ML20063F261
| ML20063F261 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 08/25/1982 |
| From: | Crouse R TOLEDO EDISON CO. |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| References | |
| RTR-NUREG-0737, RTR-NUREG-737, TASK-2.K.3.30, TASK-TM TAC-45817, NUDOCS 8208310192 | |
| Download: ML20063F261 (11) | |
Text
-..
o 1
l itXE00 EDISON Roan 0 P. Csoust Docket No. 50-346 va, e,m NA%d License No. NPF-3 m9am wa Serial No. 848 August 25, 1982 Mr. Harold R. Denton, Director Office of Nuclear Reactor Regulation United States Nuclear Regulatory Commission Washington, D.C.
20555
Dear Mr. Denton:
On July 20, 1982, Toledo Edison and the B&W Owners met with the Staff to culminate the continuing dialogue on the scope of the program for resolu-tion of NUREG-0737, Item II.K.3.30, " Revised Small Break (SB) LOCA Methods to Show Compliance with 10CFR50, Appendix K."
This letter formalizes the Toledo Edison responses to your verbal requests made at that meeting.
Toledo Edison will resolve the two separate areas identified by the Staff in the April 16, 1982. meeting. The first, assurance of core cooling (10CFR50, Appendix K), is being evaluated under an ongoing SB LOCA Methods program approved by the Staff. Toledo Edison will continue to address the NUREG-0737, II.K.3.30 staff issues in the SB LOCA methods program as identified in Attachment #1.
B&W on behalf of Toledo Edison has also prepared a number of reports as a result of the recent joint test evalu-ation with the Staff which are identified in Attachment #2.
The items in the Attachments are endorsed by Toledo Edison.
The second area outside the scope of NUREG-0737 Item II.K.3.30 deals with the analytical basis for recovery of natural circulation, long term cooling and training for these events. Toledo Edison proposes to bench-mark our best estimate codes with Integral System Test (IST) data from the GERDA (Geradrohr Dampferzeuger Anlage which is German for " Straight - Tube Steam Generator Test Facility") SB LOCA test facility. This facility was designed to provide better understanding of the longer term response of the B&W system. The inclusion of GERDA test data should also alleviate the general uneasiness regarding the need for improved understanding of the B&W design which was expressed by the staff in our meetings. GERDA
'will provide test data for natural circulation, interruption of natural circulation, the transition to boiler-condenser mode of cooling and the long term cooling of the system. This additional data should provide the S6aff with sufficient information to resolve the single non-safety related concern (i.e. " Bubble Dynamics") raised in the staff's review of Toledo Edison's program to resolve NUREG-0737 Item II.K.3.30.
THE TOLEDO EDISON COMPANY EDISON PLAZA 300 MADISON AVENUE TOLEDO. OHIO 43652 i O
I O208310192 820825 PDR ADOCK 05000346 I
Docket No. 50-366 License No. NPF-3 Serial No. 848 August 25, 1982 Page 2 Toledo Edison is not willing to commit to an open ended test program, but we do recognize that issues may be identified as data is developed which require further evaluation. We propose to evaluate any issues which arise and to take appropriate action for their resolution.
The following is more detail on the support for this position.
Background
Following the accident at TMI-2, the NRC required that further small break LOCA analyses be performed and that operator guidelines for managing small break loss of coolant transients be developed. The results of this work were documented by B&W in the May 7, 1979 " Blue Books".
In their review documented in NUREG-0565, the NRC concluded that while there was not a safety concern, certain features of the B&W SC LOCA Evaluation Model required more extensive verification.
In general, the recommendations were:
1.
Additional code model predicitions of Semiscale and LOFT experiments should be performed.
2.
The SB LOCA methods should be revised to address their specific concerns.
In addition, the licensees should verify the analysis models with appropriate integral system data.
These recommendations were implemented as requirements in NUREG-0737, Item II.K.3.30 and the following describes our actions towards resolution of this item.
Discussion Toledo Edison and the B&W Owners have taken several actions in responding to these recommendations.
In yesponse tg recommendation 1, computer gode simulations of LOFT tests L3-1 and L3-6 and Semiscale test S-07-10D were submitted. The B&W simulation results compared well with the test data and the simulations presented by other Vendors.
Since configurations tested in Semiscale and LOFT do not reflect all plant designs and arrangements, the acceptance by the Staff of benchmarks by other Vendors would seem to be also applicable to B&W benchmarks of the same tests as adequate testing of computer codes used in SB LOCA calculations.
Prior to any action to respond to the SB LOCA issues in NUREG-0565, Toledo Edison and the B&W Owners met with the Staff on December 16, 1980 to obtain a better quantification of the Staff's issues relative to NUREG-0737 Item II.K.3.30.
TgeStaff'sissueswerespecifiedintheStaff minutes of that meeting On May 12, 1981, Toledo Edison and the B&W Owners again met with the Staff, to present the program designed to address the issues of reference
Docket No. 50-346 License No. NPF-3 Serial No. 848 August 25, 1982 Page 3 4.
The Staff concluded that eight of the nine issues would be resolvbd by the implementation of the program presented but that IST data would be
- 1 rec,uired before II.K.3.30 could be signed off by the Staff. Attachment,he details the response to each of the nine items in reference 4.
During t main meeting the Staff raised a number of issues over and above those originally quantified as II.K.3.30 issues.
Following this meeting and for several months thereafter, a continuing technical dialogue was held between Toledo Edison and the Owners and the Staff in an effort to obtain and understand a complete list of specific issues.
Finally, in a meeting on October 23, 1981 with B&W Utility Executives, the Staff identified the issues as uncertainties regarding hot leg " bubble dynamics" during the transition from natural circulation to the boiler-condenser mode.
From that meating, the Staff agreed to participate in an in-depth review of the then current Babcock & Wilcox Small Break LOCA Methods Program, including the verification base. At the same time Toledo Edison and the B&W Owners agreed to participate in a joint effort with the Staff to assure that current Small Break LOCA method programs are fully understood.
The program was to include the following:
Staff to identify issues.
Code parameters, models, assumptions, etc., which are important in controlling dynamics of interest will be identified and available experimental data substantiating their validity will be reviewed. This would be done using results of the improved evaluation model in order that the most accurate dynamic response characteristics are reviewed.
Additional existing experimental data, from separate effects or integral tests, will be identified which address specific technical gaps, if any.
Identify where and how additional experimental data may be obtained, if any is required, i
Toledo Edison and the Owners B&W Owners set a meeting with the Staff for
(
December 16 and 17, 1981, to implement this commitment. Toledo Edison and l
the Owners came to that meeting prepared to address the single identified NRC concern on " bubble dynamics" and the CRAFT code. The Staff expected to be presented with a test program and the meeting ended in an impasse.
In a letter to the Staff on February 5, 1982, the Owners again set a meeting to discuss:
phenomena of bubble dynamics sensitivity of the system to decay heat, number of HPI' pumps, phase slip, and interphase heat transfer
Docket No. 50-346 License No. NPF-3 Serial No. 848 August 25, 1982 Page 4 discussion of benchmarks On April 9, 1982, six reports were hand delivered to the Staff for review prior to the April 16, 1982, meeting with Toledo Edison and the Ownere.
Attachment #2 to this letter provides a brief description of these reports.
In the period between February and April, the Staff again expanded issues outside of II.K.3.30 (reference 5).
Since Toledo Edison and the Owners were involved in an intensive effort to produce documents in response to the identified focused issue of " bubble u; amics", it was not possible to address the items in reference 5 specifically in the April 16 meeting.
The presentations in the April 16 meeting were perceived by Toledo Edison as being well received by the Staff and to date no negative comments have been received from the Staff on that meeting. We have since addressed these issues (Attachment #3). Toledo Edison endorses the items in this attachment.
At the conclusion of the April 16 meeting, the issues could clearly be separated into two parts. One part deals with the assurance of core cooling (10CFR50, Appendix K) which is safety related. The second part deals with the analytical basis for recovery of natural circulation, long term cooling, and training for these events using best estimate codes which has never been defined as a safety issue.
At this time Toledo Edison and the B&W Owners began to develop the program described below for acquiring GERDA data to benchmark best estimate codes.
Summary Toledo Edison is continuing work to address II.K.3.30 with the SB LOCA Methods Program described to the Staff and with the six reports described in Attachment #2.
We further offer to benchmark best estimate codes with GERDA test data to orovide better Staff understanding of the concerns in reference 5 which are outside of NUREG-0737 Item II.K.3.30.
We believe that GERDA is a technically acceptable test effort to address the phenomenon associated with recovery from a small break and offers a unique way to benchmark several of these phenomenon as they interrelate - that is, GERDA is an IST focused on the longer term natural circulation phenomena of the B&W design and it has been designed to specifically address the " bubble dynamics" issue on B&W designed plants. We provided the Staff with technical presentations on the design of GERDA at the Alliance Research Center on July 7, 1982, and followed with a tour of the facility.
The majority of Staff comments were favorable during and immediately following the presentation. However, a very negative comment was made by the Staff in the July 20, 1982, meeting with the Executives. We would be happy to address any technical questions the Staff or their consultants might have regarding GERDA and the test programs at each facility. B&W at the request of the B&W Owners will be sending you, under separate cover, a description of the GERDA test program.
Docket No. 50-346 License No. NPF-3 Serial No. 848 August 25, 1982 Page 5 We view our IST test program as a the final element in addressing the
" bubble dynamics" issue raised by the Staff during their review of the II.K.3.30 SB LOCA program and as a source of useful data to address other issues. Toledo Edison does not regard the IST as being needed to resolve the II.K.3.30 issue. These tests will be used as the bridge in the next logical step towards identifying any residual need for additional or modified test facilities.
We, therefore, invite the Staff to consider our test progcam as the means to minimize limited owner and staff resources while enhancing the knowledge of the B&W system.
We intend to provide a follow-up letter within the next three weeks which will provide additional details and milestones which we intend to pur;ue.
Very truly yours, RPC:TJM: lab cc:
DB-1 NRC Resident Inspector L
References 1.
"B&W's Post Test Evaluation of LOFT Test L3-1", Document No. 51-1125988-00, May 1981.
2.
"B&W's Best Estimate Prediction of the LOFT L3-6 Nuclear Small Break Test Using the CRAFT 2 Computer Code", Document No. 12-1124993-01, March, 1981.
3.
"B&W's Post Test Analysis for Semiscale Test S-07-10D", Document No.
86-1125888-00, May, 1981.
4.
Summary of Meeting with the B&W Owners Group Concernig the Abnormal Transient Operating Guidelines (AT0G) Program and TMI Action item II.K.3.30 Small Break Loss of Coolant Accident Models (Decamber 16,1980).
5.
Letter f rom Eisenhut to Mattimoe, March 25, 1982, Docket No. 50-312, Subj ect:
Need for Model Verification.
4
--rq-(
ATTACHMENT #1 Nine areas of concern for II.K.3.30 were identified in the meeting of December 16, 1980 between the Staff and B&W Owners.
These concerns are repeated below as found in the minutes of that meeting prepared by Mr. Throm of the Reactor Systems Branch.
Owner responses to each concern are also included.
1.
NEED TO VERIFY THE CURRENT NON-CONDENSIBLE MODEL AND THE CONSERVATISM 0F THE CONDENSATION HEAT TRANSFER RATE IN THE STEAM GENERATOR.
a) Report has been prepared describing a method to predict the amount of non-condensible gases in the primary system, including gas produced via radiolytic decomposition which may be released during a SBLOCA.
This report will be submitted to the NRC in August 1982.
b) A non-condensible gas heat removal model has been prepared and incorporated into the CRAFT code.
This model is described in the revision to the CRAFT Topical Report scheduled for submittal to the Staf f in September 1982.
2.
NEED TO VERIFY THE NON-EQUILIBRIUM MODEL AND TO JUSTIFY THAT THE AMOUNT OF ECCS WATER INJECTED IS CONSERVATIVE.
a) Report has been prepared and will be submitted to the Staff in August which justifies the current B&W ECCS evaluation model which utilizes CFT injection into the lower downcomer region.
b) This work was discussed with the Staff in the technical presentations on December 16, 1981.
3.
NEED TO DISCUSS THE PRESSURIZER MODEL AND THE EFFECTS OF A NON-EQUILIBRIUM MODEL.
a) A non-equilibrium pressurizer model has been incorporated into the CRAFT code.
This model will be addressed in the revised CRAFT Topical Report to be submitted to the Staff in September 1982.
This model was discussed with the Staff on December 16, 1981.
b) The surge line model was discussed with the Staff on December 16.
The open question from the Staff will be addressed in a written response in September 1982.
4.
NEED TO ADDRESS THE FORMATION OF A STEAM BUBBLE IN THE HOT LEG " CANDY CANE".
(IS IT A REAL OR CALCULATED PHENOMENON?)
EXPERIMENTAL VERIFICATION BELIEVED NECESSARY.
a) This'is addressed in several parts of the SBLOCA Methods Program:
e System modeling study (steam generator, hot leg, and reactor vessel head) e Steam generator and pressurizer model changes i
- _-. )
ATTACHMENT #1 (cont'd) b) The joint NRC/0wners testing evaluation task concentrated on this issue.
Documents described in Attachment #2 support the evaluation of this concern, and the report on " Bubble Dynamics" specifically addresses this Concern.
5.
THE STAFF INDICATED THAT A MECHANISTIC MODEL OF THE STEAM GENERATOR HEAT TRANSFER SHOULD BE DEVELOPED. A BEST ESTIMATE OR VERIFIED CONSERVATIVE MODEL WOULD BE ACCEPTABLE.
a) The steam generator model has been upgraded and will be described in the revision of the CRAFT Topical Report to be issued to the Staff in September 1982.
b) Steam generator model was presented to the Staff in the December 16, 1981 meeting.
6.
AS PART OF THE ADDITIONAL SYSTEMS VERIFICATION NEEDED, THE FOLLOWING SEMISCALE AND LOFT TESTS SHOULD BE CONSIDERED:
SEMISCALE S-07-100, LOFT L3-1, L3-5, AND L3-6.
a) The Owners considered the above tests and provided the Staff post test evaluations of L3-1, L3-6, and S-07-10D (References 1, 2, and 3 to this letter).
7.
THE OVERALL THERMAL-HYDRAULIC BEHAVIOR OF THE CORE DURING UNC0VERY SHOULD BE VERIFIED AGAINST APPLICABLE EXPERIMENTAL DATA, PARTICULARLY THE RECENT ORNL DATA.
a) ORNL data has been used to show that the current application of the Ditters-Boelter correlation is conservative.
Data was discussed with the Staff on December 16, 1981, and a report will be provided to the Staff in August 1982.
8.
THE INFLUENCE OF METAL HL1T ON THE SYSTEM PRESSURE RESPONSE, PARTICULARLY ON THE TIME OF ECCS INJECTION, WAS IDENTIFIED AS AN AREA 0F CONCERN AND SHOULD BE SHOWN TO BE PROPERLY CONSIDERED IN THE ANALYSIS MODELS.
a) The BAW ECCS Evaluation Model currently accounts for metal heat and no change needs to be made.
9.
THE BREAK FLOW MODEL NEEDS TO BE CONFIRMED.
THE USE DF COMBINED MODELS WITH VARIOUS DISCHARGE COEFFICIENTS APPLIED TO THEM NEEDS TO BE COMPARED TO A BEST ESTIMATE MODEL TO DEMONSTRATE CONSERVATISMS.
a) The existing leak discharge model has been found to produce results which are similar to yet still conservative with respect to those obtained with the best estimate model.
j b) The work was discussed with the Staff on December 16, 1981 and the report j
will be provided to the Staff in August 1982.
1 h
4 i
ATTACHMENT #2 Documents prepared and submitted to the Staff from the B&W Owners' participation in the joint test evaluation task with the NRC.
"The GERDA Test Facility" This report was prepared in fulfillment of the October 23 commitment by B&W.
" CRAFT 2 Prediction of ARC Loss-of-Feedwater Test", 12-1132544-00,
, April 1982 This report shows that the revised steam generator model adequately predicts the temporal response of key once-through steam generator parameters af ter a complete loss of feedwater.
" Auxiliary Feedwater Penetration", 12-1132513-00, April 1982 "Auxilia ry Feedwater Axial Flow Distribution, 12-1132543-00, April 1982 The first report describes the calculation model and testing basis for the penetration of the auxiliary feedwater in the OTSG, and the second report uses this model and shows how the axial flow distribution was derived f rom F0AK testing at Oconee 1.
" Benchmarks for AFW Models", 12-1132555-00, April 1982 This report contains the benchmark results of the AFW models against actual plant data fran four plant transients.
The ability to predict plant response following loss of offsite power for the extreme conditions under which the AFW system will function is demonstrated in this report.
" Bubble Dynamics", 12-1132565-00, April 1982 This report is focused on the main phenomenological aspects of steam in the hot leg "U" bend and addresses test data and engineering evaluation used to understand " bubble dynamics".
Based upon the focused Staff concern on the dynamics of a trapped steam bubble in the inverted U-bend of the hot legs, two issues were identified:
- 1. 'During the blowdown portion of the transient, does the code properly ll[
predict the fonnation of the steam bubble and its resultant
);
interruption in natural circulation?
2.
During the system refill phase of the transient, how does the trapped steam bubble behave?
.i l!
I; I
I sc2r r
ATTACHMENT #2 (cont'd)
In addressing these issues, a review of the calculated plant response was performed in order to assess the controlling phenomena. As a result of that review, it was determined that the governing phenomena were:
1.
Interruption in Natural Circulation
- Spatial heat transfer in the steam generator
- Distribution of steam flow from the core
- Phase slip within the hot leg
- Steam condensation in the steam generator 2.
System Recovery Phase
- Steam condensation on steam-liquid interface Test data supporting the modeling of these phenomena has been evaluated and reported in the documents listed above.
Further understanding of the plant response is provided in a qualitative assessment of plant behavior to various input and modeling assumptions contained in this report.
It is clear that the concern on the interruption of natural circulation is a byproduct of the Appendix K assumption on HPI flow.
Using the single failure assumption of Appendix K, it is shown in this report that phase slip modeling is important to the development of the plant response.
Phase slip modeling is a part of the current SBLOCA Methods Program.
The adequacy of current phase slip modeling was shown in the evaluation of test data discussed in the April 16 meeting with the Staff and summarized in this report.
l l
l l
l l
T wm,,
,m
ATTACHMENT #3 Responses to the Eisenhut to Mattimoe letter of March 25, 1982.
1.
Interruption of Natural Circulation e Branch Flow The effect of preferential steam flow to the hot leg or the RV head has been addressed in the "dubble Dynamic" report (see Attachment #2).
Branch flow was discussed with the Staff in the April 16, 1982 meeting.
e Hot Leg Flow Regime This was addressed in the Slip model presentation to the Staff on April 16,1982 and is discussed in the report " Bubble Dynamics" (see Attachment #2).
2.
Cold Leg Thermal Shock The concern over cold leg thermal shock was derived, as we understand, from TRAC computer calculations perfonned by LASL for the Staff wherein significant cyclic temperature variations were shown in the vicinity of the cold leg ECC injection.
We encourage the Staff to have an independent QA performed on these calculations by an organization familiar with the hardware and components of the B&W designed systen If the cyclic behavior is confinned, programs are al ready in place to address thermal shock and this item would be included in that ef fort.
3.
Hydraulic Stability Following Accident Recovery This concern is addressed in the report " Bubble Dynamics" and was discussed with the Staff on April 16, 1982.
In addition, the presentation given in that meeting, " Steam Condensation on Steam-Liquid Interface", also addresses the governing phenanenon in the recovery phase.
Other concerns in the March 25 letter were:
break isolation, steam generator tube rupture, and cooldown and depressurization following a SBLOCA.
These concerns are covered by the ATOG Guidelines and some are specific per plant type.
Further discussion on these items is expected but not as a part of 11.K.3.30.
1,
.