pools and drywell? How was the " correct total water inventory" determined? Why was an orifice not placed in the line to scale the ,

flow appropriately? l 900.3 Pages 4-1 and 4-2. Why were the loss-of-coolant accident (LOCA) tests performed with initial conditions starting at I hour? Why was l the test of the entire transient not run with venting to the sup-pression pool? How can the TRACG model be validated without first-hour data from several different breaks?

900.4 Page 4-3. Does varying the passive containment cooling system (PCCS) initial pool temperature have any effect on the heat transfer capability of the PCC condenser? For instance, does initial PCCS <

temperature affect the magnitude of the peak pressure during a i design-basis accident (DBA), the timing of the peak pressure? l 900.5 Page 4-3. Why was nitrogen injected into the PCC condenser to

. stabilize pressure? Why was there intervention in the test? ,

900.6 Page 4-4. What kind of measurements were being taken on the Drywell I Nitrogen tests? What was the test's purpose? Why was there ini-tially steam in the drywell and then nitrogen was introduced? Would not the opposite occur in the event of an accident? Was this intended to represent opening of the vacuum breakers? Was the drywell sufficiently instrumented to accurately determine the nitrogen distribution in the drywell? Can gradients be accurately simulated when the internals of the drywell have not been completely modeled?

900.7 Page 4-5. For the Phase 2 tests, the gravity-driven cooling system (GDCS) gas space to drywell line was closed off. On Page 3-3, the report indicates that the two spaces are connected by a "large diameter line to equalize pressures in a similar manner to the SBWR containment." What is the function of this line and why was it closed off for the Phase 2 tests? [ Figure 21.6.3-2 of the standard safety analysis report (SSAR) seems to indicate that the GDCS gas space is open to the drywell.]

Enclosure

RAI ON THE SBWR DESIGN 900.8 Pages 4-6 and 4-7. It is not readily apparent from the LOCA simula-tion descriptions on these pages how the facility simulated the accident scenarios tested. Provide a detailed explanation (proce-dure) as to how each simulation was performed. Also provide a discussion addressing how the scenarios were scaled (GDCS initial level,etc.).

900.9 Pages 5-1 and 5-6. What was the cause of the net pressure rise from the PCC inlet through the water box to the PCC drain line exit?

Since steam is being condensed during this time, should not pressure have been dropping? How could there be flow in the correct direc-tion if pressure is increasing?

900.10 Page 5-4 and Figures 5.2-1 and 5.2-2. Explain the oscillatory patterns in the graphed data. What causes pressure and water level to oscillate so consistently in the Phase 1, Step 2 experiments?

(Trends in these parameters are well defined, but the oscillatory .

characteristic on the graph is not explained. Temperature oscilla-t bns are explained due to nitrogen and steam mixing, but this does not explain pressure and water level oscillations.)

900.11 Page 5-15. It would be anticipated that the tube temperature would-drop as distance from the steam box increases due to the cooling and condensing of the steam. From Figure 5.1-3, it is evident that this is not the case. Provide an explanation. On Page 6-2, the report states that the cause "has not yet been explained." In order to- '

predict tube heat transfer, it seems necessary to understand this phenomena. Can the TRACG code predict this phenomena? )

l

. 900.12 Pages 4-10 and 6-21. All tests were performed twice to verify repeatability of results. Were there any significant differences in the data obtained between the two test runs on any of the experi- .

ments? Page 6-21 indicates that certain tests were run under I slightly different conditions to determine the sensitivity to certain parameters. What were the results of those tests? What insights were gained? How do TRACG sensitivity study results compare with the experimental sensitivity studies?

900.13 Page 6-6. Due to the unexpected drywell and suppression chamber pressure characteristics experienced in the tests to determine the effect of vent line submergence depth, are three data points suffi-  !

cient to develop a conclusion? The explanation provided in the I report seems waak, with insufficient data to conclusively support the explanatiu. If more data points are needed, what is GE's plan to obtain the data? If GE feels that no further data is needed, ,

explain why. '

l

l

. i RAI ON THE SBWR DESIGR c'10.14 Page 6-7. It was noted that while Figure 5.2-2 shows that different vent line submergences cause different pressure transients in the drywell, Figure 5.2-6 implies that the temperature transients are very similar in trend and timing, primarily differing in magnitude. ,

Provide an explanation.

900.15 Page 5-28. Where is the location of the temperature sensors that are shown by the two top lines on the temperature graphs in Fig-ure 5.2-67 900.16 Page 6-13 and Figures 5.5-1 and 5.5-2. The explanation for the main steamline break transient on Page 6-13 implies that temperature and pressure were level or decreasing up until GDCS flow stopped at approximately 10,000 seconds. At 10,000 seconds, the report states that there was a "small increase in the drywell pressure." Accord-ing to Figures 5.5-1 and 5.5-2, pressure and, for the most part, temperature were also rising steadily during the entire first 10,000 seconds of the test. Clarify what was actually occurring with drywell pressure during the first 10,000 seconds. Why was pressure rising with GDCS inje. tion still occurring? What happened at 10,000 seconds that caused pressure to turn the corner?

900.17 Page 6-14. Explain the occurrence of the spillage from the reactor-pressure vessel through the depressurization valve line. It is not clear from the report whether or not the spill over was anticipated or not. Is this a characteristic of the test platform or would this-be anticipated to occur in a full-size SBWR reactor?

900.18 How has the tester ensured that the external heat added to account for system losses to the environment has not affected the flow pattern in the test facility?

900.19 Have the results of the GIRAFFE tests been compared with the results from the University Condensation test and GIST [ gravity-driven cooling systems test] programs? What are the key findings? How i have the findings been incorporated into the philosophy of the l future tests at PANTHERS and PANDA? '

900.20 Discuss any inadequacies in the test instrumentation. Were there any problems with the instrumentation? Are there any lessons learned from the GTRAFFE tests regarding instrumentation that will be applied to future test facilities (e.g., PANTHERS and PANDA)?

Could the test instrumentation have been improved? If so, how? l Were there any pieces of data that would have been desirable to have  !

that were not accessible due to lack of appropriate instrumentation?

_j_

RAI ON THE SBWR DESIGN 900.21 Explain the deduction that a decline in temperature indicates a pocket of nitrogen. Is there an experimental basis for this deduc-tion other than GIRAFFE? Throughout the accident, would not it have ,

been anticipated that a lot of mixing would take place with the '

temperature of the nitrogen rising as it absorbs energy from the steam, especially since a large LOCA would create a very turbulent )

atmosphere inside the containment? Is there any other method of validating that this assumption is leading to the correct  ;

conclusions? .

900.22 How well did the geometry of the test components such as the PCCS l heat exchanger reflect the geometry of the actual design? For i instance, how well did the test platform represent the steam box, the water box, and the PCCS piping configuration? Provide a de-tailed drawing of the PCCS arrangement on the test platform.

900.23 How does the change in PCCS design since the construction of the test facility affect the test platform's ability to accurately model PCCS performance?

900.24 Why was the testing not performed using a lighter-than-air noncon- l densable gas, such as helium, to represent hydrogen? Does GE have a clear understanding of how the presence of lighter-than-air noncon-densable gases would affect the accident scenarios? If so, on what basis? [The staff recognizes that the University Condensation tests

,,rovided a good understanding of the heat transfer degradation due .

to the presence of hydrogen (using helium as a substitute), but the -!

test program, so far, has not demonstrated with an integral test how ,

the systems are affected operationally. The University Condensation '

tests did not accurately simulate the whole PCCS heat exchanger (HX), only the HX tube. It is conceivable that stratification of noncondensables in low-velocity regions, such as the steam box on the HX, could inhibit flow somewhat due to the build up of hydrogen. i In the GIRAFFE, experiments, after the initial venting, it appears that nitrogen tended to collect low in the drywell. From there, it did not really affect the PCCS heat transfer too much. Hydrogen, on the other hand, would tend to migrate towards the PCCS, and could conceivably, pass through the PCCS several times. PANTHERS will provide data on a separate-effects basis, but the staff still feels that an integral test with lighter-than-air noncondensables still needs to be performed.] )

i 900.25 Explain the GE philosophy on severe accident testing. None of the. j conditions tested at the GIRAFFE facility represented severe acci- i dent conditions (e.g., no lighter-than-air noncondensables in the i test). What would be the ramifications of modifying the facility to perform severe accident conditions, such as testing with helium to represent hydrogen? Would the test facility be available to perform ,

these tests? If so, when? What kind of testing schedule would be l required? i i

1 l

. RAI ON THE SBWR DESIGN .

900.26 A major objective of the GE test program is to validate the TRACG  :

code. How well did TRACG predict the test results in the pre-test 1 analysis? What adjustments were made to the code, or the method of.  ;

using the code, due to the test results? Provide some comparisons  ;

of test data with pre-test predictions. Were any of the tests 1 performed " blind," that is, with the analyst being given the actual test initial conditions and being asked to predict the. test outcome?  !

If so, what were the results? If not, what consideration has been '

given to doing so on future tests?

-l PANTHERS Test Program Questions 900-40 and 900-41 below relate to the PCCS test matrices in general and to the overall test program, respectively. The other following questiins are based on review of the PANTHERS Test Specification, GE document number-23A6999, Revision 1, dated September 29, 1992. ,

i 900.27 Page 9, paragraph 4.1.3.1. The test platform for the PCC is de-signed to provide helium at a sufficient rate to fill the PCC in ,

15 minutes. Does this provide the test platform with sufficient '

capacity to bound DBA conditions for hydrogen generation or severe accident conditions for hydrogen generation? (The staff recognizes that this is a system sizing criteria only, not a matrix test condition.)

900.28 Page 57. Does the helium flow rate to be tested bound both DBA and  ;

severe accident conditions? Provide a revised test matrix. (The i test matrix previously provided to the staff offers no information '

on the helium concentrations to be tested. The staff notes that the i PCCS should be tested with sufficient helium to demonstrate PCCS performance in a severe accident environment, as well as DBA >

conditions.)  !

900.29 Page 33, paragraph 5.1.1.1. The specification lists the specific thermal-hydraulic test objectives. One of the objectives listed under Item C is to "... Confirm that heat transfer and flow rates are stable and without large fluctuations." GIRAFFE test data has shown that under actual accident conditions, the heat transfer rates would not be " stable," rather they would be oscillatory due to the pres- j ence of noncondensable gases cycling through the PCCS and vacuum  !

breakers. The PANTHERS test seems to be built around setting up a l constant flow rate at a constant steam /noncondensable gas ratio. .

While these tests are certainly important to confirm heat transfer l performance, they do not demonstrate PCCS performance under antici-  ;

pated containment conditions or improve the TRACG code's ability to model post-accident PCCS performance. Explain why these tests should not be combined with transient tests to provide a complete (continued on next page) i

RAI ON THE SBWR DESIGN picture on PCCS performance. (It is the staff's opinion that GE should provide some means of simulating the containment under accident conditions and to provide flows which are more representa-tive of varying steam / gas ratios, which would be present in an actual containment under post-accident conditions. This would provide better data on anticipated PCCS performance.)

900.30 Provide a copy of the TRACG input deck on diskette for the staff's use in verifying the RELAPS and CONTAIN codes (which will be used for independent code verification in the staff's review of the SSAR).

900.31 Page 7, paragraph 3.6. Provide the complete Test Plan and Proce-dures Document (TP&P) prepared in accordance with this requirement.

[The one previously provided to the staff was incomplete (missing a large number of sections, figures and tables.) GE has not yet provided detailed test procedures to the staff.)  ;

l 900.32 Page 12, paragraph 4.1.4.2. What is the basis for selecting the  !

tubes in which the temperature sensors are to located? i 900.33 Provide a detailed description of how the condensate and nonconden-sables are being separated. From the schematic shown on Page 13, it appears that the noncondensables are vented from the top of the water box. Is this the case? I 900.34 Pages 10 and 11, paragraphs 4.1.3.3 and 4.1.3.4. The specification requires that the vent tank be arranged such that the test can be conducted with the vent line either submerged or not submerged. The condensate drain tank must have the ability to test with varying water levels. The test objectives sections on Pages 5 and 33 do not clearly specify any test objectives which would require these characteristics. Are there any test objectives that are not speci-fied on Pages 5 and 337 If so, what are they? What are the objec-tives associated with the special drain and vent tank capabilities cited above? Provide revised test matrices that include data regarding what water levels will be maintained in the condensate tank and what the vent line submergence will be and define vent tank and condensate tank pressures and temperatures. Inlet steam pres-sures in the matrices are given as very broad ranges. Provide more specific test conditions. The staff recognizes that some latitude is needed in the range of pressures, but the ranges in the test matrices are excessive. How can any pre-test predictions be per-formed?

900.35 What pre-test analyses have been performed using the TRACG code for the PANTHERS tests? What were the results?

I I

l RAI ON THE SBWR DESIGN 900.36 Page 39. What will the initial pool temperature be for the pool level tests? What is the preferred method for dropping the water  !

level? Two options are given. The first option is to let the water I boil off and the second is to drop the level by draining. The first option is more representative for demonstrating the actual heat exchanger performance expected.

900.37 Page 40. How can a LOCA cycle simulation be performed in 30 min- l utes? The cooldown/depressurization period from a LOCA is substan-tially longer than this. The PCCS may not experience peak pressures for a very long time, but the containment will be pressurized above 300 kpa (40 psi) five hours into the accident with saturated steam still being cycled through the PCCS. If the objective is to simu-late 5 times the life cycle LOCA pressure / temperature cycles, should not the whole LOCA cycle be simulated? Also, provide a discussion as to how the 5 times the expected number of lifetime cycles was selected for determining the numi,er of test runs.

900.38 Page 41. Pressurizing the PCCS for 1-2 minutes does not seem representative of the length of the pressurization cycles that the containment will experience throughout its life cycle. Typically, the containment will be pressurized during a Type A test for any-where from 6 to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Explain how 1-2 minutes was selected. In j addition, current requirements for Type A testing is for 3 tests to j be performed per 10-year cycle. This would require 18 tests'per i life cycle. Even on an accelerated schedule where there would be -l testing during every outage, the maximum number of tests would be on the order of 40. Explain why 60 tests were chosen as a baseline for determining the number of test runs. Also, explain the selection of '

110 psig as a test pressure.

l 900.39 Page 44. Does the helium flow rate specified for the test with low density noncondensable gases bound maximum DBA and severe accident ,

conditions with regard to hydrogen concentration in the containmentf l Is it anticipated that conditions might be such that both hydrogen  !

and nitrogen could be mixed with the steam? If so, why is the test  ;

not arranged so as to provide test conditions reflective of antici- j pated actual conditions?

900.40 Are the steam flow rates chosen in the PCCS test matrices bounding for all anticipated DBA conditions? For severe accident conditions? ,

4 900.41 Explain how once-through testing of the PCCS can be representative of the operational behavior of the PCCS. What information will be provided by this test platform that is different than the previous tests (GIRAFFE and University Condensation tests)?

4

RAI ON THE SBWR DESIGN PANDA Test Program 900.42 Provide test program specifications, procedures, matrices (including objectives,. test parameters, conditions, mass and energy release profiles, etc. for each test planned), the scaling analysis, and an instrumentation description (layout and characterizations, e.g.,

steady state and transients measurements).

900.43 Provide a detailed presentation on the PANDA test program. The presentation should include, but not be limited to, a description of the facility (with schematic drawings), test scenarios planned, scaling, instrumentation, the latest test matrices, a schedule update, and a description of any pre-test analyses (either completed already or planned in the future).

900.44 Provide the results of the pre-test analyses (once completed).

900.45 Provide the staff with copies of the TRACG input deck on diskette, a noding diagram, and input description for use in the performance of confirmatory calculations.

University Condensation Tests 900.46 Provide the test matrices GE agreed to transmit during the June 10, 1993, meeting in San Jose. (The matrices should be complete for all, tests run at both university test sites.)

900.47 Provide the final test report for staff review as soon as it becomes available. (The staff's current understanding is that GE intends to forward it to the staff in October 1993.)