ML20083P828
| ML20083P828 | |
| Person / Time | |
|---|---|
| Issue date: | 02/28/1982 |
| From: | Marino G, Picklesimer M, Silberberg M, Van Houten R, Wright R NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | |
| Shared Package | |
| ML20083P812 | List: |
| References | |
| FOIA-84-119 NUREG-0840, NUREG-0840-V01, NUREG-840, NUREG-840-V1, NUDOCS 8404200196 | |
| Download: ML20083P828 (60) | |
Text
l NUREG-0840 l
Vol.1 l
l Program on Behavior of Damaged Fuel Report of the NRC Fuel Testing Task Force U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research G. Marino, R. Wright, M. Silberberg, M. Picklesimer, R. VanHouten p e n%q s
s s
84j4 g 96 840307 SHOLLY84-119 PDR
NOTICE Availability of Reference Materials Cited in NRC Publications Most dc :uments cited in NRC publications will be available from one of the following sources:
- 1. The NRC Public Document Room,1717 H Str tet, N.W.
Washington, DC 20555
- 2. The NRC/GPO Sales Program, U.S. Nuclear Regulatory Commission, I
Washington, DC 20555
- 3. The National Technical Information Service, Springfield, VA 22161 Although the listing that follows represents the majority of documents cited in NRC publications, it is not intended to be exhaustive.
Referenced documents available for inspection and copying for a fee from the NRC Public Docu-ment Room include NRC correspondence and internal NRC memoranda; NRC Office of Inspection and Enforcement bulletins, circulars, information notices, inspection and investigation notices; Licensee Event Reports; vendor reports and correspondence; Commission papers;and applicant and licensee documents and correspondence.
The following documents in the NUREG series are available for purchase from the NRC/GPO Sales Program: formal NRC staff and contractor reports, NRC-sponsored conference proceedings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code of Federal Regulations, and Nuclear Regulatory Commission Issuances.
Documents available from the National Technical Information Service include NUREG series reports and technical reports prepared by other federal agencies and reports prepared by the Atomic Energy Commission, forerunner agency to the Nuclear Regulatory Commission.
Documents available from public and special technical libraries include all open literature items, such as books, journal and periodical articles, and transactions. Federal Register notices, federal and state legislation, and congressional reports can usually be obtained from these libraries.
Documents such as theses, dissertations, foreign reports and translations,and non-NRC conference proceedings are available for purchase from the organization sponsoring the publication cited.
Single copies of NRC draf t reports are available free upon written request to the Division of Tech-nical Information and Document Control, U.S. Nuclear Regulatory Commission, Washington, DC 20555.
Copies of industry codes and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, and are available there for reference use by the public. Codes and standards are usually copyrighted and may be purchased from the originating organization or, if they are American Nationa1 Standards, from the American National Standards institute,1430 Broadway, New York, NY 10018.
I GPO Printed copy pnce: $4.50 l
l
NUREG-0840 Vol.1 Program on Behavior of Damaged Fuel Report of the NRC Fuel Testing Task Force Manuscript Completed: December 1981 Date Published: February 1982 G. Marino, R. Wright, M. Silberberg, M. Picklesimer, R. VanHouten
/
Division of Accident Evaluation Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, D.C. 20555 fo..,
NhYk.
l ABSTRACT An NRC Fuel Testing Task Force was appointed in March 1981 by the Director, Office of Nuclear Regulatory Research, to review and evaluate the projected information needs, program, and test facility resource requirements for research on the behavior of fuel assemblies during severe accidents, e.g.,
severe fuel damage (SFD). The objective of the SFD program is to provide the experimental data base and analytical methodology for understanding and predicting core behavior under severe accident sequence conditions. The Power Burst Facility and AnnularCore Research Reactor represent the major reactor test facilities to be used for the SFD program. The report considers the following technical and programmatic details:
Programmatic NRC needs in severe fuel damage, phenomenological research needs, severe fuel damage program schedule and its relationship to rulemaking, role of related foreign and U.S. programs. The report has been organized into two volumes, the main report (Vol. 1) and Volume 2, which contains additional technical and background information on severe fuel damage processes, research needs, and facility descriptions. The report notes that information from the TMI-2 core examina-tion will furnish an invaluable benchmark for severe fuel damage processes and code analyses.
iii
J NRC FUEL TESTING TASK FORCE TASK FORCE MEMBERS NRC/RES G. Marino*
R. Wright
- M. Silberberg*
M. Picklesimer*
R. Van Houten*
OTHERS R. Chapman, ORNL E. Courtright, PNL P. MacDonald, EG&G D. Morrison, IITRI A. Pressesky, DOE /HQ J. Scott, LANL J. Walker, SNL
- Contributing Authors to Main Report (Volume I) 1
TABLE OF CONTENTS
1.0 INTRODUCTION
AND
SUMMARY
1.1 Introduction and Statement of Problem 1.2 Program Objective and Approach 1.3 Summary Task Force Findings, Conclusions, and Recommendations 1.4 Summary of SFD Program, Facilities, and Schedule 2.0 SEVERE FUEL DAMAGE PROGRAM - NRC NEEDS 2.1 NRC Needs - A Generic Perspective 2.2 TMI Action Plan Needs 2.3 Current Perception of Rulemaking Plans and Needs 3.0 SEVERE FUEL DAMAGE - CURRENT AND NEEDED INFORMATION 3.1 Damage Regime 1 3.2 Damage Regime II 3.3 Damage Regime III 3.4 Damage Regime IV 3.5 Damage Regime V 3.6 Core Debris Coolability - Current State of Knowledge 4.0 CURRENT AND PLANNED NRC PROGRAM ON SFD 4.1 Program Approach 4.2 Analysis 4.3 Integral SFD tests (In-Pile and Ex-Pile) 4.4 Phenomenological Separate-Effects Experiments 4.5 Foreign Programs on Severe Fuel Damage 5.0 ADDITIONAL CONCLUSIONS 5.1 Need for a SFD Program 5.2 Identification of Specific SFD Research Needs 5.3 Program Technical Approach 5.4 Program Scope and Balance of Emphasis 5.5 Selection of Program Facilities and Projections of Future Needs e
5.6 SFD Tests in LOFT 5.7 Instrument Capabilities for SFD Experiments 5.8 Program Schedule 5.9 Impact of Budget Constraints 5.10 Interfaces with other RES and NRR Programs 5.11 Foreign Programs in SFD Appendicies A.
Task Force Objective, Scope of Work, and Activities B.
Response to Comments on Draft Report vil l
FORWARD The NRC Fuel Testing Task Force was established by the Office of Nuclear Regulatory Research to evaluate the projected information needs, program, and test facility resource requirements for research on the behavior of fuel during severe accidents, e.g., severe fuel damage (SFD). The draft task force report was issued in July 1981 for review by NRC staff and experts in the U.S. and abroad, and discussed at a peer review meeting in August 1981.
The draft report was also discussed with the ACRS subcommittee on Reactor Fuels in July 1981. Written comments were received from some of the reviewers and considered in the final draft. We appreciate the interest, participa-tion, and contributions of the reviewers and look forward to our continuing discussions on the planning and progress of the SFD program.
The final version of NUREG-0840 has been organized into two volumes, the main j
report (Volume I) and Volume II which contains additional technical and l
background information on SFD processes, research needs, and facility descrip-tions which were provided to the Fuel Testing Task Force.
The program described in this report is based upon the recommendations of the j
Task Force and is provided for the information of the Office of Nuclear Regulatory Research for the purpose of program planning. The actual program carried out will be based upon many factors, including:
regulatory informa-tion needs and priorities, budget constraints, and specific results and findings obtained during the course of the research.
In this research, we intend to review the progress of the integrated SFD program annually and make adjustments when appropriate.
At the time of publication of NUREG-0840, a draft program plan for nuclear plant severe accident research (NUREG-0900) was in preparation in support of policy determinations and possible future regulatory actions.
The technical scope and estimated schedule for the SFD program, as described in Draft NUREG-0900, was derived from NUREG-0840 with some changes to reflect program-matic decisions which have evolved since August 1981.
The integrated SFD program will provide the data base and methodology to fill an important gap in our generic understanding of accidents involving severe core da.nage (e.g.,
TMI-2).
Information from the SFD program will furnish technical bases for significant developments in the following important applications:
improved PRA, severe accident sequence analysis, assessment of alternative accident management and recovery strategies, radiological source term, and the inter-pretation of the TMI-2 core examination.
Finally, I would like to acknowledge the members of the Fuel Testing Task Force and the contributing authors for their special effort and support in the development and preparation of this important document.
M. Silberberg, Chairman NRC Fuel Testing Task Force l
ix
i i
1.0 INTRODUCTION
AND
SUMMARY
1.1 Introduction and Statement of Problem The accident at Three-Mile Island Nuclear Station-2 clearly demonstrated the following points:
o Severe core damage resulting in large hydrogen and fission product releases to the containment can occur despite current regulatory pro-cedures and engineered safety systems.
o Accidents that result in core temperatures in excess of 2200 F need not result in a massive core melt or pressure vessel failure as has been tacitly assumed (because of the lack of useful data) in the past.
o Some severe core damage accidents can be managed successfully and the reactors brought to a safe shutdown, o
The fomulation of new requirements and procedures which may be needed to manage and/or mitigate the consequences of such accidents require the development of a data base and analytical methodology far beyond that needed for current design-basis-accidents.
Accordingly, the NRC is currently involved in deliberations directed at determining the need for fundamental changes to regulatory policies which have guided nuclear plant design and siting over the past 25 years.
In particular, the Commission has initiated actions for:
(1) determining require-ments for reactor sites and emergency plans and (2) determining regulatory requirements for severe accidents.
The research program of RES has an important role to play in these areas by providing data and analytical methodology to support the resulting require-ments of the evolving regulatory process.
In order to assure that the RES research program will be responsive and timely, it has been necessary to reevaluate on-going research and, where appropriate, to implement significant redirection in the planning and execution of the program. This reevaluation has been part of the NRC implementation of the lessons learned from TMI-2.
Various Commission, Office, and staff-level documents have furnished bases and guidance for program planning. Some of these documents include:
the Task Action Plan, the RES Long Range Research Plan, ACRS Reports to Congress on the NRC Research Program, and most recently NUREG-0772, " Technical Bases for Estimating Fission Product Release During LWR Accidents." The results of these programs will be strongly coordinated with the SFD program, and where possible, the maximum amount of fission product behavior data will be extracted from the SFD program and integrated with the fission product program results, thereby providing a final package of data on fission product behavior during degraded core accidents.
Since TMI-2, the research program of the Fuel Behavior Branch has been under-going significant redirection, from concern regarding fuel behavior during operating transients within the Appendix K rules to concern regarding the behavior of fuel during severe accidents, e.g., severe fuel damage (SFD).
Starting in FY 1980 and continuing into FY.1981, the SFD program has evolved 1-1
l-2 into the formative and detailed planning stages.
In order to better coordi-nate and broaden the effort, the Director of RES requested that the Director of RSR establish a task force to review and evaluate the research necessary to meet the projected needs of the agency as related to SFD studies.
1.2 Program Objective and Approach The stated objective of the SFD program, as considered by the Task Force, is to provice the experimental data base and analytical methodology for under-standing and predicting core behavior under severe accident sequence condi-tions which can be used in addressing the following questions:
1.
What are the physical and chemical states of a reactor core at any point in time af ter fuel rod surface temperatures have exceeded 2200 F?
2.
Is the resultant severely-damaged core coolable by reflooding for the relevant range of fuel damage scenarios and conditions?
3.
Which generic accident management procedures, safety systems, instruments, and diagnostic information are required to terminate the accident at different points in the accident sequence?
Given the answer to question 1, the other two questions can be answered with reasonable accuracy by the proper utilization of controlled separate effects experiments and properly verified analysis methodology.
The importance of this conclusion is illustrated by the following points:
(1) the answer to question 2 is critically important in assessing the overall risk for a given accident sequence since the Reactor Safety Study (WASH-1400) has shown that those accident sequences that result in core melt represent the greatest contributors to risk, and (2) an accurate assessment of question 3 will provide the necessary information required for rulemaking actions on accident management and minimun engineered safety features. Therefore, the initial research should be focused on question 1 with emphasis on prototypical scop-ing tests and accompanying analysis codes which will reveal the full extent and limiting states of severely damaged cores.
The almost unbounded number of possible scenarios leading to severe core damage may lead one to think that the problem is insurmountable; however, there are limits on those para-meters which determine the final state of'a damaged core, and if the research program is designed to encompass the realistic ranges of those parameters, definite bounds on the state of the core can be obtained.
Given such bounds, supplementary experiments and analyses can be applied to determine core coolability, fission product and hydrogen release rates, and the risk to the public for a given accident sequence.
The question is then reduced to assess-ing the most important parameters (and their limits) which determine the state of the core at any point in time during the accident sequence.
These parameters are discussed in detail in Section 3.0 along with the important phenomena that determine the state of the core.
1-3 As an oversimplified example of the process by which information from this program could be used in accident management, consider an event in which it was impossible to prevent core uncovery to a depth of x feet for t minutes.
Such information could be correlated with the SFD-generated information in the form of a chart depicting the expected worst-case length of the core embrittled to thermal shock as a function of x and t.
Experimentally unsup-ported calculations made today show that, in such an event, the length of core above the water level that is not embrittled may be as much as 3-4 feet.
Thus, the operator could safely reflood to (x + 4) feet without shattering any significant part of the core and, at the same time, cause an increase in steaming rate (due to covering more of the core) to reduce the temperature of the embrittled part of the core to a safe and controllable temperature (i.e.,
less than 2000 F).
The operators could then continue reflood at a very slow controlled reflood rate to minimize further damage and prevent as much as possible additional core shattering.
The only information required by the operators for such a recovery scenario would be the water level and time of uncovery; no calculations would be needed.
Charts showing the expected length of core embrittlement and the immediately refloodable distance for parametric variations in uncovery rate and decay-heat level could be con-structed today. However, such charts would require verification by in-pile prototypic testing. Moreover, the effects of slow reflooding on the presumed much-cooler portions of the embrittled length of the core would have to be determined by in-pile separate-effects tests for several initial conditions.
The above discussion in no way recommends that the NRC develop such a procedure for reactor operations.
Its inclusion in this report is only for demonstra-tion purposes on how the SFD-generated data and codes could be used in acci-dent management and post-accident recover scenarios. Many other possibilities for use of these data exist, some of which may be more tractable than the example above.
A four-part integrated program of research is proposed to provide the needed information base (data and verified models).
The first part consists of integral, multi-effects, in-pile tests in the PBF to provide early scoping data on governing phenomena, and later for proof tests of the severe fuel damage models and codes developed in the programi The second part consists of separate-effects experiments on the governing phenomena, both in the ACRR test reactor and in the laboratory, to furnish a data base for model develop-ment. An analysis package is the third part of 'the integrated program, including development of severe fuel damage models from the experimental data base and their integration into a severe fuel damage code.
There will.be continuous active interaction and feedback between the analysis and experi-mental programs.
The fourth part involves the information to be obtained from the TMI-2 core examination.
Given that the objective of the program is realized, the ultimate benefits of the program to the NRC will include:
1.
An understanding of SFD phenomena that are of importance in determining the performance requirements for engineered safety features for accident management for recovery from core damage events leading to in-vessel termination.
l
l-4 2.
A base of technical support for policy determinations on regulatory requirements for severe accidents.
3.
Guidelines for the acceptance criteria for use by the staff in reviewing documentation submitted by licensees in meeting regulatory requirements.
An additional, more general benefit from the program will be to provide the necessary SFD data base for more accurate calculations of the true risk to the public for a given accident scenario. Detailed knowledge of degraded core behavior will enable risk analysts to include risk reduction computa-tions resulting from recovery of previously failed safety systems and/or operator actions during a severe accident sequence. Also, the state of the core after a given sequence may be considerably more benign than that cur-rently assumed because of a lack of data available on core degradation pro-cesses.
Finally, it must be emphasized that although this program addresses accident management and degraded core behavior, it does not lessen the need for, nor should it detract from, strongly focused efforts in accident pre-vention to reduce the probability of a severe accident through improved systems reliability.
1.3 Summary of Task Force Findings, Conclusions, and, Recommendations 1.3.1 Findings and Conclusions o
The SFD program is focused on the end-product needs and the information needed for severe accident mar.igement procedures and requirements. This can only be accomplished if the SFD program and its important research and regulatory interfaces in the NRC are integrated in terms of real-istic objectives, schedules, and technical milestones.
o The major test facilities and key contractor personnel needed to conduct a successful SFD program are available.
Test technology is also avail-able from the current NRC/RES LWR and fast reactor safety research programs. Moreover, no.equirements were identified for new test reactor facilities for SFD testing; it appears that some modest upgrading and extension of the capabilities of several existing test facilities may be necessary, will be cost effective, and will satisfy program requirements.
o The most productive and cost-effective approach to resolving key issues is the integrated program, described in Section 1.4.
o The highest priority element for the program, at this time, can be identified: the prompt execution of the Phase 1 SFD series of five integral tests in PBF. The Phase 1 PBF test serics furnishes the scop-ing data base for the SFD program, and is on the critical path to a successful program.
o Information from the TMI-2 core examination, when available, will furnish an invaluable benchmark for SFD processes and code analyses.
It repre-sents the only whole core data in SFD and will serve to evaluate all SFD correlations as to applicability to real accidents. We are concerned about the lacx of progress towards the examination of the damaged TMI-2 core because of various delays. The TMI-2 core examination must be put
1-5 on a very high priority basis.
It should also be noted that information obtained from the SFD program during FY 82 and FY 83 can play a vital role in the detailed planning and implementation of the core examination to maximize the useful data base and to assure proper interpretation of the data.
o
'Je do not have a sufficient technical base, at this time, to judge whether later SFD proof tests in LOFT are needed and if they would be cost effective. A potentially important objective of such proof tests would be integral testing of SFD accident manage.aent effectiveness.
It is our understanding that the LOFT facility will not be available based upon current policy decisions.
o Currently available state-of-the-art instrumentation is sufficient to meet programmatic goals; however, some modest improvement of existing capability and limited development in several areas (such as fission product release) would significantly improve the results of the SFD
- program, o
Although the phenomenological issues and unknown areas in SFD are reasonably well defined, it is more difficult to assess their relative importance and priorities because of limitations in the existing data and current methodology for assessing severe accidents.
Therefore, to assure adequate programmatic balance, semi-quantitative, analytical sensitivity studies are needed early in the program and should be given high priority, o
In addition to the elements described above, it may be necessary, later in the program, to address the question of in-vessel core melt progres-sion leading to vessel failure.
o The SFD program is currently front-end funding limited (e.g., FY 82-83) because of prior budgetary process decisions.
However, efforts are continuing to provide additional funds, so that the program can provide key information for severe accident management and improved risk assess-ment for timely application to regulatory requirements for severe accidents.
o The urgency of the funding problem in FY 82 demonstrates that the program is cost effective at the projected level of about $25M per year and sub-marginal at significantly lower funding levels. A key finding is'a need l
to address the FY 82 funding issue (particularly for PBF and ACRR).
Important secondary sources of support include foreign programs such as in the FRG and elsewhere which can complement the NRC program in signi-l ficant ways through in-kind and financial contributions.
l 1.3.2 Summary of Recommendations for Action l
Management Actions o
The highest priority should be to find increased funding for accelerating l
the Phase I, PBF-SFD test program.
l o
RES Management should give high priority to urging and supporting the program for the early removal of the TMI-2 core and its examination.
1-6 RES should seek a high degree of coordination with other NRC offices to o
integrate developing infomation needs with detailed planning and execution.
RES will attempt to provide a broader resource support base for the SFD o
program from financial and in-kind participation of other research organizations in the U.S. and abroad.
Program Actions The SFD program element priorities will be established on the basis of o
information needs for policy determinations on regulatory requirements for severe accidents and accident management. This can and will be done more effectively when the total NRC research program in severe accidents is better defined and integrated.
o The analytical development effort will be given increased support with high priority with emphasis on early sensitivity studies on the govern-ing phenomena.
o The planned acquisition of significant fission product release data in the PBF-SFD tests is being reviewed in detail to assure applicability of the data, data limitations, and proper integration into RES accident source term needs per NUREG-0772.
This review is being accomplished within the framework of the existing RES fission product release and transport programs.
o The SFD program will participate in and maintain a key presence in the technical discussions and decisions in the planning and implementation of the TMI-2 core examination. This will be accomplished through FBB membership in the GPU, EPRI, NRC, 00E (GEND) advisory committee.
o The Task Force report was given an independent peer review.
1.4 Summary of SFD Program, Facilities, and Schedule The SFD program is an integrated program that includes the following elements:
o Program planning, management, and coordination.
o Integral
- in-pile tests (PBF test reactor):
Initial scoping tests to determine overall behavior and the governing phenomena (Phase I - 5 tests).
Later verification tests of the integral codes developed in the program.
- Where multiple phenomena are interactive and are integrated in a test.
i-7 Separate-effects phenomenological experiments for the development and o
verification of models of the governing phenomena:
- In-pile experiments (i.e., debris ** formation, relocation, and coola-bility) - ACRR (20-30 tests),
Laboratory experiments (i.e., oxidation, clad ballooning, and material interactions).
o Analytical model development:
The integral code Severe Core Damage Analysis Package (SCDAP) for use in the analysis of accidents and Integral experiments.
Phenomenological models for use as modules in SCDAP and in the analysis of separate effects experiments.
o Analysis and characterization of the TMI-2 core debris from the DOE /
EPRI/NRC TMI-2 core examination program, o
Research products in the form of analysis codes (SCDAP) and documentation of all results.
The major programmatic questions to be addressed by the SFD research program, the technical information needed, and the experimental facilities to be used in their resolution are given in input-output matrix form in Table 1.
This table shows the integrated nature of the SFD research program with both integral multi-effect in-pile tests and complementary separate-effects phenom-enological experiments used to address nearly all of the technical issues.
Each of the major programmatic questions involves information on several of the technical issues and the expected results from more than one set of experiments or experimental facilities.
The major facility for the integral in-pile tests in the PBF, whereas, for the in-pile separate-effects phenom-enological experiments the major facility is the ACRR.
The capabilities of candidate test reactors for the SFD program, and of potential upgrades of these reactors, are given in Table 2.
The current program includes integral multi-effects tests in the PBF and the ESSOR reactor at Ispra and also phenomenological separate-effects experiments in ACRR.
In the future, integral proof tests of the SCDAP code and its phenomenological models may prove to be needed in an upgraded PBF or NRU.
Large-scale integral tests on melt-progression to reactor-vessel failure may be performed in the large Melt Facility (LMF) which can provide pours of up to 500kg of UO2 and Tn Phase II of the PBF program. The program also includes electrically-heated-bundle experiments on damaged clad behavior and laboratory experiments on materials properties and interactions.
- The use of the term " debris" refers to a broad range of damage fuel configurations in terms of shape and size.
w TABLE 1 Input-Output Matrix for,the Severe Fuel Damage Experfeent Program
- e,,g,#
sb g*d Te Answer These Quest 1ons:
s
- ls*
s' c
' \\*
'69 6 b
4" g*
L*
t* c
- o g
t 8g*Q /
Q,<*
s' o
,\\
f,so' 8 I:femat1on. From ---- >
t*
gp l
Sep. Ef. [Y g
9 y
Multi.-Ef.
l I
i 1.
C1rd Ballooning. Burst, and PSF-1. 2 ACRR Blockage mRu X
X X
X 2.-Ouidation(Hydrogen)
P8F-1. 2 ACRR
. nRu X
X X
X t,,
3.
Fission Product Release PBF-1. 2 Lab tad Attenuation X
7 4.
Fuel Debris Characteriaatfen P8F-1. 2 ACRR
- 5.. Fuel Debris Relocation.
PSF-1, 2 ACRR siockage mRu X
X X
X 6.
Reflood Debris Character-P8F-1. 2 ACRR izstica nRu X
X X
X X
7.
Aapid Steam Generation and PSF-2 ACRR-Emplosten LMF Lab X
X X
8.
Samaged Bundle Coolability PSF-2 750 X
X
- g. - Fuel Detris Coolabi11ty PSF-2 ACRR Lab X
X X
- 10. Post-Dry-Cut Behavfor ACRR Lab X
- 11. Nelt Progress 1on TBD
- 12.. D'abris Character 1astien at TBD TBO Vessel Failure X
o its data and models frem these esperiments are integrated into the SCBAP severe core damage code l
TABLE 2 l
CANDIDATE TEST REACTORS FOR SEVERE FUEL DAMAGE PROGRAM PBF PBF Upgrade ACRR ESSOR NRU Upgrade LOFT TMI-2 Debris t
ka Part of Curreat Program yes Future?
yes Future?
yes e
e
. av Type of Experiments Integral Int./ Proof Phenom.
Integral Int./ Proof Proof Benchmark Fission Sim. Dec. Heat yes yes yes yes yes yes Full F.P. Decay Heat yes
?
yes yes yes Full Power Irrad. 0 3 a/o yes
?
yes yes yes yes T
e Bundle Size - Max.
32 rod 17 x 17 32 pin 32 rod 76 rod Multi-Assem. Whole Core
-Core Length (m) 0.9
-0.9 0.5 2.0 3.3 1.5 3.7 On-Line Optical Diag.-
yes On-Line Fuel Imaging yes Tomographic n-rad yes yes v
Removable Test Assembly yes yes yes yes yes On-Site PIE yes yes yes yes yes yes Unit Test Cost Med.
High Low Med.
High V. High V.V. High
1-10 The phenomenological research needs in the SFD program and a matrix of important experiment parameters are presented in Sections 3 and 4.
An overview schedule of the SFD program is given in Figure 1, according to major NRC program elements and programs of other organizations.
The schedule shown is based upon baseline budgets for fiscal years 1982 and 1983, the projected budgets for later years, and the assumption of support for an accelerated schedule for the Phase I, PBF-SFD tests.
The initial version of the SCDAP code will be completed in FY 82 with yearly updated and expanded versions thereafter.
Phase I of the P8F test program and related analysis will be completed in FY 84. Phase II, PBF tests (yet to be defined) and related analysis are projected for FY 86 completion, along with the ACRR program of phenomeno-logical separate-effects experiments.
The schedule shown in Figure 1 is timely and responsive to the needs of severe accident policy determinations and related regulatory actions, projected at the time of publication of this report.
lSCDAP development l
-~
a a
B 8
PBF phase I expts PBF phase II l
oc E
- ToLCDA, L
TeSCDA*
'r i
ah ACRR debris formation & relocation expts r.srD.,
r.scD.,
N l
g ACRR debris coolability expts l
Laboratory separate effects expts I " "
"T severe core
'**j"
y dar age assmt
~
capability-7 TMI-2 core examination ve sco.7 - -
t I'**c"**
FRG/KFK laboratory separate effects expts o
E ESSOR SFD tests n.
5 ESSOR.SSTF (PBF phese I equivalent tests) i5 o
FY 1982 FY 1983 FY 1984 FY 1985 FY 1986 l
1 1
I i
i i
I FIGURE 1 OVERVIEW SCHEDULE
2.0 SEVERE FUEL DAMAGE PROGRAM - NRC NEEDS A detailed review of the phenomenological needs for SFD and the research programs designed to address those needs must be considered in the context of the agency's mission, in general, and specific regulatory process needs in particular.
Accordingly, the task force reviewed the NRC's information requirements in order to provide a basis for evaluating the content, schedule, and role of the program. Twa categories of documented, closely-related NRC needs were considered:
(a) NRC generic needs and (b) needs identified in the TMI Action Plan, including rulemaking.
2.1 NRC Needs - A Generic Perspective It was clear to the NRC and to the independent groups reviewing the accident at Three Mile Island that there was a need for an improved understanding of reactor behavior under conditions that involve severe core damage.
This was reflected in the Kemeny Commission report
- which noted, "We strongly urge that research be carried out promptly to identify and analyze the possible consequences of accidents leading to severe core damage.
Such knowledge is essential for coping with the results of future accidents.
It may also indicate weaknesses in present designs, whose corrections would be important for the prevention of serious accidents."
The recently released (September 1981) " Report of the Reactor Safety Research Review Group," chaired by N. C. Rasmussen and submitted to "The President's Nuclear Safety Oversight Committee" (NSOC), chaired by Governor Bruce Babbitt, said the following concerning hRC research needs:
"The Reactor Safety Study indicated that the small LOCA is a greater contributor to risk than the large LOCA.
The ability to analyze the small LOCA has not kept up with the ability to analyze the large LOCA, in part because of tne long computing machine runs required and the resulting buildup of calculational error.
In addition, the integral LOCA experiments in the NRC program were designed to test calculations of large LOCAs, and are not well suited to small LOCAs.
The amount of damage to the reactor core fron a small LOCA will be very sensitive to the extent and duration of uncovering of the fuel. Some level's of the boiling boundary below the top of the fuel can be tolerated for some time, but the specifics are still more uncertain tnan is desired. An NRC program to clear up the uncertainty is definitely needed."
"The NRC program on the small loss-of-coolant accident should be continued, to the point where capability to answer important questions of thermal-hydraulics and fuel performance is assured."
=" Report of the President's Commission on the Accfdent at Three Mile Island,"
Library of Congress Catalog Card Number 79-25694.
2-1 i
2-2 In the event of such an accident, effective and timely action "or accident management is required by regulatory and operational agencies. The TMI-2 experience clearly demonstrated that the scientific data base upon which effective management policies and operator actions should be based was in many cases incomplete or unavailable.
Furthermore, if safety systems are to be required in future operating plants to control and mitigate the conse-quences of or otherwise deal with severe accidents, information on the response of these systems is required.
In regard to these items, the NSOC report went on to list the following emphasized paragraphs:
"The correct operator response to unexpected events depends in large measure on his understanding of the system and on how unam-biguously the information presented to him portrays the actual situation in the plant.
The accident at TMI revealed problems in both these areas, particularly with regard to understanding unusual transients."
"A national LWR system simulation program should be under-taken cooperatively by DOE and NRC.
This program would treat both PWRs and BWRs generically.
The principal goal should be the development of computational capability to study LWR behavior in real time or faster through a wide range of severe transients including accidents involving extensive core damage."
Although the scope of information and the degree cf detailed phenomenological understanding needed for this data base cannot be precisely quantified at this time, it is clear to the task force that the existing data base may not be adequate.
In addition, comprehensive guidelines are needed for severe core damage accident management and requirements.
The NRC program to deal with severe core damage accidents in the regulatory process is described in the TMI-2 Action Plan, where specific areas of needed information are identified. Areas applicable to the SFD program are discussed in Section 2.2.
Studies in severe core damage represent an important element in the NRC's research program as stated in the recent ACRS comments on the Fuel Behavior Branch SFD program.*
"As a bounding estimate, risk assessment studies have assumed that undercooling a core leads to nelting the entire core. This very conservative assumption has lead people to ignore the substantial difference between overheating and melting the oxide core.
The behavior, and margins to collapse, of an overheated core is an essen-tional yet largely unstudied question where answers are needed for decisions on accident evaluation and mitigation. The program is still in the formative stage, but has made good progress."
9
- " Comments on the NRC Safety Research Program Budget for Fiscal Year 1983,"
NUREG-0795, July 1981.
2-3 2.2 TMI Action Plan Needs The TMI-2 Action Plan (NUREG-0660), Sections I.A.4, I.C.9, and II.B.5, specifically identify information needs for the consideration of degraded or melted cores in safety review.
Specific areas to be addressed by NRC action and program implementation include:
1.
Research on phenomena associated with core degradation to support the basis for rulemaking and to confirm certain licensing decisions.
2.
Procedures for operating personnel for dealing with inadequate core cooling events.
3.
Use of additional instruments and equipment for monitoring plant variables and systems during and following an accident.
4.
Training of operating personnel in the use of plant systems to control or mitigate an accident in which the core is severely damaged.
The major focus of the SFD program stems from these four interrelated areas and it is anticipated that information -from the SFD program will be essential in defining the details of these relationships in the context of regulatory requirements for severe accidents.
2.3 Current Perception of Rulemaking Needs and Schedule The approach and philosophy for severe accident rulemaking has been evolving over the period of time of the Fuel Testing Task Force activities. At the severe accident research (Draft NUREG-0900) program plan for nuclear plant time of publication of NUREG-0840, a draft was in preparation in support of policy determinations and possible future regulatory actions.
The integrated program presented in NUREG-0840 meets the needs and schedule of Draft NUREG-0900.
~
d
3.0 SEVERE FUEL DAMAGE - CURRENT STATE OF KNOWLEDGE It was stated in Section 1.1 that knowledge of the physical and chemical state of a severely damaged core was the major prerequisite for determining the ultimate coolability of the core.
Since those scenarios which lead to core melt represent the greatest contributors to risk, the determination of core coolability at any time during a severe accident sequence will ultimately govern the risk to the public.
Section 2.6 of the Executive Summary of the Reactor Safety Study (WASH-1400) states that:
"The only way that potentially large amounts of radioactivity could be released is by melting the fuel in the reactor core." It goes on to state that:
"Thus, for a potential acci-dental release of radioactivity to the environment to occur, there must be a series of sequential failures that would cause the fuel to overheat and release its radioactivity." The methodology used in WASH-1400 was based on event tree and fault tree analysis which determined the probability of failure of certain systems during an accident sequence. Once it was determined that a system failed, no allowance was given for its ultimate return to service.
During slow accident sequencas, such as that which occurred at TMI-2, recovery of such systems was possible by proper operator intervention.
Such actions resulted in the prevention of a massive core melt and very low risk to the public. Therefore, when one considers the risk to the public for a given series of equipment failures, one must also consider the effect of the return to service of those systems and the proper mitigating actions of the reactor operators.
In effect, these considerations will cause the risk during an accident sequence to be a function of time and not a go no-go situation as is assumed in WASH-1400.
In order to properly manage an accident sequence and reduce the risk to the public during the event, an approximate knowledge of the state of the core is required at all times. Moreover, such knowledge can be used as major input to risk analyris for all sequences, and be used to determine the reduced risk prior to the occurrance of an accident.
- Finally, such information on the core state will enable risk analysts and reactor designers to determine additional safety features that will considerably reduce risk during a severe accident sequence. With these thoughts in mind, the remainder of this section will define our current knowledge of the extent of core damage to be expected as a function of maximum core temperature, and the further information required to determine the coolaMlity of such cores at each point in the accident sequence.
The current state of knowledge on severely damaged fuel is based on past experiments and analyses conducted world-wide to provide a basis for under-standing and evaluating core melt behavior and for risk assessment studies.
The only work that was focused on the early stages of severe fuel damage (as opposed to steam explosions, core melt behavior, etc.) was performed at KfK in the Federal Republic of Germany by S. Hagen from 1976 through 1978.
These experiments showed clearly that the damage state of electrically heated fuel rod simulators heated in steam to temperatures in excess of 3600*F (2255 K) will depend primarily on four major parameters, namely:
(1) the final tem-perature reached, Tmax; (2) the heating rate, dT/dt; (3) the rate of cooling, dQ/dt; and (4) the ' pressure difference between the interior of the rod and the reactor coolant, t2P.
The current knowledge can be summarized in terms of these parameters by defining " damage regimes" in terms of Tmax and expressing the effects of the other three parameters on the phenomena which occur.
The 3-1 1
3-2 following paragraphs discuss each regime in detail by focussing on (a) the physical / chemical phenomena involved, (b) the effect of the damage incurred on further damage processes, and (c) the safety issues to be addressed (if any).
Figure 2 gives a simplified schematic illustration of the damage regimes discussed.
Finally, the current state of knowledge and needs of debris coolability are discussed in Section 3.6.
3.1 Damage Regine I - T
< 1700 F- (1200 K); LP negative 100-1200 psi; any dT/dt max 3.1.1 Physical / Chemical Phenomena - Cladding buckling, collapse, and " waist-ing" on the fuel stack.
This pnenomena was studied extensively by C. S. Olsen for the NRC LOFT program and is well correlated with data.
Very little additional data are needed and modeling of the effect can proceed with confi-dence.
However, this maximum temperature is above the melting or softening point of tne neutron absorbers in the stainless steel clad control rods.
Very little is presently known as to the behavior of these materials at such temperatures or as to the probability of fracture of the stainless steel cladding under the loads applied at these temperatures.
3.1.2 Effect on Further Damage Processes - This phenomenon leads to " intimate contact" between the fuel and clad which promotes the early formation at low temperatures of molten alloys between Zircaloy and uranium (approximately 2000 F) ar.d the formation of " liquified fuel" as a result of the reaction between liquid Zircaloy and solid U02 at approximately 3600*F (2255*K).
It may preclude ballooning and rupture, and thus decrease or eliminate fission product release if the accident is terminated at this point. Also, the cladding of the control rods may collapse or burst depending on their pres-sure differential.
3.1.3 Safety Issues - Possible release of fission products (in-vessel) due to pellet / cladding interaction leading to stress-rupture or stress-corrosion cracking failures.
Reactor behavior in this regime is covered by current licensing practice for DBA events and is not considered to be " Severe" fuel damage.
3.2 Damage Regime II - T
< 2200*F (1475"K); AP Positive; any dT/dt max 3.2.1 Physical / Chemical Phenomena - Cladding ballooning and burst.
This phenomenon has undergone extensive study in the last 10 years. Plentiful data are available and preliminary models have been developed.
Final reso-lution of the effect on core coolability awaits completion of the NRU balloon-ing experiments in FY 1982, and future tests at KfK in the FRG.
Depending on funding levels, the conduct of the Deformed Core Coolability (DECCA) program scheduled to begin at ORNL in FY 1983 may also aid in the resolution of this t
l 1ssue.
Failure of the control rods is most likely in this regime. Although the consequences are not well known, the stainless steel cladding will oxidize as rapidly as Zircaloy at 2200*F and the vapor pressure of the Cd and In will be significant enough for considerable vaporization.
. - - -. -. = -. -
i i
3-3 I
1
~~
~
_5100*F (Melting of U0 )
2 (3038"K)
_ 4900*F (Melting of Zr0 )
E
+
2 (2977*K)
- )
i liquidphase) 4
__4700*F (Possible fonnation of UO -Zr02 2
(2866*K) l i
3600*F (Melting of remaining unoxidized clad (if any)."
+
E~~(2255*K)
Begin reaction of liquid clad with UO2 to form " liquified fuel.")
i g
xs 1
x
{
(Possible fragmentation of highly oxidized cladding due g
j
{j to thermal shock of quenching.)-
i i
=
c 2200*F (Current design-basis limit) i u
~
~
j (1477*K) i i
2000*F (Under clad collapse conditions, possible t _{l366*K)formationofZr-Uliquid)
=
e I
2 E
i u
E f
a
}
~~$
e t
e 8
1400*F(CladballooningandburstifAPispositive) 'g i
~
U T1033*K)(ReleaseofF.P.inventoryinrodgap) y T
E 1200 F (Begin cladding collapse if AP is negative) i o
g _~(920*K) i H
l E
2
=
E u
5 600*F ("Nonnal operating temperature")
I FIGURE 2 l
Schematic (nottoscale) 1 i
.i.
m.,
l 3-4 i
3.2.2 Effect on Further Damage Processes - This phenomenon introduces "two-
)
sided" oxidation of the Zircaloy cladding into the damage scenario.
This j
will cause an increase in the rate and total arount of oxidation which occurs
)
and, therefore, influences the amount of hydrogen released and the amount of Zircaloy available for melting at 3600 F.
Moreover, it will increase the degree of cladding fragmentation due to oxygen embrittlement if the accident is terminated by reflood in the rapid oxidation regime (2200'F-3500*F).
j Finally, it may influence the downstream axial temperature profile of the rods due to the restriction in flow caused by ballooning strains.
3.2.3 Safety Issues - Release of the rod gap fission product inventory to the reactor coolant. Research at ORNL has shown this to be a minor issue at temperatures below 2200'F (1477'K).
The ballooning process will affect the coolability of the core due to partial closure of coolant channels.
If such blockage is near 100%, partial localized melting may occur.
Current evidence indicates that the latter possibility is very unlikely.
In any case, the programs mentioned above will fully investigate this possibility. As is the case for Regime I, damage in this area is covered under current regulatory j
practice and is not considered in this report to be " Severe."
3.3 Damage Regime III - T
< 3400 F (2140 K); any AP; any dT/dt max a
3.3.1 Physical / Chemical Phenomena - Very rapid oxidation of the Zircaloy cladding.
This results in severely embrittled cladding which will fragment on reflood quenching.
The embrittlement and fragmentation of highly oxidized Zircaloy has been studied extensively for the NRC at ANL.
The limits on the maximum time-at-temperature which result in no fragmentation due to thermal shock from reflooding have been determined and can be used in our current models.
No additional work is needed or is planned in this area.
- However, the oxidation kinetics of Zircaloy are not well known above 1800'K (2800*F),
and high-burnup fuel may experience considerable swelling due to fission product release.
3.3.2 Effect on Further Damage Processes - If the heating rate dT/dt is kept below approximately 1/2 F/sec in this regime, all the cladding will oxidize to Zr02 prior to reaching the clad melting point at 3600'F.
Thus, no " liquified fuel" will form when that temperature is reached and no liquifaction of the core will occur until the melting point of Zr02 is reached (approximately 4900'F). However, if insufficent cooling is available and rapid heating
)
(dl/dt > 2*F/sec) occurs, as much as 2/3 of the cladding can remain unoxi-1 dized, resulting in formation of large volumes of liquified fuel at 3600'F.
3.3.3 Safety Issues - If the accident is terminated below approximately 3400*F, the issue becomes related to the coolability of a core containing fragmented pieces of oxidized and embrittled Zircaloy-clad fuel rods.
Determination of core coolability will be made in the early stages of the planned PBF, ACRR, and ESSOR programs to be discussed later in this report.
Another issue is the extent and amount of fission product release at these i
higher temperatures. This question is being addressed by current experiments i
at ORNL as well as by the PBF and ESSOR programs.
f r
j
45 3.4 Damage Regime IV - T
< 4700 F (2870"K); any AP; any dT/dt max 3.4.1 Physical / Chemical Phenomena - Melting of the remaining partially oxidized cladding; reaction of liquid cladding with solid U02 to form " liquified fuel;" flow and refreezing of liquified fuel to produce " candling" type (cohesive) damage and blockage; continued oxidation of liquified fuel during i
flow and after refreezing.
The only available data in this regime is that of Hagen at KfK which used rod simulators containing a core rod of tungsten (as a heater) surrounded by annular rings of U02 More prototypical tests are required which use rods of standard design that are volumetrically heated by either fission or decay heat so that the damage and debris formation scen-arios for representative fuel rods can be studied and modeled.
The effect of high burnup will also be important in this regime and in regime V below.
The release of excessive fission gases and volatile fission products will induce fuel fragmentation, swelling, and foaming.
As a result, the liquified fuel flowing and freezing characteristics my be considerably altered.
Experi-ments planned in the PBF, ACRR, and ESSOR reactors will scope the damage expected; study the phenomena involved; and provide the necessary verifica-tion of the early SCDAP models which will be based on Hagen's data.
3.4.2 Effect on Further Damage Processes - This damage regime is critical to the determination of further core damage since it may produce the first major loss of geometry imposed on the core. The character of the damage will a
determine further coolability regardless of whether emergency core cooling is 4
applied and will determine where regions of UO2/Zr02 melting will or will not occur throughout the core. A successful description and understanding of this regime is essentiel for describing subsequent fuel core melt scenarios and the interaction of such a melt wit. the pressure vessel.
3.4.3 Safety Issues - The major safety issues for this regime are core coolability (i.e., can the accident be stopped?) and fission product and hydrogen release from very hot solid fuel rods, liquified fuel, and frag-mented fuel.
The P8F, ACRR, and F MnR programs are designed to answer these questions by the performarse o. core debris coolability studies and the monitoring of fission prod.ct and hydrogen releases during the experiments.
3.5 Damage Regime V - T
> 4700*F (2870*K) max 3.5.1 Physical / Chemical Phenomena - Melting of remaining 002 and Zr02; growth of the melt; motion of the melt; foaming of molten U02 due to fission product release; interaction of the melt with the pressure vessel sides and l
lower core support plate; interaction of the melt with water in the vessel; hydrogen and fission product release; explosive and non-explosive steam generation.
Except for steam explosion studies currently being performed at Sandia, very little information is available on the phenomena mentioned above.
Some information on melt motion may be obtainable from fast reactor experiments and models, but new information is definitely required for LWR's in this area.
Current plans for studies of this sort are not yet fully defined, but may be part of the ACRR program and Phase II of the PBF program.
Investigations into the modifications required for the PBF to perform such studies are just getting underway and no detailed plans are yet in effect.
l
3-6 3.5.2 Effect on Further Damage Processes - This regime is critical for the determination of when and if a vessel melt-through will occur. The manner and time at which this happens will determine the resulting loads on the containment structure and, therefore, the possible release of radiation to the environment.
I 3.5.3 Safety Issues - The major safety issues are whether the core will melt i
through the vessel and whether the containment structure can safely house the resulting hydrogen, steam, and fission product releases.
Programs in other a
branches and divisions of NRC research are also addressing these problems and, therefore, will be fully coordinated with the research results of regimes I through V to fully integrate and understand the potential problems.
The next chapter will discuss, in detail, the current program plans for research in each of the above damage regimes. Because of the paucity of data available today and the importance of regime IV on determining whether a full core melt will occur, the bulk of the reuerch is centered in this regime, 1
both in scope and in dollarmen:. F mJlts from this research will guide later research efforts in regime V and provide information for Severe Acci-dent Rulemaking.
3.6 Core Debris Coolability - Current State of Knowledge A primary goal of the SFD program is to determine, for each state point of severe fuel damage, whether or not the core debris is coolable and what the coolant requirements are to achieve coolability.
Debris is said to be cool-able if a geometry and temperature distribution have been achieved that are stable in time. The coolability approach used in the SFD program is to determine the damage state points for which the core debris is coolable by slow reflood (i.e., stagnant pool) and, for those damage state points outside this space, to determine the coolant flow velocity and pressure necessary to achieve coolability.
The most important, most easily defined, and most easily measured coolability limit is the dry-out heat flux limit in which liquid coolant does not reach some regions of the debris.
It has been shown in fast-reactor safety exper-iments in the ACRR test reactor with sodium-cooled debris beds that stable temperature distributions and geometries are possible at decay-heat power levels with local dry-out in part of the debris bed. However, the available coolability margins and debris behavior between local dry out and progression into core melt are only poorly known. Therefore, the well-defined and rela-tively easy-to-measure dry-out limit is the best criterion of coolability to use in reactor safety assessment and research. Operation beyond this limit i
involves great uncertainty and the risk of further core-melt progression.
In the boiling zone in a packed (unchanneled) uniformly-heated debris bed, cooling is by downward flow of liquid and upward flow of vapor in the adja-cent interconnected channel between the particles in the bed. The bed behavior is essentially one-dimensional or independent of the bed diameter, so that experiments on a 10 cm scale are applicable to beds of indefinite radial extent. Dry-out is reached at the flooding limit, where the upward vapor flow interrupts the downward flow of the liquid coolant.
Coolability research and models have nearly all been on quasi-static (long term) coolability limits, not transient quenching, on the basis that the long-term behavior is governing.
3-7 A substantial data base and relatively sophisticated analytical models of dry-out coolability limits as a function of mean particle size and bed depth have been developed in the fast-reactor safety research program.
The exper-iments have included several coolants: sodium, water, and organics and several methods of heating, including fission heating of siaulated debris in the ACRR test reactor to simulate fission-product decay heating.
Lipinski at Sandia has developed a relatively sophisticated first-principal model for the dry-out limit of a packed unstratified debris bed that agrees well with the world data base for all the liquids tested.
This model includes capillary forces and both laminar and turbulant vapor flow. 1he ACRR experiments with sodium-cooled debris beds have shown that the formation of vapor channels in the bed that can occur at high subcooling can increase the bed dry-out limit by about a factor of five.
These experiments have also shown that bed strati-fication with an increase in mean particule size with distance below the bed surface can decrease the bed dry-out lih.it by a factor of five.
Verified models of these phenomena and of the onset of channeling in debris beds do not yet exist.
In the absence of verified dry-out models for LWR-specific conditions, the fast-reactor models (with water properties) were used in assessment of the coolability of the THI-2 core and.in the Zion and Indian Point-2 studies.
Verification of the fast-reactor models for LWR-specific conditions is needed for debris coolability assessment in the SFD program.
LWR-specific conditions that need investigation are high pressure, larger debris particule size, deep beds, and possibly local debris non-uniformity or blockages.
Data and models on the effect of inlet flow on the dry-out limit are also needed.
i f
i 4.0 CURRENT AND PLANNED NRC PROGAM ON SFD 4.1 Program Approach The approach planned for the research on SFD is an integrated program with three principal components illustrated schematically in Figure 3, which shows i
the internal program structure and outside interfaces.
Initially, integral in-pile tests are performed in P8F and other test reactors to provide scoping j
data on severely damaged fuel behavior and governing effects, and later for l
j verification of the integral S,evere,C_ ore Damage A,nalysis P_ackage (SCDAP) code j
as developed from current data and the results of the research program.
j j
Separate-effects phenomenological experiments, both in-pile and laboratory, i
j are used to provide a data base for the development and verification of analytical models of the governing phenomena that are then incorporated into l
the integral SCDAP code. Analytical model development is the essential third i
j part of the program, and includes both the integral SCDAP code and phenomen-
)
ological models of the governing processes for incorporation as modules into l
SCDAP.
This approach of an integrated three part program is far more effec-
]
tive than the sum of its compnnent parts, and is far more productive and cost 4
j effective than a program of in-pile tests standing alone.
4 l
The results of the SFD research program will be incorporated into accident-analysis systems codes and will also be used directly for the design and t
assessment of accident-mitigation systems.
The usefulness of experiments in the severe core damage regime has been questioned on the basis that there are so many different accident sequences and state points that only a small fraction of them may be covered by exper-j iment.
It has been argued that valuable resources should not be expended upon such a hopeless task.
If the program output were only the data from large-scale integral tests, this argument might be valid. However, with an coordinated program of integral tests, separate-effects phenomenological experiments, and analytical modeling of the governing phenomena, the accident state-point parameter space is covered by the analytical models, and it is not necessary to perform integral experiments near each accident state point of interest. A few integral experiments in significant regions of the acci-j dent-state-point p:rameter space (see Section 3.0) are needed, as scoping i
experiments to identify governing phenomena and as bounding experiments, and i
later for verification of integral models. Much cheaper, better-defined, and better diagnosed separate-effects phenomenological experiments, both in the I
laboratory with non-reactor materials and in test reactors with reactor j
materials and temperatures, will then be used to acquire a data base for the development of phenomenological models.
This approach makes tractable, although still very difficult, the research needed to acquire an understand-ing of reactor behavior during accidents involving severe core damage. ThE i
probable phenomena and data needs, and the experimental variables and ranges that must be considered are presented in Table 3.
Those actually studied j
must be carefully selected as to emphasis and priority, schedule, program-i matic interactions, and funding availability.
Some of the research needed i
j will be conducted outside NRC Research and mst be integrated by mutual j
agreement.
i i
4 j
4-1 1
b i
i
~
4-2 Figure 3 Severe Fuel Damage Research Program Interfaces and Internal Structure ACCIDENT MANAGEMENT FOR m
IN-VESSEL TERMINATION Eh tEERED REQUIREMENTS PROCEDURES FEATURES l
r i
DOE, VENDORS INTERNATIONAL
n a
V
[
i ACCIDENT OTHER
)
TMI-2 4 ):
ANALYSIS RESEARCH**
CODES h
a U
1r if U
V SEVERE FUEL DAMAGE RESEARCH PROGRAM W
t INTEGRAL IN-PILE TESTS b
ANALYSIS AND MODEL DEVELOPMENT U
(SCDAP) 7 SEPARATE EFFECTS PHENOMEtt0 LOGICAL EXPERIMENTS t
Themal-hydraulic, Risk, SASA, Human engineering and l
Instrunantation I
i t
~
4-3 TABLE 3 g!st? $fVItt Furt Om8Gt inFle!*tel P484 PETER SPAlt i
n.e.eaan r.a u t,aest wienare a,eee
- s. w esaads.nes C 8 1*
~k;- k*'
p$
-s sji *al $
EY A.
dh,9id!!!
2 L,i
- n. t.teo.f n-.iecs.ee
.%reture iOso. isso a Ii si t
- i. Ci..co.
n-in.ic.na,se B.
Fensten produce release
- Meete rate
.2%. le t/s 8
0
- 30.000 8tM/t C.
Coolatillty of rees
. An'F"*
37.a.nd 64 PeJte Il e Il ar $tse doomstrees of tiectage
-Os t detien i
~
2.
Itetaley Ostdattoa 3
3 3
3 g) I s,
3, annettes et high tegeratures
.Teeeerature IGNI. 2 300 t
.5
- 4 63/hr se, and ta steam starved enetrea.
.Stenstae rate sea ts
.Onygte and hydrogen f
8.
Hydrogen predsetten parttel pressure
.tedte length
. S.2 e
.Arrep $tae 4 = 32 I{I
'E10;E I
A A.
Elaettes of f truettaa
.Te eeretwee of Iteutd 2000 2910 E 3.
Litvif tet fuel I
S.
TheNeyneet t of reettions *
.featereture of set te 2000
- ELO E e st aa t t aa C.
Ther*E9aFDCal Drecerties *
.et sture tausest tlam
.12.,1 canettag and freeslag (neate retel beaset er 0nyven and hedregen f
9.
fissten eredvet reteettaa Partial pressures sad release (steestag rates) 4.
Saelltag sad feassag et
. tvraw 0 10.100 pho/t negn teseerstwees
.inse fee ela, 4.
tod frapeatettaa t
1 11 t) A 4.
Perttele site distratet6 oms
.Teeeerature F000 e 1100 E t.
Paestag eastte
.iteatue rate (esteetten)
.4 4 t/s C.
frapeatettaa srt tarts
.Ceeltag rate Stas to 100 t/s 0.
lire sad lotettaa of clos
.tvadie sine t to 32 rods channels around tee ts
.sette length
.6 to 2e
(.
finstem preswet reteese
. gvenw 0
50.000 pte/ t 7,
Pertte te cussesttten sad
.sete spacerg t herseere s t e r s 9
Optrts Caniatt tity DJ
{
a i
A.
Dryout testific Power
.P er
%.=,esorged tede (coeltog
-se,ttele site.distettuttom 10m.10 se 4.
al
..eria,in, noviel
- s. Posi er,.et sesseret.,es ceasesin
.. z,s., 0 and reactlene
. Sed desta
. 4
- 2e C.
elelt benettcs
.itratificattee tes erite
. Pressure
- 1. INI ate
.Coelant secaeltag 0e 100 t
. Settee teollag
- 6) P4(see and flutetied 1
KN I
3 4 Sryout Specifle Posee
. Particle stae distributtes le ese e le se bees caeled by forced S.
$trettfitattee and sesneet 404 tescosttlee es.e Ir0s ett.
flow poteattel 004 depth
.9 Ie C.
finidi s ett en
.Pressere I. If0 ate 8.
Post dryevt heat transfer
. flee rate
?
(.
Pressem Orop. flew
.Ceenaat subcesilag 0 100 E resistance
-See perestty
.5.6
.tweie stae 32
- Ill rods 1
I A.
Therempeysit. erseerties.
.fesserotere 24J0
- 3200 5 4.
melt Pyssten and M@ert strutture Attatt tandilag and freettag
.Gundle stae 8
- lol roes beneeter 0 stay meet
.5 I se/se 8.
Esteet of the smelltag
.twene 4. E 000Pte/t and feselag
.ites I ele.
- I tr.
C.
Theresseynamics of reestteas 8.
fIssten product retaattes
- ad release
(.
- t. tent of the estal and redsel rele(atten F.
Desege te care sueeert stres twres F.
Degreed sedle W
a a
A.
errest tentfic Peer
.sedle slie it.168 rods
(.ie l ael i t ty 8.
Post dryest heet transfer
.eedle length
.g*ta C.
Pressure drop a flew
.f ree and entent of the resistance flee Discoeges F
D.
Post drye.t tesseretures
. flee rates f
and reactlens
. Pressure I 170 ete
. Coolant subcesites 4
- 100 s 4.
finstem Preewst Release N
I I g
4, Asteese *etes for tapertasit
.feaguerstere 1000
- 1200 t i
and fremseert teet tes e t tes isa fee hears 5.
Cheettel ferie of released 0asiege telleestag, eatdallen, a
llowef (tien.
4 spettes C.
Cecar heel release fra pentattest. fuel D.
Piete 64et en wolle er teve salt ead rolesattaa structure
.Pressere I 110 sta
( ) Lletted Date
i 4
4-4 l
The major part of the program of integral in-pile tests is the Severe Fuel i
Danage (SFD) series in PBF.
The initial series of five tests in Phase 1 of
[
l the program, which is now underway, will provide integral scoping date for i
damage regimes III and IV as discussed in Section 3.
They will also provide j
data on hydrogen generation and fission product release from the reactor f
vessel.
The characteristics of the severely damaged fuel will be obtained i
from post-test examination.
This series is the foundation of the severe fuel
}
damage research program. Scoping data from these tests will fonn the neces-
[
}
sary base for the in-pile and laboratory separate-effects experiments on i
governing phenomena as well as for the models in the integral fuel-behavior code SCDAP.
1 i
On a longer time scale, integral SFD data may be available from the SUPER.
l SARA program in the ESSOR reactor at Ispra.
Fourteen of the twenty planned l
j tests will involve severe fuel damage.
The data will supplement the 3-foot PBF results and determine the scaling effect of a 6-foot axial length. The i
SUPER-SARA tests also cover a wider range of accident conditions than the PBF Phase I tests, and include both PWR and BWR conditions.
l Tests on reflood quenching of intact fuel rods and on cladding, ballooning, j
and blockage with full-length 12-foot bundles of 24 and 32 rods are underway
]
in the Canadian NRU reactor (damage regime II in Section 3). A new doubly-1
]
contained test loop would probably be necessary to perform a broad set of full-length severe fuel damage tests in NRU, and such a loop would have the i
capability for tests with 12-foot bundles of 76 rods.
I There has been preliminary planning for a second phase of integral severe-j fuel-damage tests in the PBF to explore the effects of high burnup fuel, i
J control rod materials, and fuel element design.
This series may also include r
1 experiments at higher temperatures and larger bundle size to explore the f
effects of using a decay-heat source built up by a 1-week irradiation of j
previously irradiated fuel rather than fission-simulation of decay heat. The larger test bundle size would also be used to determine the effects of large i
}
arrays on blockage distribution and the effects of prototypic amounts of absorber materials.
These latter tests would require a modification of the PBF to incorporate a larger test loop to accommodate up to full-diameter 1
17 x 17 PWR fuel bundles. Large-scale, out-of-pile, integral melt-progression j
tests may also be perfonned with the Large-Melt Facility (LMF), which can i
provide pours of up to 500 kg of molten U02 3
i The second major part of the research on severe fuel damage is a program of supplementary separate-effects phenomenological experiments on the dominant j
processes involved in the behavior of severely damaged fuel. This program
)
includes both laboratory experiments and laboratory-type in-pile experiments that use fission heating in a test reactor to simulate fission-product decay i
heating.
In-pile separate-effects phenomenological experiments have been l
J used quite successfully in the fast-reactor safety research program in ACRR.
l i
l I
i l
i 4
i i
4-5 l
A major objective of the separate-effects experiments is to determine the range of core conditions (if any) for which simple reflood is not sufficient to cool the debris and terminate the accident, and the flow velocity neces-sary for coolability under these conditions. Under reflood conditions, cooling in debris beds is by internal cellular one-or-two-phase convection with additional coo possible from net inlet flow.
The dry-out coolability I
limit is reached when 11guid cannot penetrate to all points in the bed against the out-flowing vapor.
Considerable data and rather sophisticated analytical l
models of the quasi-static dry-out coolability limits of beds of decay-heated I
particulate fuel debris under liquid pools have been developed in fast-i reactor safety research.
These models are based on a successful series of experiments in ACRR with sodium-cooled debris beds, but the data base includes 1
laboratory experiments with water and other liquids and no inlet flow.
Fast-J reactor-safety models were used in analysis of the coolability of the core debris in the TMI-2 accident, and in the Zion / Indian Point studies, because LWR safety research had not addressed this problem.
Beginning in late 1982, a series of seven LWR-specific core-debris coolability experiments will be performed in ACRR.
These will be extensions of the previous LMFBR safety experiments, and the purpose of the initial experiments will be to validate l
for, LWR accident conditions, the current fast reactor debris-coolability models.
The LWR-specific conditions that require experimental verif. ation, in addition to the change to water cooling, are high pressure, very deep debris beds, inlet cooling, and particularly the characteristics of the LWR j
core debris.
It is known that the characteristics of the core debris are a j
major determinant of the dry-out coolability limit under reflood conditions.
A program of separate effects phenomenological experiments has also been i
started in ACRR on the mechanisms involved in the formation and relocation of fuel debris and on the characterization of debris.
These experiments will provide visual diagnostic data continuously in time for high-probability unprotected accident sequences, as well as debris characterization for reflood quenching at various times in the accident sequences.
Data from these separate-4 effects experiments will be used to develop phenomenological models of the i
major processes for incorporation into SCDAP.
These separate-effects exper-
)
iments effectively supplement the larger-scale integral Phase 1 Severe-Fuel-Damage tests in P8F, and they will substantially broaden the data base for model development.
The cinematographic and experimental techniques used in these experiments have been developed in a series of successful fast-reactor safety experiments in ACRR.
j Laboratory separate-effects experiments are planned to determine the therno-i dynamics and kinetics of the reactions between U02, Zircaloy, and steam.
0)periments are also planned on the candling process with the ternary (U. Zr Ex liquified fuel and on debris formation in reflood quenching.
Laboratory experiments (regime !!) have been performed with ciectrically-heated multi-rod bundles of Zircaloy cladding on clad ballooning, bursting, multi-rod burst tests (MRBT)g heating transients with steam cooling.
and blockage fonnation durin These have been performed with 16 and 64-rod bundles.
i The experiments may be extended in the DECCA tests with full-length bundles to measure the coolability of the deformed, but not fragmented, bundles.
i l
4-6 Related experimental and analytical work outside the SFD program is needed on hydrogen and steam generation rates during the reflood of hot-solid and molten fuel debris.
This information is needed for accident management to determine possible upper limits on reflood rates to avoid failing the (hot weakened) reactor pressure vessel from overpressure. Work is needed on both the explosive and the non-explosive steam-generation processes.
The analytical component of the integrated program includes development of the integral SCDAP code for the detailed analysis of fuel behavior during severe accident transients including melt progression to reactor-vessel fail-ure and the development of the individual models of dominant processes incor-porated in SCDAP. The development of SCDAP and particularly of the indi-vidual models will depend upon the data base provided by the integral tests in PBF and other reactors and upon the in-pile and laboratory separate-effects phenomenological experiments.
The analytical program will also include analysis of the high-probability accident sequences involving severe fuel damage to determine the governing phenomena and uncertainties.
This analysis will be performed early in the program to guide the experimental program and also the model development for SCDAP.
The TMI-2 core examination constitutes a unique and invaluable resource on the characteristics of severely-damaged fuel.
Early recovery and adequate analysis are highly desirable to provide a benchmark for research on the behavior of severely-damaged fuel, including development of the SCDAP code, and for understanding severely-damaged fuel behavior.
The current program of the Fuel Behavior Branch includes modest support for analysis of TMI-2 core debris, but does not, of course, address the cost of the TMI-2 recovery operation.
The actual program of SFD research needed to provide a sound technical basis for accident management and licensing activities may prove to be considerably less extensive than that outlined in this report.
This program was derived from our current state of knowledge on the characteristics of severely damaged fuel and on the behavior of such fuel, for which the data base and verified models are in a primitive state.
It may well be that later, with data from the PBF Phase I scoping tests, the early ACRR phenomenological separate-effects experiments, the TMI-2 core examination, and with models developed from these data, that some of the additional program will prove unnecessary.
In any case, the SFD program requirements and program plans should be and will be reexamined periodically.
The follJwing sections outline the above programs separately and in greater detail.
4.2 Analysis Development of the Severe Core Damage Analysis Package (SCDAP) 4.2.1 Objective and Schedule The objective of this work is to develop a comprehensive code package for the detailed analysis of light-water _ reactor cores during severe accident sequences.
The code will be used to:
4-7 o
Aid in the process of formulating rules and evaluating mitigating safety systems and actions which wil! ensure the public safety during severe accident sequences, o
Aid in the analysis of the proper actions required to bring a damaged reactor system under control.
o Evaluate and plan the severe fuel damage experiments in the PBF, ESSOR, ACRR, and other possible test facilities.
o Ultimately serve as the nucleus for the development of a simplified, fast-running, code for use in parametric studies of accident scenarios.
Work began on this task in FY 1980 in the form of planning discussions and NRC program reviews.
Initial funding was begun in FY 1981 at INEL to provide the NRC with a code planning and design report.
The major milestone schedules are itemized on the chart below.
Milestone Milestone Completion Date SCDAP Planning and Design Report 4/1/81 Begin Programming of SCDAP 10/1/81 Complete Preliminary Version of SCDAP 6/1/82 Complete Production Version of SC0AP 6/1/83 Complete Final Version of SCDAP 6/1/84 Beginning in FY 1985, the SCDAP code will be on maintenance with funding for updating and model refinements as needed.
Note that a preliminary operating version of the code will be available before severe accident regulatory actions and prio_r_ to the first planned severe core damage experiment in the PBF.
This version will not be released to the public, but will be tested at EG&G against the ex-reactor TRG data of Hagen and the TMI-2 transient.
4.2.2 Description of SCDAP Code and Capabilities - See Volume 2 (section B.1) for More Detail The code will be modular in nature and be designed to predict the following quantities and phenomena in a LWR core:
fuel rod temperatures as a function of time and axial position; the total quantity and types of fission products released from the fuel; fuel rod deformation including clad ballooning and clad collapse, the total amount of hydrogen generated and released, and its axial distribution along the bundle; total amounts of liquified and resolid-l ified clad and fuel material; the amount of oxidation of the Zircaloy cladding; the total mass of rubble debris and the debris spatial distribution; and an estimate of the flow blockage expected.
In order to accurately predict these quantities. SCOAP will draw heavily on existing models already developed for l
other codes such as:
the FRAPCON and FRAP-T fuel codes; the TRAC and RELAP T/H system codes; the EXMEL melting code developed at Stuttgart, FRG, from l
Hagen's experiments; the MATPRO UO2/Zr materials properties package; and the i
4-8 entire experimental data base on fuel rod properties developed by NRC, in-dustrial, and foreign research programs over the last 6 years.
Certain outputs like the mass and character of the debris bed and its coolability wili rely on the results of the P8F and ACRR experimental prograns and cannot be considered accurate until tnese tests are perforneJ.
Other outputs requir-ing extra exper" ental confirmation include the amounts of " liquified fuel" to be expected and the amount, location, and composition of resolidified fuel and clad debris.
The code will initially model the behavior of fuel bundles and their associated nan-fueled coryonents such as control rods, canister walls, and instrunentation tubes, and later be extended for core-wide analysis.
Thermal hydraulic input is expected from existing T/H systens codes via data tapes or direct code linkage.
Rad deformation and thermal nodels will be taken f ron the FRAP code series.
Fission proJuct release fron the fuel will be nodeled using the FASIGRASS fission gas release code developed for NRC/RES at ANL.
It is intended thit the code be very fast running; this requires that the degree of sophistication needed for certain models be checked out by corparing sinplified noJel results and code running times with highly conplex, more accurate moJels.
These studies will be done in parallel throughout the development process.
Finally, beginning with nod 1 of tne code, a built-in uncertainty analysis package will be included to enable users to have uncer-tainty bands on the output calculated internally and displayed on output data plots.
4.2.3 Interface Rotween SCDAP and Sygens Code',
s All of the SCDAP planning to date has emphnited its ultimate use for core-wide coolability predictions af ter severe fuel danage has occurred.
To date, only tenative plans to do so have emerged, although an analysis to determine the nodifications rcquired to use the SIM'tER code as a parent showed it to be not feasible.
it is probable that this issue will be settled by expanding later versions of SCDAP to contain simplified core-wide thernal/ hydraulics Capability.
Such a decision requires approval by upper managenent and a strong connitment of progran (.cordination between branches and/or divisions.
In eny case, the SCDAP code will be very useful in detailed analyses of severe fuel danage processes and could even be used to ghe accurate input into a very simple code such as MARCH.
Early in the progran, analyses will be performed at EG&G and Sandia on the governing phenonena and the principal uncertainties involved in the high-prubability accident sequences that lead to severe fuel damage.
The results of these analyses will be used to guide the SFD experimental program and the nodel development for SCDAP.
Input for this work will come from analysis of the TM1-2 accident, the Zion / Indian Point study, the SASA Program, and the work of the Probabilistic Analysis Division.
Later in the program as results from the research on melt-progression to reactor-vessel failure became available, a melt-progression module will be developed at Sandia for addition to SCDAP.
4.3 Integral _SFDTests(In-PileandEx-Pild The presently-planned, integral in-pile SFD tests (regimes !!! and IV) will be conducted in the PBF and ESSOR test reactors.
Potential SFD tests of full-length fuel bundles in the NRU test reactor are being discussed, but no
4-9 definitive plans have been made. Both PBF (Phase 1) and SUPER-SARA SFD (ESSOR) tests will use fission heat in 32-rod fuel bundles to simulate decay heat, and both will examine fission product release and debris formation.
However, PBF is limited to 3-foot long fuel rods while the SUPER-SARA loop in ESSOR can test fuel rods 2 meters long.
The two programs will complement each other, and yield infomation on the scaling effects of length on debris bed fomation and severe fuel damage progression.
The NRU SFD tests (if implemented) would add the effects of full-length fuel reds (12-feet fueled length) in 24-rod bundles.
4.3.1 P8F-SFD Tests PBF Test Facility:
The PSF consists of an open tank reactor operating at ambient conditions with a driver core region having an active fuel length of 0.914 m (3 feet) and a central flux trap space containing an in-pile pressure tube.
The experirental loop of the pressure tube can duplicate most of the environmental conditions of a LWR.
The reactor can be operated in three primary modes:
steady state, natural burst, or shaped power burst.
It is readily reduced to decay heat power levels typical of LWR's.
The present pressure tube can contain test trains as large as 32-rod fuel bundles with accompanying instrumentation and themal shields.
Phase ! SFD Tests:
Present plans call for five 32-rod tests to be conducted:
two at slow heating rates less than 0.5'C/second (to fully oxidize the clad-ding and, therefore, preclude the fomation of liquified fuel), two at faster heating rates of about 4'C/second, and one approximating the estimated TMI-2 conditions. One of each of the slow and fast rate tests will be cooled slowly from maximun temperatures of about 2175K (1900'C, 3460*F) to preserve as much as possible the configuration existing at the maximum temperature, and the other will be quenched with reflood water to produce debris beds.
The experimental conditions of the fifth test have not yet been specified.
These tests will also verify the adequacy of the designs of the test train and the shroud which is expected to contain the liquified fuel.
Finally the tests will produce the debris to be used for determining the expected size-ranges, compositions, pemeability, and coolability of severely damaged fuel.
These characteristics will be used in guiding and planning ex-reactor coola-bility experiments in the ACRR.
Phase !! SFD Tests: The test matrix for Phase !! studies in PDF has not yet been defined. At a ninimum, preliminary results from Phase I tests must be available before detailed planning can begin.
HOWever, it is presently proposed that at least the early Phase !! tests extend the five tests in Phase I to include high burnup fuel, control materials, fuel rod design effects, higher temperatures, free-heating conditions (essentially $^agnant steam),andheatuptestsdrivenbydecayheat.
Later tests may also examine core disintegration, melt progression processes, and scaling effects of radial size by using up to 17 x 17 bundles (this requires a larger diameter in-pile tube, and modification of the PDF driver core and test loop).
I i
su 4-10 4.3.2 ESSOR SUPER-SARA SFD Tests SUPER-SARA Test Facility:
The ESSOR reactor is a heavy-water moderated pool-ch type reactor located at Ispra. Italy and supported by the cultinational jM European Economic Conmunity (EEC).
Several test loops can be inserted.
The
- .k SUPER-SARA test loop has an in-pile section capable of holding a 32-rod 3
assembly of PWR-type fuel rods up to 2 meters in length.
The loop can be cw]
operated to 15 MPa (2177 psi) with conditioas typical of commercial LWR's.
.M The Icop also contains a side leg fitted with an analog electrically heated f M.,k fuel rod simulator bundle for pretest establishment of the desired thermal-E ""
m hydraulic conditions.
~
.. [-I:
SUPER-SARA SFD Tests:
The present consensus test plan calls for the conduct yc p i
of 5 to 7 large-creak LOCA tests interspersed with 13 to 15 severe core bi gab damage small break LOCA tests using 32-rod bundles and the containment shroud i,-- n' design from the PBF-SFO test trains.
The small break tests will exami1e i^
conditions for severe core damage with peak temperatures to 2300K and higher.
Q, The tests will be complementary to those conducted in the PDF Phase I and Phase !! test series. One of the major features of the SUPER-SARA SFD tests
.[2 will be the greater bundle length which will allow much deeper debris beds to
..h, g..
J be formed and ruch greater axial extent of severe fuel damage and candling.
The earliest SFD test of significant interest to the SFD program will probably A.g ;s be conducted in 1985-1986.
+o
$:J.
4.3.3 NRU.SFD Tests
- ~,
ug -
NRU Test "acility:
The NRU reactor is a heavy-water-moderated and cooled
(( "3 reactor located at AECL, Chalk River. Canada.
It has an effective core e - l,y height of 10 feet. At the present time, the vertical U-2 loop is being used
',4 by PNL to conduct (for NRC) several large-break LOCA tests of full-length 32-
%5 rod bundles of cornercial-type PWR fuel elements.
Tht. use of an insulated
(.? <
d ' :[i shroud in conjunction with the cristing loop represents a possible extension Y".'Y for SFD tests.
Another test loop with double containment can also be con-I '*
structed and interted for the conduct of SFD tests of full-length 24-rod arrays of commercial PWR fuel elements to temperatures above 2300K (3700*F).
>Y %
j Preliminary discussions are underway at this time regarding the role and
' hhe extent of the use of the NRU for tests in SFD regimes.
.N 4.+g NRU-SFD Tests:
Several SFD tests would be conducted to provide well-characterized
.Y,
in-pile cata of prototypic fuel rods under realistic accident conditions.
~ (l s.j rs The data would allow an accurate evaluation of the axial relationships for 4/.
SFD data established from the PBF tests.
1."y {
f,,,G 4.3.4 TMI-2 Core Eranination V,'
- r
,n THI-2 Facility,:
On March 28, 1979, the commercial PWR reactor Unit-2 at f?
Three-Mile Island, PA underwent an accident resulting in severe fuel damage in the core of the reactor.
This is the only commercial power reactor to M
have been severely danaged in service.
Extensive plans have been made by GPU
[ [,'j (the owner). Bechtel (the contractor), and DAW (the builder and subcontractor
- 2. t " '-'
for the nucicar system) for the clean-up and recovery of the entire plant.
Several early entries into the damaged plant have been made to assess the
'off 1 Y,# l r.
.f., j,. 2
-g tW,,.
4-11 extent of damage to the containment systems and problems to be expected and solved before active clean-up can begin.
Not only must the plant be cleaned of deposited radioactivity, but the polar crane and the fuel handling bridge rust be restored to good condition before disassembly of the reactor pressure vessel can proceed.
TMI-2 Core Examination:
The damaged reactor presents an unique opportunity for bencnnarking codes, models, interpretations, understanding of fuel behavior, fission product release f rom danaged fuel, etc.
Extensive plans have been made by several technical committees associated with the TMI Working Group f ormed by GPU, EPRI, NRC, and D_0E (GEND) for the examination of the condi-tions of the reactor vessel, its internals, and the core and the fuel during the clean-up and recovery of the plant.
Details of the plans have been, or will be, published in several GEND reports available from the National Technical Infomation Service, Springfield, Virginia.
At this time, it is planned to observe the condition of the top of the core before the removal of the pressure vessel head by inserting a small TV camera and necessary lightirg into the vessel through small access pipes in the head and the side WalI of the vessel.
Af ter removal of the vessel head, the condition of the upper internals, the drive trains for the control rods, and the upper support plate will be determined.
It is planned to establish the vertical and radial patterns of damage in the core during its disassembly, with the recovery of selected sections for detailed examinations and analyses in hot cells after renoval.
Extensive documentation and photography will be the major tools of record of the state of the several sections of the system during disassembly and removal.
The plans of the FDB/NRC for the TMI-2 core examination are to observe and participate in the dissassembly of the core; to confer on the analyses of core and fuel samples to be done by EPRI, DOE, and GPU; and to conduct analyses on selected specimens of particular concern to NRC. but not to the other participants.
4.4 Phenonenoloq1 cal Separate-Effects Experiments and Model Development The purpose of the program of phenomenological separate-effects experiments is to supplement the data obtained in the large-scale in-pile integral tests described in Section 4.3.
These experiments also provide a data base for the development and verification of models of the governing phenomena in the behavior of severely-damaged fuel for use in the SCDAP code.
This program consists of:
(1) in-pile laboratory-type separate-ef fects experiments using fission heating of the fuel in the ACRR test reactor (2) Possible out-of-pile laboratory experiments on the properties of the ternary U-Zr-0 system, on fuel debris fomation and relocation, and on the coolability limits of multi-rod bundles of electrically heated cladding that have undergone severe ballooning defomation. These latter experiments (DECCA program) would be performed in the thermal hydraulics test facility (THTF) at ORNL. The in-pile experiments concern the coolability limits of fuel debris under reflood conditions, and the formation, relocation, and characterization of fuel debris for the range of high-probability severe accident sequences and for reflood quenching at different times in these sequences. Much of the progran of in-pile phenomenological separate-ef fects experiments utilizes the ACRR test reactor.
I 4-12 ACRR Test Peactor:
The Annular Core Research Reactor is an open-pool, natural-convection-cooled, test reactor, with a 23 cm diameter dry central experiment cavity into which experiment packages are inserted.
The UO2. Be0 core is 52 cm high.
The ACRR has both high-power pulsed and low-power steady-state capability.
The later node provides steady-state fission heating of 3%
enriched LWR experiment fuel at and above LWR fission-product-decay power levels.
Operation is characterized by a high experiment rate, about one-per-day for simple experiments.
Extensive use is made of visual diagnostics, and a fuel-position diagnostics system has been developed that utilizes the coded-aperture imaging of fission gamma rays from the experiment fuel.
This test reactor has been used estensively for in-pile phenomenological separate-effects experiments in fast-reactor safety research.
4.4.1 Debris Fomation and Relocation Experinents in the ACRR Few-pin E speriments on Fuel Rehavior During Unprotected Accident Sequences -
This work consists of a series of few (4-9) pin, in-pile phenonenological separate-ef fects experirents, supporting laboratory experiments, and support-ing nodel developrent on key ohenomena in the processes of debris formation and fuel relocation following fuel failure in severe accidents.
Data on fuel behavior and relocation are provided continuously in time throughout the accident transient sequence, not just at the end-of-transient time point available fron post-irradiation examination.
Important phenomena to be investigated in these experirents include the fomation of particulate core debris, streaming and free !ng of the ternary U-Zr-0 liquid, and the reloca-tion of the fuel debris.
These experiments will start in FY 1982 with completion scheduled for FY 1985.
Few-pin Experiments on Debris Forration by Reflood - These experiments will provide data fron post-test examII1ation on the characteristics of the debris produced by reflood of severely damaged hot fuel at several points in candi-date severe accident sequences.
Such data will be used directly for select-ing simul 3ted debris for subsequent experiments in the inpile core-debris coolability experiments.
The data will also be used for safety assessment and nodel developnent.
These experiments will start in FY 83 and be com-pleted in FY 85.
Phenomenological Experiments on Molten fuel Streaming and Blockage Formation, lioth Laboratory and [n-pile - The purpose of this work is to acquire a data base for the development of analytical nodels of streaming and blockage formation by molten fuel and by the fuel, clad, and clad-oxide liquid phases.
The work includes:
(1) laboratory experiments with nonreactor materials on the basic processes involved, and (2) fission heated experiments with reactor materials that can also provide continuous heating of the fuel in simple, well-characterized geonetries.
Similar fast-reactor-safety exportments, both laboratory and in-pile, are currently underway.
They will be modified and extended to cover LWR-specific conditions in FY 1983 and will be completed in FY 1986.
4-13 Model Development - Appropriate analytical models of the governing processes in debris formation and relocation will be developed from the results of the experimental program.
These nodels will be used where appropriate in SCDAP and in accident consequences codes such as MARCH 2 and MELCORR.
There will be interaction and feedback between the nodel developrent and the developing experirental program.
4.4.2 Debris Coolability Exrerirents in the ACRR In-pile Experiments on Debris Dev-Out Coolability limits - These phenomenological separate ef fects esperiments will be performed to obtalii a data base for safety assessment and model development on the coolability limits of parti-culate LWR core debris under reflood conditions over the range of applicable LWR accident conditions.
These experiments are an extension of current successful experinents in ACRR on the coolability limits of particulate LMFOR cor e debris in a sodiun pool.
A major determinant of the coolability limits of a bed of particulate core debris is the character of the debris, in particular the particle size distribution and the degree of vertical stratification by size in the bed.
00ta on debris characterizction will be obtained from the experiments on debris formation and relocation listed previously.
The initial experiment will validate the use, with water, of models derived fron sodium-cooled experiments.
LWR-specific conditions that require experiniental investigation are the ef fects of hi h pressure, inlet Cooling, and very deep debris beds.
The LWR-specific debr.s-coolability experiments in ACRR will be started in FY 82 and completed in FY 86.
Close coordination will be riintained with laboratory work elsewhere on the debris-bed dry-out coolability limit, most of which use inductive heating of beds of retal spheres.
The ACRR experiments provide check points in parameter space with 002 debris that has uniform volumetric heating regardless of particle size distribution or vertical stratification of the bed.
The ACRR experinents with fission-heated U02 debris have unique capability for inves-tigating the ef fect of netallic zirconium and Zr02 debris in the debris bed, along with the U02 fuel debris, and also the post-dry-out behavior of such beds.
The ACRR experiments are also better instrumented than current labora-tory experiments, and are the only planned debris cooldt.ility experiments to investigate the prototypic LWR high pressure range.
These experiments and the inductively-heated laboratory experinents are actually complementary and not duplicative.
Current laboratory work on debris-bed coolability is under-way at KfK, ANL, Westinghouse, and UCLA.
P)ssible Follow-on In-Pile Experinents on Post-Dry-Out Debris !!ehavior - The experieents on debris-bed coolability limits in ACRR with beds of particulate debris in a sodium pool have shown that local dryout is not the true coola-bility limit of a bed, but that stable operation at decay-power levels with part of the bed in dryout is possible. As with the LMFBR debris coolability research, it may prove desirable to perfcrm experiments in the range of extended local dryout in the bed.
These experiments would give data on
4-14 coolability limits associated with the melting of dif ferent materials and eutectics in the bed, and also on the transition into partial core melt of a bed in extended dryout.
They would furnish a base for future research on melt progression to pressure-vessel failure.
These experiments, probably in ACRR, could start in FY 84 and would continue through FY 86.
Model Development - Appropriate analytical models of the governing processes in cetermining debris-coolability limits will be developed from the results of the experimental program.
These models will be used, where appropriate, in SCDAP and in accident codes such as MARCH and CONTAIN.
There will be interaction and feedback between the model development and the developing experimental program.
4.4.3 Melt Procression There are no real data or other-than-preliminary analyses on the sequence of tvents and the governi M processes in the progression of core melt to pressure-vessel failure. Both the range of accident conditions that result in pressure-vessel failure and the nature of the failure with the melt release into the containment are of considerable importance in safety assessment.
Before substantial experirental work is undertaken, however, in-depth analysis of the melt progression to pressure-vessel failure is needed to determine what, if any, new experimental data are really needed.
If indicated, large-scale, in-pile, melt-progression experiments can be performed in PBF, particularly in its upgraded form, and smaller-scale experiments can be perfomed in ACRR.
Large-scale integral melt-progression experiments, if needed, can also be performed with the large-Melt-Facility (LMF) which can provide pours of up to 500 kg of miten U02 Laboratory experiments may prove to be needed on material properties, rele< ant reaction themodynamics and kinetics, and melt-progression mechanisms.
The capabilities for research on melt progression and reactor-vessel attack of the upgraded PBF and the LMF are largeiy complementary, and both can accommodate relatively large-scale experiments.
The PBF provides continuous heating of the melt (as does ACRR) for experiments on relatively-slow long-term melt-progression processes, while the LMF provides melts for large pours to study transient processes governed by stored heat.
Smaller-scale experi-ments or the dominant phenomena in melt progression can be performed in ACRR, as an extension of the extended dry-out experiments on debris-bed coolability.
This program element will start in FY 83 with an in-depth analysis of the dominant sequences in the progression of core nelt to pressure vessel failure and on the dominant processes.
The need for new experimental data in order to properly assess melt progression and the threat to the integrity of the pressure vessel will be detemined.
Experimentation, if needed, could begin in FY 84 4.4.4 Laboratnry Experiments The program of laboratory separate effects experiments includes:
(1)the DECCA experiments on the coolability limits of electrically-heated multi-rod bundles that have undergone severe ballooning deformation, (2) experiments on ternary U-Zr-0 thermodynamics and kinetics, and (3) laboratory experiments on debris formation by reflood quenching of molten fuel.
4-15 Deforced 9undle Coolability Experiments (DECCA)
DECCA Test Facility:
The DECCA test facility would be constructed by modifying the presently existing Thermal-Hydraulic Test Facility (THTF) at ORNL used for the conduct of the Blowdown Heat Transfer research program for the Separate Effects Branch, RES/NRC.
The required electrical power supplies, coolant pumps, control instrumentation, and data acquisition systems are already in place and serviceable.
The test chamber would have to be replaced or modi-fied to accept the differently sized, replaceable test bundles of full-length fuel rod sirulators required for the DECCA program.
DECCA Experiments:
Electrically heated, full-length, 64-rod bundles of fuel rod simulators using Zircoloy cladding would be ramped in temperature during various hign-pressure small-break LUCA boildown simulations to deform, bal-loon, and burst some of the simulators under known thermal-hydraulic condi-tions.
The defomed bundle would be removed from the test facility and rein-strumented for a set of tnemal-hydraulic tests in the same facility.
These,
tests would detemine the cladding and steam temperatures, steam flow rates, etc., both upstrean and townstrean of the deformed and blocked region of the bundle, and tnus the coolability of the bundle af ter severe deformation has occurred.
This infomation is important since it may significantly affect the production and type of debris produced at higher temperatures during SFD (see Chapter 3.0).
Ternary U-Zr-0 Thermdynamics and Kinetics - This program would include:
(1)
Eboratory-scale stucies of the tnermodynacnics and kinetics of the reactions between U02, Zircaloy, and steam over a temperature range from at least 1700 to 2600K (2600*F,1400*C to 4200*F, 2300*C); (2) the physical and oxidation properties of liquid U-Zr 0 phases at all temperatures up to about 2600K (4200'F, 2300*C); (3) the " candling" behavior of dripping, reacting, liqui-fled fuel fomed by raaction of Zircaloy, U0, and steam in bundles of fuel 2
rods; and (4) the ef fects of reflood quenching on relatively large masses of liquid and frozen U-Zr-0 phases formed in the test bundles, fhese experi-ments are planned to begin in FY 83 and by FY 85 it is expected that the initial programs on reaction kinetics will have provided enough guidance to allow interpretation of the data from the other experiments.
The program would be completed in FY 87.
Currently planned foreign programs (the FRG in particular) nay preclude the necessity of these ex-reactor studies.
Final plans for foreign-sponsored studies will not be available until 1982.
There-fore, a decision on USNRC research in this area will not be finalized until
- then, laboratory Experiments on Fuel Debris Formation and Character 1 ration - The purpose of this work is to develop a data base and, where needed, analytical models on the fuel debris fomation process.
The emphasis is on the complete-ness of the fragmentation of molten fuel debris upon reflood and on the characterization of the debris with regard to particle-size distribution in particular.
Data are needed for the full range of molten materials that occur in severe LWR accidents, including the various oxidation states and liquid phases.
Field-scale fragmentation tests in water will be performed with themite melts in the fully instrumented test series (FITS) facility to
4-16 cover a range of relevant melt-oxidation states and liquid conditions, how-ever, these tests will be quite limited.
Starting in FY 1983 and ending FY 1985, f ragmentation experiments will be performed with furn3ce-heated melts in the large Melt f,acility (LMF), where larger-scale experiments can be perforced that cover a rx;ch larger range of fuel, clad, and clad-oxide states and cutectics than is possible with the self-heated thermite melts.
4.4.5 Related Stean Generation Experiments The rapid generatian of steam and hydrogen during reflooding of a hot damaged core nay be sufficient to fail the hot, weakened reactor vessel by over-pressure, leading to full core melt.
The margin to overpressure failure of the reactor vessel during reflood in the TMI-2 accident may not have been large. Allowable reactor-vessel pressure may set limits on the pernissible reflood rates at different times in the managerent of severe accident sequences.
Of concern are both explosive and non-explosive steam generation from reflood-ing of molten or hot 5011d fuel.
Research on rapid steam generation is not part of the SfD program.
- However, sone related work on steam explosions and steam generation by solid partic-ulates is now underway in the fuel Behavior Branch and elsewhere in the Division of Accident Evaluation, but this work is largely addressed to ex-vessel conditions. Assessment is needed as to whether this current work reets the requirenents of in-vessel accident managenent.
Further work may be needed to determine stean-generation limits on reflood rates for the in-vessel nanagement of severe accidents.
4.5 roretan Proqrrs On Severe ruel Damago Our current understanding of foreign programs in SFD was obtained during an infornal recting on plans for studie', of severe fuel damage ($fD) in LWR's held at the Japanese Atomic Energy Research Institute (JAERI) Headquarters Building in Tokyo, Japan on May c5, 1981.
During this meeting, the needs, justifications, and plans for research on STD were discussed by participants fron Japan, United States, Federal Republic of Germany (FRG), and france.
The plans of the USNRC are tne subject of this report.
A summary of the discussions fron each of the other nations is presented below.
Jann - Several committees have been formed at several levels (f rom technical to political) to study the Japanese needs for research on accidents beyond the design basis accidents (DBA's).
The Japanese do not consider that they have a need for SID studies at this time, and have no research programs on SfD actively underway or planned.
They do consider that a power-cooling-mismatch (PCH) during power operation could cause a " cohesive Core" to be fonted, resulting in a large amount of core damage which could include melt formation over a large area of the core at some given Icvel, but they have not yet estimated the probability.
The JAER! committee examining long-term core coolability (problems re'.ulting from an accident beyond a DUA) is pre-paring a final report.
The staff of JAER! and the Japanese ACRS are very interosted in exchanging research infomation to obtain the SFD data and codes developed by other organitations, and in participating in discussions of data and plans for research with them.
4-17 Federal Republic of Germany (FRG) - The planning and research on STD in the TRG is beina conducted by Projekt Nukleare Sicherheit (PNS) at the Kernfor-Schungszentrum Karlsruhe. At this time, the FRG Safety Committee is apparently in full agreement with the present PNS plans on SFD research.
The PNS objec-tives for their SfD studies are:
(1) to increase their knowledge of the safety nargins presently existing in their operating reactors, and (2) to prove out the sifety systems to show they can mitigate the risk in an acci-dent causing severe fuel damage.
PNS plans no in-pile SFD studies at this time, although some discussions are underway as to the feasibility of conduc-ting SFD tests in the French reactor PHEBUS.
They are presently planning to extend:
(1) the 1976-1979 ex.ptle scoping work of S. Hagen (at KfK) on bundle damage, candling, liquified fuel fonnation, etc., using fuel rod simulators heated electrically by centerline tungsten rods. and (2) the scoping work previously conducted by P. Hofnann (KfK) on the U-Zr-0 phase diagram and recction kinetics in the system. A new test facility for the bundle studies is being designed to acconnodate bundles of up to 7 x 7 fuel red simulators 1 meter long, with an availability of quenching by water reflood to disrupt the damaged fuel rod simulators.
All work will be done at near-atmospheric pressures of both steam and water. Modeling of SFD (based on data reported by Hagen) has been previously conducted at the University of Stuttgart, producing the code called EXMEL.
PNS now plans to continue this development at KfK. using the available code SSYST for calculations to 1200*C (2200*f), and a modified and extended EXMEL above that temperaturts.
EXMEL will be modified to include new models on nelting and slumping. Start of melting, dif fusion of the Zr-UO2 liquid interface into solid U02. cladding feilure, solidification of the relt, and stop-and-go molt-freeze during candling.
The PNS proposes no time schedules as yet, except that the redesign of the original Hsgen facility is actively underway.
The PNS/fRG is also participating in the ESSOR SUPER-SARA program on SFD.
The pNS is very inter-ested in an exchange of research data and plans on SFD, as they have been in the past for research on fuel elenent cladding and codes.
France - Several of the french fuel researchers are examining the types of research they ray need on SfD, but there are no research programs presently underway or planned with the exception of participation in the ESSOR SUPER-SARA SFU program.
There are internal discussions concerning the nodifica-tions that would be required for conduct of SFD tests in PHLBUS, but there are no firm plans at this time.
Great Reitain - From recent informal discussions with knowledgeable staf f in tre'TfADE it appears that 1.he UK does not have, at this time, more than tentative and preliminary plans for STD studies in their research programs other than participation in the SUPER-SARA SfD study.
However, we expect the UK will be reviewing their needs in the near future to determine the extent of their interest and possible participation in the $fD proqram in the PBf and the other SfD studies underway and planned by the NRC/RLS.
Ital 2 - The Italians have shown great interest in the NRC SfD program as
'cifr~rently planned and may ptrticipate financially.
They will also ho parti-cipants in the ESSOR SUPLR-SARA SFD research program discussed in Section 4.3.2.
4 18 Moltinattanal - At the tirne of this report, only one cultinational research progra-an SID is knewn to tv planned or underway.
This program is the
$UPLR-5 ARA STL' staJy planned for conduct in the LSSOR reactor at Ispra.
- Italy, ibis stu-1y is one of those supported by the European Economic Cocmunity (ELC) of w5t"h the f RG, f rance, Italy, the llK, and several other European countries are partners.
Participation by the U$',RC in this progrum is currently t'ejng reviewed.
s k
5.0 ADDITIONAL CONCLUS!0NS Additional conclusions o' the task force are sunnariled in this Section.
These evaluations focus on a nunt'er of irportant Criteria for anciting the approach, content, and schedule for the program and the cf fcctivenc%s of resource utilization.
5.1 inn Need for a sovere Fuci Damy Proot an A program of research on the characteristics end behav!or of $cverely darna Jcj fuel 15 clearly needed for decisions ami action'. In the NRC regulatory procen.
The infomation required includes:
o A data baso and analytical rethodology for developing regulatory requirc-c.cnts for severe accidents.
O Infomation for severe accident Nnagement plant and operation 4 such at:
Data for the " Accident Signatures Kindbook" developed by the $A$A prograr.
fast-running accident nanagement analysis codes for use in train-ing operatort and improving sinulator sof tware.
Core con 11tions for which $1rrple reflood will tominate the accident.
The neccenary coolant flow rate to teminate accidents under various core conditions.
Tho$e conditioni for which reflood wortent the accident.
o K;re accurate in-ventel radioactive source term and hydrogen generation r a t e '..
o Refinerents in the design of engineered gafety features and operating practices for tovere accidents.
The consequences of not doing the research on *,cvere fuel darnage are the following:
o Infomation which it the cornerstone of key accident ranagenent analyset and strategies for in-vettnl termin4 tion of a irportant levere accident teguences will not be available.
o rey phenomenological infomation will not be available for auctling the risk reduction potential of accident mnagencnt and recovery strategies for early temination prior to attuming core melt.
$1
2 l
l 5-2 r
1
\\
y S.2 Idenjif tcaQn of Specific Up Roscarch Needs
{
The areas of uncertainty and related safety tuucs in 5fD phenomenology have been renonably well defined; however, anessments of the relative importance 7
of these urcertainties an1, hence, their priority it presently not certain.
This situttlon is largely a result of the very limited data base and analytical rothodol01y available in 5FL. Moreover, spectfic research requirements for
'J D in terns of 'WC pro;ramitic needs, c.9.. accident management, and safety features, cannot be precitely delineated at the present time because the objecttve%. scope, overall approach, and interf acc5 for these other activities are tming defined.
3?
5.3 Me n Techn'ral Approach Although Sf D sequences may be numerous and hany of the goynrning phenarena 5
technically corples, the a;)proach currently outlined appears to be viable in 2
tern nf resolvin] key Itsuc% with reasonablo atturance of succe%% for achiev.
ing a bournf able, quan*itative, research product.
?
liy far the rmt productive and cost cf fcctive technical approach it an intrgrated program involving:
g o
Integral in Dile tettt!
]A
-a Init141 tcuping te$tt to determino overall behaylor and governing prenor ena.
Later vertf tcetion tests of the integral codes developed in the 7
progre.
i Separate etfccts phenomenological esperiments for the devolopment and f;
o verlf tcet ton of no tels of the gnverning phenortnal 2;
i' In pile espertrent' (1.c., debris formation rolocation,anit t,colab111 ty ),
y l
Laboratory ceperiments (1.c., outdation ons clad ballooning).
]'
e o
Analyt1 Cal Podel developW nt!
f 2
the Integral %vore fore { Wage Analyttg Package (SCOAP) for hic in the analyt16 of Accidents and Integral espnritncntt, Phenonenological twielt for ute 4% fvfulpt in $CDAP and in tho
=
analygtt of separate cffects esperipents.
.i 0
_.a
.w_-
m_
a Although the progran technical approach 14 4 viable onn, it is not cicar whether thn various steget of 4 MD accident prpience (e.g., early fuel damagn, govero fuel damagn, and core molt progrettfon) are rocciving balanceil
.c 5-3 enphasis.
The current erphasis evolves from our perception of the SFD phen-1 omenolcyical research needs as derived from a very limited data base.
The m*
research needs and priorities must be continually reevaluated because per-spectives and insights gained fron new analyses.ind experiments may signi-y ficantly elter our thinking.
One area which appears to require additional emphasis is analytical model development.
The highest priority element of the SFO program is the prompt execution of
[
the SfD (Phase 1) series of five integral tests in the PDF.
L O
lt is tne only reasonably near-term source of integral scoping data 7
available to provide:
fuel debris characterization, both with and without reflood, high N
and low heating rates, and under TM1 conditions.
=
The data base required for scoping severely damaged core character-15 tics for separate-ef fects phenenenological experiments, and for A
development and verification of both the Severe Core Danage Analysis Package (SCDAP) and the phenomenological nodels for use in SCDAP.
r Significant data on in-vessel fission product and hydregen source i
terms during a SFD sequence.
7g o
Without these five 5fD tests in Puf, there will be no firm base for a J
STO tescarch program or severe accident assessemnt.
j In addition to the above points, early conduct of the initial PDF series is i
also of critical ictortance to the demonstration of the test train Shroud del 1 9 for the SUPLR SARA test program in EnUR and possible follow-on 5fD 9
tetts in the NRll, in tu+rtaryde pur phaso ! tories is on the critical
- Vn purg,
- a n ur tno lhe 9 0 prograN, 45 currently planned, does not significantly address the question of nelt progretston to failure of the reactor vessel.
A program on this queation may be needed and thould progrett only af ter completion of an carly, realistic in depth analysis of the problem to show which additional osperimental data are required to provide an adequate attestrwnt of the melt progret, tion threat to reactor-vettel integrity as needed for improved PRA.
K Analysis and characterization of the TM! 2 core debril, when it becomes Available, will furnish an invaluable benchnark for severely damaged fuel
^
procettet and for guidance of future esperirmnts and analysis.
It represents our only opportunity for a whole core assentrent of SID predictive nothodology.
When we h4vo results of the Phase 1 PDF tests and the early ACRR experiments, we my find that a lost entensive 5fD program than currently planned will P
rent program needt.
Ihe program objectivet and plant will be reesamined at that tin.
2
5-4 S.5 Selection of Program Facilities and Projection of Future Needs The major in-pile test facilities currently planned for use in the program are the PBF and the ACRR.
The major ex-reactor test facility being considered thus far is the THTF (DECCA) facility at ORNL. We do not see a need for new facilities for SFD testing.
It appears that some upgrading and extension of capabilities of existing test facilities may be programmatically desirable.
Examples of projected extensions and approximate estimates of their cost r
include:
o Upgrading the PBF reactor and test loop for testing full 17 x 17 bundles for Phase 2 program - $3M-4M, o
Upgrading the NRU test loop to include double containment for SFD test-ing of full-length (12 feet), 76 rod bundles - $10M.
Existing out-of-pile facilities which may be useful to the program include:
o Large Melt Facility (LMF) at Sandia - 500 Kg for melt progression, 5.6 SFD Tests In LOFT The LOFT Special Review Group (LSRG) concluded (NUREG-0758) that the SFD tests, originally scheduled for FY 1985-1986, were peripheral to the LOFT mission.
In addition, they noted that the facility does not have the neces-
[
sary equipment to properly handle severely damaged cores, and procurement *
(
costs for such equipment were not included in the present LOFT budget. The
/
LSRG did note, however, that "the DCC decision making that is involved with the multiple rulemaking ventures may create the need for additional tests in the degraded core area." However, it must be realized that current policy precludes the use of the LOFT facility beyond its current program.
5.7 Instrument Capabilities for SFD Experiments Currently available state-of-the-art instrumentation is adequate to meet basic objectives of the SFD experimental program.
However, relatively modest instrumentation development in several areas has the potential for signifi cantly improving the results of the SFD research program. One area is'on-line real-time imaging of the fuel-damage process, both by optical a~nd by fuel-source-imaging techniques. Another area is trans-axial tomography of the post-test debris by neutron radiography. Reinstrumentation of the post-test debris for subsequent in-pile measurement of debris-coolability limits appears promising. Potential improvements in temperature measurement include measurement of surface temperatures by' actual cinematography and film densi-tometer traces and measurement of temperature distributions in debrissby ultrasonic thermometry. Measurement of hydrogen production and the release of both radioactive and non-radioactive gases during fuel-dama'ge transients are important and planned techniques can be improved.
In all of,these areas, the fundamental research has been done and only work on' specific applications in the SFD experiment program is needed.
A-
\\
5-5 5.8 Program Schedule Assuming the accelerated schedule for the SFD Phase 1 tests in PBF, it 1s projected that the program can be completed in FY 1986 (See Figure 1). This would exclude the most important of the SUPER-SARA tests on SFD in ESSOR.
The pacing item in the program is the PBF - Phase 1 test series.
Their schedule is important for the development of severe accident policy guidelines.
Another critical item of considerable concern will be the continuing delays in the TMI-2 core examination as well as the possibility that it will never be completed.
Completion of the TMI-2 core examination at the earliest possible date is important to the SFD program and to severe accident safety assessment and accident management in general.
5.9 Impact of Budget Constraints The most critical existing budget constraint involves the FY 82 shortfall for the Phase 1 PBF-SFD tests. Additional funding is required in the FY 82-to-FY 83 timeframe to carry out the a::celerated program, e.g., one test in CY 82 and four tests in CY 83 for completion of Phase 1, as opposed to one test in CY 82 and one test in CY 83 with Phase 1 completion in CY 84.
Thus, if the accelerated program is not funded, there would be about a 1 year delay in completing Phase 1, along with the concommitant delays and impacts in other SFD program elements, e.g., SCDAP.
In addition, a low test rate of one test per year is not cost-effective for the operating expense associated with the PBF.
In the area of progression of severe fuel damage to core melt and reactor vessel failure, no funding is currently projected for FY 82 and that in FY 83 is uncertain.
Funding for the analytical development element, particularly SCDAP, is marginal in FY 82.
The sensitivity of the SFD program to budget constraints can be summarized as follows:
o Large reductions in program funding, would have a severe impact on the PBF tests because they are the highest cost element of the program.
The PBF program would not be viable and without PBF the value of the entire SFD program is questionable.
If the reductions are taken in the phenom-enological, separate effects experiments element -or the analysis element, the advantages of the interfacing SFD elements are lost, without which the value of the SFD program is also in doubt.
o As stated earlier, the program is very sensitive to small reductions in funding.
They would impact the program by eroding the phenomenological, separate effects experiments and, hence, the viability of the program approach. An alternative scenario would probably involve elimination of participation in foreign programs, e.g., ESSOR.
i 5-6 On the other hand, small increases in funding above the accelerated o
program, e.g.,10%, would bring a high return for the increment because the program could be upgraded with improved support in such areas as analysis and planning for possible NRU full-length SFD tests.
5.10 Interfaces With Other RES and NRR Programs Important interfaces between the SFD program and other programs in RES and NRR which impact TMI-2 Action Plan responses, such as Severe Accident Rulemak-ing are either non-existent or poorly defined.
These other programs include:
o SASA Program, Systems codes such as MARCH, TRAC, RELAP, o
Whole core thermal hydraulics research, o
o Hydrogen program, o Human Factors o Improved operator training, o Improved operating procedures, o Improved ESF design.
Plans are presently lacking for incorporating SCDAP into one of the severe accident system analysis codes (or vice versa). This is of some concern since plans being made now for SCDAP development should take into account code interfacing and compatibility at an early date.
5.11 Foreign Programs in SFD At the present time, there are only two foreign programs of direct interest to the SFD program.
One is the program in the FRG sponsored by BMFT and carried out mainly at KfK by PNS; most of this effort is the extension of the ex-pile work of Hagen and Hofmann. The other foreign SFD research is the SUPER-SARA SFD test program planned for the ESSOR reactor at Ispra, Italy.
The planned FRG SFD work is complementary to the NRC-SFD program in the ex-pile separate effects and modeling area, and will be timely as well.
}
The tests-currently planned in SUPER-SARA could make a significant, cost-i effective contribution to the SFD program provided the program focuses on the SFD tests (where NRC is focusing its support) at an early date.
In the current schedule for the SFD tests in SUPER-SARA, the first test with condi-tions equivalent to the first SFD test in PBF would not be run until about early 1986 (almost 4 years later). Our concern is that the SUPER-SARA SFD test schedule starts to provide information when the NRC-SFD program is essent Mlly completed. Based on the current schedule, the SUPER-SARA SFD information will not be timely.
l l
1 i
APPENDIX A Task Force Objective, Scope of Review, and Activities An NRC Fuel Testing Task Force was appointed on March 27,1981, by the Director of the Division of Reactor Safety Research (RSR), Mr. O. E. Bassett, to evaluate ways for resolving the various problems discussed in Section 1.1.
The Task Force consisted of the following RSR staff:
M. Silberberg, Chairman M. L. Picklesimer, Vice-Chairman G. P. Marino R. W. Wright R. Van Houten The following consultants, largely representing RSR contractors, were made available to the Task Force:
Name Affiliation R. Chapman Oak Ridge National Laboratory (ORNL)
E. Courtright Battelle-Pacific Northwest Laboratory (PNL)
P. MacDonald EG&G Idaho, Inc. (EG&G)
D. Morrison Illinois Institute of Technology Research Institute (IITRI)
A. Pressesky U.S. Department of Energy (D0E)
J. Scott Los Alamos National Laboratory (LANL)
J. Walker Sandia National Laboratory (SNL)
The Task Force commenced its evaluation at an initial meeting at Headquarters on April 16 and 17,1981.
Subsequent meetings were held at Headquarters on April 30 and May 1 and June 10 and 11,1981.
During May 26 through 29, the Task Force visited the PBF at INEL and the ACRR at Sandia to review the in-pile test programs currently planned at each of these laboratories. The PBF and ACRR represent the major U.S. reactor test facilities to be used for the SFD program. The LOFT facility was also visited during the INEL review.
In order to meet the objective and charter for our evaluation, it was necessary for the task force to consider the following technical and programmatic details in the scope of the review.
o Programmatic NRC needs in SFD and the role of the SFD program, o
Phenomenological research needs in SFD (our state-of-knowledge).
o Technical scope and approach of the SFD program as planned.
o SFD program schedule and its relationship to degraded core cooling rulemaking.
o Role of related programs in U.S. and foreign countries.
o Facility and instrument needs.
o Interfaces with the SFD Program, i
1
APPENDIX B - RESPONSE TO COMMENTS ON DRAFT REPORT B.1 Summary of Reviews of Draft NUREG-0840 On July 10, 1981, copies of draft NUREG-0840 were sent to members of a special Peer Review Group which included scientists and engineers from various national laboratories *, government agencies, universities, foreign countries, and private industry.
The Peer Review Group convened on August 6, 1981, to discuss the draft report. Those present are listed in Enclosure 1.
At the Peer Review Meeting, the authors of NUREG-0840 requested that the reviewers submit written comments on the draft report. Those who submitted written comments arc listed in Enclosure 2.
Copies of the July 30, 1981 draf t and the written comments can be inspected in the Public Document Room (PDP.) at 1717 H Street, Washington, D.C.
We have considered the comments and have factored them into the revision of the draft report as appropriate.
B.2 Responses to Major Comments Most comments indicated general agreement with the contents of Draft NUREG-0840.
However, because of the strong peer reviewer interest in several topics, those items which received special support will be listed; no response is needed beyond this listing.
1.
Recommendations for In-Depth Examination and Evaluation of the Damaged Three Mile Island-2 Reactor Core - Almost all peer reviewers agreed with the NRC Fuel Testing Task Force Members that the comprehensive, proper, and timely examination of the Three Mile Island-2 core was a key part of any adequate SFD test program.
In fact, most reviewers who submitted written comments felt so strongly on this topic that they elected to make a special plea for this activity.
The almost unanimous special support for comprehensive early examination of the Three Mile Island-2 core was perhaps double both the breadth and the depth of special support for the next three items.
2.
Recommendation for comprehensive theoretical and experimental study of SFD test bundle length relationships.
3.
Recommendation for incorporation of control rod materials into the SFD tests.
4.
Recommendation for special evaluation of both the positive and negative aspects of fission heat simulation of decay heat (it is recognized that, e.g., migration and volatilization of radioactive fission products can I
cause a measurable change in local heat generation rates. This, in turn, can influence time-related spatial assessments of fission product source terms.
National Laboratories who were represented on the Task Force were excluded from the peer review.
B-2 A fifth item received frequent comment and its support was generally coupled with an expression of some type of special reviewer concern.
The item was:
5.
Support of SCDAP, the modular Severe Core Damage Analysis Package computer code.
Reviewer comments variously reflected concerns that:
a.
The code modules could become ready tools for hampering accident management by the addition of artificial conservatism.
(It was feared that these conservatism fettered code modules could lead, in turn, to invalid or improper restrictions on allowable accident evaluation and/or accident management devices and procedures.)
b.
Improper user attention could be given to code predictions and extrapolations, ignoring the fine details which could be derived from independent examination of the test data (i.e., that the code might become a harmful crutch).
c.
The code would be too complex.
d.
Using the code might prove too expensive and cost ineffective.
e.
The code might not run fast enough.
f.
The code development and control might be too arcane.
These comments largely reflect the check list which should be considered during the development and early use of any new computer code. One of the best reasons for reviewer reassurance that the SCDAP code develop-ment early use and assessment will vigorously and effectively avoid these pitfalls is the excellent record of the team of code developers and their very recent experience and success with the development of the FRAP and FRAPCON fuel behavior computer codes.
Several other SFD program items, e.g., the DECCA (Deformed Core Coolability) program and program pacing showed a wide spectrum of recommendations, with less consensus and at least a few contradictions. With respect to the proposed DECCA studies, future decisions on this work will be based on evalu-ation of nearer term test results from the REBEKA and NRU programs, as well as budget constraints. With respect to program pacing, the current schedule should provide major support for the preparation of a proposed Severe Accident Rule and its confirmation.
Future submittals from the Peer Review Group or from others will be considered at the quarterly or semiannual meetings of the SFD Program Review Group.
ATTENDEES Severe Fuel Damage Task Force Review PEER REVIEWERS (AT THE TABLE)
Name Organization M. Silberberg*
NRC/RES R. Van Houten*
NRC/RES G. P. Marino*
NRC/RES M. L. Picklesimer*
NRC/RES R. W. Wright
- NRC/RES G. E. Culley HEDL G. R. Thomas EPRI/NSAC L. E. Hochreiter Westinghouse /NES L. W. Deitrich ANL/ RAS R. Gehlberg EPRI S. M. Gehl ANL/MSD M. Ishikawa JAERI/ Japan T. Hoshi JAERI/ Japan P. M. Lang 00E/HQ 3
J. S. Tulenko B&W R. A. Proebstle GE D. L. Burman Westinghouse /NFD G. D. McPherson NRC/RES L. G. Williams U.K.N.I.I.
E. D. Hindle U. K. A.E. A.
J. C. Janvier CEA/ France N. Duco CEA/ France A. W. Fiege KfK/PNS/ Germany W. Gulden KfK/PNS/ Germany P. Griffith MIT W. Y. Kato BNL
- Authors and Task Force Members
-- (Cor.tinued)
ATTENDEES Severe Fuel Damage Task Force Review GUESTS Name Organization A. S. Rao GE J. B. Rivard SNL J. E. Hanson EG&G F. J. Erbacher KfK A. J. Pressesky DOE /HQ E. T. Laats EG&G J. V. Walker Sandia R. R. Hobbins EG&G R. W. Miller EG&G W. A. Spencer EG&G S. L. Seiffert EG&G B. J. Buescher EG&G H. Bairiot Belgonucleaire J. M. Broughton EG&G P. E. MacDonald EG&G T. S. Kress ORNL J. M. Kramer ANL/E D. J. Osetek EG&G G. Sdouz 0FZS/ Austria C. H. Bowers ANL/E K. Hassmann KWU J. Duco CEA A. C. Marshall Sandia M. S. El-Genk EG&G P. S. Pickard Sandia E. L. Courtright PNL R. H. Chapman ORNL P. R. Davis Intermountain Tech.
List of Commenters on Draft NUREG-0840 i
(As of November 6,1981)
Individual Organization Date G. Culley HEDL Undated H. Denton NRR 10/29/81 L. Deitrich ANL 9/11/81 P. Griffith MIT 8/20/81 E. Hindle UKAEA 8/21/81 L. Hochreiter W
9/25/81 R. Oehlberg EPRI 10/19/81 J. Taylor B&W 9/3/81 L. Williams UKNII 8/26/81
ACRONYMS, INITIALISMS AND GLOSSARY AC Alternating Current ACPR Annular Core Pulse Reactor (Pre-ACRR, Sandia)
ACRR Annular Core Research Reactor (Sandia)
ACRS Advisory Committee on Reactor Safegards (U.S.)
/
AECL Atomic Energy of Canad Limited AHL Argonne National Laboratories (Chicago, Illinois)
ART Aerosol Release and Transport B&W Babcock and Wilcox (Reactor Manufacturer)
BDHT Blowdown Heat Transfer (Facility, ORNI,)
Be0 Beryllium 0xide BMFT (Reactor Safety Licensing Commission, FRG)
CAIS Coded Aperture Imaging System Cd Cadmium CD Core Disintegration CONTAIN Accident Containment Computer Code Cs Cesium CY Calendar Year DAE Division of Accident Evaluation (NRC)
DBA Design Basis Accident DCC Degraded Core Cooling DECCA Deformed Core Coolability DOE Department of Energy (U.S.)
dQ/dt Cooling rate (Kelvins /sec)
DRA Division of Risk Analysis (hRC) dT/dt Heating rate (Kelvins per second) e Uranium 235 enrichment ECCS Emergency Core Cooling System EEC European Economic Community EG&G Contractor (Idaho Falls, Idaho)
EPRI Electric Power Research Institute (U.S.)
ESF Engineered Safety Feature ESSOR Essais Orgel Reactor (Ispra, Italy; contair.s Spor Sara Test Loop)
EXMEL Fuel Melting Computer Code (FRG)
F Temperature in Degrees Farenheit FASTGRASS-Fast Running Fuel Fission Product Release Computer Code (NRC)
FBB Fuel Behavior Branch FITS Fully Instrumented Test Series (Facility, Sandia)
FPDS Fission Product Detection System FRAPCON Fuel Behavior Computer Code (Steady State, NRC)
FRAP-T Fuel Behavior Computer Code (Transient Conditions, NRC)
FRG Federal Republic of Germany FY Fiscal Year (October 1 to September 30) g Gram G
Mass rate of flow of coolant (H 0)
TMI related reports LGPU, ~EPRI,2NRC, DOE)
GEND GPU General Public Utilities (TMI Owner) -
H Hydrogen 2
I Iodine I&E Office of Inspection and Enforcement (NRC)
ICP In-Core Phenomenology IITRI Illinois Institute of Technology Research Institute In Indium INEL Idaho National Engineering Laboratory (D0E; Contractor is EG&G)
.. J Joule JAERI Japanese Atomic Energy Research Institute K
Temperature in Kelvin (K = Centigrade + 273.1)
Kg Kilogram Kr Krypton KfK Kernforschungszentrum Karlsruhe (FRG)
LANL Los Alamos National Laboratory (New Mexico)
LMF Large Melt Facility (SNL)
LMFBR Liquid Metal Fast Breeder Reactor LOCA Loss of Coolant Accident LOFT Loss of Fluid Test (Facility, INEL)
LPSE Low Power Safety Experiment LSRG LOFT Special Review Group LWR Light Water Reactor (s)
M Million MARCH Accident Computer Code MATPR0 Materials Properties (Computer Input)
McGuire Commercial Power Reactor MESF Minimum Engineered Safety Features MP Melt Progression (Program)
MPa Mega Pascal (Pressure unit = 147 psi)
MRBT Multi-Rod Burst Test (0RNL)
MWD /T Megawatt day per long ton (1000kg)
NRC U. S. Nuclear Regulatory Commission NRR Office of Nuclear Reactor Regulation (NRC)
NRU National Reactor Universal (Chalk River, Canada)
NUREG NRC Report 0
0xygen ORNL Oak Ridge National Laboratory (Tennessee)
AP Pressure Difference Between fuel rod fill gas and reactor coolant PBF Power Burst Facility (INEL)
PCM Power Cooling Mismatch PHEBUS French Test Reactor PIE Post-Irradiation Examination PNL Pacific Northwest Laboratory (Contractor, Richland, Washington)
PNS Projekt Nukleare Sicherheit (Nuclear Safety Project, FRG)
PWR Pressurized Water Reactor R&D Research and Development RELAP Thermal Hydraulics Computer Code RES Office of Nuclear Regulatory Research (Research and Standards, NRC)
RIA Reactivity Initiated Accident RSR Division of Reactor Safety Research (now DAE)
Sandia Contractor (Albuquerque, New Mexico)
SASA Severe Accident Sequence Analysis SCDAP Severe Core Damage Analysis Package SFD Severe Fuel Damage SIMMER Core Melt Computer Code SNL Sandia National Laboratory (Albuquerque, New Mexico)
Sr Strontium
_ _ _ _ SSYST Fuel Behavior Computer Code (FRG)
Super Sara - EEC Reactor Test Loop - Ispra, Italy T
Temperature (in Kelvins = Degrees centigrade + 273)
TCE Total Clad Elongation Te Tellurium THTF Thermal Hydraulic Test Facility THI Three Mile Island Reactor (GPU, Pennsylvania)
TRAC Thermal Hydraulics Computer Code U
Uranium UCLA University of California at Los Angeles UK United Kingdom (Great Britain)
U0 Uranium dioxide reactor fuel 2
w Watt WD Water-Debris Xe Xenon Y0 Yttrium oxide Zh Zirconium Zr0 Zirconium oxide Ziok/ Indian Point - Commercial Power Reactors (Illinois, New York) t
- I U.S. NUCLE AR REGUL ATORY COMMISSION I
1 BIBLIOGRAPHIC DATA SHEET 4 T6TLE AND SUBTITLE (Add Vo/urne No.,,f aopicer<aret
- 2. (Leave e/ane)
PROGRAM ON BEHAVIOR OF DAMAGED FUEL Report of the NRC Fuel Testino Task Force
- 3. RECIPIENT'S ACCESSION NO.
I
- 7. AUTHORtS)
R Van Houten u^o'N rs lvEAR G. Marino, R. Wright, M. Silberberg, M. Picklesimer December 1981
- 9. PE RFORMING ORGANIZATION N AME AND MAILING ADORESS (/nclude I,a Code)
DATE REPORT ISSUED MONTH lY(AR February 1982 Fuel Behavior Branch, Division of Accident Evaluation 6'*"'*"
Office of Nuclear Regulatory Research O. S. Nucl ear Regulatory Commission
- 12. SPONSORING ORG ANIZ ATION NAME AND M AILING ADDRESS //nclude 2,0 Cooe A
- 10. PROJECT / TASK / WORK UNIT NO.
awe
- 11. CONTR ACT NO.
Task Force Report - Program Plan April - December 1981
- 15. SUPPLEVENTARY NOTES j4 ft,,,,of,,,j
- 16. ABSTR ACT (200 woras or irss)
An NRC Fuel Testing Task Force was appointed in March 1981 by the Director of RSR to review and evaluate the projected information needs, program, and test facility resource requirements for research on the behavior of fuel assemblies during severe accidents, e.g., severe fuel damage (SFD). The objective of the SFD program is to provide the experimental data base and analytical methodology for understanding and predicting core behavior under severe accident sequence conditions. The PBF and ACRR represent the major reactor test facilities to be used for the SFD program. The report considers the following technical and programmatic details: Programmatic NRC needs in SFD, Phenomenological research needs, SFD program schedule and its relationship to rulemaking, Role of related foreign and U.S. programs. The report has been organized into two volunes, the main report (Vol 1) and Volume II which contains additional technical and background infonnation on SFD processes, research needs, and facility descriptions.
The report notes that infonnation from the TMI-2 core examination will furnish an invaluable benchmark for SFD processes and code analyses.
- 17. KE Y nOR DS AND DOCUMENT AN ALYSiS 17a CESCRIPTCR$
Severe Fuel Damage Fission Products Debris Coolability Power Burst Facility Model Development Severe Accident Rulemaking Core Melt Accident Management Annular Core Research Reactor 17b. IDENTIFIE RS' OPE N ENDE D TE RMS 19 Unclassified 56 CURITY CLASS (TAs recorr/
21 NO OF PAGES 18 AV AILABILITY STATEVENT Unlimited
- 20. SE CuRiTY CL ASS (T6s page) 22 PRICE S
NEC FORM 335 17 77) e U.S. GOVERNMENT PRINTING OFFICE: 1982 361-302/512
UNITED STATES 2
NUCLEAR REOULATORY COMMISSION f
]
C W ASHIN GTON, D. C. 20555 POST AGE AND P E ES P AID OFFICI AL BUSINESS u S. N'JC LE A R R g Ou L A TOR Y PEN ALTY FOR PRIV ATE USE, $300 COMMISStDN g
k_
d P
0 M
LD o c//
'YU o:?, ; yyn
. UNION OF CONCERNED SCIENTISTS i34e c_-ti. me_..s.w.. s. noi. wa.hi.,,.. nc 2eo3e. <2o2,2,e 5eoo 20 February 1984 Mr. J. H. Felton, Director T RE I
Division of Rules and Records
[OZA pg/f Office of Administration U. S. Nuclear Regulatory Commission kg /g g Washington, D.C.
20555 RE:
FOIA Ecquest for NUREG-0731, NUREG-0840, SECY-81-651, and SECY-82-3 [Sholly FOIA Request Number-84-05]
Dear Mr. Felton:
Pursuant to the Freedom of Information Act, please make available at the Commission's Washington, D.C.,
Public Document Room copies of the following listed documents:
A.
NUREG-0731; B.
NUREG-0840; C.
SECY-81-651; and D.
SECY-82-3.
Should there be any questions about this request, please have your staff contact me at (202) 296-5600.
Your cooperation in responding to this request will be appreciated.
Sincerely, l
Steven C. Sholly Technical Research Associate cc:
Ellyn Weiss, Esq.
l v?S gyo V/ fM 6-Stain Office: 26 Church Street. Cambridge. 51amchusette o2238. (617) 547-5552