ML20248E199

Informs That Staff Has Identified Six Ltrs (Enclosed) to GE Re Potential Proprietary Disclosure.Ltrs Specifically Requested That GE Review Encl to Determine If Contained GE Proprietary Info.Encl Withheld

ML20248E199
Person / Time
Site:	05200004
Issue date:	05/28/1998
From:	Scaletti D NRC (Affiliation Not Assigned)
To:	Sawyer C GENERAL ELECTRIC CO.
Shared Package
ML20013J847	List: ML20248E199
References
NUDOCS 9806030212
Download: ML20248E199 (38)

Text

_ __ _ - -

May 28, 1998 Dr. Craig D. Sawyer, Manager Advanced Reactor Programs GE Nuclear Energy 175 Curtner Avenue, M/C 754 San Jose, CA 95125

SUBJECT:

SIMPLIFIED BOILING WATER REACTOR PROPRIETARY CORRESPONDENCE i

Dear Dr. Sawyer:

The staff has identified six letters (enclosed) to GE regarding potential proprietary disclosure. The letters specifically r6 quested GE to review the enclosures to the letters to determine if they contained GE proprietary information. The staff does not have a record of any GE's responses; therefore, we request that you review the enclosed letters in order to clear them to be made available to the public, if you have any questions regarding this matter, please contact me at (301) 415-1104.

Sincerely, original signed by:

Dino Scaletti, Project Manager Standardization Project Directorate Division of Reactor Program Management Office of Nuclear Reactor Regulation Docket No.52-004

Enclosure:

As stated cc w/o encl: See next page DISTRIBUTION:

Docket File PDST R/F TQuay w/o encl .

PUBLIC DScaletti JNWilson w/o encl.

ACRS (11) w/o enci. JMoore,0-15 B18 DOCUMENT NAME: A:\ PROP.DCS To reesive a copy of this document, indicate in the box: "C" = Copy without attac'iment/ enclosure "E" = Copy with attachment / enclosure "N" = No copy OFFICE PM:PDST:08PM l D:PDOT:DRPM l l L l

NAME DCScaletti V)W TRQuay -q%4 DATE j/15498 & 9 /2f/98 OFFICIAL RECORD COPY t

TXDb  !

_N k gJ =

9'806030212 980528 i PDR ADOCK 05200004 e A PDR i

Dr. Craig D. Sawyer Docket No.52-004 GE Nuclear Energy

( cc w/o enclosure:

Mr. Rob Wallace Mr. Brian McIntyre i

GE Nuclear Energy Westinghouse Electric Corporation 1299 Pennsylvania Avenue, N.W. Energy Systems Business Unit Suite 1100 Box 355 l Washington, DC 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. Department of Energy NE-42 .

Washington, DC 20585 I-l Mr. Robert H. Buchholz GE Nuclear Energy 175 Curtner Avenue, MC-781 l

San Jose, CA 95125 Mr. Steven A. Hucik GE Nuclear Energy i 175 Curtner Avenue, MC-780 l San Jose, CA 95125 Mr. Tom J. Mulford, Manager SBWR Design Certification Electric Power Research Institute 3412 Hillview Avenue Palo Alto, CA 94304-1395 4

4 9 Mr. James E. Quinn, Projects Manager LMR and SBWR Programs GE Nuclear Energy 175 Curtner Avenue, M/C 165 San Jose, California 95125

SUBJECT:

STAFF EVALUATION OF uENERAL ELECTRIC'S (GE's) TEST AND ANALYSIS PROGRAM DESCRIPTION, NEDC-32391, REVISION C

Dear Mr. Quinn:

In response to your letter dated March 4, 1996, the staff has prepared the enclosed report on its evaluation of the GE's Simplified Boiling Water Reactor (SBWR) Test and Analysis Program Description (TAPD).

Overall, the staff notes that GE has made significant progress in addressing previous issues and questions identified by the staff in the Draft Safety Evaluation Report (DSER) and the requests for additional information (RAls).

The staff concludes that, with the exception of the PAR and PIRT issues TAPD Revision C can be accepted as a framework for the SBWR testing program a,nd the TRACG qualification process if it fully implemented as described. However, a final approval of the adequacy of the test program for qualification of the TRACG code and for design certification of the SBWR is not possible without completing topical reports, a detailed and GE's review of thethereof.

analysis test data, scaling report, TRACG licensing 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.

Sincerely, Theodore R. Quay, Director Standardization Project Directorate Division of Reactor Program Management Office of Nuclear Reactor Regulation Docket No.52-004

Enclosure:

Introduction and Background

's c Mr. James E. Quinn Docket No.52-004 j GE Nuclear Energy I

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 I Mr. Steven A. Hucik I GE Nuclear Energy Mr. Rob Wallace 175 Curtner Avenue, MC-780 GE Nuclear Energy San Jose, CA 95125 1299 Pennsylvania Avenue, N.W.

Suite 1100 i 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 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  ;

i i

l

INTRODUCTION AND BACKGROUND In 1992, General Electric Nuclear Energy (GE) submitted to the NRC an applica-tion for design certification of the Simplified Boiling Water Reactor (SBWR).

L The SBWR is a " passive" plant design, in that operation of safety systems does not require " active," ac-powered components. To support design certification, GE developed a Design Certification Testing Program, to satisfy the require-

.ments of.10 CFR 52.47(b)(2)(1)(A), which states that for a plant that

" utilizes simplified, passive, or other innovative means to accomplish its safety functions," the applicant must demonstrate that:

1. The performance of each safety feature of the design has been demonstrated through either analysis, appropriate test programs, experience, or a combination thereof;

2. Interdependent effects among the safety features of the design have been

.found acceptable by analysis, appropriate test programs, experience, or a combination thereof;

3. Sufficient data exist on the safety features of the design to assess the analytical tools used for safety analyses over a sufficient range of normal operating conditions, transient conditions, and specified accident sequences, including equilibrium core conditions.  !

The NRC staff began its review of GE's design certification test program in 1991, prior to GE's formal design certification application. In October 1992, the staff issued its preliminary review of the test program in SECY-92-339,

" Evaluation of the General Electric Company's (GE's) Test Program to Support Design Certification for the Simplified Boiling Water Reactor (SBWR)." In this document, the staff indicated that it had several concerns regarding the proposed test program that needed to be resolved. These concerns involved such issues as the design of test facilities, scope and range of test matri-ces, and GE's classification of certain programs as " confirmatory" rather than required for design certification.

Between 1992 and 1994, the staff and GE met several times to attempt to resolve the issues discussed in SECY-92-339. In addition, the staff continued its detailed review of GE's' test programs, both those completed prior to the submission of the application and those The detailed test review raised additional concerns. planned In Marchfor1994, the future.

by letter from Dennis M. Crutchfield, Associate Director for Advanced Reactors and License Renewal, NRR, to Patrick W. Marriott, Manager, Advanced Plant Technologies, GE, the staff detailed testing-related issues, both those remaining from SECY-92-339 and new concerns, that it believed needed to be resolved to permit the design certification process to continue, j' On April 1, 1994, in response to the staff's letter, GE committed to perform a

, reassessment of the testing and analysis programs for the SBWR, and to report the conclusions of that reassessment to the staff. The outcome of GE's i

Enclosure

reassessment, entitled "SBWR Test and Analysis Program Description,"

NEDC-32391, hereinafter referred to as TAPD, was submitted to the staff on

. August 10, 1994. The staff issued a draft safety evaluation report (DSER) on the initial version (Revision A) of TAPD to GE in November 1994. In that review, the staff reached the conclusion that, while the document provided a good framework for the assessment of testing and analysis requirements for the SBWR, additional information was required on several. aspects of the material in the report, and that additional testing in two major areas was required to support SBWR design certification: containment-related integral-effects testing with lighter-than-steam non-condensible gases; and integral systems testing covering the late blewdown and early ECCS injection phases of SBWR design-basis accidents. The staff said that, if the required modifications to the test program were made, and if the additional information requested was provided, satisfaction of the requirements of 10 CFR 52.47(b)(2)(1)(A) was  !

feasible.

]

The staff met with GE in December 1994 and January 1995. In addition, GE presented the TAPD to the Advisory Committee on Reactor Safeguards (ACRS) in j January 1995. As a result of these meetings, a number of action items were  !

identified. GE committed to address these items in subsequent revisions of  !

the TAPD. GE informed the staff that this would be a "two-stage" process, j with Revision B of the TAPD addressing about half of the action items, and l Revision C addressing the remainder. In response to the staff's conclusion i regarding additional testing, GE also committed to perform two series of tests in the GIRAFFE facility at Toshiba's Nuclear Engineering Laboratory, in Kawasaki, Japan: an "H" series, using helium to simulate hydrogen behavior in the containment; and a series of " system interaction tests," denoted GIRAFFE / SIT, to examine the late blowdown /early ECCS injection phases of SBWR LOCAs.

TAPD, Revision B, was submitted to the NRC in April 1995. While a number of the action items were resolved, the staff found that there were still many unresolved issues and questions that GE needed to address. The staff issued Requests for Additional Information (RAls) 900.102 - 900.181, concerning TAPD, Revision B, in June 1995. GE agreed to address these RAls in Revision C of the TAPD.

TAPD, Revision C, was submitted to the NRC in Augus't 1995. It comprised a  ;

substantial revision o. the previous versions of the report, and included a supplement describing in detail the process of developing Phenomena Identifi-cation and Ranking Tables (PIRTs) for the SBWR. In addition, a separate {

report, entitled " Scaling of the SBWR Related Tests," NEDC-32288, Revision 1, 1 was issued in October 1995 to supplement the TAPD in the area with regard to I the scaling approaches for the various SBWR test facilities. )

In March 1996, during the staff's review of TAPD, Revision C, and the associ-  !

ated scaling report, GE announced that it was withdrawing the SBWR from the l NRC's design certification program. GE committed to complete the " technology i phase" of the SBWR program, which comprises the test programs and associated I reports, plus a very limited description of SBWR modeling using the TRACG l L.'..

i .

computer code. GE has completed all of its planned testing activities and has

= issued all test reports with the exception of the PANDA test reports. The NRC staff agreed to complete its review of the TAPD, Revision C and also to provide feedback on the scaling report.

Summary of Staff's~ Evaluation of the SBWR TAPD Revision C Report and Ma.ior Conclusions Evaluation of the SBWR TAPD Revision C Related to Reactor Systems Area This assessment of the SBWR test and analysis program covers only the TAPD descriptive material (primarily PIRT development) and the test matrices for relevant programs. Detailed evaluation of the test data has not been per-formed, and no conclusions can be drawn about the final acceptability of the test programs to provide data for code validation, to satisfy the requirements of 10 CFR 52.47 (b)(2)(1)(A).

The staff has reviewed TAPD, Revision C, and has determined that GE has essentially fulfilled its commitments to address (1) the " actions items" from the DSER, except as noted below, and (2) non-scaling-related RAls from its review of TAPD, Revision B.

In the DSER, the staff stated, "The staff requires considerably more informa-tion than is available in the TAPD repot t on the details of the code qualifi-cation program for TRACG. Neither the TAPD report nor the code qualification documentation for TRACG that GE has submitted...provides sufficient informa-tion on code models and correlations and their applicability over the range of SBWR thermal-hydraulic conditions, nor has the staff been able to determine from these documents how the test data will be used to quantify uncertainties and biases in the analyses, especially for LOCAs." In February 1996, GE submitted the TRACG Model Licensing Topical Report (LTR), Revision 1 to the NRC for review. However, on March 13, 1996, GE requested that staff suspended the review of the TRACG Qualification LTR Revision 1 and the TRACG Application Revision 0, and also requested that staff provide written summary of status of the review of these reports. The staff has committed to issue status reports on-the reviews of the Application and Qualification LTRs, and also to issue a letter documenting an acceptance review of the Model LTR (letter, Crutchfield to Quinn, " Status of the Simplified Boiling Water Reactor (SBWR) Test Programs and TRACG Review," Apr.; 12,1,96).

The staff concludes that TAPD, Revision C represents a substantial improvement over Revisions A and B, and that most of the staff's non-scaling-related comments and questions have been satisfactorily addressed. The remaining outstanding issues related to TAPD Revision C concern the PIRT, and are addressed in more detail below. Most significantly, the staff notes that GE has complied with its commitment to perform late blowdown /early ECCS integral systems testing in the GIRAFFE facility; the GIRAFFE / SIT series of experiments l was' completed in October 1995. -The staff reviewed GE's test matrix for l GIRAFFE / SIT, and determined that it likely would address the technical issues for which the testing was required.

l

-- -- y I

GE also revised the test matrix for the PANTHERS /IC test program, performed at SIET Laboratories in Piacenza, Italy,'in response to staff comments in the TAPD, Revision A, review. The staff reviewed the revised test matrix, and determined that.the changes satisfactorily addressed the staff's concerns.

GE.has addressed most of the staff's comments and questions regarding discus-sion of the PIRT'as presented in Revisions A and B of the TAPD. The staff notes that, the PIRT process, as documented in the TAPD and the PIRT supple-ment, represents a commendable effort that addresses in detail the thermal-hydraulic phenomena relevant to the SBWR in the context of safety analysis.

However, the staff still has several comments and questions related to the PIRT in TAPD, Revision C, many of which concern consistency in phenomena identification and ranking it different sections and/or tables in the TAPD, or the rationale for choosing certain rankings during various phases of SBWR accidents. Additional comments relate primarily to coverage of specific phenomena in data from test facilities and programs outside of those performed as part SBWR design certification testing program. These comments and questions.are listed in Appendix A to this report.

In the DSER, the staff indicated that it would also review the implementation

.of quality assurance (QA) in'the conduct of the test programs, to determine if GE and its partners in the SBWR test program fulfilled GE's commitment to meet NQA-I requirements for SBWR design certification testing activities. -The staff has conducted QA inspections'of all of GE's major SBWR design certifica-tion test programs (GIST, PANTHERS /PCCS, PANTHERS /IC, GIRAFFE, and PANDA), and .

has concluded that, for GIST,~ PANTHERS, and GIRAFFE, NQA-1 requirements were met, or that appropriate remedial actions were taken to correct deficiencies found during those inspections. The PANDA QA inspection was conducted in March 1996,.and two non-conformances were reported to GE as a result of that review. The staff therefore requires that GE implement corrective actions to close the deficiencies identified during the PANDA QA inspection.

Summary and Conclusions When the TAPD was originally submitted to the NRC, GE posed three major questions to.the staff for review:

'3. Is the test program adequate for qualification of the TRACG code?

2. Is.the test and analysis program adequate for design certification of the SBWR7

3. Is construction of a new integral test facility required for additional SBWR testing?

The staff's original evaluation stated that, if GE provided the required additional information on TRACG and modified the test programs per staff recommendations, accomplishment of the first two objectives was " feasible."

l i

l l The staff further determined that the required additional testing could likely be accomplished in an existing SBWR test facility (i.e., GIRAFFE), so that j construction of a new facility was not required.

The conclusions reached in the current evaluation of the revised TAPD are consistent with those cited above. 'Since the staff will not have the opportu-l nity to review detailed information regarding TRACG qualification using data from the SBWR test program, a final conclusion cannot be reached on Ques-

tion 1. The response to the second question must also be a preliminary one.

l As stated above, the staff's evaluation of the test matrices for reactor-systems-related testing is that the specified test conditions acoear to be l adeouate to provide necessary data for design certification and TRACG qualifi-cation. However, final conclusions concerning the adequacy of testing to j satisfy the requirements of 10 CFR 52.47(b)(2)(1)(A) cannot be drawn without a detailed review of the test data and GE's analysis thereof.

l l Evaluation of the SBWR TAPD Revision C Related to Containment Area The staff has reviewed TAPD, Revision C and has determined that GE has sufficiently addressed many of the previous staff's concerns and that TAPD, Revision C defines a systentic and comprehensive plan to test and analyze the ,

l SBWR-related phenomena. The staff concludes that the plan, as described, can '

l be accepted as a framework for the SBWR testing program and TRACG qualifica-tion process. However, in that GE has not fully dealt with all of the previous staff's concerns in Revision C, the following discussion is provided.

Some of the containment issues identified in the DSER (letter from J. H.

Wilson to P. W. Marriott, November 29,1994) are not addressed in the Revision  !

l C, i.e., (a) PCCS and containment response with a stuck open vacuum breaker-l (b) degradation of PCCS performance through ingestion of debris in the l drywell, and (c) potential influence of the passive autocatalytic recombiners i- (PARS), including interaction between the PARS and PCCS. In the subsequent l meetings GE presented a position that these issues can be addressed analyti-l cally using qualified TRACG code. In principle, the staff can accept such an approach; however, the final acceptance of this approach can be made only l after the qualification of the TRACG code is completed.

.The staff remains concerned with the exclusion of the PAR testing from the TAPD framework. In the response to Question 900.135, (which is not included in the Revision C), GE claims that 'An extensive database already exists on the PAR units." The staff is not aware of such a database; therefore, GE should provide additional information with a detailed description of the PAR l technology. Without this additional information, the staff believes there may l be a need in the future for additional tests.

I The review of TAPD, Revision C led to a number of additional questions listed in Attachment 1. These questions do not invalidate the acceptance of the TAPD, Revision C as a framework document. Rather, it is a list of detailed i issues that need to be addressed before final approval of the program can be  !

granted. i I

l I

i The staff's review of TAPD, Revision C has focused on the three general areas: i

1) scaling of the test facilities, 2) quality of the test data, including i adequacy of the test matrices, and 3) physical models incorporated into the {

TRACG code. These three areas are discussed below. -l l

1 The staff has previously pr.ovided an evaluation of the GE TAPD scaling analysis. The staff's major concern was the application of the scaling analysis within the frame-work of the testing program. The only other related comment the staff would make in this evaluation is regarding the statement in Revision C Section 1.2.2 that "GE has used a procedure similar to CSAU methodology .... and submitted to the NRC." This statement refers to a Rash to Jones letter in 1992 that submitted a set of view graphs from a presenta-tion, essentially outlining what is now TAPD, including initial PIRTs. It contains a few of the elements of CSAU, but it does not constitute a procedure l similar to CSAU. In particular, there were no sensitivity and uncertainty.  !

analyses included.

1 Regarding testing, the staff has performed an evaluation of the PANTHERS /PCC  !

data and is reviewing the GIRAFFE data. The PANDA results have not yet been l submitted for the staff's evaluation. In general, the test matrices described i in Revision.C seem to be adequate, i.e., the range of parameters expected for

.the SBWR operation seems to be well covered. However, by witnessing these tests the staff noted that in the case of the PANDA program some test proce-  :

dures were modified. In some cases the tests were not performed as long as i originally-intended. In another case, GE claims that-the objectives of two i different tests are met with a single test, e.g., tests M6 and M8. The staff  ;

is concerned that such late modifications may affect the implementation of the i framework document.  ;

The TRACG Modgl (LTR NEDE-32176P) review is pending and its evaluation will be ,

included in a separate SER. The TAPD, Revision C identifies the SBWR-related ,

phenomena and a need for associated computer models. In particular, Table 2.3-2 provides a good cross-reference between the SBWR containment phenomena and the code physical model. The list seems to be comprehensive and, as presented, is acceptable.

However, the staff is concerned with the application of the TRACG as a best- l estimate code. This issue hrs already been raised in RAI 900.134. The staff  ;

was informed verbally (August 1995 meeting at San Jose) that GE is now proposing a new " hybrid" approach, i.e., a mixture of best estimate and bounding (conservative) calculations. This new approach may have a signifi- l cant impact on some TAPD items. Discussion of this change is missing from the i document.

SUMMARY

AND CONCLUSIONS The staff's review of TAPD, Revision C has determined that it defines a systematic and comprehensive plan to test and analyze SBWR-related phenomena.

The acceptance of the plan is based on current knowledge without having the l

i i

opportunity to evaluate many of the test results. As a result, it should be understood that final acceptance or approval can only be provided after having l completed the evaluation of all of the program elements, i.e., the scaling report, the test data evaluation, the TRACG model, and the TRACG qualification report.

While the staff has accepted the plan, there are certain areas that have not been addressed which are necessary to have a complete program. These areas '

are identified in the form of RAls from the series 900.102-900.181. Examples of such areas include the effect of noncondensible gases (900.132), or the one  ;

regarding best estimate scenarios (900.134), or regarding TRACG nodalization '

(900.141). These examples are issues that have not been addressed within TAPD, Revision C and need to be resolved before TRACG qualification can be j used for design certification.  !

l l

l i

4

^ - - - - - - - - - - - - . _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _

l i

APPENDIX A l l l l Note: Most comments are related to the PIRT Tables, and generally deal with '

l identified phenomena and/or their rankings in those tables. References to tables and page numbers are keyed to the main TAPD document and Supplement 1 on PIRTs.

j 900.182 p. S1-2, critical flow is described as the " primary parameter l of interest" in the early phase (s) of a LOCA. However, in Table 2.3-1 (p. 2-36), phenomenon E8 (break flow) is ranked only "5" (medium for SBLOCAs.

primary para) meter of interest.This appears to be inconsistent for a

900.183 Also with regard to E8, the table of phenomena definitions I (Table S1-1) describes " break flow" in terms of critical flow I l only. However, there are times in the transient, especially after l blowdown, during which flow out the break would not be critical.

000.184 The ranking rationales in Tables SI-2 through -8 are somewhat ambiguous, since they do not break down phenomena ranking into different phases of the accidents covered. Also, some of the accident categories are quite broad (e.g., SBWR LOCA), and do not appear to differentiate between LOCAs of different types (e.g.,

MSLB, GDCSLB), for which the phenomena and their rankings may not be the same.

900.185 The handling of the specific phenomenon of break flow is somewhat confusing, since both E8 and L1 - L4 deal with it. Although the

, table does not limit E8 to liquid breaks (or non-MSLBs), since the "L" phenomena cover steamline phenomena, the implication is that .

E8 does not. Some clarification is needed of what is meant by l these phenomena, and further justification of the importance  ;

rankings is indicated.  !

l 900.186 There appears to be somewhat of a " disconnect" between identifica-l tion of phenomena from the " top-down" process and those from the l " bottom-up" process. Section 2 of the TAPD does not irlude any l

isolation condenser-related phenomena, although they are later  !

l. included in Section 4.

900.187 The cross-reference to the TRACG model report Section 3.1 for '

phenomenon L2X, acoustic effects, is misleading, since the TRACG report does not have any details on the phenomenon beyond tk- i basic general conservation equations.

900.188 The discussion of phenomenon C8 (p. 51-17) refers to its impor-

.tance in operating BWRs. It is difficult to see why it should be rated even " medium" for the SBWR (Table 2.3-1, p. 2-34). l 900.189 With regard to Section 3.1, while in the context of the TAPD, it is appropriate to focus on SBWR " unique" features, it is still i necessary to validate TRACG for phenomena that occur in the operating fleet,'as well,. since the code has never been approved as a licensing-basis tool.

l

l r

l l -

l 1

l 900.190 . Although this version of Table 3.2-1 is somewhat clearer than previous versions, and the table column headers are more descrip-tive, the way in which the entries are classified and the right most two columns, i.e., what qualifies as "SBWR Unique" and what is in the " Existing Fleet" is still somewhat hard to follow. In addition, there are still a number of apparent inconsistencies and other unclear attributions in this table and other related ones (e.g., Table 3.3-1, the tables in Section 4.1, and Table S19).

Some examples are given below.

t. The reference to PIRT Item C12 for Bil/7 in Table 3J-1 ,

(p.3-5) is hard to follow. It would appear that F1 is more I relevant.

b. For B32/2 (p. 3-7), why is " stratification in IC drums" under

" condensation in tubes?"

c. It is difficult to see why, for B32/4, " impact on IC unit heat transfer correlation" is a " phenomenon." In addition, this "high" ranked entry does not appear to be carried into Ta-ble 3.3-1 on p. 3-33.

d. The entries for E50/10 in Table 3.2-1 and Item 32 in Ta-ble 3.3-1 appear to be misleading. PANDA and GIRAFFE / SIT are noted as sources of data to address this issue, which involves passive / active systems interactions. The staff is not aware of any data taken in either of these test programs that would address this issue.

e. There appear to be some inconsistencies between the TAPD discussion of TRACG qualification and information that appears in the TRACG documentation itself. For instance, the TAPD cites data for chimney void fraction up almost to SBWR operat-ing pressure. However, no data are cited in the TRACG Quali-fication Report below about 20 bars. Since low-pressure L phenomena are of substantial interest in the SBWR, the appar-ent lack of low-pressure data is of concern.

f. The staff disagrees with assertions that TRACG models have

, been ' qualified" for SBWR applications. For instance, in l

Section 3.3.1.2, this claim appears to be made for predictions i of SBWR stability. Not only is this statement apparently at  !

odds with information in the TRACG Qualification Report, but i until the staff accepts TRACG as being qualified for SBWR, I claims that it is qualified must be viewed as preliminary, at  ;

best. i l

g. In Table 4.1-la, the issue listed for L5 is not clear; it is also not clear why this issue is ranked in the GDCS phase, I

L

. i I

rather than the blowdown phase. The rank of "high" for a ,

large break is also not clear, since depressurization through l the break would appear to be dominant.

h. In Table 4.1-lb, the medium ranking in the GDCS phase for IC phenomena appears to be inconsistent with the information in Tables 3.2-1 and 3.3-1, which show "high" ranking in the blowdown phase.

1. Item C27 in Table SI-9 is ranked "high" for stability, but does not appe n at all in Table 4.5-la.

900.191 The description of the systems interactions study described in  !

Section 4.2.2 and the results discussed in Appendix C do not i appear to be entirely consistent. It would be useful to number the cases or to use some other unique identifier to be able to determine the correspondence between these two parts of the report.

900.192 There appear to be some inconsistencies in Table 5.1-la, where specific test facilities are cited for phenomena that are not appropriate. Examples include:

a. Phenomenon A4 (p. 5-3): how do these facilities address lower  ;

plenum axial void distribution?

b. Item El: Neither Edwards test data nor Marviken tests include uncovery of a horizontal break line.

900.193 In a similar vein to #11, the treatment of CCFL does not seem ,

consistent in the tables in Section 5 when compared to comparable l information in the TRACG Qualification Report. The applicability i of SSTF to SBWR'for this phenomenon (Table 5.2-la) is open to question, while areas in which SSTF might be useful (e.g., Phenom-enon C9--parallel channel flow distribution) are not shown.

900.194 While indirect confirmation of phenoment can be derived from some separate-effects testr, that does not necessarily mean that they are S-E ter's for those phenomena. An example is interfacial heat transfer (Item C2Ax), where information on' interfacial heat transfer might be inferred from test results and associated analyses on the Edwards and Marviken experiments; that was not, however, the main objective of the tests.

900.195 Please clarify Refs. 71 and 72 (p. 8-4), which are called out in i Table 5.2-la. They have different titles, but the same number and date, 900.196 For Item A9, bottom head stratification, Table 5.5-la indicates ,

this will be evaluated by analysis. Considering the "high" 1 l

ranking for this phenomenon during start-up transients, it is not clear that evaluation in the apparent absence of data is appropri-ate. Please clarify how this phenomenon will be assessed.

900.197 Reference to coverage of the issue of " decay heat" (C24) by integral systems' tests is ambiguous. Certainly, integral tests can assist in evaluating the effect of different decay heat levels on-system behavior. As far as validating the decay heat model is concerned, however, that cannot be accomplished using an out-of-reactor test facility. ,

900.198 Several phenomena are listed as having only plant data available for " test coverage." These include C3 (gap conductance); C9 (par-allel channel flow distribution; C3BX (pellet heat transfer); C8X (void collap;e); F2 (chimney flow distribution); and 12 (separator inertial pressure drop). It is not clear for some of-these phe-nomena whether existing plant data are sufficiently detailed to provide the basis for code model assessment. For some cases, at least, it would appear that reasonably detailed local data are needed, while plant tests tend to be more " globally" instrumented, and local data may not be available.

l l

l l

i

APPENDIX B SIGNIFICANT QUESTIONS 900.199 Pg. 1-1 & Hydrogen generated by 100 percent metal-water 4-3 reaction is said to be addressed as DB require-ment, with reference to Table 4.1-2c. No further details are given, except that apparently the potential effect of light NCs on the PCCS is considered in the Interaction XC6. What is the effect of the 500 knol of hydrogen on the contain-ment pressures and on the volume fractions of H, and 0 in the contair.mnt during a LOCA? How do these, compare to the requirements of 10 CFR Sec-tion 50.34(f). Rough estimates appear to indicate that the containment design pressure would be ,

exceeded if 500 knol of hydrogen were to be added.

Or does 100 percent here refer to the clad surface only, with the clad layer to be oxidized, being the one used in Section 6.2.5.3 of the SSAR (0.00023 in.)? That would correspond to less than 1 percent of metal-water reaction.

900.200 Pg. 1-8 The first sentence of Section 1.2.2, Major SBWR Test Facilities, states "GE has used a procedure similar to CSAU methodology .... and submitted to the NRC", referring to Rash to Jones letter of 1992. That letter submits a set of view graphs from a presentation, essentially outlining, what is now TAPD, including initial PIRTs. It contains a few of the elements of CSAU, but it does not constitute "a procedure similar to CSAU". In particular, there is no sensitivity and uncertain-ty analysis and many of the steps _ leading to that analysis are also missing. Please revise the text accordingly.

900.201 Pg. 1-8 The remainder of the first paragraph of Sec-tion 1.2.2 appears also to overstate the case.

Whera in the PIRT are 900 data sets defined? (A PIRT ranks phenomena and does not define data sets). And if it were correct that of the impor-tant phenomena "each has been qualified", then there would be no need for a TAPD. All that would still be needed is a documentation of that effort rather than a TAPD! Please revise as appropriate. I I

i L _ _ _ ____ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - - - - - - - - - - - - - - - - - . . _ _ _ _ _ - - - - - _ .__ __ __ _ _ . _ _ _ _ - - .

2 900.202 Pg. 2-3 a At the beginning of Section 2.2.1.1 the piping and 2-4 valve sequence for the GDCS lines is prescribed, and the break is assumed to occur between the squib valve and the nozzle entering the vessel.

Under " Blowdown Period" in Section 2.2.1.2 it says

" simultaneously the pool side of the broken .line drains ....". The squib valves only are actuated 150 s after ADS actuation, about 300 s after the break occurs (see Figure 6.3-54 of the SSAR), i.e.

towards the end of blowdown and close to the beginning of the GDCS Refill Phase. (There are several later references to "almost immediate" tank draining also in Sections 3 and 4, for in-stance Page 4-2, 2nd paragraph.) Please clarify.

900.203 Pg. 2-26 & The first paragraph of Section 2.3 apparently 28 describes the procedure applied for the LOCA in-vessel PIRT of Table 2.3-1, while the second paragraph on Page 2-28 describes the procedure used for the containment PIRT, Table 2.3-2. The procedures are not quite identical, and .this should be clarified. The team cf experts was apparently not the same, consensus was used on Page 2-26, while averaging is used on Page 2-28.

Either procedure is appropriate, but it should be clarified with the first paragraph of Page 2-26 that the procedure, described there, was not used for all subsequent PIRTs.

900.204 Sect. 2.3 It is surprising to find that the isolation con-densers are not at all mentioned in the current PIRT, not even under transients. (For instance, note inclusion of Transient 15.5.1, Inadvertent Startup of ICs in Table 3-1 of the Applications Report (NEDE-32178P)). This would appear to be an omission and should be corrected. In particular with the three uses of the PIRTs outlined on Page 2-26, their inclusion is important for all three uses, but in particular for the third one of determining model bias and uncertainties. With the extensive Pt.NTHERS test data and the universi-ty test data of Reference 19, there would appear to be a sufficient data base, and one would not expect that the inclusion of the ICs here would significantly affect the tables of Sections 5 and 6.

1 i

': o 3

900.205 Sect 2.3-1 At the top of Section 2.3.1 the overall transient is broken down into three time periods, followed by a statement, that rankings were performed for "each of these periods". While this was done for the containment, it was not done for the in-vessel events of Table 2.3-1. At least this statement should be corrected, . including a justification, why rankings for the long term cooling period are not expected to introduce further significant in-vessel phenomena.

900.206 51 Tables While the phenomena descriptions of ' Supplement 1 SI-3 & 4 are extremely helpful in understanding the PIRTs, there are two major deficiencies, which should be corrected:

1. The phenomena brought in via Section 3 into the combined PIRTs of Section 4 are not covered at all.

2. The ranking rationales of Tables S1-3 and S1-4 do not address individual rankings and are much too brief to be of help.

900.207 4th 1 of The PIRTs of Revision C have been slightly revised Section from the previous PIRTs. We assume that this was 2.3.1 done without the team of experts, by one or few in-house contributors. While this is, strictly speaking, not within the rules for PIRTs, from a  !

practical point such revisions are unavoidable and should be acceptable, if done cautiously by ex-perts. We recommend that this be pointed out in the report, with a description of the procedure used for PIRT modifications.

I 1

I 4

900.208 2nd 1, Page In accordance with the focus on the first 72 hr, 2-28 given in Section 1, the deletion of PC6 (PCCS vent I fan) from Table 2.3-2 is justified, since the fans l are apparently only available after 72 hr. Howev-er, in Section 3, several items with up to 100-day i

time scale are being retained. This significant

! discrepancy in focus is not clear to us. Also, l details about this PCCS fan are still not known to' i us. It is not shown in the SSAR. What is its purpose? Can it become an obstruction? Can it l push excessive steam and/or NCs through the PCCS vent lines into the suppression pool? Is it considered in the modeling of the pressure drop in the PCCS intake lines, which was ranked high? Why is this fan not included in the PANTHERS and/or PANDA experiments, if it is a resistance in the flow path? We would like to know more about its location, its function and purpose.

900.209 Table 2.3-2 Another phenomenon which we would consider to be of interest only after the 72 hr focus, if at all, is 001, heat transfer to safety envelope. The safety envelope appears to begin at the outside surface of the containment concrete walls. Appar-  !

ently free convection on the outside of the 2 m l thick containment concrete walls is considered l here; it takes approximately one week for 1 per-cent of the temperature rise inside the contain-ment (about 50 to 100 C) to be felt on the out-side of the wall. Why is this included?

900.210 Section 3 In Revision C the contents of the QDB sheets is described in more detail. Apparently they include specification of the required data range for the important process variables. This information '

would be of great value in assessing the complete-ness of the currently available data base. We again would like to recommend that this informa-tion should be made a'<ailable, maybe as an appen-dix or a supplement to the :eport.

900.211 Sect. 3.2 Such system structure tables as Figure 3.2-1 Table 3.2-1 generally have further detailed subsysten struc-tures associated with them, leading from the systems to the components. A " system and compo-nent perspective" was promised in Section 3.1.

l Without seeing the further breakdown from such

! major systems as Nuclear Boiler System, one cannot ascertain the completeness of the features and phenomena selected in Table 3.2-1.

t 5

l 900.212 Table 3.2-1 From the information given, it is not at all clear, what belongs to the RPV System and what to the Nuclear Boiler System.

900.213 Table 3.2-1 The term " unique SBWR design feature" is now well defined in this revision of the report. We concur with the definition, but we still have some prob-lems with it:

! 1. TRACG will have to be validated for all phenom-ena, not just for the SBWR unique ones. While I

we would expect there to be a sufficient data base for the non-unique ones, we do not see, ,

why they should be excluded from r# consideration for code validation.

2. It remains unclear, how under QDB Screening (Columns 10 and 11 of Table 3.2-1) many phenom-ena are to be qualified against BWR fleet data.

In the text on Page 3-2 the third paragraph er.d the end of the last sentence on that page appear to be in conflict. If validation can be based on fleet data, then the phenomenon is not SBWR unique as defined in the third paragraph.

3. It is also often not clear, why a feature is l- considered unique. For instance, Entry T10/0 l in Table 3.2-1 lists " pressure suppression type containment" as a unique concept. There have certainly been previous pressure suppression type containment designs. If uniqueness here now refers to the data range (Item 2 in the third paragraph of Section 3.2), then fleet data would not be available to assess this j phenomenon.

l 900.214 Table 3.2-1 In many cases, the reader cannot appreciate what is meant by the brief entries in Columns 5 and 6.

l For instance, Page 3-5, System / Issue No. B11/7

and 8, the entries in the first row are clear, but what do the next two rows mean? (Core flow ea-surement/ determination and Potential for Bundle flow ma1 distribution...) What are their rankings?

In the next row (Bll/8), what is " nuclear heatup" and why does it enter under RPV chimney and not under core?

In Row Bil/7 reference is made to PIRT No. C12, which is part of the core bundle. Wouldn't F1 be more appropriate?

l

1 6

l l '900.215 Table 3.2-1 The indexing of the entries in Table 3.2-1 can

! often not be followed. There appear to be incon-l sistencies. It should be revised and/or fully 1 l explained. For instance, for System MPL j t E50, GDCS, we find the following sequence of l numbers and letters:

l System MPL Unique Issue important

/ Issue No. Features Pheno.neos l

t E50 l)

E50il a) 1) 1)

'E50/8 2) 1) l E50/9 3) 1)

E50/2 b) 2) 1)

! ISO /3 2)

E50/10 3) 1)

l. E50/4 c) 3) 1) '

l a)

E50/5 b) 2) 2)

E50/6 a) 1) 1) b)

E50/7 2) 900.216 Section 3.3 The subsections of Section 3.3 provide mainly details to the

& Table 3.3- entries of Table 3.3-1. Ilowever, the user must at times I count the table items marked by asterisk to find the appro-priate subsection. The System MPIJIssue No. should be l given with each subsection, and the titles should remain identical to the entries used in Table 3.3-1 (Many ue, but not all.)

900.217 Table 3.2-1 Under E50/1, the GDCS period is specified, but one of the E50 concerns is equalization line draining, which would only be anticipated for the long term cooling phase of some LOCA scenarios. Please explain or modify.

Table 3.2-1 Similarly, it would appear that Time Phase 7 would also be 900.218 E50 relevant for Item E50/8, since ADS and FW interactions l would not be expected after the blowdown period (See also the time scales given in Table 4.1-Ic.) j l

900.219 Table 3.2-1 We do not understand why Item E50/9 has been added: with )

E50 the check valves between GDCS pools and the RPV, mano- i metric oscillations would not appear to be possible.  !

i 1

7

, 900.220 Table 3.2-1 The E50 entries in Table 3.2-1 repeatedly reference PIRT l E50 Ref. Nos. GD3 & GD4. No such PIRT entries are found l in Sections 2, 4, or 5 of the report, nor are they defined in i Table SI-1.

900.221 Table 3.2-1 E50/4 appears to consider equalization line flow which E50 would occur in Phase 9 only. In this context, the entry in Column 5, Issue, is completely unclear: What does equal-ization line flow have to do with interactions with ADS and l l FW7 Please explain or modify. l l.

l 900.222 E50 The 250 entries E50/1 to E50/5 are apparently divided into l Table 3.2-1 different time ranges in Column 4, Unique Features, which .

l is not really a " unique feature", but the subdivision appears !

to make sense. However, the definitions of short, medium I l and long term cooling are not clear, and the terms are apparently not used in the same context as the Time Phas-  !

es 7 to 9 of Column 8. In particular, since this item is classified in Section 3.3, below, as a key item, a more l thorough definition of what is meant here should be provid-ed. Apparently more than just equalization line flow is considered in Section 3.3, while the text of Column 4, here, really only applies to equalization line flow. What is

" System post-LOCA heat transfer", as listed under E50/27 Please clarify.

l 900.223 Table 3.2-1 The Squib valves and the biased open check valves are listed ,

E50/6 as components to be developed, apparently for the GDCS lines as well as the equalization lines. Their flow and pressure drop characteristics apparently remain to be estab-l lished. The potentially relatively low flow rate through the l equalization line is mentioned. When and where are these data to be developed and where will they be reported? The Table 3.3-1 entry " design specification must be met by prototype hardware" implies that future testing is planned.

What are the design specifications, in particular loss coeffi-cients over the total operating range, including low flow rates in the equalization lines?

i

i 8

I 900.224 Table 3.3-1 An essentially new item is the possible effect of stuck open '

No. 34 equalization line check valves, with sloshing between the SP

& RPV coolant levels. 'Ihis is listed as E50/7 of Ta-

ble 3.2-1, with PIRT Ref. No. EQ2. Being ranked high (7),

i it is carried forward to Table 3.3-1 as Item 5c, where it is marked as item for further evaluation. However, no further entries for this item nor a definition of EQ2 are found in l

Sections 4 or 5. Rev. B listed this item in Table 5.3-2, to be covered by PANDA tests. It is not listed in the corre-sponding Section 5 table of Revision C. However, it was finally found in Ts.ble 5.3-2b under medium ranked phenom-ena and without any more reference to PANDA. Please explain.

Why does Column 4 of Table 3.3-1 refer to MAAP calcula-tions and Column 5 to TRACG evaluations?

l 900.225 Subsection The statement is made that the "TRACG models have been l 3.3.5 qualified against GIST test data". However, GIST is not  :

considered part of qualification data base. GIST data can be used, but they cannot be the only data source. (The same actually also applies to App. A.3.1.4.) Mention should be made, most likely in Appendix A, that GIST data are being used, but only as additional assurance, not as part of qualifi-cation data base.

l .

_ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .__m

9 900.226 Sect. 3 There were ten GDCS phenomena identified in Table 3.2-1

& Sect. 4 (System E50). Nine of these were ranked h 4 and were carried forward to Table 3.3-1, even though some of the wording and the sequence was changed in this process. In particular, four items were grouped together in No. 32 and declared to be the " single most significant phenomenon" (top of Subsection 3.3.5). The four had ranks 7, 5, 4, and 7 in Table 3.2-1, while three other items were ranked 8.

Why and how are items ranked seven and below suddenly more significant than items ranked 8? Why are some of them lumped in No. 32 and then appear scptely again later in Section 4, like for instance E50/4 which enters as EQ1 in Table 4.1-2a. What happened to E50/2, referred to PIRT Ref. No. GD4 and ranked 8 in Table 3.2-1, it is No 31 in Table 3.3-1, but never shows up in Sections 4 or 5? Frequently, Table SI-9 permits tracking of items, but in this case there is no entry there either. A complete accounting of were items go from Section 3 to Section 4 is required. '

900.227 Table 3.3-1 What is the vague reference to " Russian data" in No. 30 of Table 3.3-17 The BWR test data in the same entry are also not referenced properly.

900.228 Table 3.2-1 Several entries under G21/2 remain unclear (what is Post-  ;

LOCA decay heat removal?), and PIRT Ref. No. FPSib is not found anywhere else.  ;

900.229 Table 3.2-1 Considering the entries for the Containment System (TIO) in System T10 Row T10/0, what is unique about a pressure suppression type containment concept? Under T10/25 a time limit of 100 days is introduced. An explanation appears required here, why inis time' period is considered, since Figure 1.1-1 stated that the TAPD focus extends to 72 hr. Such a time limit would certainly also require additional action, like replenishing the PCCS pool. Nothing said about that. What does containment heat transfer for times beyond 6 h have to do with the 100 day time span?

10 900.230 Table 3.2-1 The concern about heat conduction through the 2 m thick T10 & TlI concrete walls is not clear. First of all, the PCCS is sup-posed to be designed for 100 percent decay heat removal after the GDCS refill period. In that case, even iflong term heat conduction into containment structures were zero, the containment pressure would remain within bounds. And over very long time periods, like the 100 days mentioned in T10/25, as the decay heat slowly decreases one would expect the pressure to decrease as well. Over such long time periods, concrete conductivities could indeed change and so could the liner to concrete resistance. However, the effect on containment pressure should not be significant as long as the PCCS is operating as designed.

Also, under T10/25, considering 100 days, this effect is ranked 7, while under Til/1, with considerations over 72 h only, the effect is ranked 8. Please explain.

Several of the sub-categories of Imponant T/H Phenomena in Til are not clear. What is " loss in ignition" CO 2content in Til/5? CO release 2 from concrete generally only occurs at much higher temperatures than those driven by a drywell long term temperature of about 140 'C. Please explain, what is meant with these four sub-categories, in particular a) and d).

900.231 Table 3.21, item Unique Features 1-b, T10/7 in Table 3.2-1 has four T10; subcategory entries in Column Issue. These appear to Table 3.3-1, correspond to Entries 49 to 52 in Table 3.3-1. The connec-Entries 49- tion between the first three entries and the last entry is not 52 clear. Considering the wording in Table 3.2-1, Item 4 is "DW response to LOCA break". But Items 1 to 3 also deal with exactly that subject and the phenomena listed under Item 4 are also phenomena occurring with Items 1 to 3.

Why are Items 1 to 3 not considered as sub-categories of Item 47 This structure is not clear at all.

Why does Item 52 refer to ABWR horizontal vent tests, when we are discussing drywell Th/H phenomena? The horizontal vent system (and suppression pool) was handled under Unique Feature 1-a, T10/1, Entries 43 to 48 in Table 3.3-1.

11 900.232 Table 3.2-1, The Table 3.2-1 Entries T10/31 & 32 consider drywell to T10; wetwell pressure differences and bypass flow. This is an Table 3.3-1, extremely important item, and it is ranked 9. The corre-Entry 56 sponding entry in Table 3.3-1 is No. 56. But why is it described there as " gas space temperature distribution / wall heat transfer ...."? That description is not clear and the bypass flow should be primarily driven by pressure differ-ences. "Ihe text of the corresponding Subsection 3.3.8.7 describes the problem well, but does not explain the Is-sue / Phenomena entry in Table 3.3-1.

900.233 Table 3.3-1 Entry 53 contains a second vague mention of " Russian studies", without providing any reference or description.

900.234 Table 3.3-1 This item was carried forward apparently from T10/29 No. 59: Table 3.2-1. There it was ranked 8. Here it is said to be not an issue. It does not appear to be carried forward from here, even though it is said to be " covered by PANDA tests." These statements appear to be contradictory.

(WW3, as described in Section 4 or in Table S1-1 does not include submergence.) Please explain and/or revise.

l l

(

l l

12 900.235 Sect. We fully agree that there is no general pool of data to cover 3.3.8.4, stratification and mixing of steam and non-condensibles in Table 3.3-1 the drywell. However, this subsection also states that the No. 51: TRACG mixing model has been " verified" in the Qualifica-tion Repon. (We assume that not " verified", but "validat-ed" or " qualified" was meant.) The Qualification Repon (Sect. 3.8) compares TRACG predictions to radial void fraction data in core configurations from EBWR and VK-50 experiments, to validate it's turbulent mixing model. The l comparison was of limited success, with the discrepancies l convincingly attributed to a lack of available information on several details of the test data. This does not mean that

" reactor plenum data are well predicted", or that the TRACG model "has been verified", as is claimed here.

There is really not much connection between the stratifica-tion of gases in a large volume, like the drywell, and the radial void distribution in the chimney section of the reactor.

A crucial point here is, that the TRACG drywell stratifica- i

tion has lately been proposed to be handled via an empirical i l upper bound model. Once such a model, including uncer-l tainty estimates, has been formulated and submitted, it must l be evaluated and accepted. The modeling and qualification l

effon of the TRACG mixing model, described here, may ,

then no longer be required. Please corrment. '

900.236 Sub-Sections The two subsections consider suppression pool stratification 3.3.8.5 and for the shon term blowdown period as well as for long term l 3.3.8.6: cooling. For the first period a " conservative interpretation" (Table 3.3-1 of the data is referred to, but the text implies that a future No 54 & 55) mechanistic model may be substituted. For the second time i period a mechanistic model, to be qualified, is implied. We were under the impression that conservative empirical models would be used for both time phases. Please explain what time scales are considered for these two items and how suppression pool stratification will be modeled for each phase.

L 900.237 Sub-Sections Of the two entries, or,1y one corresponding entry is found in l 3.3.8.5 and Table 4.1-2a (WW6), i.e., all the previous detail, presented 3.3.8.6: here, is lost. WW6 is carried forward into Tables 5.2-2, (Table 3.3-1 5.3-2, 5.5-2 & 6.1-1, but with reference to PANDA, &

No 54 & 55) PSTF & Mark III data only. No more reference to "ABWR horizontal vent tests" is given, as in Table 3.3-1. Please explain.

L

17 900.238 Table 3.2-1, Entry T15/4 lists potential detonation of H2 /02 with steam.

T15/4 This was assigned an Importance of 5. H2 (but not 0 2)can really only be produced from massive fuel clad / water reaction, while both can be produced to a moderate degree by radiolytic decomposition of water. With PARS the presence for either should have an infinitesimal probability.

The minimum required O2 concentration for deflagration in H 2/02systems is about 5 vol percent, with a corresponding-ly higher detonaticin limit'. O2 is originally uniformly mixed with nitrogen at 3.5 vol percent. Thus, its concentra-tion after adding H2 O and H2 can only be lower than that value. If detonations are not credible, as the above argu-ment implies, than they cannot have a rank of 5. However, if they are possible at all, then they will have to be a top priority! A detonation in a PCCS unit could leave a gaping hole in the containment boundary! Either the item should have a rank of 0, since incredible (which we hope it is), or it must be carried forward. (Table 3.3-1 implies that it is included in No. 65, but detonations or even deflagration burning are not mentioned there at all.)

' H. F. Coward and G. W. Jones, " Limits of flammability of gases and vapors", Bulletin 503, Bureau of Mines, Department of the Interior, AD-701575,1952.

900.239 Table 4.1-2a Why are the suppression pool hydrodynamic loads being Table 5.3-2 brought in here (WW9)? The current GE position is to use a previous " approved methodology" for these and not TRACG. We assume that the entries in Table 5.3-2 that TRACG qualification against these data has been completed are erroneous?

)

'H.F. Coward and G.W. Jones, " Limits of flammability of gases and vapors", Bulletin 503, Bureau of Mines, Department of the Interior, AD-701575,1952.

I

18 l

900.240 Table 4.1-2a The entries considering conduction through the concrete walls and convection on it's outside would not be items for TRACG validation (OC1 and CW2). However, if wall heat transfer is indeed important, then property data for items of CW2 would certainly be of interest. Also of interest would then be the liner / concrete gap conductance.- This was carried in Rev. B as PIRT Ref. No. CWl, but has essential-ly disappeared in Revision C. CW1 is used in Table 3.2-1 for conduction through walls, which was ranked low. Table 3.31, in No. 60 mentions it, referring back to Item T10/25 in Table 3.2-1, which has no reference to liner / concrete gap

.. conductance. It does not seem to be used at all in Section

4. We would expect that the analysis would be done with a simple model or code for heat conduction, analyzing the process with fine nodalization over very long time periods.

Please explain and confirm.

900.241 Table 4.1-2a At this time it is not clear what the assumed conditions are, for the operation of the PARS during LOCA scenarios. We would assume that they would also not be covered by TRACG analysis, and the threr, entries for System T49 in Table 3.3-1 would also imply this. Please confirm or explain.

900.242 Table 3.2-1 Typical flammability control systems use hydrogen igniters System T49 (glow plugs) to cause controlled recombination (combustion) of H 2and O .2 Such a system is described in Section 6.2.5 of the SSAR. Reference there is to " igniters" an:! " glow plugs" while the phrase " Passive Autocatalytic Recombiners" (PAR) is never used. This report exclusively uses the term PAR. Are they the same, or is another system being considered now? Use of PARS is listed as unique in T49/1 in Table 3.2-1. The glow plugs mentioned in the SSAR would not appear to be unique for flammability control. If the PARS are different from the glow plugs of the SSAR, can they also remove H 2 alone, i.e., without the presence of O 27

19 900.243 Table 3.2-1 Without a definition, what comprises this system, it is not System U71 clear how these entries interface with those of System Til, Table 3.3-1 Containment Vessel. By the SSAR, Section 6.2.1.1.2, the No. 74 safety envelope aaoears to be the outside surface of the containment conevie walls, while the reactor building structure would be outside the safety envelope. But item 74 in Table 3.3-1 refers to heat transfer to the safcy envelope.

Why is FP holdup mentioned for regions outside the con-tainment when there has not been a scenario identified, which resulted in FP release from the containment (U71/2 in Tah 3.2-1)7 A more detailed description is required as to what is meant'and how Tl1 and U71 are related.

900.244 Section 4.0 At the beginning of Section 4 one gets the impression that

& 4.2 the interactions, to be consid: red here, will be the ones identified in Tables 4.1-Ic and 4.1-2c. (The list of interac-tions is screened in Section 4.2 ....). Actually that is not so. Of the LOCA/ Containment interactions of Table 4.1-2c parts of XC2 and XC4 are the only ones considered in Section 4.2. It is suggested that this should be made clear in Section 4.0 and at the beginning of Section 4.2.

900.245 Table 4.1-2h Page 4-2, end of second paragraph, states that the " integral response of the containment for the DBA" with 100 percent I

metal water reaction "is included as an interaction (XC6) in Table 4.1-2c". From the references to XC6 in the remain-

! der of the report, we assumed that XC6 only covers the effect of H 2on stratification and PCCS performance.

)

l l

l

20 900.246 Section 4.2.2 In Section 4.2.2 six interaction scenarios were listed as Section C.3 having been analyzed, with more details to be given in Appendix C. However, there does not appear to be a one to one correspondence between the scenarios of Section 4.2.2 and C.3. The text, justifying why these six cases were chosen is almost identical in the two sections, but the scenarios are not! As far as we can determine, there is a partial correspondence:

Sect. 4.2.2 App. C.3 1 1&7 2a (inadvertent DPV) 3 2b (IC drain before and after DPV)  ?

2c (connection IC & DPV severed)  ?

3  ?

4  ?

5 6 6 4 Please revise the corresponding report sections, so that a reader of Section 4 can indeed readily find the additional detail in Appendix C, without extcasive detective work.

900.247 Section 4.2.3 The fifth paragraph of Page 4-26 states "but the only system for which possible adverse interactions were identified was the FAPCS". Close to the top of the next page it says "results for these cases all show the use of the FAPCS has a favorable effect on containment pressure and temperature".

These statements appear to be contradictory. After study of Appendix C one sees what apparently is meant: After running the first three interaction cases, the possibility of an adverse effect with limited u*e of the FAPCS as drywell spray was suspected. Running that case, it was found that no adverse effects were observed. Please confirm and/or modify the text in Section 4.2.3.

I 1

I

21 M)0.248 Table 4.1-2c There does not appear to be a systematic development of the interactions of Table 4.1-2c and without a complete indexing system one often cannot understand where they come from.

In particular in view of the fact that GE claims in Section 7 that Section 4 " identified" the important interactions to satisfy 10 CFR 52.47 requirements, we feel that the follow-ing further information is required:

a description of the method to define all impor-tant interactions, e a key of their progression from' Section 3 to I Table 4.1-2b, 4 a more detailed description of the interactions, preferably at the point where they are first intro-duced, and

• r Description of the interrelations between the interactions of Table 4.1-2c and those of Sec- 1 tion 4.2.

900.249 Table 4.1-2c Many of the interaction descriptions are very unclear and change from entry to entry. Consider for instance Interac-tion XC2, first introduced in line T10/24 of Table 3.2-1.

Eight items of interactions are listed there, nothing is said, beyond a listing of acronyms. 'Ihen, in Table 3.3-1, Item  ;

58, three of those items are mentioned only and PANDA  :

experiments are listed as data base. The corresponding more detailed text mentions most of the compo-nent/ phenomena list of Table 3.2-1 and adds the main vents.

Finally, the Table 4.1-2c entry lists only IC/PCCS/GDCS of the original eight entries and adds the DPVs. This is certainly not consistent, and the reader is at a loss to follow.

900.250 Section 4.3.1 The third paragraph of Page 4-30 repons in the summary  ;

section some results from an apparently analytical study on the effects of hydrogen on PCCS performance. This would appear to correspond to Interaction XC6 of Table 4.1-2c.

No details of this were provided in Section 4.2 or Appen- -

i dix C. Since XC6 is carried on to Table 5.3-2a (GIRAFFE /

Helium), we anticipate that the details will be provided with  !

GIRAFFE / Helium reports? This again points to the per-plexity which is created when new items are first presented in a summary section and not in the body of the section >

which is being summarized.

l l \

! l

22 900.251 Table 4.3-3 Interactions XC8 and XC9 are not listed here, without any justification. Please modify orjustify.

900.252 Interaction In Section 4.3.1 the further deletion of interactions XC1 and XC3 XC3 is justified. 'Ihis appears reasonable for XC1. How-ever, it is by no means clear what is meant with XC3, and what is meant by " heat sink". Returning to Table 3.2-1, and searching for XC3 in Column 7, one gets to G21/1, which has many cryptic and unclear entries in Columns 4 and 5, while Column 6 implies FAPCS/PCCS interactions.

Going forward to Table 3.3-1, one finds under G21/1 only.

FAPCS/PCCS interactions discussed, the same applies for the description in Subsection 3.3.6.1. Nowhere does one ever find out which heat sink is meant. Therefore, one cannot follow the argument in Section 4.3.1, to drop XC3.

This example also shows how laborious it is to track such an item. Please provide clarification.

900.253 Table 4.3-3 While we consider it reasonable, to remove the CW items from funher considerations, we do not un6tand, why OC1 was retained. Natural convection on the outside of the containment walls is not of interest for weeks, and is mean-ingless, when uncoupled from heat conduction through the

]

wall. l 900.254 Section 4.3 now contains an added discussion of Interaction XC8, finally making it clear that contrary to the apparently erroneous entries in Section 3 (specifying Time Period 7 in Table 3.2-1), the long term effect of unlikely horizontal vent opening during time phases 8 and 9 (GDCS refill and long term cooling) are the focus of this interaction. Nothing is said, what would cause such drywell pressure increase of at  ;

least 0.2 bar, to open the horizontal vents, but the phenome-  !

non is carried forward to Section 5, even though it was i omitted in Table 4.3-3. We would expect funher details to be ,iven with the GIRAFFE / Helium, GIRAFFFJ SIT and PANDA tests,'since these are referenced in Table 5.3-2a for this interaction. Further detail should be provided up front,  !

were the interaction is first introduced, i.e., in descriptive text, detailing Table 3.2-1 entries. A summary section should not present information, which was not introduced in i the body of the section being summarized.

i 4

23 900.255 Section 5.5 'Ihe value of the overall tables in Section 5.5 could be greatly enhanced, if the notation of "T", "Q" and "X" of the preceding tables were to be retained, which could be done very easily. If more than one apply, there is enough room to show them side by side or show only the symbol indicating most work required ("Q" supersedes "X", etc.).

900.256 Section 5 Phenomenon DPV1, depressurization valves mass flow, is referenced in Table 5.5-2a as being validated against compo-nent tests. No corresponding entry is given in Table 5.2-2a.

Please correct or explain.

! 900.257 Section 5 Interaction XC4, which was considered in previous versions i of the report has disappeared here. It was last mentioned in this version in Table 4.3-3. Validation against PANDA test data were planned in previous versions. This could be an oversight, or it might have been intended to combine it with )

FPSI, and the corresponding entry was missed. Please clarify.

900.258 Table 5.3-2a For suppression pool stratification as well as for drywell 3-d effects GE has stated that bounding models are to be ap-plied. In all likelihood, that would appear to be the most practical approach. However, the claim to use PANDA and GIRAFFE data for the 3-d effects of DW3, WW6 and WW7 appears questionable. The geometry of the containment cells differs so drastically from the tanks of the test models.

Therefore, we cannot see how one can compare the 3-d flow l pattems between them, or use models of a simplified tank to I

confirm 3-d flow patterns in much more complex geome-tries.

900.259 Table 5.3-2a The PANDA entry for EQ1 states via a footnote that it "may" be included in the PANDA Phase III tests. Please finalize.

l 900.260 Sect C.3 The first sentence of the Subsection Isolation Condens-er/DPV Interaction does not appear to make sense. The previous paragraph dealt with IC/GDCS interactions during the refill phase. We are now in the blowdown phase. What is the "the same two cases"?  !

24 l

l 900.261 Table 6.1-1 Considering all containment items marked either "Q" or "T,Q" in Section 5, the following questions remain:

• The following entries from Tables 5.3-2a and 2b

! for the PANDA test facility do not show up in l Table 6.1-1: WW3, EQ1, XC8, XC9.

l

• None of the GIRAFFE / SIT entries of Table 5.3-l i

2a and 2b are found in Table 6.1-1 (12 items).

• The following entries from Table 5.3-2a and 2b l for the GIRAFFE / Helium tests do not show up l in Table 6.1-1: WW7, PC2 to PC4, XC8.

Please correct or justify.

900.262 Table 6.1-1 Interaction XC4 (FAPCS/PCCS) was marked as "T,Q" in Table 5.3-2 of Rev. B, to be qualified via PANDA test data.

l It was omitted, apparently as an oversight in Section 5 of I

Revision C, and it also is not shown here. Section 4.2.3 l stated that for four selected cases of FAPCS interaction with other systems, primarily the PCCS, its effect was always beneficial. PCCS operation would stop with spray actuation and would resume if the sprays stopped. However, these are calculations and not experimental validation. Please i correct or explain.

l 900.263 Page 7-1 Under " Test Plan" on Page 7-1, GIST is listed as completed tests for GDCS flow, GDCS initiation times and RPV levels. We are under the impression that, as for the early -

GIRAFFE tests, GIST data are only to be used as support l and not as sole qualification basis. Please explain and/or

modify the text accordingly.  !

900.264 Sect. With the GIST TRACG Analysis Plan (Section A.3.1.4.4)it A.3.1.4.4 should be mentioned that GIST data are to serve as support only and not es sole validation data base. i 900.265 Sect. PANDA test data are referenced for qualification of phe- '

l A.3.1.3.3 nomenon WW8 (suppression chamber spray condensation) in Table 5.3-2a. The test program of Table A.3-9b shows only drywell spray actuation in Test M5, unless this is planned for tests M9 or M10, which were not defined yet in Revi-sion C of the report.

l l

l 25 I

900.266 Sect A.3.2.3 Component development for the DPVs and the vacuum

&4 breaker valves is described in detail in Sections A.3.2.3 and A.3.2.4. In Section 3 (Tables 3.2-1, System MPI/ Issue i No. 50/6, and Table 3.3-1, No 33) the need for flow pres-sure drop characteristics of the squib and check valves fer i

the GDCS drain lines and equalization lines was mentioned, emphasizing in particular the low flow regime. We would have expected a corresponding section for these valves here l too. Please expidn.

l l

l i

i 1

i l

t <

• 26 APPENDIX B MINOR QUESTIONS 900.267 Pg. 2-6 Spillover holes in the main vents are mentioned in Line 8.

Where are these holes located? We have found no other reference to these holes. Please state their elevation, num-ber and size. j 900.268 Table 2.2-1 Bypass valve closure at 6 s is not explained. (Apparently, f due to loss of power to condenser circulating pumps, as described under transients in Section 2.2.2.1?) i 900.269 Pg. 3-3 There are 13, not 12, safety systems (see also Table 3.2-2).

2nd1 900.270 Fig. 3.2-1 Figure 3.2-1 is too small in its current 8 % by 11 inch presentation. Many of the details, in panicular the system identification numbers, cannot be identified, even with a j magnifying glass.

900.271 Table 3.3.1 Under Issue / Phenomena it is not stated that this is the sup-No. 55 pression pool temperature distribution / plume. 4 900.272 Table 3.3-1 The item is carried forward as WW3 and is marked in No. 57 Table 5.3-2a as "T&Q". We assume it should then also be marked as Q here.

900.273 Table 3.3-1 We assume it should read " vent line". The vent line, not No. 68 the vent sparger, passes through the gas space.

900.274 Figure 6.0-1 The two black circles are still almost illegible. Please correct.

900.275 Table 3.2-2 The T10 Entry should apparently be 33 and not 34; Item T10/33 is shaded, i.e., not evaluated.

j 900.276 Table 3.2-1 It is interesting to note that the three sub-phenomena T15/10 ,

to 12 have a ranking of 8, while the preceding row, a sum-  !

mary entry for these three sub-phenomena, has a ranking of j

7. Please correct or explain.

900.277 Page 4-26 Third paragraph, Line 9: "...the PCCS was able to resume the decay heat load..."; we assume it should be " remove" instead of " resume".

900.278 Table 3.2-1 Acronyms SPTMS and PSWS are used but nowhere de-fined. Please correct

27 900.279 Table 5.5-2a The x mark for DPV1, Component Tests, should be re-moved, or an appropriate entry in Table 5.2-2a should be l provided.

{

l I

i l

F l

l j

.__ . - _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ - _ _ _ _ _ - _ _ -