ML20210K549
| ML20210K549 | |
| Person / Time | |
|---|---|
| Issue date: | 06/30/1975 |
| From: | NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | |
| References | |
| NUREG-75-058, NUREG-75-58, NUDOCS 8702120259 | |
| Download: ML20210K549 (147) | |
Text
--
NUREG -75/058 REACTOR SAFETY RESEARCH PROGRAM l
A DESCRIPTION OF CURRENT AND PLANNED REACTOR SAFETY RESEARCH SPONSORED BY THE NUCLEAR REGULATORY COMMISSION'S DIVISION OF REACTOR SAFETY RESEARCH l
JUNE 1975 UNITED STATES NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REGULATORY RESEARCH DIVISION OF REACTOR SAFETY RESEARCH WASHINGTON, D.C. 20555 elR22E8!!! " '
75/059 R PDR
7_
NOTICE THIS REPORT WAS PREPARED AS AN ACCOUNT OF WORK SPONSORED
.BY.THE UNITED STATES GOVERNMENT..NEITHER.THE UNITED STATES NOR' THE UNIT D STATES NUCLEAR REGULATdRY COMISSION, NOR ANY OF THEIR EMPLOYEES, MAKES ANY WARRANTY, EXPRESSED OR IMPLIED, OR ASSUMES ANY LEGAL LIABILITY OR RESPONSIBILITY FOR THE ACCURACY, COMPLETENESS OR USEFULNESS OF ANY INFORMATION, APPARATUS, PRODUCT OR PROCESS DISCLOSED, OR REPRESENTS THAT ITS USE'WOULD NOT INFRINGE PRIVATELY ~.
0WNED RIGHTS.- -
i Available from National Technical Information Service Springfield, Virginia 22161 Price: Printed Copy $5.75 ; Microfiche $2.25
NUREG-75/058 1
i REACTOR SAFETY RESEARCH PROGRAM A DESCRIPTION OF CURRENT AND PLANNED REACTOR SAFETY RESEARCH SPONSORED BY THE NUCLEAR REGULATORY COMMISSION'S DIVISION OF REACTOR SAFETY RESEARCH t
i June 1975 Prepared by personnel in the Division of Reactor Safety Research i
l I
(
United States Nuclear Regulatory Commission Office of Nuclear Regulatory Research Division of Reactor Safety Research
~
l l
s Y
i i
CONTENTS Page 1
1.
I NT RO DUCT I O N...................................................................
1 l
L 2.
DESCRIPTION OF RESEARCH PR0 GRAMS................................................
-2 3.
WATER REACTOR EXPERIMENTS AND ANALYSES..........................................
2 3.1 Systems Engineering........................................................
2 I
3.2 Fuel Behavior..............................................................
8
.3.3 Analysis Development.......................................................
10 3.4 Metallurgy and Materials...................................................
12 3.5 Environmental and $1 ting...................................................
15 3.6 Other Studies in Water Reactor Sa fety Research.............................
15 4.
FAST REACTOR SAFETY ASSESSMENT..................................................
16 4.1 Fission Product and Fuel Release and Transport.............................
18 4.2 Analytical Methods.........................................................
20 4.3 Fuel Interactions..........................................................
24 4.4 Safety Test Facility Studies...............................................
25-a 4
4.5. Monte Carlo Analysis and Planning for Safety-Related Criticel Experiments..
26 j
4.6 Ga s-Cool ed Fa s t B reeder Reacto r............................................
28 1
4.7 Plant Systems..............................................................
28 5.
GAS-COOLED REACTOR SAFETY ASSESSMENT............................................
30 5.1 Phenomenol og i cal Re search..................................................
33 5.2 Analytical Research........................................................
35 5.3 Proof Tests................................................................
36 5.4 Relationship to the Development Program....................................
36 l
6.
Nucl ear Safety Tes t Facili ti es..................................................
40 l
APPENDIX A - Branch Program Plan for Systems Engineering Branch................. A-1 to A-17 APPENDIX B - Branch Program Pl an for Fuel Behavi or Branch....................... B-1 to B-24 APPENDIX C - Branch Program Plan for Analysis Development Branch................ C-1 to C-25 APPENDIX 0 - Branch Program Plan for Metallurgy and Materials Branch............
D-1 to D-20 APPENDIX E - Branch Program Plan for Environmental and Siting Branch............
E-1 to E-14 4
i l
i j
I-1 1
'h
LIST OF FIGURES Page FIGURE 1 - Safety Research Process...................................................
17 FIGURE 2 - Radioactive Source Release and Transport..................................
19 FIGURE 3 - HTGR Safety Research Process..............................................
34 LIST OF TABLES Page T ABLE 1 - C ode Ve ri fi cati on...........................................................
12 TABLE 2 - Gas Cooled Reactors - Major Concerns........................................
31 TABLE 3 - Gas Cooled Reactor Safety - Phenomenological Research.......................
32 TABLE 4 - Gas Cooled Reactor Safety - Analytical Research.............................
33 TABLE 5 - Comparison of RSR and RRD HTGR Program P1ans................................
36 TABL'i 6 - RRD HTGR Safety R&D Program Man-Year Requirement for 1975-1984..............
40 TABLE 7 - Nuclear Safety Test Facilities.......................
41 Table 1.1 - Index to Siting Questions in Environmental and Siting Safety.........
E-3 Table 1.2 - Outline of Program Plan for Environmental and Siting Branch..........
E-5 List of Projects Underway or to be Initiated During Fiscal Year 1976 (Environmental and Si ti ng Safety)....................
E-14
i 1
ABSTRACT THE REACTOR SAFETY RESEARCH PROGRAM, SPONSORED BY THE NUCLEAR REGULATORY COMMISSION'S DIVISION OF REACTOR SAFETY RESEARCH, IS DESCRIBED IN TERMS OF ITS PROGRAM OBJECTIVES, CURRENT STATUS, AND FUTURE PLANS. ELEMENTS OF SAFETY RESEARCH WORK APPLICABLE TO WATER REACTORS, FAST REACTORS. AND GAS COOLED REACTORS ARE PRESENTED TOGETHER WITH BRIEF DESCRIPTIONS OF CURRENT AND PLANNED TEST FACILITIES.
l
t a
v REACTOR SAFETY RESEARCH PROGRAM 1.
Introduction It is predicted that expanding electrical energy needs in this country over the next few decades will be met in most part by a continued growth of nuclear power. At least over the next decade, light water reactor systems are expected to constitute the largest fraction of nuclear power commitments by the utility industry. The development and use of gas-cooled thermal reactors for power production are expected to increase to some extent during the coming decade, in competition with light water reactor systems. Economically competitive and environmentally acceptable fast breeder reactors are expected in the early 1990's.
The Commission's policies and programs are directed toward ensuring that this growth of nuclear power takes place with adequate assurance of reactor safety. This objective is accomplished in large measure through an intensive safety review and licensing program assisted technically by an active program in safety research.
4 I
The Division of Reactor Safety Research was established in the AEC in May 1973, to provide increased emphasis on reactor safety research associated with water reactors, and to introduce an independent capability for safety assessment with respect to advanced reactors. In January 1975, the AEC was reorganized into.NRC and ERDA with the Division of Reactor Safety Research becoming a j
part of the Office of Nuclear Regulatory Research in the Nuclear Regulatory Commission.
The overall objective in the direction of reactor safety research programs is to provide the basis and means for reliable and credible analysis of the course of events in hypothetical acci-dents to nuclear reactors, and the estimated consequences of such accidents. More specifically, the objectives of the Reactor Safety Research Program are:
t To develop understanding of the basic phenomena involved in hypothetical accidents, as needed for their description in analytical models.
1 To develop the basic data on these phenomena so that the analytical models will have the appropriate realism or conservatism over the desired range of application.
To integrate these data and models into complete analytical descriptions of the hypothetical accidents, accounting for those quantities and parameters important to review of safety.
4 To provide integrated experiments that are designed to test the adequacy of the analytical descriptions in their predictions of accident sequences and consequences, to ensure their accuracy and their completeness.
To apply analytical models to analysis of generic questions as needed for their resolution.
i To develop further the methods of reliability analysis required in the probabilistic analysis of the safety of nuclear reactors.
1 To refine the current conclusions from probabilistic analysis of reactor safety.
To apply probabilistic analysis of reactor safety to other kinds of reactors and other situations than have been analyzed so far.
2.
Desc, ption of Research Programs Research programs under the direction of the R'eactor Safety Research Division are divided into four major activities:
Water Reactor Experiments and Analyses Fast Reactor Safety Assessment Gas-Cooled Reactor Safety Assessment Nuclear Safety Test Facilities The following sections sumarize the principal areas of current and planned research, the specific goals and objectives of the work, descriptions of various tasks in terms of data needs, and the application of the results in the development of predictive analytical models for the safety analysis of full scale nuclear plants.
3.
Water Reactor Experiments and Analyses The Water Reactor safety research program is directed at providing a capability for an independent confirmatory assessment of the safety of nuclear plants under postulated accidents. The research data and the analysis methods are iteratively applied to the assessment of hypothetical nuclear plant accidents to gain confidence that the margins of safety identified in the licensing review are well defined and quantified.
The Water Reactor safety research program is divided into five categories which are functionally identified with the Branches responsible for their planning, implementation, and management.
The five categories are:
Systems Engineering Fuel Behavior Analysis Development Metallurgy and Materials, and Environmental and Siting Detailed descriptions of the program plans for each of these categories are provided in the Branch Program Pl ns, Appendices A to E of this report. A brief description of the principal tasks and objectives of the research under each program category is given in the following sections.
3.1 Systems Engineering In licensing reviews of nuclear plants, considerable attention is given to the analysis and evaluation of emergency core cooling systems (ECCS) which are designed to cope with postulated loss of coolant accidents, including accidents that are based on the assumption of major pipe failure.
l 1
3 l One of the goals associated with the licensing review of emergency ccre cooling systems is to seek inorovement in the reliability and effectiveness of such systems through design innovations and ccns!deration of alternate ECCS concepts.
3 Other postulated accidents which are analyzed and reviewed during the licensing process can be categorized under the heading of Reactor Transients. These include postulated accidents such as loss of coolant flow without pipe failure, reactivity insertions accideats, loss of steam turbine load, and anticipated transients without scram.
s Systems Engineering research programs are directed to three areas of research corresponding to the needs for plant safety systems analysis and accident definition just identified. The three areas of research include: 1) Loss of Coolant Accidents and Emergency Core Cooling, 2) Alternate ECCS, and 3) Reactor Transients.
Study of a hypothetical loss of coolant accident to a reactor must encompass all of the themal, mechanical, chemical, and hydraulic processes which would occur in a nuclear plant if the highly pre murized coolant were suddenly depressurized and rapidly ejected from the primary system bcdndary as a result of a large pipe break. All of the components and structures of the primary circuit would be involved. A detailed study of mass and energy transfer processes affecting the reactor c. ire, pumps, steam generators, emergency cooling circuits, and the containment system is required.
Emergency core cooling water would be injected into the primary vessel during the early phase of the accident sequence, to prevcnt overheating and melting of fuel as a result of nuclear decay heat generation. The interaction of cooler emergency coolant with flowing steam, the heated walls of the primary vessel components, and the overheated core must be studied and understood to provide information needed to assess the effectiveness of the system in reducing and controlling fuel temperatures.
The research program associated with the postulated loss of coolant accident places heavy emphasis upon experiments that develop information on fundamental phenomena that would appear in the course of such an accident. For example this will include fluid behavior during rapid depres-surization (blowdown) in the absence of heat suurces such as tne decay heat from fuel. Omitting such heat sources serves as a means of reducing the numbe c ' variables and simplifies understanding of transient fluid flow and phase transfomations. Tk int rmation developed in the course of such experiments can then be used in partial dev. m :
,t r.odels that can be used for interpre-ting sinilar phenomena in more complicated expeh. t t.
h might involve integral system behavior.
The experimental programs and the development of analytical models pertinent to the hypothetical Loss of Coolant Act ' dent and Emergency Core Cooling Systems are grcuped into six separate programs:
Blowdown Heat Transfer, Reflood Heat Transfer, ECC Mixing and Bypass, LOCA Pump Behavior, LOCA Integral System Tests, and Alternate ECCS Investigations. The research underway and planned in these programs is sumarized in the following sections:
l 4_
3.1.1 Blowdown Heat Transfer Research in this area will provide data to determine hydrodynamic behavior, time to Critical Heat Flux, and transient heat transfer rates during coolant depressurization and blowdown, as influenced by variations in power, system pressure, coolant flow, and break location.
The experiment most directly related to the PWR in this area is the PWR-BDHT program at HNL.
This program will include use of a 49-rod, 12-ft long electrically heated bundle of rods, to investigate time to CHF and transient heat transfer regimes, both under blowdown and controlled pressure and heating transients. Preliminary data will become available for model development in late 1975, with final results obtained by December 1977. Complementary data will be available from the Semiscale and LOFT Programs, which are discussed later. Blowdown heat transfer measure-ments pertinent to the BWR are being performed by the General Electric Co. (joint NRC/GE/EPRI funding) through use of a 2-loop facility containing a full-scale heated bundle (49-rod, 12-ft heatedlength). The initial tests in this program have already been pe; formed providing data which became available near the end of 1974. Extension of this program to include a 64-rod configuration and ECC heat transfer is under consideration.
Supporting model development was initiated at ANL and other sites (universities, etc.) to provide correlations for transient CHF and transition and film boiling heat transfer. Preliminary correlations should be available by the end of 1975, with final verification by December 1977.
3.1.2 Reflood Heat Transfer The research in this area is intended to provide data for use in developing models that can be used to analyze reactor core bottom flooding, fluid mechanisms in quenching hot fuel, and transient heat transfer during the post-blowdown period, and at the time of initiation of reflood.
The central experiments in this area are those comprising the FLECHT program at Westinghouse (joint NRC/W/EPRI funding). The initial systems effects tests that were done to determine the effects of intact loops on core reflood and heat transfer rates were suspended in May 1974 in view of concerns as to correctness of geometric scaling and the existence of core oscillations which are believed to be atypical of a PWR.
Initial studies in the revised program will include experiments at luw reflood rates, to provide data on heat transfer and liquid carryover. Analytical studies will also be performed, to explore the relationship between experimentally observed oscillations during reflood and any that night be predicted to occur in PWRs. Need for any subsequent systems tests will be evaluated on the basis of these analyses.
3.1.3 ECC Mixing and Bypass Work in this area is required to provide models and data for the description of steam condensation during steam-water mixing, when ECC water is injected into the cold leg of a PWR system prior to end of blowdown.
Included is the experimental investigation of the effects of steam produced when the cold water strikes the hot walls in the downcomer annulus, investigation of the rate at which water reaches the lower plenum, and study of steam-water mixing in the lower plenum region. These studies will
_ _ _ _ _ improve the basis for analyses and estimation of the fraction of ECC water which bypasses the lower plenum region during blowdown and is subsequently lost through the broken loop, thereby reducing the total effectiveness of the water injected from the pressurized accumulators.
The major experimental program in this area is the Plenum Fill Experiment (PFE) being conducted by the Battelle Pacific Northwest Laboratories. This program will utilize both 1/5-scale and nearly full scale (4/5) vessels, with full 4-loop simulation, and representative PWR hydraulic flow conditions. The PFE facility will be instrumented to obtain data on steam / water mixing in the ECC injection downcomer and lower plenum regions. Initial testing is now scheduled for July 1977. The experimental program is expected to be completed by December 1978.
The steam / water mixing program at CE (joint NRC/CE/EPRI funding) was completed in January 1974.
The results show that steam plugging did not occur at the ECC injection region. Limited tests in the 1/5-scale vessel facility (one intact and one broken leg) revealed that ECC bypass should occur when the steam flow down through the core and up through the downcomer is high in rate.
Experimental and analytical investigations at CREARE are conducted to provide data to be used in calculating delay times for ECC delivery, that may be induced by hot wall effects and by oscil-latory behavior in the ECC injection region. Data obtained to date have revealed that one-dimensional theories are inadequate for cal mi xing ECC bypass if the downcomer has a large gap, and application of the data to downcomers 01 such sizes indicates delay times of 1-5 seconds.
Additional investigations underway include effects of steam up-flow, multiloop geometries, and oscillations irduced by ECC injection.
3.1.4 LOCA Pump Behavior Current pump models utilized in LOCA analyses are based on homologous single phase pump theories and assume that these scaling laws, with some modification, can be used to calculate transient, two-phase pump behavior.
Two-phase pump experiments carried out on the Semiscale pump in FY 1975, in both steady-state and transient modes, resulted in head degradation correlations, thereby providing a means to establish a modified transient pump performance model. The Semiscale pump data is currently utilized in the Regulatory EM (Evaluation Model) pump model.
Current program approach is to utilize data obtained from the EPRI-CE cooperative program. Current EPRI-CE plans call for testing a 1/5 scale CE pun in a modified CE loop the first half of CY-76.
CE's current approach is predominantly a two-phase pump performance steady state characteristics testing program, followed by a limited amount of transie.,t pump tests.
3.1.5 LOCA Integral System Tests Integral studies are required to test the use of system codes in predicting the detailed behavior of systems undergoing conditions typical of all or part of a postulated loss of coolant accident.
System behavior under partial LOCA conditions will be studied in the Semiscale and LOFT integral test facilities.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ 3.1. 5.1 Semiscale The Semiscale Mod-l facility, which was completed in the early pcrt of FY 1975, is a one-dimensional, non-nuclear representation of the themal hydraulic aspects of a pressurized water reactor facility.
The 1.6-megawatt core consists of 40 electrically heated fuel pins 5.5-ft long and 7 inches in diameter. The primary caolant system is designed with the same system elevations and the same ratio of total volume to core power as exists in LOFT. System subvolumes, e.g., inlet plenum, core region, etc., are also designed with relative volumes similar to those in LOFT.
The unbroken PWR coolant loops are simulated by a single unbroken circulating loop in the Semiscale primary system, and the postulated broken loop is simulated by a blowdown loop.
Semiscale Mod-l operation was initiated in late August 1974. The initial test sequence duplicated the non-nuclear test sequence schedule of LOFT. These tests are being followed (start in Spring of 1975) by separate effects testing in the area of blowdown and reflood heat transfer. This will complete the separate effects testing in the Semiscale size with the exception of transient pump testing. Integral tests will be conducted in the beginning of CY-1976 in which current and alternate ECC injection concepts will be investigated and system response to small breaks assessed.
A twelve-month test program to detemine the effect of core length (12-ft versus 5.5-ft) will also include an assessment of active versus passive components in the blowdown loop on integral system response. Detailed break studies will be conducted following core length studies to investigate the system response due to: small versus large break, hot versus cold leg break, commuaicative versus non-comunicative type breaks.
3.1.5.2 LOFT LOFT is a 55-MW(t) pressurized water reactor facility intended to simulate the major behavior of generic 1000 MW(e) PWRs in carefully conducted loss-of-coolant experiments (LOCE). The nuclear core is approximately 5.5-ft long and 2-ft in diameter, and contains 1300 fuel pins and four i
control assemblies of typical PWR design. The primary coolant system is designed with a similar l
primary system volume to core power ratio as exists in typical PWRs. Primary system subvolumes, e.g., cold leg, core region, and hot leg also are designed with relative volumes similar to PWRs.
The unbroken PWR coolant loops are approximately simulated by the single unbroken circulating loop in the LOFT primary system, and the postulated broke. PWR loop is simulated by the LOFT blowdown loop having passive components.
The non-nuclea-test sequence in LOFT (approximately five tests) is expected to commence in November 1975. These tests will provide an information base to be used in testing the predictive capability of computer codes and to assist in the understanding of the more complex nuclear tests.
The follow-on nuclear test sequence in LOFT will simulate the full effects of a postulated loss of coolant accident to a PWR, progressing from initial low power level to initial high power level.
The objective of the nuclear tests, which should comence in late 1976, is to provide data to be used in testing analysis capability of the following phenomena important to a LOCA.
Effect of cold leg to hot leg ter:perature differential on subcooled and saturated blowdowr..
Effect of primary-to-secondary heat transfer on blowdown and reflood.
~~ '
g
..y r
A
- '..~
,y,
-e ;,
Effect on primary coolant pump performance of increased hot leg temperature (relative to 7.-[- l., j ~._?
isothermal blowdowns).
(. j h i it ;%. -
Effect of decay heat on blowdown and reflood.
3 qL..;
(l*.. _ ;[
I The integrated effects of decay heat, primary pump performance, metal-to-water heat transfer, p.;V-?.,f.{ ;
and steam generator heat transfer on fuel integrity.
- h.. %.'. E A.
7y ;. -
.m 3 7:,tf@.
Effects of alternate E',C injection locations.
3
.E M.. w. :3 " 7 3.1.6 Alternate ECCS Investigations In the ECCS Criterial the Comission "....(urged) the pressurized water reactor manufacturers to (f;i j ~ }. r.
seek out design changes that would overcome steam binding." The Advisory Comittee on Reactor f.T'S..'..i N..
u-Safeguards in its September 10, 1973 letter to the Comission on the same subject stated that 9{..[i!
".... reactors filing for construction permits after January 7,1972, should have significantly k{(f4.-
fq'N,e..l s?
improved ECCS capability." The Regulatory Staff reiterated this position on October 18, 1973, in REG:RSR 108 in which it was stated that " implementation of such (alternate ECCS) studies is i;l].'((J
~-
considered by Regulatory to be an A priority task." In response to the above requests, the N:
'p;;!
Division of Reactor Safety Research has initiated studies of alternate ECCS concepts to be (ll, 'i',il l conducted in both the Semiscale and LOFT integral test facilities.
h.-} 'ig
- c.
g N. ' -(
The major portion of this test program will be conducted with the Semiscale facility beginning in
- y., 7 O *s (.
FY 1976 and concluding in FY 1977. Tentatively, concepts to be tested include lower plenum
[ -[ f injection, upper plenum injection, simultaneous upper and lower plenum injection and combined f,.Q1N(
p [.m-:pg..
lower core rv.gion and lower plenum injection.
e;r.h
,-e g: y The ordering of ccncepts to be tested will be established after review of analyses. Those designs
-- 'f: f L which are tested in Semiscale and which demonstrate an improved benefit will then be included in g a s;,
the LOFT facility for testing during the low and high power nuclear tests, m>..
$f,k.p;., -
e.
V Results from recent Semiscale isothermal tests have demonstrated a faster refill time by injection g p.g{.,]
into the lower plenum than by injection into the cold leg. Current plans call for similar testing
/
during the LOFT nonnuclear test series.
'.,,g g,..17.,pg-.
J ?*$ '.
3.1.7 Reactor Transients h%fi.[
The Regulatory Staff has recently (September 1973) documented its licensing position on Reactor
. x.
Transients in WASH 1270: Anticipated Transients Without Scram (AfwS) for Water Cooled Power h.-
Reactors. This document identifies a series of postulated transients which may occur and states that "an analysis should be made of the consequences of (these) anticipated plant transients in M % %.(
%S
- h. - Q the event of a postulated failure to scram."
f l!.';;
- .. p
. ? C '* ':.
f..
[
C,i a 110 CFR Part 50, Licensing of Production and Utilization Facilities, " Acceptance Criteria for o'd CN Emergency Core Cooling Systems for Light Water Cooled Power Reactors " December 28, 1973.
_ The basic approach in treating reactor transients is to provide a sufficient data base to assess the ability of the code to predict the combined thermal-hydraulic and neutronic behavior attendant to a particular off-normal transient. These transients involve the nuclear characteristics of the facility, i.e., feedback conditions such as void coefficient, Doppler coefficient, etc. For this reason, tests of potential off-normal or accident conditions outlined in WASH 1270 can best be performed in the LOFT nuclear facility. Again, complete understanding and resolution of the atypicalities in LOFT will be resolved prior to the conduct of ATW5 tests. Tests under consid-eration in the LOFT facility are:
Loss of Feedwater - These tests involve the loss of one or all of the feedwater pumps.
Luss of Primary Flow - These test transients include the loss of one of two pumps.
Primary System Depressurization - This transient covers the opening of the largest single safety relief valve in the primary system.
Small line break - These transients cover the failure of an instrument, drain, or sampling line connected to the primary system.
Rod withdrawal - Transients in this category include control rod withdrawal when the reactor is in the hot critical condition or at full power.
Such tests are being considered for performance during the high power nuclear series in LOFT.
3.2 Fuel Behavior The purpose of the fuel behavior program is to provide a detailed understanding of the response of nuclear fuel assemblies to hypothetical off-normal or accident conditions. This information is then used to develop physical models which are incorporated into fuel analysis codes. The fuel codes are verified through integrated in-pile tests. The understanding of the release of fission products from the damaged fuel and their transport to the containment area is also being studied.
The program is separated into eight areas:
3.2.1 Review of Material Property Corre'ations Physical models and materials properties of the fuel rods and their components are reviewed and statistical estimates of their uncertainty are made. Appropriate correlations are recommended for incorporation into the fuel behavior codes. New information will be compared with the existing correlations and improvements will be incorporated until evidence is derived from code verification that further development is not necessary.
3.2.2 Cladding Properties Programs in this group are expected to improve understanding and reduce uncertair.R. in the rate of Zircaluy oxidation in steam and in the degree of embrittlement of the cladding. An embrittle-ment criterion will be derived. The deformation of bundles of rods during burst testing will be measured and a failure correlation will be produced. Experiments will be conducted on cladding creepdown and collapse and clad strain-to-failure to improve the steady state mechanical properties models. In-pile experiments are included as part of each of the above programs.
_ ___ 3.2.3 Fuel Pellet properties Fission gas release will be measured under steady-state and transient conditions in a reactor and af ter irradiation to approximately 40,000 MWD /MTU. The infomation will be used to improve models for gas release, gap conductance and internal gas pressure. Pellet growth, cracking and expansion data will be taken to improve understanding of fuel rod gap closure and pellet-cladding mechanical interaction. A test assembly was installed in the Halden reactor in February 1975.
3.2.4 Fuel Rod Properties Experiments conducted under this category have as their objective the improvement and verifica-tion of models for calculating stored heat to a fuel rod. Addition:lly, the effect of pellet relocation on transient gas flow along the length of a fuel rod will be measured in-pile. The Ross-Stoute contact heat conductance data will be extended. In-pile experiments will include short term parametric studies and instrumented irradiation of UO2 and Pu recycle compositions to s?5,000 MWD /MTV. Assemblies were installed in Halden in February and May 1975.
3.2.5 Integrated Tests of Fuel Models Under Accident Conditions The Power Burst Facility, which is an oxide-fueled, water-cooled and-moderated, open tank reactor with a pressurized loop, will be used to produce experimental conditions in which fuel rods can be tested to failure in a nuclear environment.
These tests will pennit detailed examination of the mechanisms of failure and of possible propaga-tion of failure from one rod to another. The PBF is expected to be the principal facility to verify the transient fuel behavior code (FRAP-T) being developed. In FY 74 the preliminary Over 60 planning and documentation was completed for the first three years of PBF operation.
experiments were identified. Test types include power coolant mismatch, flow blockage, loss of coolant, reactivity insertion and gap conductance. Plutonium-containing fuel rods are included among the test fuel rods.
3.2.6 Reactor Decay Heat The heat generated by the fission products in the first 100 seconds following reactor shutdown is an important source term in LOCA modeling. New data will be obtained and analyzed to reduce the statistical uncertainty, Calorimetric and spectroscopic techniques are being utilized. Decay 239 u will be measured. The ENOF-IV nuclear data file serves as the base for the of 235U and P
analytical effort.
3.2.7 Fission Product Transport Programs conducted in this category are designed to obtain detailed information on the chemical and physical states of fission products released from fuel under conditions of normal operation, terminated and nonterminated hypothetical LOCA, and spent fuel handling and transportation The information will be used to improve the existing fission product transport nodels accidents.
and develop new ones where indicated.
The behavior of semivolatile fission products, such as iodine and cesium, released into proto-typical accident environments will be emphasized in experiments applicable to terminated LOCA and less severe accidents. A scaled facility to verify the ar,alytical models will be built. New information on the release of fission products into postulated core meltdown environments from
_ prototypical molten core components will be made available. In addition to iodine and cesium, the release of less volatile fission and activation products and their characteristic behavior as aerosols will be emphasized.
3.2.8 Core Meltdown There is a small but finite probability for failure of engineered safety features which are provided to prevent melting of the core in the event of a postulated LOCA or severe transient.
The postulated consequences initiated by such an event are considered potentially to impact public safety and merit an improved physical understanding of the phenomena involved.
The Reactor Safety Study (WASH-1400) presented scenarios of physical occurrences postulated to occur during core meltdown accidents. Based on a study of the state-of-the-art of experimental infonnation applicable to the analyses of such accidents, early research topics will include chemical and physical interactions involving the molten core, behavior of fission products, steam explosions, and heat transfer in molten pools. The knowledge gained from these programs will be integrated with the results of related core meltdown research sponsored by the Federal Republic of Germany and coordinated through a bilateral infonnation exchange agreement. The long range goal of such a task is the development of analytical codes to describe realistically the conse-quences associated with hypothetical core meltdown accidents. These studies will provide insight to further research that might be required for the design of additional engineered safety features or for the implementation of countenneasures to cope with such hypothetical accidents should they occur.
3.3 Analysis Development Experimental programs in Water Reactor Safety Research are directed toward improving ability to understand specific phenomena and the results of tests, with the goal of applying this under-standing to the interpretation of "similar" phenomena that might take place if a reactor accident were to occur.
This is done through the development and use of complex digital computer codes. These codes are used to compute the time-dependence of important parameters during total or partial accident sequences, and equivalent effects whicn might occur if a full scale reactor accident ever took place. lhe credibility of use of such codes for reactor safety assessment is based on the success achieved in using the codes to predict results of various separate effects and integral experi-ments conducted in the Reactor Safety Research Program. Code development and application, there-fore, represents one of the Research Division's highest priority programs.
3.3.1 System Analysis Codes Top priority is given to the improvement of the present intermediate level system code, RELAP 4.
This code has two versions. The " Evaluation Nodel" (EM) version is used by the Regulatory staff in its licensing activities. This version provides conservative analysis through incorporation of the Commission's Acceptance Criteria. The "Best Estimate" (BE) version, on the other hand, incorporates realistic (not necessarily conservative) mathematical descriptions of the system.
The "best estimate" version of the code will be especially useful for verification by means of comparisons between code predictions and test data.
Development of an Advanced System Analysis code represents an important facet of the code development program. The approach is to employ parallel path development in which a number of
_ _ _ _ - _ - _ - _ different models and numerical techniques are studied to obtain a timely resolution of this very complex problem. The advanced codes currently under development for LOCA ECCS analysis are grouped into two categories: (a) the Systems (or loop) Codes, and (b) the Component Codes. The former category is comprised of SLOOP (ANC). THOR (BNL), and TRAC (LASL) codes while ANC's SCORE and SPLEN, LASL's KACHINA, and PNL's COBRA-4 belong to the component code category. All advanced codes will account for thermal non-equilibrium and for unequal phase velocities. All component codes will employ multi-dimensional treatment. LASL's TRAC code will employ mixed (1-D, 2-D and/or 3-D) geometry in the System (loop) Code.
As soon as the best modeling and calculational procedure can be established through parallel path development, a single program effort will be effected, involving interlaboratory specialists.
The reactor core noding will allow a detailed view of the hot assembly and a coarser view of the remainder of the core with allowance for the inclusion of secondary flow patterns. The FRAP (Fuel Rod Analysis Package) code will be coupled with the core code to account for energy transport within the fuel rod and its mechanical deformation. The advanced code will be of the "best estimate" type, and its primary purposes are to (a) improve the capability to define the degree of conservatism built into the present Evaluation Model (used for licensing purposes), (b) help to define possible reductions in excessive conservatisms that may be present, and (c) update the Evaluation Model to reflect more advanced analysis techniques.
3.3.2 Containment Analysis Codes A development effort is planned to improve the capability to model and analyze conditions that would exist during the early and later stages of a LOCA in commonly used containment buildings.
For containment transients during the early stage of a hypothetical LOCA, a multi-dimensional view will be taken of the compartment in which the rupture is located. Emphasis will be placed on pressure loads and on structural loads caused by the impinging flashing jet. Other compart-ments will be treated in the lumped parameter mode. The long term containment transients will be dealt with by the lumped parameter codes, applicable to existing containment concepts.
3.3.3 Fuel Behavior Codes These codes are developed to describe:
physical changes in the reactor fuel over the period starting with fuel loading and ending with initiation of a postulated accident, via the FRAP-S code.
physical changes in fuel rods during a transient caused by a postulated accident (LOCA, anticipated transient without scram (ATWS) or reactivity initiated accident (RIA)), via the FRAP-T code.
3.3.4 Code Verification Activities Most Peactor Safety Research experimental activities are directed toward obtaining test data useful for (a) developnent of analytical models, and (b) testing and improving where desirable the analyses applied for evaluation of reactor safety. This process of using experimental data to test and improve the predictions made using computer codes is called corputer code verifica-tion. ine first important activity in verification is to define the types of measurerents required for testing, and to define the accuracy required for these measurerents.
__ The second activity is to perform comparisons between test data and code predictions. The third activity is to perform code sensitivity analyses. The last two activities lead to estimates of the code prediction uncertainty bands, and thereby, to quantification of the conservatism in the analyses. Table 1 sunmarizes the code verification activities.
TABLE 1 CODE VERIFICATION 1.
MEASUREMENT REQUIREMENTS & ACCURACY 2.
COMPARISONS: CODE RESULTS VS. TEST DATA FOR MODELS, SUBC0 DES, CODES 3.
SENSITIVITY ANALYSES ON MODELS, GEOMETRICAL REPRESENTATION 4.
METHODS DEVELOPMENT FOR UNCERTAINTY (ERROR PROPAGATION) ANALYSIS 5.
CODE RESULTS UNCERTAINTY (ERROR PROPAGATION) ANALYSES a.
OF SYSTEM COMPONENTS b.
OF INTEGRAL SYSTEMS 3.4 Metallurgy and Materials One of the major elements of the Water Reactor Safety research program encompasses studies of metallurgy and materials as they relate to improvements in primary coolant system integrity, since the failure of the primary coolant system would cause a loss of the normal coolant for reactor operation and would require subsequent dependence on emergency core cooling systems.
The criteria developed by the Regulatory staff of NRC for licensing of nuclear power plants require that nuclear plant primary systems (pressure vessels, piping, pumps, heat exchangers) be designed and fabricated in strict conformance to prescribed nuclear codes and standards to provide a high degree of assurance of primary system integrity over the operating life of the plant. The intact primary system will ensure that there will be no hazard to plant personnel or to the public from postulated reactor accidents which could result in the release of fission products from the fuel elements.
The purpose in undertaking safety research in the area of Primary System Integrity is to generate a more confident basis for criteria and analytical procedures for design, fabrication, and opera-tion of the pressure vessel, piping, and associated components of the primary system pressure boundary of water reactors, so as to improve identification of hypothetical failure modes and l
estimation of the probability of failure, and to establish ways by which failure probability can be reduced. The pressure boundary of current water reactors consists of a steel pressure vessel cf thickness approaching 12 inches in some cases, and primary piping of thickness as much as 4 inches. The typical materials for construction of vessels, including A533B plate and A508 Class 2 forgings, and for piping, including A106-B and 304 stainless steel, have been studied extensively to develop information on trends for mechanical property behavior under appropriate test conditions of temperature, stress, neutron irradiation, and reactor environment. It has been necessary to conduct these studies with laboratory-scale test specimens because irradiation of massive, thick reactor components would be either prohibitively expensive or technically unfeasible. Thus behavior of the full-section-thickness materials and components must be predictable from criteria and trends developed largely with small-scale laboratory test specimens. Despite knowledge already attained concerning properties of primary system component materials, improvements in information are still sought to round out the basis for judgments affecting continuing reactor i
safety.
_ _ _. The present areas of particular interest to water reactor safety include:
Elastic Plastic Criterion Fatigue Crack Growth Inspection and Flaw Detection 3.4.1 Elastic Plastic Criterion The overall objectives of this work are to develop and validate analytical models for fracture toughness in the elastic-plastic and fully plastic regimes, including fracture-analysis procedures and design criteria for prediction of stress levels and flaw sizes required for crack initiation and for the subsequent propagation and/or arrest in thick section water reactor pressure vessels and primary piping. This also includes development of the mechanics of loading under static, dynamic and thennal transient stresses. The effects of neutron irradiation upon the fracture toughness of pressure vessel steels are studied in various programs to detennine the basic causes of radiation-induced degradation of shelf level toughness and notch ductility and to understand recovery of these properties through postirradiation annealing.
Experimental phases of this work include the HNL program of Intennediate Test Vessel tests under a variety of temperature and flaw conditions; this will be completed in FY 75. Following the ITV program, is a new study to evaluate the effects of pneumatic loading in thick-section pressure vessels; this program will lead to tests in FY 76 and preliminary results in FY 77. The effects of neutron irradiation on reactor vessel steels will be studied by developing a KIR curve for irradiated thick-section specimens of plate, forging and welds, and by correlations of smaller specimen results to large specimen results; this work is underway and will be largely completed during FY 77-78.
The initial K results, and those of the following program on crack arrest, will be of great IR value to the test program to study effects of thermal shock on pressure vessels. This program should result in the availability of engineering test data on steel cylinder specimens in the latter part of FY 76. The crack arrest program alluded to above will develop the theory of crack arrest and establish the invariant nature of an arrest parameter using a variety of different test specimen configurations and materials, with additional verifications from static and dynamic computer cnalyses, and from photoelastic studies. In a parallel study with other contractors under the Electric Power Research Institute (EFRI), it is expected that the crack arrest theory will be validated in late FY 76, and similarly, that test specimens and methods will be validated shortly thereaf ter 50 that an unirradiated and irradiated materials data base can be built up by Mid-FY 78. The effects of neutron irradiation on shelf level toughness will be investigated and energy-level limits for fully ductile fracture will be established by Mid FY 77. A data base of quantitative fracture toughness resulting from postirradiation annealing of RPV steels will be available by Mid-FY 79 for development of code rules and REG guides on annealing of reactor vessels.
3.4.2 Fatigue Crack Growth The objective of this task is to establish the magnitude and characteristics of crack initiation and propagation from cyclic fatigue of water reactor vessels and piping, using plate, forging, piping, welds and weld HAZ in unirradiated and irradiated materials under BWR and PWR coolant environments. This knowledge should pemit improvements in the code rules for fatigue design for unirradiated materials and fomulation of code rules for materials following in-service irradiation.
The experimental programs in cyclic fatigue include environmental exposure in the Dresden-I BWR rear. tor coolant loop under the GE Primary Coolant Rupture study of numerous piping and vessel steel specimens for the study of crack initiation, crack growth under static loads, and crack growth under dynamic loading; these studies will be completed in Mid-FY 77 and will be forwarded to the appropriate ASME code comittees imediately following. The increased rate of crack growth resulting from slow cycling will be studied extensively in both irradiated and unirradiated materials at rates of 0.1 cpm in PWR coolant water at 550*F. Critical experiments to evaluate the influence of radiation will be completed in FY 75, and the preliminary results for a variety of material forms will be available by late FY 77. Additional studies will qualify the effects of strain rate, (especially in the low strain range) tension hold-time (especially very long time) mean K (at small AK amplitudes) and the scatterband resulting from heat to heat variations for different product forms.
3.4.3 Inspection and Flaw Detection The first objective of this task is to upgrade pre-and in-service ultrasonic inspection capabilities to be more compatible with the stringent allowable flaw size requirements of ASME Code Section XI (Summer 1973 Addenda) for In-Service Inspection of Reactor Vessel Components. Next, additional pre-service inspection capability is required for field-identification of the susceptibility of welds to intergranular stress corrosion cracking (IGSCC). Finally, the use of acoustic emission must be qualified and incorporated into the code for flaw detection during welding, and for detection and flaw-growth monitoring during regular operation of nuclear components.
The experimental program for improvements in ultrasonic testing capability will center on intense developmert of all aspects of information resulting from a pulse-echo test including phase, frequency, anplitude and search unit position; interfacing of this data production with a minicomputer is also to be accomplished so as to facilitate data processing, storage and retrieval to permit more meaningful and accurate compariso.3 of previous inspection data from new inspections.
Feasibility will be demonstrated in FY 76, and a 3D system made operat' 9al by FY 79. The development of a field test for determining the susceptibility of welds to intergranular stress corrosion cracking must be preceded by establishment of the correlations between furnace or processing treatmen:s and susceptibility to IGSCC, with measurements of the electrochemical j
potential of materials at these same steps being taken to formulate the basis for the field test.
The goal is to conduct this field test by the end of FY 77. The monitoring of nuclear welding by acoustic emission will be proven feasible in FY 75 for piping and in FY 76 for reactor vessels.
A "no hands" system will be built and proven under shop welding conditions by FY 77. The use of acoustic emission will also be studied for on-line monitoring of operating reactors in FY 76-77.
To be proven is that signals can be recorded and understood amid the background noise of an operating system. Signals will also be recorded from fracture toughness and fatigue specimens of krown properties, and the test signals will be integrated with the operating reactor system to produce a "no hands" acoustic emission monitor for continuous flaw detection and growth monitoring, including an alarn function to indicate approaching criticality of a flaw. The goal of these studies is to develop a proven system by FY 78 and details forwarded to the appropriate ASME cofe comittee and the REG staf f for application to operating reactors on a routine basis.
__ _____ 3.5 Environmental and Siting To minimize the potential for external and environmental influences which might cause a reactor accident or affect the course and consequences of an accident, nuclear plants are sited at locations that have been carefully studied in the light of need for protection against earthquakes, floods, tornadoes, and other natural phenomena. The Environmental and Siting safety research program undertakes work in this area.
The main purposes of the Environmental and Siting safety research program are to provide information to help ensure the safety of plants as affected by the environmental conditions at specific sites, and to ensure that future research needs for site assessment in all regions can be met on a timely basis. Accomplishing these purposes involves the understanding and prediction of the effects of severe natural phenomena such as earthquakes, tornadoes, and floods. Also involved is the development of engineering methods to help ensure the adequacy of plant structures, systems, and components to withstand imposed environmental loads safely.
Environmental and siting safety requirements are closely related to the licensing process. A further objective of the research and development program in this area is to help speed up the nuclear facility licensing process, both through development of improved site evaluation methods and by the collection of baseline information.
The program of Environmental and Siting research includes regional studies of earthquake, tornado, and flood potential and assessment of alternative concepts of nuclear facility siting (offshore, underground, and floating platform concepts). Engineering and design studies will be supported under this program to develop additional information and criteria to ensure the resistance of plant structures, systems, and components to the potential loads of severe natural phenomena (earthquakes, floods, tornadoes).
The new work being initiated in Environmental and Siting research includes meteorologic and seismic studies, evaluation of alternative siting concepts, and engineering-design studies of structures with improved resistance to potential damage from natural phenomena. The development of such information and its application to siting criteria and design are needed to reduce delays in preconstruction licensing reviews and thus to contribute to alleviation of future energy crises.
3.6 Other Studies In Water Reactor Safety Research 3.6.1 Reactor Safety Study The Reactor Safety Study is being conducted to estimate the public risks that could be involved in potential accidents in comercial nuclear power plants of the type now in use in the U.S.
The objective of the study is to make a realistic estimate of these risks and to compare them with nonnuclear risks that already exist in society. The study is also aimed at the development of methodology needed to evalute the probability and consequences of potentially severe accidents, the collection of data needed for the analysis, and a detennination of estimates of the probabilities and consequences of potential accidents.
A draf t report, MASH-1400, An Acecement of /ccident hieko in U.S. Gmercial Mlear Tcuer TZants, was issued in August 1974 in order to obtain cocinents from the public sector. The report
_ is being revised in view of the comments received and will be published in final form in 1975.
Extensive revisions are being made in the consequence model to eliminate errors and to improve the modeling techniques used.
4.
Fast Reactor Safety Assessment The objectives of the programs under this activity are to establish an independent capability for safety assessment of early fast reactors by the Regulatory staff of NRC, and to establish a program leading to the timely resolution of any safety problems that might be associated with advanced fuels and systems. This differs from the program of the Reactor Research and Development Division of ERDA which has the objective of establishing a sound design basis for the introduction of commercial plants.
The program is defined in large part on the basis of Regulatory needs, expressed in both fonnal documents and staff contacts, and in response to recommendations of the Advisory Comittee on Reactor Safeguards.
The program is organized into seven generic categories:
Fission Product and Fuel Release and Transport Analytic Methods Fuel Interactions Safety Test Facility Studies Criticality Experiments Gas Cooled Fast Breeder Reactor Plant Systems The logic process that is generally used to develop the programs in these categories is illustrated in Figure 1.
In addition to the generic programs which are being carried out at national and industrial laboratories, a number of programs have been initiated at universities. A fairly broad-based program dealing with fuel interactions, post accident heat removal, and fuel failure dynamics as well as a scoping study of the safety problems associated with advanced fuels has been initiated at UCLA. It is our intent that this be a long range program not only to provide guidance for more complex and technically detailed studies at national research institutions, but also to aid in the educational program. Studies are starting at the University of New Mexico on modes of fuel fragmentation and detailed modeling of fuel-coolant interactions. A program of analysis on gas cooled fast reactor problems has been started at Northwestern University.
This analysis is also applicable to the study of clad motion following a loss of flow in an LMFBR.
A brief description of the principal tasks and objectives under each program category follows:
1l i
lJlI 8
0, l_
Ig I!llglI!llIlg!IlllIIIgll F I p D
E S I S I
R A EB V
7 Y
J_
llg Ig D
E S E
T T R H N
D C L A TM p
l I U P
IWx q
E DS M
T E
EE O
A g\\
RR C
R P
G v
E TN S
I S
3T
/
[
S I
f 2S S
NS NE E
OY I
F IF S
OT C
IL LS TA AY O S "'
CO T
T O
CN N L O E I A R T '*
SR E
RP A
P H
A C
_1 R
_ E Il g
II1III!ll!glIIlllIIlI 1l R
A U
ig Il1lgilllIlgII;lgIlIgi l
J_ G E
Y m'
I S
F E
N_
S R
D L
I O C
E D
SH O
Y A T B E M
M Y
1 G
1L t
O F
m A
NO ON S
\\
I E
\\
TM I
O S
C Y
E N
S SN CA R
T G Y I T P I E
S
& O xS H
AA E
EE A
BD H
P D
N A
T Y
A 4
[
N S
O T
A M"
L^
AS LTI S uCs L
N I
T' CT EC T OUn C
G L'
p P^
mP iD T
P I T U A S
T SA E*
x X
x FR E i
E sox E
R*
P*
ED D
E RE E
rR
\\
C T
P I
N L
M r
Il Ill
=
L
-t llIll Il1IIII!3 Il3IIl1lIIII1 l
_ _ _ - _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.. _ _. 4.1 Fission Product and Fuel Release and Transport Fission product and fuel release and transport studies were initiated at Oak Ridge (HNL), and Battelle Memorial Institute-Columbus (BCL) with the objective of developing an experimental base for defining the source term for aerosols generated by an HCDA (Hypothetical Core Disruptive Accident), which may be transported into large containments, and providing an analytic method capable of extrapolating sources and aerosol' characteristics to containment systems typical of large LMFBRs. The program overview is shown in Figure 2.
This project is regarded as a cornerstone of assessment of doses from any accident involving a release of radioactive material to containment.
4.1.1 Analytical Studies of Aerosol Behavior Prediction for Fast Reactor Safety The objective of this program at BCL is to develop a theoretically sound model for aerosol behavior, to verify its accuracy, and to demonstrate its conservative nature by comparison with experimental data. The major effort since program initiation in February 1974 has been to compile, compare and evaluate available aerosol data applicable to analysis of fast reactor safety and to improve the currently available analysis models for aerosol behavior. Ihe AI code HAA-3 was selected for detailed study, and improvements were considered in the areas of agglomerate volume, density, themophoretic deposition, particulate collision efficiencies and size distribution. In addition.
BCL has been providing technical assistance to the Division of Reactor Safety Research and HNL by perfoming parametric sensitivity analyses and some pre-analysis of experiments.
An operational version of an aerosol behavior code, HAARM-1, has been developed for NRC use and a topical report was published (1975). The program to improve models of aerosol behavior and ta provide technical assistance to Reactor Safety Research is continuing in FY 75.
4.1.2 Transient Release from LMFBR Fuel The objective of this program at HNL is to investigate the radionuclide release from LMFBR fuels under power transients in ex-reactor experiments that simulate postulated accident conditions.
This program was initiated in February 1974.
In this program it is planned to measure the amounts of fuel released in aerosol fom and to detemine the chemical and physical state of the released radionuclides as a function of ambient conditions, rate of application of power, starting temperature, burnup, fission gas plenum pressure, presence or absence of sodium, temperature and pressure of cover gas, and size and history of vapor bubbles. Prior to experiments in sodium, initial experiments will be performed in the dry condition to test methods and to provide data for establishing upper limits of release of activity to containment during hypothetical accidents. Analytical models for prediction of scale effects are being developed at BCL in conjunction with this program.
In FY 74 a detailed plan for this program was prepared and facility modifications largely completed.
The program for FY 75 includes initiation of cold proof tests of V0 aerosol release in the absence 2
of sodium and the initial dry (absence of sodium) hot cell tests of U0 aerosol release. The pre-2 paration for the HCDA fuel vaporization tests without sodium (dej) are to be completed at the end of FY 75.
In addition, the design and assembly of the HCDA fuel vaporization tests with sodium will be completed at that time.
l m
GE NG E
I A E S D
TO K
L I
~
L 3
1
~
Y YT TI P
I I
V V
T UM T
I C
N E C
ADE AT ADE OEC ES OEC iI I KR LY D AU CS I KR D AU AEO AEO RLS RLS C
_~
T
, 1 I l g l
l l
I I l
i R
O P
J S
R
&T T
N L O A
OI R
R R
EE RAA RA 1
AB N
EER ER A
ALT AT L
E S
T A
2
, i l
i l
l i ! I E
L r
-l L
E L
'P N
C R
R X^
qI TOI U
E g
Il L
H O
I R
E YSV G
_^ LOA E
T I
C L
R ARH L
F E
NEE C
LO R
AAB B
OP S
V ON S
f 1
- RA
_ r N ER E
T I AT V
NE E I
I S T
S A TE M
N CS C
NE UEO UA A
RE TR L
t SC SOFPR D
A OR R
=
~ ll I
M&S Y
A U L A L
D DL R I
i C
I H
OPP SSS r
YRM AE MT I S RY PS
___.. _ _ _ 4.1.3 Experimental Studies of Aerosol Transport Behavior Associated with LHFBR Low Probability Accidents The objective of this program at HNL is to provide data from intermediate scale aerosol tests to verify existing models for aerosol transport and to investigate the potential for resuspension of particles. The experiments will be carried out in the NSPP (Nuclear Safety Pilot Plant) facility.
This prog-am was initiated in January 1975. Studies will concentrate on the behavior of aerosols containing radionuclides and sodium. The prime questions in these tests are of aerosol agglomeration and scale-up with respect to vessel height and free volume.
4.2 Analytical Methods Analytical programs are now in progress at three laboratories: Argonne, Los Alamos, and Brook-haven. The Argonne work includes the important task of technical coordination in which a senior investigator with a supoorting staff is responsible for the coordination of technical efforts in the analytical area, with a view to securing agreement on the parameters and processes to be represented in models, ranges of values to be investigated, and generally setting the technical sct,e of the work.
The Los Alamos work is directed at developing a modularized code (SIMMER) for the analysis of extended motions of the core, such as might follow a hypothetical accident in which the core is no longer coolable by nomal processes but not so widely dispersed that it could not again achieve criticality.
The Brookhaven work is in direct support of the Regulatory assistance programs carried on at Brookhaven and is centering on the development of system response codes and codes for the analysis of pipe breaks.
In detail these programs include the following:
M. 2.1 Technical Coordination of RSR Safety Analysis Activities This program at ANL provides coordination for research on fast reactor safety analysis supported by RSR, and includes the following tasks:
Task 1.
Study of Basic Problems in Accident Analysis This task focuses on the adequacy of the mathematical models of physical phcnomena relevant to fast reactor safety and also an the consequences of uncertainty in the parameters used in such models, including uncertainties in reactivity coefficients. It will study in turn the earlier stages of an HCDA and later the final disassembly phase of HCDA. Also, studies will be made of milder accidents, to assist in probabilistic studies.
Part of this task is a review of possible initiating conditions of reactor accidents, to assure that conditions of possible importance are not overlooked, to analyze the consequences of a variety of possible initiating conditions, and to detemine the sensitivity of consequences to assumed parameters. This task will also include consideration of the varying safety aspects of different reactor designs, sizes, fuels and operating conditions.
_ _ _ - _ _. The sensitivity of calculated accident histories and estimated consequences to choices of models
-and key parameters will be analyzed. The ability of models to predict the results of available experiments will be reviewed. If the results are found to be too sensitive to the choices of parameters which cannot be specified with any certainty, the need for further experimentation and/or model development will be evident.
Studying the effect of varying the models is likely to involve a certain amount of computer programing. This will be kept within reasonable limits, however, and it is assumed that any extensive programing will be handled elsewhere. If appropriate, simplified codes will be developed to study the effects of some model changes.
This subtask also includes any programing or other activity required to implement computational capability on computers available to NRC personnel, t
Task 2.
Coordination of RSR Safety Analysis Research The coordinator establishes mutual agreements among the contractors on recommendations for interim and end results of the studies and on the milestones and schedules of the various programs to meet program requirements. The coordinator works with the code developers to set requirements for models to be used, options to be provided, for equations to be solved in implementing these models and for ranges of parameters to be included. The coordinator evaluates proposals received by RSR in the field of safety analysis research from industries and universities as well as from national laboratories and submits the evaluation results and recomendations to RSR.
Task 3.
Evaluation of Progress in Safety Research Important new developments in safety research are reported and evaluated. New insight that has been gained as a result of ongoing work is reported from time to time. Critiques are made of the adequacy of the available analytical methods to evaluate the safety characteristics of plants currently under construction or design and of larger future plant. Recomendations will be made for any additional capability that appears to be needed beyond that under development in ongoing programs.
4.2.2 Extended Core Motion and Recriticality The objective of this program at LASL is to produce a calculational capability for describing the evolution of hypothetical accidents involving the melting of cores of liquid metal fast breeder reactors with subsequent large scale motions of the molten core. Orderly development of this calculational capability requires that the development of initial models and computer programs be consistent with existing comouter capability yet include the most important features of the accident being described. TW program,will be developed with sufficient flexibility 50 that additional effects can be included when they are deemed important.
The scope of this program includes development of models and computer programs in the areas of neutronics, material dynamics and heat transfer, and equations of state.
Included are: the accurate treatment of extended motions of core materials, simulation of the significant material interactions, and prediction of total energy release and partition. The output of the calculational capability is to be structured so as to facilitate the design of diagnostic experiments.
_ Task 1.
Development of Extended Core Motion Code A comprehensive code with the acronym SIMMER (SN, implicit, Multi-component, Multi-phase. Eulerian, Recriticality) is being constructed with a projected effort of 30 man years.
Work in neutronics will include a computer program for solving the two-dimensional R-Z geometry time-dependent neutron transport equation. This program will utilize methods developed for the one-dimensional time-dependent TIMEX program. Arbitrary numbers of groups of neutrons and delayed neutron precursors will be allowed. Doppler feedback will be treated by use of effective absorp-tion cross sections. Initial methods will be tested extensively for problems in which no material motion is allowed. Research will be undertaken to refine and improve these methods as necessary.
Finally, this computer program will be coupled with the material dynamics and heat transfer algorithms developed as indicated below. Cross sections for each mesh cell will be developed by multiplying a set of reference microscopic cross sections by densities that are output from the fluid dynamics computations and then sununing over all reactor materials.
The material dynamics and heat transfer algorithms will describe the motion of multi-component vapor (fuel, sodium and steel) moving through globules of molten or solid fuel and globules rep-resenting molten or solid steel in the form of superstructure or confining walls. The following phenomena will be included in the initial model:
Interpenetration and interdiffusion of the various materials, Phase changes from the molten globules to the vapor state of fuel or steel, Heat transfer amon1 the various materials and between the materials and their surrounding container walls, Drag effects of the vapor on the globules, Anisotropic forces representing the effects of subassembly wall strength, which may act to retard the radial motion of the fuel, Fuel-coolint interaction when globules enter prescribed regions, and l
Coalescerce of the droplets into large ones, and the entrapment of vapor because of collective e f fec t? 'n the drag (i.e., marked decrease in permeability).
These ef fects will be included in the initial computer program with various degrees of sophistica-tion, in subsequent studies, more refined treatments of the above processes will replace the initial models, and the code will be continually updated to include the latest developments.
The Implicit Continuous-fluid Eulerian (ICE) solution procedure will be used to solve the fully two-dimensional equations of motion that govern the dynamics. This implicit formulation of the solution procedure will permit the investination of a wide range of theoretical core meltdown accidents from those that involve low speed or incompressible flow to those in which the speeds of materials are comparable to or greater than the speed of sound in the medium.
,- -. ~
_ _. ~ - -
1,
Theoretical techniques developed for prediction of equations of state in the mixed-phase and
-imperfect gas regimer will be applied to sodium, steel, and the oxides of plutonium and uranium.
Complete tabular equations of state will be developed for these materials. Concurrently, the literature will be reviewed and modifications to the initial models will be devised. These modt-fications will produce better agreement with experimental data where available and hence better j
predictions where the data are not available. This program will produce accurate tabular equations of state for the materials of interest. These tabular equations of state will then be used with the material dynamics algorithms.
Initial results of the work described above will be monitored to ensure applicability in realistic 1
situations. The more complicated phenomena in large scale disassembly will be studied so that they may be included where appropriate as the initial program evolves. These phenomena include restric-
[
tion of flow passages by freezing and plugging, miscibility of molten steel and fuel, release of fission gas, and strength and integrity of the subassembly and plenum steel. Methods for
[
fncorporating such effects in the model, as well as improvements in the initial approximations, will i
be explored. Sensitivity studies will be undertaken when the initial computer program is complete, i
Such studies will guide future research.
4 Task 2.
Evaluation of Code Results j
The $1MMER code is to be applied to trial problems and compared to VENUS-!!, FX2 and improved diffusion theory methods to detennine sources of bias and error.
Predicted core motions are to be compared with such experiments as are available to aid in modeling 1
j material interactions, and mock experiments are to be performed to guide design of new experiments.
1 J
]
Task 3.
Consistent System of Core Disruptive Accident Analysis j
The entire accident analysis sequence - SAS-type codes, VENUS-SimER codes, and damage codes of the
]
REXCO type - is to be put on an internally consistent basis to minimize the introduction of errors stensning from changes in reactor description arbitrarily dependent on code input requirements.
t Task 4.
Coding Standards l
A cooperative effort is needed to estabitsh uniform coding standards to promote comparison of j
computed results, improvement of methods, removal of error arising from arbitrary input require-ments, and to improve documentation.
l 4.2.3 System Analysis i
The objective of this program at BNL is to develop a capability to analyze the response of an LMFBR l
system to failure at any point in the system. The complexities of a reactor system preclude the
{
development of a single analytical tool which can treat each of the components in all required detail, consider the entire interactive system, and yet permit many cases to be examined with a reasonable expenditure of computer time. For this reason, a code must be developed to treat processes important to accident conditions in sufficient detail while keeping running time to a i
reasonable level.
}
l l
Research is required to develop a code ar system of codes which can model the major features of the I
primary, secondary, and tertiary heat transport loops while incorporating features important in j
accident delineation and analysis: sodium boiling, flow reversal, and natural convection.
l It is necessary to include some mechanism for incorporating secular changes in reactor power arising either from feedback of effects of changes in primary flow or from the decay of fission product heat sources. In the latter case, flow in control rod, blanket, and reflector elements may be as significant as flow through fuel elements.
In the case of natural convection, representation of parallel flow through damaged channels and undamaged channels is needed to assess the ability to cool in place a slightly damaged core.
Task 1.
Modify Existino Code Work was initiated in FY 1974 to achieve a simplified capability for predicting the short term LMFBR system response af ter a major pipe treak in the primary heat transfer system. The Aerojet Nuclear Company's system code for water reactors, RELAP-3 (M0036), was adapted for this purpose.
Two versions have now been created NALAP for the LMFBRs and HELAP for the GCFRs. These codes are operational, test cases are being run, and final modification will be complete in FY 75.
During the first half of FY 76 a User's Manual will be published to facilitate use by external groups.
Vask 2.
Super System Code (SSC)
This task will be directed toward the development of a computer code for the analysis of LMFBR plant response to transient perturbations. The code will be structured to permit the study of:
(1) the plant respcnse to perturbations in normal operation, (2) " Anticipated Transients Without Scram " (3) the consequences of a major break in the primary coolant loop and (4) the long term heat removal under reactor shutdown conditions. The code will make full use of existing and proven models and methods. Key areas of simulation under transient conditions such as potential reverse flow and sodium boiling will be incluied in the modeling. The transient calculations will be started from consistent steady state conditions initialized by the code. The code will be highly user oriented and adaptable to other computing systems. The initial version will model plants of a loop-type but will be structured with a modular design so that plants of significantly different design (e.g., pot type) can be modeled by restructuring a small number of program modules.
Task 3.
Analysis of System Integrity This task is an analytical effort in the area of system integrity and will be concerned with the interaction between fluid and structure. Investigations will include pressure wave propagation in the primary coolant loops of LMFBRs from any off-nonnal event including (1) sudden closure of a checkvalveand(2)apressurepulseproducedthroughenergeticinteractionbetweenfueland coolant under off-normal conditions.
4.3 Fuel Interactions A program is being started at Sandia Laboratories on the interaction of fuel with coolant and structural material and on the thermophysical properties of fuel, and structural and coolant materials.
4.3.1 Post Accident Heat Removal (pAHR)
Two sets of in pile experiments are planned:
1.
Internally heated fuel debris bed. This will augment and determine the applicability of laboratory fluid heated and bottom heated experiments that have been conducted at ANL in the
__ _ _ _ _ _ - _ _ _ past. The dry-out limit of coolability of the debris-bed is essential in assessing the range of applicability of in-vessel post accident heat removal.
2.
Fission-heated U02 pool convective heat transfer experiment. This in-pile experiment will determine upward, downward, and lateral heat fluxes with real reactor materials and realistic temperatures. The applicability of results of existing laboratory experiments to the reactor accident case will be detemined. This upward, downward, and lateral distribution of the decay heat generated in the pool is necessary to detemine the rate of penetration of a fuel pool into either core-catchers or unprotected structures and materials.
4.3.2 Prompt-Burst Excursions and Fuel Coolant Interactions Millisecond-period fuel-pin meltdown experiments will be performed in sodium. Single pin experiments are to be perfomed in ACPR (Annular Core Pulse Reactor) using a stagnant-sodium autoclave at first, and later a flowing sodium once-through loop. Oxide fuel will be used first. Later carbide and nitride fuel will be used. Both fresh and irradiated fuel pins will be used.
4.3.3 Equation of State and Themophysical Properties The objective of this program is to measure thermophysical properties of fresh and irradiated U and U-Pu oxide, carbide, and nitride fuels at temperatures up to about 5000'K. A variety of in-pile and out-of-pile techniques will be used, with emphasis on dynamic measurements at the higher tempera tures. Properties to be measured include vapor pressure and total pressure, enthalpy, heat capacity, heat of fusion, thermal conductivity, and thermal diffusivity.
4.3.4 Annular Core Pulse Reactor (ACPR) Upgrade The objective is to upgrade the ACPR by increasing both the fluence in the pulse and the steady state power by a factor of about 3.
This will increase the value of use of the ACPR for LMFBR safety experiments on prompt-burst disassembly and post-accident heat removal. The upgrade consists of a new core with an inner zone of high-performance U02-Be0 fuel and a small forced-flow heat removal system.
4.4 Safety Test Facility Studies A review is under way at Los Alamos Laboratory on the needs for new in pile facilities for testing the safety of fast reactors. The review takes into account important experimental needs which cannot be met by use of existing or planned in-pile or out-of-pile facilities.
Consideration is being given to possible adiabatic burst facilities and facilities which combine some steady state cooling with a burst capability, and to possible simulations of postulated whole-core disruptive accidents. One at morc, alternative facilities will be recomended for possible construction, and preliminary costs and schedule estimates will be prepared. The review is being conducted by members of the los Alamos Scientific Laboratory staff with extensive assistance from consultants and other experts from the fast reactor safety comunity. This review is directly related to fast reactor safety work being performed at Los Alamos and to similar studies that have been underway at Argonne National Laboratory under the sponsorship of the Reactor Research and Development Division.
This review includes consideration of a draf t report of a study performed at the Argonne National Laboratory on needs for safety test facilities.
~
-. _ - - - - - _ - A preliminary report has been prepared which assesses the following questions: (1) What are the experimental requirements which cannot be satisfactorily met by use of existing in-pile or out-l of-pile experiments? Situations involving loss of fuel pin integrity, together with the inter-action of hot fuel or cladding with the coolant, are of primary interest. (2)Whatcapabilities must a facility have in order to simulate the accident situations of interest? In particular, how many fuel pins, how rapid a burst, what kinds of reactivity control are required? (3)To what extent do detailed characteristics of the fuel (enrichment, oxide vs. carbide, thennal and irradiation history) affect the likely course of an accident, and how important is it for test fuel to resemble reactor fuel in all these characteristics? (4) What sort of experimental methods should be used to study transient behavior of fuel, clad, and coolant.
In light of answers to these and similar questions, a variety of possible designs of burst and t
cooled facilities will be examined and their merits assessed. Various driver fuels will be considered, and, in particular, some giving a fairly fast spectrum will be reviewed to see if
}
they would provide reliable reactivity shutdown.
l The study is aimed at defining functional requirements that will ensure sufficient flexibility and versatility of the safety tests required for establishing an independent assessment capability
[
J for commercial LMFBRs; and translating those functional requirements into a detailed experimental program.
l 4.5 Monte Carlo Analysis and Planning for Safety-Related Critical Experiments The objective of this work at ANL is the Monte Carlo analysis of safety-related critical experiments
{
which have been performed in past years and planning for a similar experimental program aimed at
[
providing safety-related data for a core prototypical of an LMFBR.
The Monte Carlo technique provides the best available method for solving the Boltzman equation in complicated re-entrant geometries where streaming is an important component of the neutron balance, j
Since geometry can be simulated with no substantial approximations in Monte Carlo, the comparison of Monte Carlo predictions with experimental data provides a test of the basic nuclear cross-section data and in certain cases builds up experience in the use of specialized Monte Carlo I
sampling techniques, i
Task 1, Monte Carlo Analysis of Safety Related Criticals This task consists of Monte Carlo analysis of safety-related critical experiments using the VIM code to verify the Monte Carlo algorithms and basic nuclear data sets. It includes the develop-ment and validation of special computational techniques for the utilization of Monte Carlo in the analysis of such experiments.
VIH will be used for preliminary analysis of ZPR 3 assemblies 27 and 28. and of the ZEBRA assemblies I
8G and 12, in which the configuration changes involved in voiding and meltdown are simulated.
These experiments were conducted several years ago and constitute the currently available experi-mental infonnation.
4 l
l i
I i
, Task 2.
Planning of Safety-Related Critical Experiments This work encompasses in scope the planning for a program of critical experiments bearing on LMFBR safety. This includes the establishment of the objectives of the experimental program, the evaluation of the experimental alternatives, the identification of material requirements for core construction, and the identification of experimental and analytical techniques requiring develop-ment for the successful completion of the program.
An experimental program in the reactor physics aspects of severe hypothetical accidents on an LMFBR will provide experimental data on properties of a plausible sequence of instantaneous (snapshot)configurationsachievedduringahypotheticalmeltdown.
At each of the configurations, experimental results will be obtained for:
a.
Criticality b.
Worth distribution of fuel, clad, coolant, and control materials c.
Fission rate distribution d,
Spatial variations of the spectrum The goals will be to:
l 4
i a.
Provide direct experimental values for Items b and c above b.
Provide results against which calculational models can be tested:
1 j
1.
Monte Carlo - The best technique for treating reentrant geometries where streaming is j
an important component of the neutron balance.
2.
Standard perturbation / diffusion theory for generating material worths and fission l
distributions for use in kinetics codes.
I The experiments are most meaningful if they are prototypical of a power reactor under considera-i tion, e.g., the LMFBR DEMO. Specifically, the relative magnitudes of all components of the neutron balance should be prototypical. While the previously-conducted experiments which form the models of Task l provide useful data, they lack the property of prototypicality.
i Snapshot meltdown configurations will be selected in consultation with ANL experts on reactor 4
safety and will be guided by meltdown calculations using the SAS3A code. Configurations will be selected based on their suitability for testing the V!M Monte Carlo code. Alternate strategies will be assessed for maintaining the experimental configurations near critical while mocking up 1
snapshot configurations which vary from super prompt critical to highly subcritical. A Design Basis Accident calculation of the critical assembly experimental configuration will be initiated if an assessment indicates it is required.
l This task will produce a detailed experimental program plan with complete specification of program steps and schedule. It will be supported by preanalysis of the experimental configurations recommended.
i I
--.~_-,--.._m,~.,_,.__.._,_-..,..-m,
,, ~. _,... _ _ _ _ _.. _ _ _, _, _ _. -
4.6 Gas-Cooled Fast Breeder Reactor A program has been started at Northwestern University to study the motion of cladding and fuel in a gas cooled fast reactor subjected to a loss of coolant. Analysis at Argonne National Laboratory has indicated that, under appropriate postulated circumstances, the motion of clad in a large GCFR under loss of coolant may yield reactivity inputs of the order of a dollar or more and might carry the reactor into the prompt critical regime.
The purpose of the project at Northwestern University is to develop models for the melting of cladding and fuel in flowing gas, appitcable to postulated loss of coolant accidents for GCFR.
The project includes models for (1) clad melting and relocation as a function of flow and energy input. (2) melting and stress patterns in fuel including thermal stress and hydrostatic stress due to fission gas expansion and release. (3) reactivity feedback and its effects on melting rates. Experiments will be perfonned to develop the parametric relationships needed to model cladding, melting, and fuel relocation, and channel blockage effects. The models will be used to prepare computer routines which can be incorporated in computer programs for fast reactor safety analysis. Applicable subroutines for transient heat transfer, fission gas motion and release, and stress in solid fuel bodies, currently part of existing computer programs, will be adapted for use in conjunction with the models under development.
Experiments will use low melting temperature alloys as simulants of the cladding. Special heaters with shaped axial power distribution will be designed and fabricated.
Related work and interfaces include studies supported by the Reactor Research and Development Division (RRD) of ERDA and technical coordination provided to RSR by Argonne National Laboratory.
In addition, liaison will be maintained with the large study supported by RRD at Argonne and to establish independent analytic programs directed at analysis of accident initiators.
4.7 Plant Systems Several problem areas are under review for future programs. They include:
4.7.1 Fission Product Control Followina low Probability Accidents in Fast Reactors (Engineered Safety Feature Assessment)
The objective of this program is to develop practical means for rapidly removing radioactive aerosols from fast reactor containment systems following hypothetical accidents.
An evaluation will be made of possible methods for removing aerosols from the containment vessel of a fast reactor. Based on this initial evaluation, one or more promising systems will be evaluated experimentally, including a large scale proof test of a prototype system.
Upon completion of the evaluation of known potential systems for aerosol removal, the exploratory experimental programs for aerosol removal will be planned. Small-scale scouting tests will be conducted to characterize the performance of candidate systems and identify problem areas. These programs will lead to larger scale tests to refine the model and to select the system offering the most promise.
l l 4.7.2 Caustic Corrosion of LMFBR Components i
The objective of this program is to characterize the attack on stainless steel under the complex environmental conditions that will be found in an operating LMFBR. The potential for pipe ruptures and the probability of such a fracture leading to large releases of sodium should be studied to provide assurance of primary system integrity. The fracture mechanics of sodium / sodium hydroxide-contaminated pipe structures under stress will be evaluated.
Initially, test samples will be designed and methods evaluated for characterization of the failure initiation and propagation processes. In addition, arrangements will be made for review of examinations in progress at other laboratories and for the examination of selected parts of decommissioned reactors which may provide ovidence not available in the laboratory.
The experimental program will be aimed at developing data suitable to predict the probability of occurrence of a large pipe break. In adriitton, it will be extended into a caustic corrosion study with maximum relevance to materials, operating conditions, and configurations in LMFBRs.
4.7.3 Integrity of Equipment Cell Liners for LMFBRs The objective of this program is to characterize the deformation of steel liners in the concrete equipment cells of LMFBR power plants as a result of possible major accidental sodium spills. In addition, a study will be made of special problems of the concrete in the cell which functions as either structural support or as high temperature insulation.
A state-of-the-art report on liners for cells containing sodium bearing equipment was completed in FY 1975. If a further program is required, a program plan will be developed and a conceptual design study of various cell liners will be performed to serve as a frame of reference for the program. Analytical techniques and design methods will be evaluated. Construction will be started of a test for subjecting large scale steel lined concrete cells to a range of sodium spill conditions.
4.7.4 High Strain Rate Behavior of Components and Systems The objectives of this program are:
To provide dynamic analysis methods and material behavior data for assessing response of reactor systems and components to impulsive loading.
To provide means for assessing structural damage as a result of impulsive loading in terms of useful operating life of reactor systems and components.
Initially the character of the loading on the systems and components will be defined (HCDA, pressure pulse, or seismic induced). Experiments will be conducted to obtain the baseline mechanical behavior data of the materials. Once analytical methods have been made available they will be refined in selected areas based upon sensitivity analyses to identify where more meaningful results may be obtained; this improvement will also be accomplished in conjunction with verifica.
tion tests using scale models, to provide the basis for establishing the useful life remaining in a component or system subjected to a series of abnormal conditions.
4.7.5 Steam Generator Safety Assessment The objectives of steam generator safety assessment are to develop methods for analyzing the behavior of LMFBR plant steam generator systems to unscheduled events and the response of the pressure relief system. The assessment will also cover the safety consequences following steam generator tube failures.
Of major consideration is the potential for a failure of the tubes in the steam generator leading to a sodium-water reaction. The safety and reliability of the steam generator viewed as a complete system are to be examined.
A review and identification of system design margins and the recommendation to conduct analytical or experimental programs to aid in assessing the safety of steam generators is a major initial task. The examination would cover the assessment of stability during the various phases of the sodium-water reaction, and the character of events during and after depressurization for the complex structural shapes in steam generators. The tasks include investigation of materials, leak detection, inspection, vent systems, tube wastage, and failure propagation.
4.7.6 Effects of Sodium Fire on Systems and Structures important to Safety The purpose of this program is to verify and improve current models for calculating the effects of potential large sodium releases on systems and structures important to safety. This investiga-tion will address sodium spills, sprays, and jets, and detennine if further work is needed to complement present programs.
Upon completion of a state-of-the-art survey, the problems of scale effects in models along with simulation of a dual containment environment will be evaluated. Characterization of sodium fires with respect to thermal parameters and fire extinguishment will be directed to both assessment and development of methods to minimize the extent of damage to vital systems.
5.
Gas-Cooled Reactor Safety Assessment The objectives of the programs under this activity are to improve the basis for independent safety assessment of commercial High Temperature Gas-Cooled Reactors (HTGRs) by the Regulatory staff, and to establish a program capable of timely resolution of safety problems that may be associated with advanced fuels and systems. This program differs from that of the Reactor Research and Development Division of ERDA, which has the objective of developing a sound basis for the design of commercially viable HTGRs and their associated fuel cycle. The program is defined in large part on the basis of Regulatory needs expressed in both formal documents and staff contacts, and through recommendations of the Advisory Committee on Reactor Safeguards.
Gas-cooled reactors have a long history, dating from the first air-cooled piles of the nuclear age. They were used in the United States solely for research purposes until the installation of the first commercial power-producing reactor, the 40-MWe Peach Bottom No. I unit, by the Philadelphia Electric Company in the mid-1960's. The sole vendor for gas-cooled reactors in this country, the General Atomic (GA) Company, since supplying the Peach Bottom plant, has constructed a 330-MWe reactor (Ft. St. Vrain), which is undergoing startup tests, and has on order six reactors rated at 770 MWe, or better. The recent and relatively rapid increase in interest in power generating, gas-cooled reactors has prompted a parallel increased interest in safety research for this reactor type and the formulation of this research plan.
1
r The HTGR is a thermal reactor with an oxide or carbide fuel incorporated in a graphite core cooled by helium. Heat is removed from the primary helium coolant in steam generators. The core and the entire primary coolant system, including helium circulators and steam generators, are located within a prestressed concrete reactor vessel (PCRV). The PCRV is a multicavity vessel tJith a steel liner. The normal helium operatirg temperature and pressure at the core outlet are 1366*F and 700 psig. Since the HTGR differs radically from water reactors in many ways, e.g.,
fuel, cladding, moderator, coolant, reactor vessel, etc., safety assessment questions for HTGRs differ considerably from those for water reactors.
The fomer Atomic Energy Comission has carried out a substantial program of work aimed at developing a commercially viable HTGR. Work by the vendor (GAI and by Holifield National Laboratory has included both base technology and safety programs. A planning guide in the safety area related to the ERDA-RRD HTGR development program has been issued: " Planning Guide for HTGR Safety and Safety-Related Research and Development" (ORNL-4968). The plan reported here is RSR's independent program in this area. Emphasis is placed on the consequences to public safety of possible abnonnal and accident conditions. Some attention is given to the related subjects of long tenn reliability and failure modes arising during long term operation. A special aim of the program is to provide assistance to Regulatory in its evaluation of safety questions arising in I
the licensing process.
The objective of the program plan is to describe a systematic approach to the attainment of-program goals in a timely manner.
Major areas of attention (Table 2) are defined by the novel nature of the HTGR design, and the exposure of materials at relatively high temperatures (1366*F) in a helium atmos'phere and radiation field.
TABLE 2 l\\
GAS COOLED REACTORS MAJOR CONCERNS 1.
ACCIDENT DELINEt. TION 2.
FISSION PRODUCT RELEASE 3.
STRUCTURAL RESPONSE j
CORE PCRV 4.
MATERIAL PROPERTIES HIGH TEMPERATURE LONG TERM 5.
COOLANT CONTROL The areas include:
Fuel failure--the fuel is in the form of particles of UO2 or UC and Th02 coated with two or more layers of carbon compounds.
b PCRV Integrity--Although there has been considerable experience with prestressed concrete pressure vessels abroad, use of such a vessel as a reactor vessel, especially a multicavity vessel, is'new in the USA.
Seismic response is of concern with respect to two features. The PCRV is a relatively rigid structure and will transmit seismic loads to the components located in the cavities. Secondly, the core support structure (as well as the core itself) is constituted of graphite. As a load bearing material, graphite has good high temperature properties, but it is a new material in this sense.
The integrity of major components in the primary system is of concern because of the reliance on i
the reliability of this system to masntain a degree of cooling under possible accident conditions even though the system has been operating at high temperatures and pressures.
The question of possible rapid oxidation of graphite is of interest because in the case of air ingress, potential fires could lead to fission product release and high themal load.
There are questions on the effects resulting from moisture leaking into the primary system from the steam generators. This moisture would degrade the graphite. It is necessary to have reliable and accurate monitors to detect moisture in small amounts, and also to be able to estimate the effects of attack by low concentrations of moisture over extended periods of operation.
Current efforts to resolve these questions are being carried out in ten task areas whose objectives are defined in Tables 3 and 4.
As a result of work in the past year it has been possible to define the program plan presented here.
TABLE 3 i
GAS COOLED REACTOR SAFETY PHENOMEN0 LOGICAL RESEARCH t
i TASK OBJECTIVE Fission Product Transport Extent and consequences of fission product release in accidents.
Primary Coolant Impurities Determine effects of reactions between primary
~l-coolant impurities and materials within the primary coolant circuit (core, graphite.
l structures).
i Rapid Graphite Oxidation Insure and provide data for safety assessment of consequences of accidental ingress of massive quantities of air / steam into the core.
Structural Evaluation Define failure conditions and determine margin of safety of PCRV, PCRV internals, core, and core support under seismic, aerodynamic, and thermal l
loadings.
Materials Technology Provide information on properties of structural materials to assess integrity and performance of components under normal and accident conditions, i
Safety Instrumentation Determine suitability and viability of safety I
and Control System instruments and control systems under noma 1, upset, and accident conditions.
TABLE 4 GAS COOLED REACTOR SAFETY ANALYTICAL RESEARCH TASK OBJECTIVE i
Accident Delineation Identify and examine for completeness accident I
sequences having potential impact on public
[
safety.
Probabilistic Analysis Provide a measure of the relative likelihood of a given sequence for all accident sequences delineated.
Phenomena Modeling and Provide validated analytical methods, models System Analysis and codes describing response of system to postulated events and conditions - document descriptions of phenomena and process models and results of accident analyses.
Proof Tests Requirement Studies Identify accidents, and suitability of existing facilities for performing proof tests - conduct tests.
The plan divides the research into three interrelated sections dealing with phenomenological research, analytical research, and proof tests (See Figure 3). Investigation of a specific problem generally starts with an analytical research study in which available data, design methods and theory are used. Gaps in theory, data, etc., are thereby identified and are then filled by phenomenological research. Iterative interaction of the analytical and phenomenological studies results in a model that may be subjected to a proof test before acceptance as a tool for making safety assessments.
Brief summaries of the content of the three research sections follow.
5.1 Phenomenological Research 5.1.1 Fission Product Release and Transport Topics of interest include coated fuel particle failure, and fission product diffusion, adsorption, plateout and lift-off, and transport into the secondary containment and to the environment.
These topics are of interest to the extent that they would be enhanced by abnormal or accident conditions. Anticipated fuel failure and behavior of released fi m n products occurring during normal operation are base technology topics.
5.1.2 Primary Coolant Impurities Tasks include characterization of reactions of graphite or fuel with coolant impurities such as steam or air. Impurity concentration levels of interest are the intemediate levels that would be found under moderate accident conditions, rather than the low levels anticipated during normal operation or the very high levels that would be found in postulated severe accidents. Also included is the study of reactions of impurities with primary loop materials and their effects on the long tem reliability of these materials.
HTGR SAFETY RESEARCH PROCESS i
PROGRAM INPUT r
n BASIC THEORY MATERIAL &
REACTOR COMPONENT DESIGN DATA ANALYSIS METHODS DATA Jl U
U 1 f SECTION 1 ANALYSIS MODELS! & METHODS PHENOMENOLOGICAL l
\\
RESEARCH I
DEVELOPMENT OF MODELS I
i l
DESIGN INTERPRET t
EVALUATION OF ADEQUACY q
OF MODELS, METHODS & DATA lEXPE'11MENTS EXPERIMENTS g
g+
EXPERIMENTAL MEASUREMENTS gg I
jl lJk I
il 1
FUEL FAILURE &
l' COOLANT MECHANICAL MATERIAL
\\#
FISSION PR DUCT IMPURITY EXCITATION CHARACTERISTICS OK?
v v
t j
L. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.\\p.___..________________________._..i gCg,g N
RESEARCH g
n INTEGRATED SYSTEM MODEL g
n I
I DYS2N 05 AN LSMF h,,,
i eROOF TESTS eROOF TESTS j
I 1[
SECTION 3 I
l l
l AGREE VERIFIED HTGR l
PROOF TESTS
=
=
SAFETY INFORMATION DISAG-
& ANALYSIS METHODS l
REE g
_____________J u _ _ _ _. y t
FIGURE 3 5.1.3 Rapid Graphite Oxidation l
Interest is centered on the rapid reactions that might result fron. postulated ingress into the core of massive quantities of steam or air and on the consequences of these reactions. Phenomena to be studied includes the production of combustible mixtt.res of oxidation products and loss of strength of the core support graphite.
5.1.4 Structural Evaluation Structural Evaluation tasks deal with the response of the core, core support structures, primary loop components, and the PCRV to major short tem mechanical excitation, e.g., to seismic events, and with long term reliability in the face of normal vibrations or other nomal excitation. Also included are tasks dealing with the monitoring of permanently installed components.
5.1.5 Materials Materials tasks include determination o' properties of metallic and control materials, graphite, and concrete useful in evaluating response to postulated abnormal or accident conditions and in evaluating long term reliability.
5.1.6 Safety Instrumentation and Control Systems Tasks in this area deal with instrumentation and associated devices and systems used for surveillance, monitoring of plant parameters, detecting potentially unsafe conditions, and initiating appropriate protective actions. Included are questions concerning the ability of instruments and devices to remain operable subsequent to transients that might be imposed by hypothetical abnomal or accident conditions.
5.2 Analytical Research The program elements of this section are iteratively interrelated, and their ordering below is significant in the sense that Accident Delineation plays a guiding role for the entire safety research process, it is recognized that initial starts will be subsequently modified as a loop is traversed, generating multiple iterations before the output is acceptable.
5.2.1 Accident Delineation The major task is the comprehensive and detailed description of specifically postulated accident sequences, including events and conditions leading to serious accidents and the consequences of these accidents.
5.2.2 Probabilistic Analysis Probabilistic Analysis tasks include the calculation of the probability of the sequential or simultaneous occurrence of events along an assumed pathway, utilizing various probabilistic methodologies such as decision analysis, event trees, fault trees, etc.
5.2.3 Phenomena Modeling and Systems Analysis Tasks in this area include the development of appropriately validated mathematical models and associated computer proi, rams that will provide adequate descriptions of the behavior of HTGR core neutronics, hydrodynamic components and subsystems, physical and chemical processes, etc., and, ultimately, the response of the HTGR plant to transients imposed during postulated abnomal or accident conditions.
__ _. =
. - ~
5.31 Froof Tests An integrated model arising from the iterative processes of Sections 1 and 2 (Figure 3) will require further. validation in those instances in which there is doubt of the adequacy of the
^
analytical solution to describe the outcome of a postulated major accident.
It is recognized that a proof test may involve the construction of facilities that are expensive and require long lead times for design and construction. Thus, not only is an early determina-tion of need desirable, but a study of the availability and suitability of existing facilities is l~
required.
5.4 Relationship to the Development Program Although some duplication of effort is needed to establish an independent safety assessment capability, the aim of this section is to make a comparison of the Reactor Safety Research plan with the Development programs to identify those areas where there are common interests and those where interests diverge. This identification will enable work to be planned to avoid needless duplication and enable productive coordination of effort.
Components of the RSR (NRC) and RRD (ERDA) program plans are shown in Table 5 below. The RRD plan is described in ORNL 4968 and consists of " Safety R&D" and " Safety-Related R&D." The plan cutline followed in Table 5 is that of the RSR plan. The RRD plan components are compatible with this outline except in Section B, Analytical Research. Equivalent program elements are present in both plans in this section, but in the RRD plan they are not conveniently listed according to the RSR plan outline.
TABLE 5 COMPARISON OF RSR AND RRD HTGR PROGRAM PLANS i
R_S R, RRD Safety-Related R&D S
RRD Safety R&D i
i Section A.
Phenomenological Research
]
(
l.
Fission Product Release and Transport Need assessment Fuel & control material migration: mechanisms, code development.
Coated fuel particle failure -
Fuel particle failure:
i
& release rates: fuel temperature mechanisms, failure transients, data review, code criteria, irradiation development, effect of capsules and irradiated hydrolysis, experimental fuel examination. Fission studies.
product release from fuels:
1-mechanisms, source terms for codes, distribution in in fuel bodies Fission product distribution Fission product behavior in in core structure: transport graphite: experimental model review, code develop-characterization, code ment, vapor pressures, development, postirradiation oxidation and hydrolysis, examination of fuel element diffusion, adsorption, fuel
& reflector graphite.
elements of advanced design I
_ _ _ _ _ -.. ~. _ _ _, _ _
. TABLE 5(Cont'd)
COMPARISON OF RSR AND RRD HTGR PROGRAM PLANS RRD Safety R&D RSF_t RRD Safety-Related R&D Fission product behav-Fission product distribution Fission products in the ior in coolant circuit, in the coolant loop: chemical coolant circuit: surface engineering scale specification, effects of steam chemistry-behavior on dynamics: f.p. distri-ingress or depressurization, coolant circuit surfaces, bution and reentrain-role of dust, loop adsorp-adsorption and evaporation, ment code development, tion capacity, spatial behavior in Peach Bottom, experiments on re-distribution.
Ft. St. Vrain, and large entrainment and on HTGRs.
f.p. and dust distribution.
Fission products in Fission product transport in containment; review containment and in the of fodine behavior, atmosphere: chemical code development for specification in containment iodine behavior, building clean-up system assessment, deposition rate in contaimnent, plateout and retention in concrete, effects of helium on transport.
release, behavior in graphite, code development, behavior in Peach Bottom.
Ft. St. Vrain, large HTGRs.
2.
Primary Coolant Impurities Need assessment Need assessment: graphite-air / steam kinetics and mechanisms, reactions of CO, H, CH4, etc., codes 2
for core corrosion l
Thermochemical studies:
Steam-core reactions and mass transport, localized coolant composition, lab corrosion, hydrolysis, studies (graphite oxida-volatile or unstable f.p.
tion kinetics, reactions forms.
involving fuel kernels, reactions involving f.p.),
modeling steam ingress, code validation.
Code validation in out-of-pile and in-pile loops.
3.
Rapid Graphite Oxidation Review and evaluation of Flammability Studies existing data and models, code development, experi-mental studies, code validation in loops.
4 Structural Evaluation Need assessment (Seismic
& vibration technology; confinement components)
Soil-structure dynamic PCRV model tests, codes for Model studies and PCRV interaction vessel, liner and attachments surveillance, Analytical PCRV failure Failure analysis methods studies Liners & penetrations.
TABLE 5(Cont'd)
COMPARISON OF RSR AND RRD HTGR PROGRAM PLANS RRD Safety R&D R2 RRD Safety-Related R&D Core seismic response:
Core-fuel element collision Reactor internals:
collision dynamics, dynamics: 3D codes, model vibration-response tests, codes, model tests, tests.
codes.
PCRV support concepts:
tests, analysis.
Containment convection studies: forced & natural convection cooling.
Code rule development.
5.
Materials Need assessment Metal components: data Primary loop metallic library, property measure-components: status report, ments, welds, failure modes, properties and failure code development.
criteria, design analysis methods, Peach Bottom examination, alternate steam generator materials.
Graphite components:
Graphite components:
mechanical properties, anneal-creep coefficient ing, creep coefficient determination, high determination.
temperature properties, thermal history effects, annealing, cyclic loading, constitutive equations; reactor surveillance.
PCRV literature review:
Concrete properties: high property changes at high temperature strength, long temperatures, tendon term creep, shear strength, relaxation, liner materials development, properties, code develop-Prestressing components-i ment.
tendon corrosion & failure, strip-wour.d prestressing systems, tendon installation, Linars and penetrations-materials, failure modes, design analysis methods.
Other materials: boron oxide Control materials: reactions formation and effects, thermal with coolant impurities, barrier resiliency & wear.
structural stability, cladding, integrity, model-ing, Ft. St. Vrain surveil-lance, Thermal barrier-testing, materials development.
6.
Safety Instrumentation and Control Systems Safety instrumentation assessment Moisture monitors, temperature Failed fuel location, sensors, neutron detectors, moisture monitors, PCRV surveillance, radiation thermometry, flux monitors, monitors, additional surveillance by noise requirenents, analysis, DAP, post accident monitoring plant controllability.
.. t TABLE 5(Cont'd)
COMPARISON OF RSR AND RRD HTGR PROGRAM PLANS RRD Safety R&D RSR RRD Safety-Related R&D I
Section B.
Analytical Research 1.
Accident Delineation
' Component qualification:
Safety studies evaluation:
reliability, design Safety significant event qualification, applica-sequences, specific accident s
tion in systems, descriptions, response evaluation.
Effect of design.
Library of accident sequences & consequences.
2.
Probabilistic Analysis Analysis of event Component reliability:
sequences: systems Model selection, functional logic, acci-Code development, dent initiation.
Calculation of likelihood confidence analysis measures for consequences of methods, common mode accidents.
failure, logical sequence models.
3.
Phenomena Modeling and System Analysis Modeling of plant Review and evaluation transient behavier:
of state of art, neutronics, components Testing of current models.
and subsystems, physical and chemical processes, behavior models, transients.
Safety analyses and Model development.
Qvaluation: review of Establishment of code available analyses,
- library, analyses and sensitivity studies, definition of additional needs.
Section C.
Proof Tests Determination of need and development of specifications for proof tests.
Conduct of proof tests.
In general, although there is not a one-to-one correspondence of elements in the two plans, the overall content of the two plans in a given subject area is similar. The RRD plan is somewhat more detailed and contains some elements missing from the RSR plan. One notable exception is the Rapid Graphite Oxidation task in Section A of the RSR plan which is essentially absent from the RRD plan. Another is the explicit provision in the RSR plan for Proof Tests (Section C).
Particularly noteworthy in the RRD plan is the distribution of work between the Safety R&D and Safety-Related R&D Categories. Quantitative measures of this distribution are given in Table 6 where the man-years of effort devoted to the two categories in Sections A and B are given for the ten-year life of the plan from 1975 through 1984 All, or substartially all, of the work on
7. Seismic and Vibration Technology and on Analytical Research is designated as Safety R&D. With these two exceptions, the rest of the work, amounting to two-thirds of the total, is designated as Safety-Related R&D. This distribution of so much work in the latter category places a heavy demand on RSR to establish particularly strong programs in these areas, so that a basis for independent safety assessment may be made available.
TABLE 6 RRD HTGR SAFETY R&D PROGRAM MAN-YEAR REQUIREMENT 1975-1984 (SOURCE: PLANNING GUIDE FOR HTGR SAFETY AND SAFETY-RELATED RESEARCH AND DEVELOPMENT." ORNL 4968, MAY 1974) a Program Area Safety P&D Safety-Related R&D Total Section A.
Phenomenological Research - Effort in man-years Fission product release and transport 50.5 339.5 390 Primary coolant impurities 7
90 97 Structural evaluation Seismic & vibration technology 205.9 13.3 219.2 Confinement components 15 226.1 241.1 Materials 30.5 226.7 257.2 Safety instrumentation
& control systems 12.5 79 91.5 Section B.
Analytical Research Accident delineation Probabilistic analysis Phenomena modeling
& system analysis 134.5 0
134.5 Totals 455.9 974.6 1430.5
' Safety-Related R&D effort determined by difference.
. Source gives total effort in " relative man-years
- and Safety R&D effort in " man-years."
6.
Nuclear Safety Test Facilities The validation of computer codes for reactor accident analysis and the requirement to obtain experimental data in a realistic geometric, nuclear, and physical environment places heavy emphasis upon the 3vailability of large scale engineering and nuclear test facilities. The Semiscale facility for PWR system separate effects testing and the BWR Blowdown Test Facility for simulating BWR blowdown conditions are examples of engineering scale facilities. The LOFT and PBF facilities at the Idaho National Engineering Laboratory were designed and constructed to under-take reactor accident experimentation in a characteristic nuclear environment. The programs identified with these facilities have been highlighted in the preceding section of this document and are further detailed in Appendix A and B.
Studies are also underway to establish the criteria and requirements for Safety Test Facilities for performing tests in a nuclear environment characteristic of postulated fast breeder reactor accidents. An evaluation of the feasibility of use of existing facilities together with the definition of experimental requirements for gas cooled reactor safety research has also been initiated as mentioned in the preceding sections of this report.
A list of planned and existing safety research facilities required to support the Reactor Safety Research program is provided in Table 7.
N,
TABLE 7 NUCLEAR SAFETY TEST FACILITIES Facility Description Status Location LOFT 55 MWt PWR LOCA-ECCS Tests Pretest system checkout ANC, INEL PBF Open tank reactor, pressu-In use '
ANC, INEL l
rized test loop, transient fuel tests Semiscale 1-1/2 Loop PWR Model of LOFT, In use ANC, INEL Mod-l 40 electrical heaters, 5-1/2 ft long Forced Convection Pressurized blowdown loop.
In use HNL Test Facility single rods up to 12 ft long Thermal Hydraulic PWR loop for blowdown heat Startup and pretest HNL Test Facility transfer, 49 rods, 12 ft shakedown long BWR Blowdown Two loop BWR simulator for In use GE, San Jose Test Facility BDHT tests, 49 rods, 12 ft heaters PWR-FLECHT Forced convection loop for In use
-W, Pittsburgh PWR reflood testing, 100 rods,12 ft length 10 ft 00,100 psia vessel, Construction PNL, Hanford PlenumFilling)
Experiment (PFE 4.5 ft OD, 600 psia vessel, ECC Steam-water mixing HSST Facility Pneumatic and thermal sho d Modification of existing HNL tests of Model reactor facilities vessels Safety Test Fast Reactor, transient in-Plann=4 Facilities reactor tests Containmtnt Research Radionuclide Source Tenn Modification of existing HNL Installation (182)
Testing, Fast Reacter facilities Fuels Annular Core Pulsed Fast Flux, short period Modification of existing Sandia Reactor (1) tests of single pins facility Laboratory (LMFBR)
Annular Core Pulsed Fast flux, short period Planned modification Sandia Reactor (2) tests of fuel cluster of ACPR (1)
Laboratory (LMFBR)
Gas Cooled Reactor Safety tests of Gas Cooled Planned review of existing Safety Tests Reactor fuel and facilities components
APPENDIX A BRANCH PROGRAM PLAN FOR SYSTEMS ENGINEERING BRANCH 1.
OBJECTIVES The objective of the safety research sponsored by the Systems Engineering Branch is to provide sufficient experimental data for establishing and verifying analytical methods which will i.
pemit assessment of the thermal-hydraulic response of the reactor primary coolant system, components and reactor containment to possible off-normal and accident conditions. Current and planned research areas included in the Program Plan of the Systems Engineering Branch are 1) Loss of Coolant Accident (LOCA) and Emergency Core Cooling (ECC). 2) Alternate ECCS, and 3) Reactor Transients. The following three sections outline the research needs in each of the three areas.
1.1 LOCA/ECCS The research effort in the LOCA/ECCS area is devoted to the generation of experimental data to provide an improved understanding of the phenomena that would be involved in a LOCA and to provide improved analytical methods for predicting the response to LOCA with ECCS in current water reactors. The objectives of LOCA/ECCS studies are:
1.1.1 Reactor Core Blowdown Flow, Heat Transfer, and Systems Effects Tests To detemine the estimated time to CHF and the heat transfer rates during pre-and post-CHF phases of t. lowdown as influenced by variations in power, system pressure, coolant flow and breaklocation(PWRgeometry).
To investigate the estimated time to CHF, the hydrodynamics of lower plenum swell, and the post-CHF heat transfer in sufficient detail to evaluate thermal behavior phenomena prior to and after the availability of emergency core cooling (BWR geometry).
To provide data for use in evaluating the validity of analytical predictions of core flow rate as a function of system volume, coolant conditions, hypothetical break location and size.
1.1.2 Reactor Core Reflood Flow, Heat Transfer and System Effects Tests To determine bottom flooding flow rates and core temperature during one-dimensional (1-D) rod bundle reflood experiments.
To analyze the relationship between 1-D reflood test results and 3-D core behavior in PWRs.
A-1
To initiate fundamental investigations which will permit improved prediction of transient reflood heat transfer processes.
To develop heat transfer correlations which would be applicable during PWR reflood.
To provide carry-over flow rate data for use in evaluating the validity of analytical predictions.
1.1.3 ECC Bypass and Steam-Water Mixing Programs To provide data for evaluating the potential for ECC Bypass which might occur during the latter stages of blowdown and the early portion of reflood under hypothetical LOCA conditions, thereby establishing the conditions when " bypass" terminates.
To experimentally investigate the influence of scale (or size), hardware effects (e.g.,
multiple loop configurations, downcomer and lower plenum), pressure, system feedback effects, and ECC inject!)n modes on the amount and rate of ECC water that would penetrate to the lower plenum.
To provide data for determining the effectiveness of steam-water mixing and condensation effectiveness in the cold leg ECC injection section, and to establish whether steam plugging would occur.
To develop two-phase flow models for prediction of ECC injection effectiveness, ECC bypass and downcomer penetration, and lower plenum entrainment; to verify these models with key experiments.
1.1.4 Reactor Coolant Pump Transient Characteristics Tests To determine pump transient, two-phase coolant behavior under test conditions simulating the blowdown portion of a hypothetical LOCA in both the broken and intact loops, including tests i
addressing the pump overspeed potential induced by the blowdown.
To utilize the results of these subscale pump tests to derive transient pump models which can be extrapolated to full size conditions and then used for LOCA calculations.
e 1.1.5 Integral System Behavior Tests To ensure the adequacy of codes for predicting the integrated coupling effects of the above j
components in water reactor systems obtained through " separate effects" or " systems effects" testing.
1.2 Alternate ECCS i
To investigate and test ECCS configurations which would:
enhance plant protection during a LOCA.
A-2
.n be insensitive to reactor design parameters and proper functioning of other components such as steam generators or reactor containment.
have redundancy, diversity and abundance of flow such that adequacy can be detennined with-out unduly complex evaluation techniques.
be unquestionably reliable.
1.3 Reactor Transients i
To provide experimental confirmation that current analysis methods conservatively describe.
effects of such hypothetical off-normal conditions as:
loss of flow reactivity insertion 4
steam line break loss of load, and anticipated transients without scram.
3 2.
PRESENT STATUS Planning of research programs must be based upon a background knowledge of the current research status and state-of-the-art for each subject area as well as the knowledge of the current licensing criteria for each subject. The background information in each of the three main research areas is presented in the following sections, i
2.1 LOCA/ECCS I
2.1.1 Reactor Core Blowdown Flow, Heat Trar.sfer and System Effects LOCA assumptions applicable to water reactor blowdown are defined in the NRC's Acceptance Criteria for ECCS.E These assumptions are based upon the use of CHF correlations derived primarily from steady-state test conditions. A common feature of the correlations for PWRs is a dependence on coolant mass velocity. Since the calculated response for a cold leg break LOCA in a PWR would have a significant flow reversal during the first second of the blowdown, the use of a steady-state CHF correlation predicts a very short ($1 sec) time delay to CHF which is probably unrealisti-l cali, conservative. This approach does not consider the tnermal inertia pf the liquid flim nor is credit allowed for the cooling effect of the liquid droplets in non-equilibrium flow. For current BWR designs no flow reversal is anticipated during a hypothetical LOCA. The time delay 3
to CHF would be longer in this ca;e.
I M10 CFR Part 50 (Docket No. RM50-1), Licensing of Production and Utilization Facilities.
" Acceptance Criteria for Emergency Core Cooling Systems for Light Water-Cooled Power Reactors "
December 28,1973.
A-3 l
4 Post-CHF heat transfer correlations are also primarily based on steady-state results without considering the cooling effect of liquid droplets which would exist because of the thermodynamic non-equilibrium effect. Additional transient data from systems during depressurization are needed.
2.1.2 Reactor Core Reflood Flow Heat Transfer and System Effects Qualitative understanding is available ta explain the interactions of physical phenomena which determine the rate of heat transfer from heated reactor fuel pins to ECC water during reflood.
Information from FLECHT tests is available to determine heat transfer coefficients for various flooding rates, rod temperatures and system pressures. FLECHT data (but not correlations) are currently used in licensing evaluations for all PWR reactors. A limited amount of heat transfer infonnation is available from FLECHT tests in which a horizontal plate was used during several constant rate flooding tests to study " blockage" effects on heat transfer. Additional one dimensional (1-D) experiments such as FLECHT-SET and Semiscale are used to study system effects.
The oscillatory reflooding behavior exhibited in the 1-D analysis or experiment may not be typical of large PWRs, hence, the heat transfer benefit of flow oscillation is not credited in PWR LOCA analyses, t
For BWRs the reflood heat transfer data were obtained by top spray and bottom flooding in the BWR-FLECHT program and in other GE sponsored testing programs. Additional analyses are required
' to relate the 1-D test data to power reactor analysis.
2.1.3 ECC Bypass and Steam-Water Mixing i
Current Evaluation Model LOCA calculations conservatively assume that water plugging can occur at the ECC injection region and that all ECC injected would bypass the core region during the blow-down phase. Steam relief tests run at Combustion Engineering in CY-73 on 1/5 and 1/3 scale cold leg ECC injection hardware have demonstrated that water plugging is not likely to occur and that local pressure drops on the order of -2 psi to +6 psi exist. Limited ECC/downce'ner penetration tests were run at CE in early CY-74 in a 1/5 scale vessel, followed by more extensive testing at BCL in a 1/15 scale transparent vessel model having multiple cold and hot leg simulation. In i
addition, hot wall delay effects were investigated in FY-75 at Creare utilizing a flat plate geometry with multiple leg simulation.
BCL has utilized the CE data to derive an ECC injection model. The results were published in
[
BMI-1931 "A Simplified Steady State Analytical Model for Prediction of the Pressure Differential in Outlet Fluid Properties of the ECC Injection Section of a Typical PWR During a LOCA " January 20, 1975. Recent results of BCL's 1/15 scale multiple leg ECC downcomer penetration results were reported in BCL's Quarterly Progress Report for January thru March 1975. Results obtained to date from the BCL tests have correlated well with the limited CE 1/5 scale tests, and provide an extensive data base for purposes of establishing best estimate models. The results of these small scale tests have revealed that partial ECC delivery can occur at high steam upflow velocities i
in the downcomer and that simplified Wallis type flooding correlations are too conservative in the prediction of this behavior.
A-4
m__
The hot wall effects tests run at Creare are an extension of previous FY-74 tests investigating top flooding and single intact leg simulation. Initial Creare results have been published in TN-202, " Hot Wall Experiments in a Simulated Multi-Loop PWR Geometry," February 1975. The initial results support previous hot wall effects findings, namely, that delay times for large gap sizes
_ generally tend to be small. The program is still in progress and will investigate the effects of steam upflow and multiple leg ECC steam mixing in FY-75.
The Plenum Filling Experimental program is currently in the final stages of test bed design.
This program will provide facilities for evaluating the potential for ECC bypass and the type of two-phase flow phenomena which might occur during the latter stages of blowdown and the early portion of reflood. Experiments will be conducted in 1/5 scale (linear) and 4/5 scale vessels with the capability to investigate the influence of geometric scale, annulus geometry configurations, 4
pressure, system feedback effects and ECC injection modes on ECC/downcomer penetration. PFE experiments will measure both the bypass and the rate of injected ECC water delivered to the lower plenum. - Test bed construction will be initiated in FY-76, followed by emplacement of a j
500,000 lb/hr steam supply system. The current target date for start of testing is the Fall of 1977.
2.1.4 Reactor Coolant Pump Transient Characteristics I
Current pump models utilized in LOCA analyses are based on homologous single phase pump theories and assume that these scaling laws, with some modification, can be used to calculate transient, two-phase pump behavior.
l Two-phase pump experiments carried out on the Semiscale pump in FY 1975, in both steady-state and transient modes, resulted in head degradation correlations, thereby providing a means to establish a modified transient pump performance model. The Semiscale pump data is currently utilized in the Regulatory EM (Evaluation Model) pump model.
CE has a 1/5 scale pump (N = 4200) and is currently under contract in a joint program with EPRI 3
to modify a steam-water facility to carry out steady-state and transient two-phase pump experiments.
i B&W has a 1/3 scale pump (N = 4200) which has recently been tested in air-water mixtures. This 3
pump may be made available to EPRI under a joint funding program for testing in the CE facility.
Finally, there are 1/4 and 1/5 scale pumps available through KWU in West Gennany which will ur.dergo testing at the planned CE facility.
In addition, questions related to overspeed potential, and associated effects, have been raised by ACRS. Plans to test this potential are under consideration.
2.1.5 Integral System Behavior i
System behavior throughout the entire hypothetical transient is currently predicted with one or a combination of computer programs referred to as system codes. Component portions of system codes are currently being developed and verified based upon data from the separate effects test programs.
]
i A-5
f Confirmation of the ability of a system code to predict the entire transient behavior can only be otained from facilities in which all known LOCA phenomena are expected to occur in their proper
-time sequence. The Loss of Fluid Test (LOFT) and Semiscale test facilities at the Idaho National Engineering Laboratory (INEL) are designed to fulfill that requirement as well as can be expected at their reduced scale.
2.2. Alternate ECCS IntheECCSCriteriaE the Commission "... (urged) the pressurized water reactor manufacturers to
[
seek out design changes that would overcome steam binding." The ACRS in its September 10, 1973 letter to the Commission on the same subject stated that ".... reactors filing for construction permits after January 7,1972, should have significantly improved ECCS capability." The Regulatory Staff reiterated this position on October 18, 1973 in REG:RSR 108 in which it was stated that
" implementation of such (alternate ECCS) studies are considered by Regulatory to be an A priority task." In response to the above requests, the Division of Reactor Safety Research has initiated l
studies of alternate ECCS concepts to be conducted in both the Semiscale and LOFT integral test j
facilities (see 3.1.5 Integral System Behavior and 3.2 Alternate ECCS).
2.3 Reactor Transients i
t In September 1973 the Regulatory Staff documented its licensing position on Reactor Transients in WASH 1270: Anticipated Transients Without Scram (ATWS) for Water Cooled Power Reactors. This i
document-identifies a series of postulated transients which could occur and states that "an analysisshouldbemadeoftheconsequencesof(these)anticipatedplanttransientsintheevent of a postulated failure to scram." The vendors are required to supply the limits for the following:
a.
Reactor Coolant System Pressure The maximum acceptable calculated transient reactor coolant system pressure should be based on the system boandary pressure limit or the fuel pressure limit, whichever is more restrictive:
(1) peactor Coolant System Boundary Pressure Limit The calculated reactor coolant system transient pressure should be limited such that the maximum primary stress anywhere in the system boundary is less than that of the
" emergency conditions" as defined in the ASME Nuclear Power Plant Components Code,Section III.
1 j
(ii) Fuel Pressure Limit w
The calculated reactor coolant system transient pressure should not exceed a value for which tests and analyses demonstrate that there is no significant safety problem with the fuel.
't i
A-6 t
.--a
.,,,,-----,n,,
---,--,---e
-e
b.
Fuel Themal and Hydraulic Pe'rformance (1) The calculated average enthalpy of the hottest fuel pellet should not result in signif-icant cladding degradation or significant fuel melting.
(ii)'; A calculated critical heat flux event should not occur.unless the calculated peak cladding temperature can be shown not to result in significant cladding degradation, c.
Containment Conditions Calculated maximum containment pressure should not exceed the design pressure of the con-tainment structure. Equipment located within the containment that is relied upon to mitigate the consequences of ATWS should be qualified by testing in the combined pressure, tempera-ture, and humidity environment conservatively predicted to occur during the course of the event.
Experimental confirmation of the ability of the current calculational tools to predict such limits shall be provided through an experimental program.
9 3.
RESEARCH PROGRAM A description of the existing and planned safety research programs in the three areas previously discussed is presented in the following sections.
3.1 LOCA/ECCS 3.1.1 Reactor Core Blowdown Flow, Heat Transfer and System Effects t
To satisfy the objectives contained in Section 1.1.1, programs have been initiated to investigate l
blowdown heat transfer under a spectrum of PWR and BWR conditions. The requirements established for each of the major programs are similar in a number of respects:
pressurized facilities are utilized to provide the coolant pressure and temperature condi-tions representative of EWRs and PWRs.
the length of heater rods, the rod diameters, and coolant channel spacings utilized in the test facilities are the same as in BWRs and PWRs.
sufficient eledrical power is available at the test sites to pemit initiation of blowdown tests with 49-red bundles operating at power levels corresponding approximately to that of similar 49-rod bun 11es in PWRs or BWRs.
principal interest in the large facilities is in studying thermal transient behavior of the rod bundle during blowdown.
each program studies the initial phase of the hypothetical LOCA, prior to the availability of ECC flow.
A-7
The PWR-Blowdown Heat Transfer (PWR-BDHT) program will be conducted at HNL, and the BWR-BDHT
~
program is underway in facilities located at the GE-San Jose, California, site. Principal differences in the two programs are:
PWR systems differ among manufacturers and no attempt has been made to provide a scaled PWR representation with.the testing system used at HNL. However, this loop facility is capable of providing a range of _ test conditions.
For BWR tests, a test facility scaled to a BWR is provided at the San Jose facilities.
The PWR-BDHT experimental-programkhich utilizes 49-rod bundles is intended to study the LOCA i
[
relationships among the principal reactor variables'which can alter the rate of blowdown, the presence of flow reversal and re-reversal, the time delay to CHF, the rate at which dryout progresses radially and axially along the bundles, and similar time related functions which are:
important to accident analysis. HNL estimates indicate that each additional second of nucleate boiling during depressurization could afford a delay of about 10 seconds in ECC injection, with-out exceeding prescribed temperature limits on the fuel cladding during a LOCA. The HNL 49-rod experimental program will be conducted within an experimental matrix which explores separately each of the variables over a range of conditions of interest to the NRC in analyzing the blowdown j
response of PWRs. Some of the principal variables are listed below:
Coolant inlet temperature
'F 540 to 560 Coolant enthalpy rise, Btu /lb,*
50 to 150 i
6 Coolant mass velocity, Ib,/hr ft 0.6 to 3.0 x 10
. System volume / power ratio, ft /MW
>3 l
System Pressure, psia
- up to 2250 Blowdown receiver pressure, psia up to 75 2
Break area, ft up to 0.03 Break location inlet, outlet varied flow split l
The types cf tests to be conducted will include:
steady-state tests which systematically examine the onset of critical conditions as loop 4
and power parameters are varied.
4 blowdown with flow reversal and immediate power ciecay, using a range of initial power levels.
blowdown with flow reversal and delayed power decay, to account for the time delay between a
inlet pipe rupture and initiation of a PWR power transient.
e J
l
- Identifies variables defining the coolant thermodynamic state at the time of blowdown initiation.
1
)
A-8
blowdown from off-design (lower than rated) pressures, to consider effects of subcooling.
blowdown under loop conditions which will permit study of flow reversal and flow re-reversal.
blowdown using test bundles having a range of power profiles, including cosine, uniform radial power, non-uniform radial power, top skew and bottom skew.
A single-pin pressurized loop is to be utilized at HNL to provide supporting services to the larger-loop program (such as heater pin evaluations) and as a simpler geometry in which blowdowns can be conducted to test instrumentation techniques and analysis methods. This loop also pro-vides the potential for performing tests which would compare the perfomance of different heater designs in a comon loop environment. For example, two principal techniques are utilized in fabricating electrical heaters to simulate nuclear fuel pins. The direct " skin heating" technique utilizes a voltage drop directly across the vertiaal length of the sheath material. When utilizing this technique, thermocouples cannot 'r applied directly to the sheath-clad, but must be attached through an insulator. Heaters of this type are utilized in the BWR-BDHT program.
The indirect method utilizes an internal heating filament and themocouples may be attached directly to the internal or external surface of the sheath which surrounds, but is insulated from, the filament. Heaters of this type are utilized in the Semiscale program. The HNL heaters will contain an indirect cylindrical heater which is surrounded by an external unheated sheath.
Thermocouples will be attached to the inner surface of the external sheath.
Utilization of the single-pin loop at HNL to compare different heater types under sinilar blow-down conditions would pemit:
an identification of the differences in themocouple readings associated with attachment methods and internal insulation.
the ability to compare predictions of clad temperatures as a function of electrical power added, versus the measured values.
the ability to ensure that the different methods of electrical programming of nuclear decay heat utilized to represent nuclear pin power will match the calculated nuclear pin surface temperature in PWRs.
the ability to develop relationships describing blowdown heat transfer dependence on heater length and axial heat profile for use in analysis of Semisca4, LOFT and PWR response.
The BWR-Blowdown Heat Transfer Program will investigate BWR system response to a postulated LOCA and incorporates BWR features such as internal jet pumps, steam separator, a representative core bundle, etc., and is intended to represent the thermal hydraulic performance which is characteristic of the BWR under both nomal and postulated LOCA conditions. The test apparatus volume has been distributed proportionally to BWR region volumes, all BWR hydraulic functional equipment which might significantly influence BDHT are included, and all significant BWR flow paths during both normal and postulated LOCA conditions are present to the extent possible in the scaled two-loop test apparatus.
A-9
Significant major parameters to be investigated in the BWR-BDHT program for hypothetical re-circulation line LOCA's include: bundle power, break size, bundle power decay, bundle bypass
)
flow area. pump coastdown time ar.d lower plenum geometry.. Steam line breaks will also be examined. I Consideration is to be given to BWR-BDHT program extensions which would include skewed bundle profiles. Zircaloy clad. 8 x 8 bundles and BDHT and ECC reflood interactions.
Each of the major BDHT programs contains a companion analytical effort utilizing RELAP-4. By comparing RELAP predictions with results from selected tests, and understanding the reasons for differences observed, a basis will be established for further refinement of analytical repre-sentations of the experiments. In addition to "best estimate" calculatisns based on the most realistic assumptions available, " licensing model" calculations are provided which are based on assumptions contained in licensing criteria. A comparison of results from the two types of cal-culations provides a measure of the degree of conservatism of the licensing models.
Specialists in heat transfer, under the jurisdiction of ANL, are to be utilized to review facility design, instrumentation methods and test results. Based on the infonnation available, including heat transfer coefficients produced from the major programs, correlations are to be developed which will incorporate the best estimates of PWR and BWR core response during the blowdown phase of a LOCA. A program was recently initiated at Lehigh University to model transient transition and film boiling heat transfer.
During 1975-76, the Semiscale facility at INEL is to be utilized in support of LOFT Core-I (5.5 feet in length). Blowdown heat transfer relationships developed from Semiscale tests will provide a basis for the LOFT best estimate calculations during the blowdown phase.
No adiitional multirod bundle BDHT tests are considered nccessary beyond those planned at the GE, HNL, and Semiscale facilities. Additional smaller scale tests under ANL cognizance have been incorporated to improve the data base for " microscopic" or local heat transfer data for a single pin or single channel.
Initial Results frum BWR-BDHT Program
- 1/74 Initial Results from PWR-BDHT Program
- 9/75 Initial Results from Semiscale, 5.5-ft heaters
- 7/75 Initial Results from Semiscale, 12-ft heaters
- 6/77 Complete BDHT Correlations
- 12/77 3.1.2 Reactor Core Reflood Flow, Heat Transfer and System Effects To satisfy the objectives contained in Section 1.1.2, reflood programs such as the Scoping Emergency Cooiing Heat Transfer (SECHT), and Full Length Emergency Cooling Heat Transfer (FLECHT) were initiated to study the interaction between ECC water and heated pins in simple, multi-rod geometries. The FLECHT program was conducted in both BWR (spray and flooding) and PWR (flooding)
A-10
modes. Both of the FLECHT programs utilized controlled ECC flows and studied variables such as heater bundle power, ECC flow rates. ECC water temperature and system pressure over ranges calculated to represent or bracket BWR and PWR ECC and core conditions following a postulated LOCA. Because of postulated interactions among system geometry, ECC flow rate, and core decay heat during PWR reflooding, the FLECHT System Effects Test (FLECHT-SET) was built to incorporate scaled or simulated system components such as steam generators, broken and unbroken loops, and downcomer.
This new configuration was attached to the existing FLECHT bundle housing and upper plenum.
Detailed examination of the FLECHT-SET facility design by an NRC Task Force plus the results of initial testing in FLECHT-SET have indicated that FLECHT-SET, as designed, cannot meet all of the established program objectives. A major problem was created by the presence of an oscillatory l
reflooding behavior which, although likely to enhance reflood heat transfer in the experiment, is not currently considered to be representative of PWR behavior during a LOCA. Causes of the oscillatory behavior are traceable primarily to the 1-D nature of the test facility. Other effects of atypical condensation causing substantial pressure drops in the system were also observed.
It is believed that for a PWR LOCA, any local reaction force caused by steam generation in the hotter regions of the core would be distributed radially as well as axially and would not result in sustained flow oscillations. Therefore, plans for FLECHT-SET were revised to provide piping modifications and a revised approach which includes:
additional forced feed ECC tests, including flooding rates at and below 1 in/sec, with and without blockage.
an improved understanding of how system variables such as upper plenum volume, steam generators, upper support plates, and containment pressure affect entrainment.
hot bundle vs. average bundle studies.
the use of skewed, as well as cosine, axial power profiles.
Operation in this revised mode will provide a better understanding of how system variables affect core cooling at fixed flooding rates, and is expected to provide nore useful infomation for the refinement of 1-D computer models such as the FLOOD code utilized in conjunction with RELAP-4.
Longer range plans include development of a 3-D analytical capability which will provide a basis for calculating the interactions among ECC reflood flow, system components and core heat transfer.
This 3-D analytical capability is considered to be the end product of reflood test infomation from FLECHT, FLECHT-SET, and Semiscale, with additional larger radius test infonnation to be provided by LOFT.
Work is presently underway at ANC, LASL, and at PNL to develop a capability for calculating reflood transient behavior in multidimensions.
A-11
Milestones Initiate testing in revised FLECHT-SET facility
- 12/74 Initial reflood results available from Semiscale
- 12/75 Initial reflood results available from LOFT
- 6/77 Completion of 3-D reflood model development
- 6/77 3.1.3 ECC Bypass and Steam-Water Mixing The overall program approach is a combination of phenomenological model development and supportive small scale experimentation leading up to large scale experiments in PFE. This can be displayed as follows:
Experimenter (s)
Modeling ECC. Injection Region CE & PNL BCL & Creare Downcomer and ECC Bypass CE, BCL, Creare & PNL BCL & C-eare Lower Plenum Entrainment BCL, Creare & PNL BCL & Creare The intent is to develop correlative models from small scale tests and attendant phenomenological models where possible so as to predict ECC bypass and penetration in large systems. Additionally, ECC injection region models will be pursued further to evaluate the pressure oscillation effects on LOCA analyses. Models will be based on first principles where possible and correlative models will be adjudged against the data obtained from 1/15,1/5 and 4/5 scale experiments. The data will be made available for LOCA code incorporation in the form of best estimate correlations and predictive models.
Milestones CE's 1/5 and 1/3 scale steam relief test data have been reported in reports CENPD-101, 127, 128 & 129. BCL has published an ECC injection AP model in BMI-1931.
CE's annulus penetration data (single intact leg, 1 broken leg simulation) is reported in CENPD-130. 1/15 scale model testing at BCL confirmed the results obtained in the CE 1/5 scale tests and established a scaling point in FY-75. BCL will complete (in FY-75) the 1/15 scale, multiple leg ECC downcomer penetration tests utilizing heated water and steam upflow and ECC injection with steam water mixing plus steam upflow tests (at near atmospheric con-ditions), thereby establishing an ECC downcomer penetration correlation. The results will be used to initiate "best estimate" models for predicting ECC lower plenum penetration.
Creare will complete (in FY-75) in a flat plate test apparatus, multiple leg ECC penetra-tion tests to evaluate delay times induced by heated walls. Creare's experiments will parallel BCL's experiments, but will concentrate on transient penetration behavior and " hot wall" delay times through a series of experiments utilizing multiple leg hot water injection, followed by multiple leg ECC - steam mixing plus steam upflow in the downcomer region.
A-12
The PFE program, which is scheduled to commence testing in September 1977, is the major pro-gram for obtaining the scaling and system feedback effects on ECC bypass, downcomer flow and lower plenum entrainment data. An initial downcomer penetration model (based on PFE scaled experiments) is now scheduled for June 1978, followed by a refined model in December of 1978. Large scale model verification via tests in the large vessel will not occur before December of 1978. Lower plenum entrainment models and ECC mixing models will follow a schedule similar to the downcomer ECC penetration modeling schedule.
3.1.4 Reactor Coolant Pump Transient Characteristics Current program approach is to utilize data obtained from the EPRI-CE cooperative program.
Current EPRI-CE plans call for testing a 1/5 scale CE pump in a modified CE loop the first half of CY-76. CE's current approach is predominantly a two-phase pump perfonnance steady state characteristics testing program, followed by a limited amount of transient pump tests.
Milestones EPRI-CE 1/5 scale pump testing early or first half of CY 1976, data becoming available third quarter of CY 76.
Verification of, or revision of, current pump head degradation model and development of torque characteristics model in first half of CY-77.
3.1.5 Integral System Behavior LOCA model's are developed from various scaled tests in an effort to establish the phenomenological dependence on size. This size dependence may or may not be continuous as the size changes. If the size effect is not continuous, extrapolation of separate effects test results from small scale to prototype size becomes difficult and further testing at different size or by variation of test parameters may be required.
Through the application of analysis and engineering judgment, phenomena expected in the prototype can then be properly modelled and understood. An integral system test is one in which the entire system is represented in order to combine all of the separate effects phenomena in their proper relation to one another. Tests conducted in an integral system facility are designed to meet the following objectives:
Verification of LOCA System Predictions: Integral tests will provide data to ensure the code adequacy for predicting the integrated coupling effect of the overall system. The postulated LOCA transient behavior of components not specifically addressed in other areas of the Branch Plan will be assessed from integral system test data.
Identification of Unexpected LOCA Phenomena: Interaction between variables, not otherwise predicted from separate effects testing (such as flow discontinuity or hydraulic instabil-ities) which may be size dependent, may be revealed during integral system tests.
A-13
-~
Scaling Laws of Integral System in a Transient LOCA: Integral system test data is required in at least two sizes to confirm the applicability of current scaling laws of the system
-(such as power to volume ratio) and to demonstrate an understanding of integral system size
{
effects.
The following approach is being pursued to satisfy the objectives outlined in the Introduction.
First, separate effects tests will be conducted in Semiscale in the area of blowdown heat transfer, reflood heat transfer, steam water mixing and pump behavior. This information will be used to determine the applicability of previously developed correlations and models for pre-prediction of test results and to provide a basis for modifying these models if necessary. The resulting analytical description of the Semiscale system, supported by separate effects test data, will provide initial information regarding interactive effects which may occur during an integral test. Similarly, an analtyical representation of LOFT will be developed, based on separate j
effects tests conducted in various size facilities, to identify those interactive effects not identified in Semiscale. Even though these analytical representations of LOFT and Semiscale may i
be applicable only to these facilities, failure to develop such representations may result in interactive effects being masked by modeling deficiencies. Second, to identify interactive dependency on size, corresponding tet.ts will be conducted in both LOFT and Semiscale. Both facilities will be instrumented to allow a one-to-one behavioral comparison of each system component in each test. Third, the initial integral test sequence will be oriented towards experimental confirmation of the minimal effect certain facility design compromises (which were based on analysis) have on the overall system behavior in LOFT and Semiscale. Once the effect of these design compromises are understood, appropriate modification in either the experimental i
program, the analytical description, ot both will be made. The effect of each atypicality will be quantified and the departure from predicted LWR behavior dacumented. Fourth, resolution of these atypicalities will allow attention to focus on accident behavior other than DBA LOCA such as tests to establish the system response to:
A variation in break size and location during a postulated LOCA.
1 l
A LOCA calculated to have variations in average and peak power distributions corresponding to different periods in core life.
Existing ECCS and alternate ECCS concepts.
Semiscale Semiscale is a one-dimensional representation of a pressurized water reactor facility. The 1.6-MW core is 5.5-ft long and 7-in in diameter and contains 40 (electrically heated) fuel pins. The primary coolant system is designed with the same system elevations and volume to core power ratio as exist in LOFT. System subvolumes, e.g. inlet plenum, core region, etc.,
also are designed with relative volumes similar to LOFT. The unbroken PWR coolant loops are simulated by a single unbroken circulating loop in the Semiscale primary system and the i
postulated broken loop is simulated by a blowdown loop. A 12-ft core will be installed for later tests (1977). The water chemistry in Semiscale will be monitored.
A-14
LOFT LOFT is a 55-MW(t) pressurized water reactor facility intended to simulate the major behavior of generic 1000 MW(e) PWRs in carefully conducted loss-of-coolant experiments (LOCE). The nuclear core is approximately 5.5-ft long and 2-ft in diameter, and contains 1300 fuel pins and four control assemblies of typical PWR design. The primary coolant system is designed with a similar primary system volume to core power ratio as exists in typical PWRs. Primary system subvolumes, e.g., cold leg, core region, and hot leg also are designed with relative volumes similar to PWRs. The unbroken PWR coolant loops are approx-imately simulated by the single unbroken circulating loop in the LOFT primary system, and the postulated broken PWR loop is simulated by the LOFT blowdown loop having passive components.
Milestones Semiscale Mod-1 operation was initiated in late August 1974. The initial test sequence duplicated the non-nuclear test sequence schedule of LOFT. These tests are being followed in the Spring of 1975 by separate effects testing in the area of blowdown and reflood heat transfer. This will complete the separate effects testing in the Semiscale size with the exception of transient pump testing. Integral tests will be conducted in the beginning of CY-1976 in which current and alternate ECC injection concepts will be investigated and system response to small breaks assessed. A twelve-month test program to determine the effect of core length (12-ft versus 5.5-ft) will also include an assessment of active versus passive components in the blowdown loop on integral system response. Detailed break studies will be conducted following core length studies to investigate the system response due to: small versus large break, hot versus cold leg break, communicative versus non communicative type breaks.
The LOFT non-nuclear test sequence of five tests is expected to commence in November 19?5.
These tests will provide an information base from which core nuclear and heat transfer effects, steam generator heat transfer effects and primary system fluid temperature differential effects can be established. The non-nuclear tests will also establish base-line system behavior without ECC delivery to differentiate the effect ECC delivery has on
(
system behavior. This test sequence will last approximately four months.
l l
The follow-on nuclear test sequence in LOFT will proceed from low power to high power testing to minimize risk to the plant operation. The initial low power nuclear tests are expected to commence in January 1977 after core loading and initial criticality checks.
These tests will provide a basis for detemining whether the analysis is capable of predicting:
The effect of cold leg to hot leg temperature difference on the subcooled and saturated blowdown.
The effect of primary-to-secondary heat transfer, with core heat generation, on blowdown, pressure suppression system response, and reflood.
A-15
. The effect of increased hot leg temperature (relative to isothennal blowdown) on primary coolant pump performance.
The effect of nuclear fuel decay heat generation on blowdown, pressure suppression system response and reflood.
The integrated effects of nuclear decay heat generation, primary pump performance, metal-to-water heat transfer, steam generator heat transfer and pressure suppression system performance, on fuel integrity.
The effects of alternate ECC injection locations.
Infonnation on the behavioral response of the fuel under possible LOCA conditions will also be assessed during this test sequence which can provide test conditions for PBF LOCA testing. This test sequence will be followed by a series of high power experiments which will subject the nuclear core to conditions simulating a PWR full power loss of coolant. These are currently scheduled for initiation in September 1977. Testing of advanced ECCS concepts will continue during the high power nuclear test series.
3.2 Alternate ECCS The LOFT and Semiscale integral test facilities currently have the capability for injecting ECCS into:
Cold leg Hot leg Lower Plenum Additional ECCS concepts being evaluated for testing include 1) lower core support plate injec-tion in which the ECCS fluid path is directed through the core support plate flow 2) Upper plenum injection concept employing both slug and spray flow, 3) simultaneous upper and lower plenum ir.jection, 4) upper annulus injection and others. Parametric tests will be conducted in the Semiscale test facility to identify the most effective injecticn pressures, flow rates, and distribution method. An analytical assessment will be made for each configuration and postulated scaling criteria for extension to large PWps will be established.
Satisfactory confinnation of above criteria will be obtained from both analysis and data in the LOFT integral test facility. Testing will be conducted in LOFT to substantiate the analysis of the most effective injection methods identified in the Semiscale system. The ECCS scaling rationale will then be assessed as to its applicability to water reactors. At this point additional tests may be required to develop a) continuous model with scale or b) a model based on near full scale experiments. ReliabilityandredundancystudieswillalsobeconEfuctedtoensurethatthe ECCS system functions in a manner which will minimize the potential for connon mode failures.
A-16
Assessment of BWR alternate ECCS concept will also be conducted.
Milestones Preliminary alternate ECC injection concept tests are scheduled for initiation during FY-75 to 76 in the Semiscale facility with testing of lower plenum injection, upper plenum and simultaneous upper and lower plenum injection to be completed in early FY-77. Similar design modifications are currently being considered for incorporation into the LOFT facility. Initially, lower plenum injection will be tested during the non-nuclear isothermal tests scheduled for initiation in November 1975 and the low power nuclear tests scheduled for January 1977. Additional concepts will be tested in LOFT after initial assessment is made in Semiscale. Upon completion of the initial phase of the program, an assessment will be made as to the improved effectiveness and reliability of these concepts by the end of FY-78.
3.3 Reactor Transients The basic approach in treating reactor transients is to provide a sufficient data base to assess the coupling ability of the code to predict the combined thermal-hydraulic and neutronic behavior attendant to a particular off normal transient. These transients involve the nuclear charac-teristics of the facility, i.e., feedback conditions such as void coefficient,' Doppler coefficient, etc. For this reason, tests of potential off normal or accident conditions outlined in WASH 1270 can best be performed in the LOFT nuclear facility. Again, complete understanding and resolution of the atypicalities in LOFT will be resolved prior to the conduct of Anticipated Transients with belayed Scram tests. Tests under consideration in the LOFT facility are:
1.
Loss of Feedwater - These tests involve the loss of one or all of the feedwater pumps.
2.
Loss of Primary Flow - These test transients include the loss of one or both pumps.
l 3.
Primary system Depressurization - This transient covers the opening of the largest single safety relief valve in the primary system.
4.
Small line break - These transients cover the failure of an instrument, drain or sampling line connected to the primary system.
5.
Rod Withdrawal - transients in this category include control rod withdrawal from hot critical condition or from full power.
Such tests are being considered for performance during the high power nuclear series in LOFT.
Milestones Preliminary results from Anticipated Transients with Delayed Scram testing in LOFT will become available by August 1978 with final results available by August 1979.
A-17
g q
APPENDIX B BRANCH PROGRAM PLAN FOR FUEL BEHAVIOR BRANCH 1.
OBJECTIVES The program of the Fuel Behavior Branch is directed at providing a detailed understanding of the response of nuclear fuel assemblies to hypothetical off-rormal or accident conditions. This understanding is expressed in physical and analytical correlations which, under the program of the Analysis Branch are incorporated into computer codes. These will be available to Regulatory agencies in licensing nuclear reactors.
The Nuclear Regulatory Commission requires that the safety of the public be assured through use of a " defense in depth" in design and operation cf nuclear reactors. Defense in depth includes the use of multiple barriers against any' escape of radioactivity from nuclear power plants. The first such barrier around the nuclear fuel is the cladding, which contains not only the fuel pellets but also the fission products which are generated along with the useful heat from the fission process. An understanding of the ccnditions, both internd and external, which influence the ability of the cladding to retain its integrity during postulated nuclear accidents is the prime goal of:the fuel behavior program.
Other barriers to escape of radioactivity include the primary reactor system and the containment building surrounding the reactor. In order to insure that unforeseen events affecting these barriers are not omitted in reactor design and s'afety analysis, several hypothetical accident sequences called " Design Basis Accidents" have been selected t,y the NRC for use in evaluation of designs of plant safety systems. The hypothetical accidents selected for study are nol based upon frequency of occurrence, for none have occurred and the probability of any actually happening is exceedingly small. The intent is instead to ensure tnat the situations considered are indeed extreme and inclusive.
The analysis of each type of accident must provide assurance that adequate safety features have been engineered into the plant to limit the occurrence and the consequences of any subsequent fission product release from the fuel, to ensure conformance with the guidelines of radiological dose specified in 10 CFR Part 100.
The hypothetical luss-of-coolant accident (LOCA) initiated by rupture of 4 large primary coolant pipe has been selected as the Design Basi [ Accident for evaluation of many of the safety features of light-water-cooled nuclear power plants. Other postulated accident sequences which would a#fect f jel element behavior, and which, therefore, must be addressed in this program, would also lead to reduction or total loss-of-coolant flow. This impairing of cooling is comr-aly called Power-Cooling Mismatch (PCM). PCri would result from loss-cf-coolant, over-power transients, reactivity B-1
initiated accidents (RIA) from such hypothetical causes as control rod ejection, and anticipated transients without scram (ATWS).
The emergency core cooling system (ECCS) is a principal safety feature installed to protect the fuel cladding from a hypothetical loss-of-coolant accident. The Acceptance Criteria for ECCS for light water reactors recently promulgated by the Commission are intended to ensure the effectiveness of the ECCS if it should ever be needed in maintaining the structural integrity of -
the cladding. Two of the criteria supply direct quantitative guidance to planning of the fuel behavior program:
peak Cladding Temperature. The calculated maximum fuel element cladding temperature shall not exceed 2200*F.
Maximum Cladding Oxidation. The calculated total oxidation of the cladding shall nowhere exceed 0.17 times the total cladding thickr.ess before oxidation. As used in this sub-paragraph total oxidation means the total thickness of cladding metal that would be locally converted to oxide if all the oxygen absorbed by and reacted with the cladding locally were i
converted to stoichiometric zirconium dioxide. If cladding rupture is calculated to occur, the inside surfaces of the cladding shall be included in the oxidation, beginning at the calculated time of rupture. Cladding thickness before oxidation means the radial distance from inside to cutside tre cladding, after any calculated rupture or swelling has occurred but before significant oxidation. Where the calculateu conditions of transient pressure and temperature lead to a prediction of cladding swelling, with or without cladding rupture, the unoxidized cladding thickness shall be defined as the cladding cross-sectional area, taken at a horizontal plane at the elevation of the rupture, if it occurs, or at the eleva-tion of the highest cladding temperature if no rupture is calculated to occur, divided by the average circumference at that elevation. For ruptured cladding tne circumference does 4
not include the rupture opening.
Subsequent discussion of these criteria in the document identified the need for additional information and improved analysis to narrow the statistical spread of exis' ting data and to relieve excessive conservatism which might have been adopted in setting the criteria. These data needs are related to:
a.
Oxidation of the zircaloy and its snbrittling effect.
b.
The estimated deformation of multirod clusters under the hypothetical accident conditions, c.
The stored heat calculated to be in the fuel at the start of the accident.
d.
The dependence of the fission product decay heat on time.
- e.
The effect of cladding swelling on the gap conductance.
Each of these concerns is addressed in Section 3 of this plan. In Section 2 the LOCA and the above-listed questions are not specifically discussed in order to avoid repeating the corre-sponding section of the Commission's document providir.g the Acceptance Criteria. The discussion B-2
of Section 2 is intended to suninarize the information presently available to describe the state of a fuel rod at the start of any hypothetical accident sequence and to follow the themal and mechanical state of the fuel iod in any situation which might result in a temperature excursion.
including LOCA's,'RIA's and other postulated accidents.
2.
PRESENT STATUS
~ 2.1 Fuel Performance 2.1.1 Introduction The condition of the fuel element at the time of initiation of the hypothetical accident would greatly influence the course of the accident. The principal initial parameters which must be known for analysis are the stored heat in the fuel, the gas pressure within the cladding, the extent of contact between the fuel and cladding, and the prior cladding strains. These param-thermal con-eters are interrelated and dependent upon a number of other properties such as 002 ductivity, UO themal expansion, UO cracking and restructuring, gap width, fission gas release, 2
2 i
. cladding creep, and fuel rod ratchetting. It is therefore necessary to understand the current trends in fuel design and to discern either by analysis or experiment their influence on the above mentioned parameters. Recent examples of such trends are the prepressurization of the fuel rods, improved stabilization of pellet density during irradiation and changes in fuel rod diameters instituted by all reactor vendors in the past year.
t 2.1.2 Cladding Properties Most light water reactor fuel is clad with a zirconium alloy because of the low neutron absorp-tion cross-section of zirconium and good resistance to water corrosion at high temperatures.
Stainless steel cladding is still used in some of the older designs. Zircaloy-2, formed by the addition of small amounts of tin, iron, chrome and nickel to zirconium is widely used in boiling i
water reactors. Zircaloy-4, whose physical and mechanical properties are almost indistinguishable l
from Zircaloy-2, has reduced hydrogen pickup at the temperature and in the more chemically reducing water chemistry of the pressurized water reactor. The composition and minimum acceptable properties for reactor grade Zircaloy are specified by ASTM 353-72.
2.1.2.1 Texture and Heat Treatment The importance of texture and heat treatment on the properties of the Zircaloys has been recognized in recent years and has been the subject of numerous out-of-reactor tests and a few conducted in-reactor. The highest strength and best ductility under biaxial loading is associated with a texture where the basal poles are oriented parallel to the radial direction of the tubing. This texture may be determined by x-ray techniques but it is most commonly obtained by requiring that the orientation of zirconium hydride platelets formed by adding hydrogen to the tubing be parallel to the circumference of the tubing. Each tubing vendor has a proprietary means of producing the desired orientation. The strength and ductility of Zircaloy of a given texture may be varied over an extensive range of values by controlling the degrees of cold work B-3
and stress relief. There is no industry wide consensus of desirable properties and each vendor selects on the basis of desired performance. In practice, irradiation effects tend to bring the properties toward common values and no definitive evidence has been advanced that one set is superior to another in an accident situation.
2.1.2.2 Mechanical Properties of Unirradiated Zircaloy The tube-mechanical properties that are of importance include yield stresses, ultimate tensile stress, total elongation, uniform elongation, reduction of area and thickness at fracture, fatigue endurance, creep rate (tensile and compressive) and collapse pressure. A satisfactory data base exists for most of these properties except creep rate and collapse pressure for the most common textures and heat treatments up to about 600*C. A limited number of untaxial tensile tests and biaxial burst tests have been conducted at 1000*C. Measurements have been made up to about 600*C of the effect of oxygen on ductility in Zircaloy, but little of the data has been obtained on specimens with homogeneously distributed oxygen. There is a need for additional mechanical property data obtained under well defined multiaxial stress conditions as a function of strain rate, and oxygen content up to 1300*C.
Hydride fonnation causes considerable loss of ductility in tests conducted at room temperature and makes the material sensitive to shock loading during handling. Above about 200*C the embrittling effect of hydride precipitates is reduced, and at reactor coolant temperatures uniformly distributed hydrides are not considered deleterious. Localized hydride formation (so called sunburst failures) caused either by a high internal source of hydrogen or moisture entry through a defect causes blister formation and ultimately, under operating stresses, the hydride cracks leading to thru-wall penetration.
2.1.2.3 Mechanical Properties of Irradiated Zircaloy Test information obtained either in-reactor or on post-irradiation tests of irradiated material is available for most of the properties listed in the previous section. The scatter of the data on irradiated materials is generally greater, and as tests are more expensive, there are fewer data points for statistical evaluation. Until recently, most irformation was obtained at BWR operating temperatures or lower.
Irradiation results in reduction of ductility, a loss in impact properties, and an increase in strength. The yield strength of annealed Zircaloy-2 doubles and remains essentially constant 20fc,2 and the yield and ultimate strengths are effectively above a fluence of approximately 5 x 10 21 coincident. The consensus of opinion is that strength changes tend to saturate between 10 and 22 20 2
10 n/cm2 (> 1 MeV) but that the effect on ductility saturates between 3 x 10 and 10 n/cm
(> 1 MeV). The strain at which plastic instability sets in is reduced with increasing irradiation.
20 Most of the work performed in this area involves relatively low integrated fast fluxes (< 10 2
n/cm ; E > 1 MeV) and temperatures ranging from 300-600*C. There are some data taken at irradiation temperatures as high as 750*C.
In all of the low fluence studies made to date, the authors conclude that higher irradiation temperatures do result in less radiation damage.
B-4
' However, some authors show evidence that, while this is true at low fluences, there is a cross-20 2
over of properties in the range of I-2 x 10 n/cm, above which the yield strength is greater at the higher irradiation temperature. Furthermore, the damage at higher irradiation temperatures persists to higher out-of-pile annealing temperatures. This type of behavior has been found by others to be common to several other structural materials, most notably the stainless steels.
Preliminary indications are that since recovery increases significantly beyond 300'C, the temper-20 22 2
ature of maximum radiation damage for exposures in the range of 10 to 10 n/cm is somewhere near and below this temperature. However, the recovery anneals are conducted out-of-pile so the extension of those results to in-pile applications may not be warranted.
Most of the mechanical property data available are based upon uniaxial tensile data which have been shown to be not well related to the multiaxial stresses encountered in service. In one test series, decreased ductility was measured under biaxial loading as in a closed end burst test and even less ductility is expected under the 1:1 loading in a blister or where a fuel pellet has formed a ridge in the cladding. Most of the data were obtained at 300*C or less.
Burst tests on irradiated tubing conducted at high temperatures where the radiation damage is expected to anneal out appear to give results similar to tests on unfrradiated tubes. Irradia-tion greatly increases the tensile creep rate of Zircaloy. The majority of the of the published work on creep of Zircaloy has been on cold worked Zircaloy-2 in the stress range of 10,000 to 20,000 psi in uniaxial or biaxial tension and at temperatures around 300"C. There is very little information available in biaxial out-of-pile compressive creep (external pressurization) of Zircaloy over the range of stresses of interest to evaluate clad creepdown and none on compressive biaxial creep in-pile.
One limitation on the mechanical perfonnance of the cladding is the lack of ductility under the multiaxial stress conditions which the cladding experiences as a result of interaction with the fuel. This is not well understood under steady-state conditions and has been studied only a few times under overpower transients. Additional data are needed to determine the amopnt of con-servatism in the currently used strain-to-failure value of 1%.
The effects of the temperature gradient across the clad on the distribution of radiation damage have not been investigated. In the hottest zone of a peak fuel rod in a PWR, the temperature difference of inner and outer clad surfaces is typically 60'C.
Variations of the tensile prop-erties across the clad may possibly affect the failure mode. Irradiation does not appear to affect high cycle-low strain fatigue. Low cycle-high strain fatigue endurance seems to be strongly affected by irradiation due to the decreased ductility, but the data are quite sparse.
No data arit available on thermal fatigue or stress relaxation, both of which may be important P.echanisms in accelerating clad creepdown and clad collapse under irradiation and changing power loads.
2.1.3 UO Properties 2
U0 has become the basic fuel material for the nuclear power industry. It is used in the form 2
of cylindrical, sintered pellets with finished densities of 92-97% of theoretical. Most current usage, because of concern over densification, is at the high density end of the range. U0 is 7
selected over other potential fuel material because of its excellent combination of chemical B-5
stability, compatibility with Zircaloy, dimensional stability under irradiation and very high melting point. The properties of UO have been studied.for over 20 years and extensive irradia-2 tion experience has been accumulateJ.
2.1.3.1 Thermal Conductivity, The principal disadvantage to using UO is its low thermal conductivity. The poor conductivity 2
limits the power density attainable without developing unacceptable internal temperatures and results in steep temperature gradients within the pellet. An accurate knowledge of the UO 2 thennal conductivity is of importance since this property determines the temperature at each position within the fuel pellet and the temperature, in turn, is the principal variable governing all other physical and mechanical properties and all dynamic processes occurring in the U0
- 2 The thennal conductivity is almost universally expressed as the integrated conductivity of a solid cylindrical rod of UO, 95% of theoretical density between 0*C and the melting point. The 2
accurate measurement of the thermal conductivity in-pile or out-of-pile is a complex undertaking and has been the subject of many investigations and much controversy, particularly as regards the temperature range from 1500*C to the melting temperature, where thermometry becomes diffi-cult. Many of the differences between different views were settled by a panel of experts from the IAEA in 1965. The panel members were unable, however, to choose between two alternate evaluation techniques which yielded conductivity values differing by slightly over 10%. Since 1965, the lower, generally more conservative figure has been preferred by most evaluators, and this value of 93 watts /cm has been selected by the Regulatory staff for use in all calculations involving UO ttermal conductivity. Above about 500*C, irradiation has not been shown to affect 2
U0 thermal conductivity. Further work in the area of radiation effects would produce only 2
incremental improvements and is, therefore, accorded a low priority at the present time.
2.1.3.2 Fission Gas Release during irradiation has The release of the gaseous fission products, krypton and xenon, from UO2 also been studied for many years. The fractional release of the fission gases is known to remain quite low ('3-4%) as long as tne temperature of the pellet remains below grain growth temperatures. Very little gas is released during a temperature transient from U0 with low 2
burnup until approximately 1500*C is reached. As the temperature increases, and the size and orientation of the grains change primarily from fine to coarse and from random to equiaxed to columnar, the fission gas release fraction increases to 30-50% and then to nearly 100%. There is no discernible effect of burnup on fission gas release below about 15.000 MWD /MTU. Above this burnup level and particularly above 30,000 MWD /HTU there is growing evidence for an increased rate of fission gas release. A quantitative understanding of fission gas release has not ueen achieved, both because several mechanisms for describing the movement of the gas atoms within the solid are currently proposed and because the detailed operating history of the fuel rod which is the source of the experimental data is often not available. The empirical gas release correlations used for safety valuations adopt conservative expressions based on experimental data for fission gas release from fuel. Experience has often shown that the gas release from commercial rods is less than that predicted from post-irradiation gas release experiments. This is believed to be related to some combination of fission fragment e e-sulution of the gas atcss and lower than calculated operational fuel temperatures.
B-6 L
2.1.3.3 Fuel Swelling matrix to expand, ine generation of the solid and gaseous fission products will cause the UO2 and this expansion must be allowed for if the fuel rods are to attain high burnups. The volume expansion due to fuel swelling is accomodated by fabricating the pellets to less than theoretical density and providing additional width to the gap between the pellet and the cladding. Under normal irradiation conditions, fuel swelling is not a particularly important parameter and a nominal value of 0.5%.
- t. V/V per 1% burnup is used. If conditions are such that an appreciable l
volume of the fuel reached approximately 1000*C or higher, this position cannot be maintained, because swelling data from irradiations of fuel with plate geom.etry have shown that at higher UO operating temperatures, appreciable volume increases due to the gaseous fission iroducts are 2
initiated even at low burnup. As fuel swelling accompanies gas release, most of the mechanisms proposed to describe u s release also include a provision to describe the accompanying swelling.
The need for additional data applicable under steady state operation is not very great except as an adjunct to better understanding of the expansion (cracking and thermal expansion and swelling) which normally takes place in a pellet. There is, however, very little information available on the magnitude of the swelling during a temperature transient and on the visco-elastic properties of pellets when contained by cladding.
i l
2.1.3.4 U0 Densification p
Densification is thought to be augmented by neutron and fission fragment action, through in-pile sintering of active fuel. It is dependent on fluence. Tests have shown that densification can densification
)
be controlled by stabilizing the fuel. Several large in-reactor test programs on UO2 are underway, some in the U. S. and some in foreign countries. It is expected that through these programs UO densification will become well enough understood to preclude any need for 2
additional work. A reliable out-of-pile test for UO2 pellet stability which can be correlated with in-pile results is urgently needed. At least one research program to provide this correla-tion is in progress.
2.1.3.5 Sner Properties of U0 2 Other intrinsic parameters besides temperature and irradiation which affect the thermal con-ductivity of UO are either second order in importance or are well enough known that imediate 2
infonnation is not necessary. These parameters include density, melting point, change of stoichiometry, burnup coefficient of thermal expansion, heat capacity and certain mechanical properties such as Young's modulus and Poisson's ratio.
2.1.4 Fuel Rod Properties
?
A nuclear fuel rod comprising UO2 pellets enclosed in a Zircaloy tube has some integral proper-ties which cannot be derived separately from the intrinsic properties of the UO and the Zircaloy.
2 For example, the calculation of the stored heat in the fuel rod requires knowledge of the gap B-7
i conductance between the fuel and the cladding in addition to the knowledge of the thermal con-ductivity of UO and Zircaloy and the heat generation rate. Other such properties are the j
2 stresses on the cladding from mechanical interaction with the fuel and the gas pressure in the rod.
2.1.4.1 Gap Conductance With pressed and sintered pellets there is usually an appreciable resistance to heat transfer bet. ween the pellet surface and the cladding. The interfacial resistance may be the result of a gas filled gap or 00 in actual contact with the cladding. Several in-pile measurements of gap 2
conductance have been made and a large number of investigators have attempted to infer gap conductance from the examination of fuel rods which were irradiated for other purposes. Most of the reliable experiments utilized small diametral gaps. There is very little well characterized dath for thermal reactor fuel with large diametral gaps, especially in the 10 to 15 kW/ft operating power range. As the fuel burnup progresses, pellet cracking and relocation, pellet swelling and cladding creepdown combine to close the gap and to improve the gap conductance.
The rate of closure of the gap has been shown in some experiments to be dependent on such operating variables as rate of power increase, number of power cycles, and power level. There is need for additional investigation of these variables in tenns of their influence on the cracking and expansion of the U02 pellet to close the gap and on their degrading the gas gap conductance through enhinced fission gas release.
2.1.4.2 Gas pressure Knowledge of the pressure of the gas in the fuel rod is necessary for use in calculating the stress on the fuel cladding. The internal pressure is expected to increase with burnup, and at end of life, depending on the degree of prepressarization and internal void volume, it may approach the pressure of the coolant. The change in internal pressure depends, e.g., on fission gas release, helium absorption, pellet swelling, clad creepdown, and differences in coefficients of thermal expansion of the fuel and the cladding.
Although some data at low burnup have been obtained in-reactor, most information on internal g pressure of fuel rods during reactor operation is obtained by calculation from pressures measured on rods during post-irradiation examination. The uncertainty in the calculation is believed to be about 25%. This accuracy is satisfactory for current LOCA calculations where rapid internal equilibrium with plenum pressure and positive pressure differantials across the clad would be expected. However, the present uncertainty leaves too much unreliability in predicting the direction and mode of cladding deformation during hypothetical power-cooling mismatch or reactivity initiated accident because of the expectation that the pressure differential would remain small.
2.1.4.3 Fuel-Cladding Mechanical Interactions Mechanical interactions between fuel and cladding resulting in elastic and plastic deformation have been measured directly in-reactor as a function of rod power level, ramp rate and power B-8
2 cycling, for moderate burnups. These interactions result in changes in both rod length and diameter. A number of post-irradiation measurements of plastic deformation of comercial fuel rcJs resulting from fuel-cladding interaction have also been made. The radial defonnation results principally from the gap closure caused by pellet relocation, and at high fuel rod powers, from thermal expansion of the pellet against the cladding. A number of failures of fuel rod cladding have been traced to this interaction and the shape of most fuel pellets in use in current reactors has been altered to reduce the incidence of failure. Elongation of the fuel rod during power cycling can also be caused by this interaction. Although the contact is usually mechanical, instances of apparent bonding of the pellet to the cladding have been reported.
The bulk of the data which are available on dimensional changes to cladding and to fuel as functions of power has come from tests in which the fuel operated at high heat ratings and at low burnups. It is necessary to repeat these measurements at power rating more typical of current operation, and as a function of burnup.
2.2 Fuel Rod Performance During Abnormal Operation The only information available in this area has been obtained by conduct of small scale experi-ments in test reactors and by out-of-pile tests in which separate properties such as cladding deformation of fission gas release are measured. There are no known incidents durf ng operation of a light water reactor in which reactivity increases or coolar'. flod abnormalities have l
resulted in enough damage to fuel assemblies to impair continued operation.
2.2.1 Overpower Transients A few tests of cladding deformation following a rapid (5 min) power increase above steady state operation have been conducted on large diameter rods. The important parameters are cladding ductility and degree of overstraining. Power increases of 20% did not cause failure of rods with '17,000 MWD /MTU. An additional 30% increase in power to -20 kW/f t) caused failure within 5 minutes. The critical heat flux was not exceeded. Rods containing high density pellets with small gaps are most susceptible to this type of failure. Typically, cracks in cladding are oriented axially and start at ridge locations (pellet interfaces) on the inner cladding surface.
The ridges were not present prior to the overpower transient.
This type of failure, white it results in extensive hairline cracking and release of volatile fission products, does not seem to cause exposure of the fuel to the coolant.
2.2.2 Reduced Flow Tests in-reactor tests under BWR conditions during wnich coolant flow was reduced while the rod power was maintained have been reported by at least two laboratories. In the reported experiments, the rods were brought to burnout by valving off the coolant flow. The cladding temperature rose to a value between 600*C and 800'C and remained there for up to 5 minutes without failure. In later tests, different rods were subjected to 3, 33 and 60 burnout excursions respectively, over periods of several years. They remained in the reactor core without failure to exposures as high as 23,000 MWD /MTU.
B-9
In more recent experiments, a cluster of 36 fuel rods, four-ft in length, was subjected to dryout conditions. The fuel cluster experienced more than 120 such onset-of-dryout events at.
~
ratings of up to 900 W/cm. and also experienced nine post-dryout transients at ratings up to 780 W/cm(-24kW/ft). The maximum surface temperature recorded was 600'C and the integrated period of time under post-dryout conditions was not less than 15 minutes. The cluster was irradiated i
4 te 900 MWD /MTU before discharge. The most serious damage to the rods found during post-irradiation examination was a reduction in strength to the annealed condition.
2.2.3 Reactivity Initiated Accident Tests The pcssibility that a reactivity transient might be induced in a nuclear reactor by the sudden-removal of a control rod was probably the first safety concern to be identified. The TREAT reactor and a series of SPERT reactors were built to study this possibility. Experiments with these reactors led to very full understanding of the inherent stabilizing features of LWRs. Tests on Zircaloy clad fuel rods with UO2 pellets have been conducted in both TREAT and SPERT to determine the failure thresholds for such rods and the magnitude of any metal-water or fuel-water reactions which could result from a sudden increase in power generation. These tests provided initial conditions of cold startup and zero flow. Although the reactor periods for the SPERT and TREAT tests were very different (less than 10 msec and 39 to 238 msec respectively),
both gave similar results relative to energy deposition in the fuel at the point of failure. In single rod tests with unirradiated rods, the failure threshold was determined to be at an energy deposition of about 225 cal /g while prompt dispersal of the fuel began at energy depositions of s
about 340 cal /g. Transient tests were performed on a small number of irradiated rods with burnup 4
4 between 1000 to 32,000 MWD /MTV. The data exhibited considerable scatter but indicated that the threshold for failure was lowered somewhat. Transient pressures considerably larger than those encountered in the single unirradiated rod tests were also observed. The SPERT tests also indicate that the failure threshold for fuel rod clusters may be lower than for single rods.
The energy conversion ratio for a seven rod cluster was about 1.6 times that of a single rod, but it was still low. Based primarily on the data from these transient tests, threshold energy limits have been established for incipient fuel rod failure and prompt dispersal. The current criteria for satisfactory light water reactor perfomance for a rapid reactivity initiated accident sets an energy deposition of 280 cal /g as an upper limit to prevent prompt dispersal of the fuel.
With each successive generation of LWRs power generation and fuel burnup have increased and the margin between the design limits and failure thresholds has potentially decreased. Additional testing with irradiated rods and clusters is desirable to improve the statistical basis of the criteria and to include the simulation of transients from hot standby and power operating conditions.
2.2.4 Fission Product Release and Fission Product Decay Heat 5
Fairly accurate estimates of fission product inventory in a reactor core can be obtained from a i
knowledge of the coerating history. However, the amount and form of radionuclides released during a hypothetical accident can only be approximated by extrapolation of limited experimental j
data and application of highly simplified analytical models. Experimental investigations have 1
B-10
been conducted primarily out-of-pile on small samples (1 to 100 g) of relatively low burnup (trace to 4000 MWD /MTU) with rare instances of higher burnup (up to 20,000 MWD /MTU).
The study of release by diffusion exceeds that devoted to all other release mechanisms combined (melting,oxidationandvaporization). While early diffusion experiments were conducted almost exclusively in helium, more recent investigations have included air, steam and various gas mixtures. The fission products released from fuel rods following hypothetical clad rupture may be divided into four classifications. Based on the relative volatility in the elemental state, there are: a) noble gases (Xe, Kr), b) halogens (I, Br), c) volatile metals (Cs, Rb, Te) and d) non-volatiles. It is estimated that under conditions not involving excessive temperatures, clad rupture releases 3-5% of the total inventory of fission product categories a, b, and c contained within the fuel.
The fission product release from accident sequences which are calculated to lead to core melt-down is estimated using data from laboratory experiments involving melting U0 in inert 2
atmospheres and a few in-pile experiments in a variety of atmospheres. Under these conditions the noble gases and the halogens would show almost complete escape. The alkali metals and Te could also be expected to show nearly total release but release may be hindered by the lower volatilities of the pure metals or by formation of compounds. Oxygen activity in the melt would be a detemining factor in the release of alkaline earth, noble metals, and rare earths since the volatilities of the oxides and elements vary considerably. On a best estimate basis, the release of these elements from the melt would be 10% or less. If the distribution in yield of fission products is considered, a conservative estimate would then imply that less than 25% of the fission product inventory would be released from molten fuel. Releases of fission products heated in air at temperatures below 1500*C are similar to meltdown during the oxidation of UO2 releases with the exception of an almost complete release of Ru.
The uncertainty in chemical composition of a totally molten core and in the chemcial and physical properties of fission products at very high temperatures and temperature gradients have hindered the fmnation of detailed analytical models and have led to the use of conservative assumptions regarding fission product release. Experiments on the transport of fission products show that deposition along the path to the containment space would tend to reduce the release to the containment even further.
The fission product decay heat is calculated from the American Nuclear Society standard which is based upon work dating from the late 1950's. The uncertainty which has been estimated to be approximately 15%, is mostly due to short half-lived decay products. For this reason the ANS value + 20% has been prescribed by the Commission for use in calculating the decay heat term in a hypothetical LOCA.
2.3 Core Meltdown There is a very small probability for failure of engineered safety features which are provided to prevent melting of the core in the event of a LOCA or severe transient. The possible con-sequences of such an occurrence have been considered severe enough to warrant attention. The B-ll
.=
Reactor Safety Study, which recently focused on such " Class 9" accidents, has reportea that the probability of a core meltdown in present generation reactors is 6 x 10-5 / reactor-year but that -
its consequences would not be as great as previously thought.
2.3.1 Core Meltdown Experimental Review A' current program has evaluated the existing experimental data applicable to analyses of postulated core meltdown accidents. Key concerns include steam and hydrogen explosions, behavior of fission products, thermal-hydraulic behavior and the nature of complex molted masses. A final report is being organized so as to outline additional experiments which are indicated.
The principal finding of the report is that in general a sizable body of useful data exists but i
that the experimental conditions are usually such that the results are not always directly appli-cable to the core meltdown case. This is particularly true for steam explosions and molted core interactions. The report also suggests that sensitivity studies using state-of-the-art models of core meltdown should be conducted to determine the importance of physical phenonena relative to the overall consequences of the postulated meltdown accident.
2.3.2 Physical and Chemical Processes Analytical models have been used to predict the times required for molten mass to fom and to penetrate the primary vessel and containment building. The likelihood of breach of containment as a result of overpressurization or steam explosion at some time during the accident is considered to lead to the " worst case" condition. Uncertainties in the composition and high temperature physical and chemical properties of the melt as well as in thennal-hydraulic behavior have hindered the development of detailed analytical models. Recent experiments conducted in the Federal Republic of Gennany have begun to shed light on the nature of the physical and chemical behavior of the core.
2.3.3 Fission Products The behavior of fission products under hypothetical meltdown circumstances is important because they provide a principal heat source whose decay heat must be dissipated in order for the melt-down scenario to reach an end and their release represents a potential biological hazard to the environment, in these respects their rate of release from and their spatial distribution in the molten mass are critical. Studies have indicated that a major amount of release from the uncooled core would occur early in the meltdown when there is a large free surface area from which volatile fission products may escape. Other aspects of the release of fission products during core melt-down conditions are described Section 2.2.4.
Uncertainties in the behavior of a molten core and in the chemical and physical properties of fission products at high temperatures have led to the use of conservative assumptions regarding fission product release. Experiments on the transport of fission products show that deposition along the path to the containment space would tend to reduce the release to the containment.
B-12
2.3.4 Containment Considerable effort has been expended in recent years to examine the behavior of airborne fission products in the containment. Scaled facilities having up to slightly over 1% of the volume of the containment in a typical PWR have been employed to examine atmospheric conditions, airborne concentration, physical and chemical states of various species, mechanisms of removal of fission products from the airborne state, and thermal effects.
Particular interest has been paid tc iodine, and one outcome of these investigations has been the establishment of Regulatory Guidelines of Licensing specifying the partition of available iodine in the containment into discrete physical forms. Elemental vapor and active particulate forms of iodine are readily removed from the air by chemical sprays and filtering systems. Methyl iodide (CH I), other organic iodides and possibly hypoiodous acid (H0I) have been identified 3
as persistent airborne species and are conservatively believed to make up no more than 4% of the total iodine released into the containment. Fission product and fuel aerosols are effectively removed from the containment space by agglomeration and gravitational settling and also by incorporated safety features. Analytical models have been developed to predict the airborne concentration of fission products as a function of time for a given set of input data including fission product concentration, particle size distribution, containment a'tmospheric conditions and geometry.
3.
RESEARCH PROGRAM 3.1 Method of Approach The Fuel Behavior Branch works in close cooperation with the Analysis Branch to provide fuel behavior codes which will integrate smoothly with the advanced whole core codes being developed.
The Branch also is responsible for providing material properties data related to analytical and empirical correlations which are combined into mathematical codes to describe (a) the steady state thermal and mechanical properties of a fuel rod, (b) the thermal and mechanical response of fuel rods when subjected to postulated accidents, and (c) a model for the release and trans-(
port of fission products. Many of the correlatfor.s in current use are satisfactory and can be f
used without further development. One part of this program includes review of the physical and mechanical property values in current use, comparisons with the data base, and estimation of l
the statistical uncertainties. Where a value or correlation appears to be satisfactory and no additional data are needed, it is recommended for use in code development. In Section 2 the need for additional experimental data was indicated and in Section 1 data needs specifically identified in the Acceptance Criteria for Emergency Core Cooling Systems are mentioned. A second line of research provides for obtaining further experimental data and organizing them in a statistically valid correlation which can be recommended for use in code development. A third line includes the conduct of diagnostic experiments in such facilities as pBF or LOFT, in which single rods or clusters of rods are subjected to conditions similar to those postulated to be encountered by the fuel rods during a LOCA, RIA, flow blockage, or other design basis accident.
These experiments are important to assessment of the possible consequences of these hypothetical accidents as well as to serve as means of testing the progress of analytical code development.
B-13
o*
Program review groups made up of technical experts from the different Offices within NRC have been organized to review the progress of the program, to make recommendations of priorities and to review the completed correlations before they are released for incorporation into the codes.
Consultants from universities and national laboratories are also utilized in the review process.
3.2 Review of Material Prcperty Correlations The strategy of code development adopted by the Analysis Development Branch
- calls for any new code to model all of the phenomena which the completed code is expected to describe even though many of the models will be initially very crude. This makes it possible to conduct sensitivity analyses of the code predictions in relation to the constituent models and provides guidance as to the amount of development required of a given model. The Fuel Behavior Branch, after review of the literature and surveys of current usage, recommends appropriate correlations for :naterials properties for incorporation into the fuel behavior codes together with an estimate of the uncertwinty. New information developed in the experimental program will be compared with the existing correlations and improvements will be incorporated until evidence is derived from sensitivity analysis of the code that further development is not necessary.
Milestones Recommend interim correlations for cladding failure, strength, and ductility - June 1975.
Models for transient gas flow, models for steady state fission gas release - Preliminary December 1975 - Final December 1976.
Complete review and update of model for transient gas release and of measurement of gap conductance - June 1976.
Correlation for Zr oxidation - January 1977.
Complete verification and materials properties portion of steady state code (FRAP-53) -
j March 1976.
Complete verification and materials properties portion of transient code (FRAP-TS) - August 1978.
3.3 belBehaviorExperiments The experimental program is divided into six research areas.
3.3.1 Cladding Properties at High Temperatures The understanding of the progress, effects, and consequences of the oxidation and deformation of Zircalo cladding during a hypothetical LOCA (or other reactor event which would result in
- Since this was written the Fuel Behavior Branch has assumed responsibility for the development of the steady state and transient fuel codes named FRAP-S and FRAP-T, respectively. The approach and schedule are contained in the Analysis Development Branch Plan.
B-14
temperature increases in the cladding) must be improved to allow improved evaluation of the conservatism of the Acceptance Criteria for Emergency Core Cooling Systems for light water reactors, and to improve the statistical basis for the Criteria, as has been emphasized by the Commission. The objectives of the experiments to be conducted in this area are to determine
- 1) the effects of oxidation, rapid heating, and irradiation on the strength and ductility of Zircaloy in the beta phase, 2) the consequences of those effects on the properties of the Zircaloy cladding after the oxygen contaminated beta phase has transformed to alpha phase on cooling to the temperatures that would be expected during a LOCA reflood, and 3) the potential of methods or modifications that may be proposed to eliminate or alleviate the worst of the effects and/or consequences. The data detemined should also allow more adequate failure criteria to be deter-mined for cladding behavior than those now used and improve the data base for the models used in the steady state and transient codes.
3.3.1.1 Zircaloy 0xidation l
The rate of oxidation of Zircaloy will be determined in steam at temperatures between 1500*F and l
2600*F (from about the start of the formation of beta phase from alpha to the probable maximum temperatureofinterest). The program includes isothermal oxidation and oxygen diffusion studies under steam environments representative of LOCA behavior. The data will be used to predict the oxidation rate and oxygen penetration of the beta phase Zircaloy during thermal transients bracketing postulated LOCA conditions. The predictions will be checked by steam oxidation tests conducted under transient conditions with a variety of heating rates.
i Milestones The isothermal data should be obtained by September 1975.
)
Preliminary correlation based on isothermal data only available by April 1976.
I Scoping of the variables and obtaining the transient data should be completed by July 1976.
A verified correlation should be available by January 1977.
3.3.1.2 Mechanical Properties of Zircaloy Containing Oxygen a
The mechanical properties of Zircaloy will be detemined as functions of oxygen distribution 6
and content, strain rate, biaxial stressing, microstructural morphology, texture, and temper-atures over the range between 300*F and 2600*F. The strength and ductility of the Zircaloy cladding at any temperature are importantly dependent upon strain rate, oxygen content, micro-structural morphology, texture, and state of stress, factors that cre important in producing and controlling cladding deformation during LOCA and PCM type accidents. Zircaloy properties under thermal stress will be studied.
B-15
.... =
Milestonts_
The uniaxial mechanical properties of homogeneous, low o ygen Zr alloys completed and biaxial testing started - June 1975.
The uniaxial mechanical properties of high oxygen, homogeneous alloys, low total oxygen composite specimen data, and biaxial low oxygen alloy data detennined - June 1976.
Tentative embrittlement criteria and test method proposed - August 1976.
Mechanical property data determination completed - June 1977.
Clad embrittlement criteria and test method for critical review and examination for general acceptance recommended - June 1977.
Clean-up experiments and verification of correlation finished - June 1978.
3.3.1.3 Strength and Ductility of Irradialed Zircaloy Cladding as Functions of Irradiation and Temperature To the extent possible, cladding from spent fuel having various burnups and various textures will be tested. The post-irradiation examinations and measurements will help define a correla-tion for strain to failure of Zircaloy when burnup is high and failure due to mechanical inter-action is probable at relatively low temperatures. Tests will include tensile, bend, burst, and expanding internal mandrel at temperatures ranging from about 300' to 700*C. An in-pile test of strain-to-failure of Zircaloy will be conducted to compare with the out-of-pile data.
Milestones The isothermal and transient annealing temperatures for irradiated rods having commercia)
)
textures and exposures should be detennined by October 1975, by the tensile and bend tests.
and by the biaxial stress burst and mandrel tests by June 1976.
The data involving strain rates at the lower temperatures typical of power transients and with stress geometries produced by ridges and resulting from clad collapse should be available by December 1976.
In-pile tests should be completed by June 1977.
Verified correlations should be available by December 1977.
3.3.1.4 Cladding Deformation During Single and Multirod Burst Testing of Unirradiated Pressurized Zircaloy Tubing The defonnation and extent of flow blockage of the coolant channels of a fuel assembly require further study to reduce the statistical uncertainty of the present data and to evaluate the B-16 n
,-w
, ~. - -
potential of methods or modificati)ns that may be proposed during the study to eliminite or alleviate the worst of the effects or consequences. Experiments will be performed on electrically heated rods with flattened temperature gradients comparable to those in PWRs and BWRs. with internal pressures from 100 to 1800 psi, and heating rates to 100*F/sec. Single rods and clusters of 16, 64 and 121 rods with typical PWR grid spacings will be studied. Rod lengths will be based on grid spacings (about two ft.).
The data from the single rod tests will be compared to the data from the cluster tests, and the development of a correlation attempted. The ccrrela-tion method should allow the prediction of multfrod perfonnance from single rod tests, and greatly decrease the cost of evaluation of various cluster configurations and cladding modifications.
Milestones Data from 16 to 64 rod clusters of PWR tubing should be obtained by October 1977.
Data analyzed and data obtained on the 121 rod clusters by July 1979.
The study should be completed on PWR rods by July 1979.
The study for BWR (scoping tests only) completed by July 1978.
A correlation cor.sidering rod interaction, scaling factors, flow blockage, heating rate, initial and burst pressures, etc., should be available by June 1978.
The comparison of the experimental results and the correlation with predictions made with data from the other studies (1.2,1.4,1.6, etc.) will be completed by December 1978.
I 3.3.1.5 Burst Te sts of Irradiated cladding The limited data available show that irradiated Zircaloy tubing exhibits less deformation during burst testing than does unirradiated tubing. Single tubes of Zircaloy (irradiated with and without fuel pellets) will be burst tested under conditions of pressure and* heating rate similar to those used with the unirradiated data. These tests should separate the effects of radiation damage and pellet-clad mechanical interactions on the deformation of the cladding.
Milestones The tests on material from one lot sho11d have been completed by June 1976.
The analysis should be completed by September 1976.
Testing and analysis on material from a second lot should be completed by December 1976.
Testing and analysis of tubing for a third lot should be completed by June 1977.
A comparison of data on irradiated and unirradiated tubing should be completed by December 1977.
B-17
3.3.1.6 Cladding Creepdown and Collapse True stress-true strain measurements and separation of thermal and radiation effects on creep and rate of creep of Zircaloy are needed to improve cladding creepdown models and understanding of the mechanism by which a bamboo structure develops. These are important in the interpreta-tion of the effects of gap conductance and gap closure as deduced from in-pile experiments and from performance of commercial fuel.
Milestones Technique development for anisotropic Zircaloy will be started in FY 1975 and major deter-minations of uniaxial data should be available by June 1976 and completed by June 1977.
Technique development for biaxial testing will start in FY 1977.
Testing and analysis of experiments on rod collapse should be completed out-of-pile by October 1975 and in-pile by September 1977.
Compressive creep studies conducted out-of-pile should be underway by December 1975 and completed by December 1976.
Compressive creep studies conducted in-pile should be underway by September 1977 and completed by June 1978.
3.3.2 Fuel Properties Experiments to be conducted in this category are expected to provide additional experimental infomation on changes to fuel pellets in steady state and transient operation. Emphasis has been placed initially on obtaining infomation on fission gas release and fuel swelling during temperature transients, as a function of burnup, and on studying the cracking pattern and expansion of pellets during startup. The information will be used to improve models for gas release in the steady state and transient codes, and the pellet expansion portion of the gap closure routine of the steady-state code - needed for stored heat calculation.
3.3.2.1 Transient Release of Fission Gas An out-of-pile experiment using an electrical heating technique which provides a temperature gradient similar to that in a fuel pellet in a reactor is proposed. The rate and amount of gas released, the distribution of gas within the pellet,,and the change in volume of the pellet will be measured as a function of burnup to 40,000 MWD /MTU.
Milestones Basic experiments completed by January 1976.
Preliminary equation for gas release ready by June 1976.
B-18
3.3.2.2 Transient Gas Release - In-pile An in-pile instrumented experiment with 24 test rods will be undertaken. Information will be obtained on absorption of He and release of fission gas during an overpower transient following steady state operation, to obtain transient release of fission gas and data on fuel swelling.
Milestones Experiment installed in reactor - May 1975.
Remove 12 rods for the absorptfor. measurement - September 1975.
Interim sunnary of data on release of fission gas - June 1976 Transient overpower test - February 1977 Verified correlation for gas release - June 1977 3.3.2.3 Fuel Pellet Cracking Pattern Information on the relative contribution to expansion of fuel pellets from thermal expansion, cracking, and fission product swelling is needed to improve the gap closure models in the steady state code. An out-of-pile experiment using electrical heating to observe cracking patterns of UO2 and pellet expansion during power cycling is proposed. Realistic temperature gradients will be imposed upon the pellet by self-heating.
Milestones Completion of experiments - June 1976.
l.
l Completion of preliminary correlation for cracking - December 1976 Comparison with in-pile experiment - June 1977.
3.3.3 Fuel Rod Properties The experiments conducted under this category have as their objective the improvement of the models for calculating gap conductance in a fuel rod, to determine whether expansion of fuel pellets and bonding of pellets to the cladding affects the axial flow of gas within a fuel rod.
3.3.3.1
_ Gap Conductance The knowledge of the gap conductance is a major factor in calculating the stored heat in' the fuel pellet at the onset of a hypothetical LOCA. The heat transfer across small gaps is poorly characterized and is apparently higher than current analytical models predict. Out-of-reactor experiments include extension of the Ross-Stoute data on contact conductance of Zircaloy-UO2 B-19
.)
interfaces to higher interfacial temperatures and gas pressures. Parameters to be varied include j
ghs composition and characteristics of the fuel-clad interface. The effect of eccentricity of fuel pellets on gap conductance will also be studied experimentally
- In-pile instrumented experiments will be conducted to provide measurements of initial gap closure, gap conductance and fuel temperature. Three gaps calculated to be closed at pre-determined power levels will be selected. Rods will be irradiated in groups of three to about 5000 MWD /MTU.
A long term irradiation of six rods will be initiated, with temperatures of fuel and cladding measurement. This experiment will provide a life history of well characterized rods out to approximately 25,000 MWD /MTU. These rods will have operated nominally at 15 kW/ft with periodic operation at 10-12 kW/f t.
Post-irradiation examination of the rods will provide additional data for the gas release and fuel swelling correlations. The experiment should serve to, verify the code used to calculate properties of fuel in steady operation. A third series of brief irradia-1 tion experiments will be conducted in which a parametric determination will be made of the effect of pellet density, gap, composition and pressure of the fill gas, and power level. These experiments will provide additional data for verification of gap conduction models over a wider j
range of variables than the high burnup test.
1 i
Milestones Begin in-pile measurements of gap conductance in PBF - February 1975 Complete measurement of contact conductance - June 1976,
]
t Recommend revised correlation for contact conductance - September 1976.
Begin measurement of gap conductance in the Halden reactor - May 1975 Preliminary data on gap closure measured in-reactor - August 1975.
)
1 Begin high burnup irradiation test - September 1975.
l l
i Revised correlation for gap closure - September 1976.
4 Verified correlations for variation of gas pressure, gas composition, rod power - June 1977 i
Compile verification of steady state code - July 1978 3.3.4 Integrated Tests of Fuel Models Under Accident Conditions Tests will be conducted initially in the Power Burst Facility and later in such additional 1
I facilities as ETR and LOFT, to improve understanding of response of fuel rods under experimental conditions in which fuel is expected to fail. These tests will permit more detailed examination j
than the current store of information permits of the mechanisms of failure of fuel rods and of f
possible propagation of failure from one rod to another. With a more realistic view of the 1
i l
B-20 1
1
,..--n.....-.-.,~-n,-,
.____,_..___,.v_.
,wn.,,-,,
accident sequence, some of the present conservative assumptions in the analysis may be replaced by others that are closer to best estimates. The features of experiments that will be performed on fuel rods in these reactors are not similar in all respects to those found in large reactors.
Several of the experimental parameters are not identical to power reactor conditions, e.g., -
length of fuel (PBF accepts 3-ft rods, LOFT takes 5.5-ft rods), certain conditions of the coolant, and the neutron spectrum. The experimental data will be used to verify analytical models of describing behavior of fuel under accident sequences. Once these models have been verified, they can be incorporated into codes that will then provide greater confidence and understanding in application to calculating the consequences of postulated accidents in large power reactors.
Tests to be conducted before June 1978, when the first core of PBF will be replaced with a core that will permit higher power levels, include: power-cooling mismatch, flow blockage, LOCA blowdown and heatup (multirod) and RIA. Over 60 tests requiring nearly 400 fuel rods are planned.
The tests will be applicable to models for stored heat in fuel, internal gas pressure, Zircaloy oxidation and strain to failure, fuel-water interacticn, and rod-to-rod propagation. Experiments with the second core of PBF and with LOFT will focus on verification of models for large clusters, including the effect of burnup on propagation of failure from one rod to another during LOCA and flow blockages. Clusters of 64 BWR-size rods and approximately 120 PWR-size rods can be accomodated in the PBF. LOFT testing will include a still larger cluster 5.5-ft long.
Milestones Initiate:
Tests on power-cooling mismatch - May 1975.
PCM tests with irradiated rods - September 1975.
Tests to determine parameters of importance in a PCM - August 1976.
LOCA heatup,16 rod cluster - July 1976.
Flow Blockage Tests - February 1977.
LOCA Blowdown - June 1977.
Reactivity Initiated Accident tests (single rod) - February 1976.
PCM cluster tests - September 1976.
RIA cluster tests - June 1977.
Complete initial PCM, BWR and LOCA heat up testing program - December 1977.
(Rods containing Pu will be interspersed with PCM-RI A tests.)
B-21
3.3.4.2 plutonium Recycle Tests will be conducted to determine the effects of the use of recycled plutonium fuel on the course of LOCA, RIA and PCM type accident sequences. The physical and mechanical properties of UO c ntaining 3-4% of added plutonium vary only slightly from those of UO. However, small 2
2 variations in sintering rates and cracking patterns of fuel pellets may alter the gap conductance and other burnup-dependent parameters affecting stored energy in the fuel. Little quantitative infonnation is available to compare with data obtained using UO alone. Among that which exists, 2
some information from commercial programs in the San Onofre, Dresden and Big Rock Point reactors is becoming available and also some information from earlier experiments in the Saxton reactor.
Post-irradiation examination of rods containing plutonium will be added to PCM and RfA tests in the PBF beginning in FY 1976, for a direct comparison with results obtained using U0 rods, and 2
to test predictions based on the new thermal data.
3.3.5 Reactor Decay Heat The conservative position taken in the ECCS Acceptance Criteria on the calculation of decay heat in the fuel following a hypothetical LOCA reflects the lack of recent experimental information on energy release from short-lived fission products, and points up the desirability of obtaining new estimates of the heat generated in the first 1000 sec. following reactor shutdown.
Programs in this area include:
235 Experimental determination of decay heat in UO by direct calorimetric measurement and 2
j by calculation from beta and gamma spectroscopy.
Analysis of existing decay schemes, statistical treatment of data on decay, and computerized capability to predict integral decay heat for various reactor conditions.
A continuation of the program to measure the fission product decay heat from fuel con-taining plutonium.
Milestones Preliminary results reported - September 1975.
Final correlation - September 1976.
Final correlations for plutonium - December 1977 3.3.6 Fission Product Transport Because detailed analytical models do not exist, conservative interpretations of existing infor-mation on release of fission products are normally used in estimating release to the containment volume. Most existing data were obtained in inert atmospheres and they neglect possible B-22
interactions of the fission products with fuel, Zircaloy and each other and with a steam environment which may modify the release. Most such effects would reduce the release. The questions must be explored.
An experimental program aimed at providing information useful in safety analysis will be initiated in FY 1975. Emphasis is to be placed on iodine, cesium and other semi-volatile fission products. A supportive analytical program dealing with the development and testing of models for the migration of fission products from fuel pellets into the gap and from a failed fuel rod into the containment vessel is planned for FY 1976. The improved model should include provisions for various mass flow paths, fission product release and escape fractions, chemical state, particle size and thermodynamic effects.
Milestones Complete initial experiments to determine the chemical and physical states of released fission products - September 1975.
Determine release and escape fractions from simulated irradiated fuel rods and extend experiments to include irradiated fuel rods - December 1975.
Develop simplified fission product transport model - April 1976.
Conduct experiments on aerosol behavior - June 1976 Develop improved transport model - July 1977.
Construct scaled facility for model testing - December 1977.
Complete basic studies in scaled facility - June 1978 3.3.7 Core Meltdown 3.3.7.1 Steam Explosion During a hypothetical core meltdown, it is likely that molten core materials would contact water.
To assess the outcome of such an interaction, a probability of the occurrence of a steam explo-sion is assigned based on the body of existing experimental data and industrial experience, most of which involves materials other than those of interest here.
A series of experiments will be conducted to understand and to define more clearly the physical conditions required to initiate and to propagate explosive interactions between water and various molten LWR core materials.
B-23
A concurrent analytical effort will be directed at predicting more accurately the consequences of such explosions in terms of the potentially destructive pressure pulses generated.
Milestones Complete small scale experiments and define effects of scaling - June 1976, s
Verify preliminary model of thermal to mechanical energy conversion - December 1976.
Perform large scale experiments if necessary - December 1977.
3.3.7.2 Molten Core Interactions Uncertainties in the physical and chemical behavior of the molten core have made it difficult
+
to fonnulate detailed analytical models of the meltdown process. In particular, the nature of the interaction between molten core materials and the concrete in the base of the containment is not well understood.
Scoping experiments will be conducted with prototypical high temperature molten materials and various forms of concrete in order to define the key physical and chemical phenomena involved in the interaction. Scaling effects will be investigated.
Milestones Complete small scale experiments and define effects of scaling - June 1976.
Develop preliminary model of interaction - August 1976.
Decision for large scale test - September 1976.
3.3.7.3 Foreign Exchange A comprehensive effort in an analytical and experimental core meltdown program has been instituted in the Federal Republic of Germany. Principal areas of investigation include molten core' interactions, fission product release and transport, thermal-hydraulic behavior, vapor explosions, modeling and sensitivity studies, and determination of physical properties. Through a bilateral exchange agreement it is expected that objectives of research related to core meltdown will be reached most efficiently through a coordinated effort involving exchange of personnel and technical reports.
B-24
l APPENDIX C j
BRANCH PROGRAM PLAN FOR ANALYSIS DEVELOPMENT BRANCH 1.
OBJECTIVES 1.1 Versions of Existing System Analysis Codes The existing system analysis code system, used for calculating the consequence of a hypothetical loss of coolant accident, consists of two parts: the Evaluation Model (EM) and the Best Estimate (BE) version. The former is to be used by the NRC Regulatory Staff and must, therefore, embody the Comission's Acceptance Criteria defined in the document in Docket No. RM-50-1 issued on December 28, 1973. The BE version, on the other hand, will embody the present best estimate for such processes as heat transfer and pressure drop, which will make it applicable to pre-prediction of tests, comparison with test data, and verification of models. The EM version should always provide conservative predictions. The BE version should enable prediction of the results of safety research experiments and estimation of the degree of conservatism in the EM analysis. The system analysis codes in their present form emphasize the description of LOCA.
Parallel development has been initiated to increase the scope of these codes to describe other postulated accidents such as the Anticipated Transients Without Scram (ATWS) and Reactivity InitiatedAccidents(RIA).
1.2 Purpose of Containment Analysis Codes The reactor containment building provides a third barrier between the fission products and the external environment. The integrity of each barrier, be it fuel cladding, the reactor vessel ar.d piping walls, or the containcent walls, is subjected to close scrutiny. The purpose of Con-tatnment Analysis is to predict the time histories of pressure and temperature for the mixture of air / steam / water along the containment building wall. These, in turn, are to be used to predict the structural response of the wall and to assess the safety of its design.
Most PWR containment designs feature the Dry Containment concept in which adequate containment volume is provided to prevent the peak quasi-steady pressure from exceeding a prescribed value during blowdown, with no dependence on pressure suppression. These containment structures may be of a single or of a multi-compartment design. The next most widely used concept (mainly for BWRs) is based on pressure suppression by condentation of steam rather than expansion. The steam is ducted to enter a large pool of water which condenses it. Another less widely used containment concept employs an array of ice columns instead of a liquid pool for the purpose of condensing steam, thereby reducing the required containment volume or the peak containment pressure. Containment code development will address itself first to advances in description of C-1
flow of steam / water / air mixtures in multiple dry compartments; this capability is generic to the analysis of all three containment concepts. Ability to describe the wet well and the venting process will be added, followed by the development associated with heat and mass transfer in the ice condenser and the effects of delay in the opening of ice condenser doors.
Apart from quasi-steady pressure and temperature loads the containment walls could also experience shock loads (from violent initial coolant flashing, in the case of a large break), or local jet impingement loads (both mechanical and thermal), in addition, one must consider whether there could be local mechanical loads caused by impact of solid objects (missiles), and conflagration or detonation of pockets of hydrogen / air mixtures.
The Containment Code will address itself only to calculation of pressures, flows and temperatures of the air / water / steam mixture resulting from the influx of fluid into the containment structure (through the ruptured second barrier); also heat transfer, and the operation of Plant Safeguards Systems. Seismic and missile impact effects are dealt with in the Environmental and Siting Branch Program Plan. Effects of hypothetical conflagration or detonation of hydrogen-air mixtures are to be evaluated by separate calculations and the resulting overpressures superpose 1 on those predicted by the containment code.
1.3 Fuel Behavior Codes l
l The fuel rod constitutes the first barrier to any release of fission products as a result of any postulated accident. The main purpose of the fuel codes in accident analysis is to predict whether this barrier would be breached and if so, to estimate the time, the extent, and the location (s) of the breach. Analytical description of fuel behavior during a hypothetical accident requires knowledge of the initial conditions for all the f.el variables that would undergo change. These involve temperature distribution (both radial and axial) in the fuel pellets and in the cladding, geometry of the cladding, gap and fuel pellets, volume of the fuel plenum, restructuring and the extent of cracking of the fuel pellets, gas pressure and composi-tion in the gap and the fuel plenum, and the mechanical strains in the cladding. All of these are affected by the degree of fuel burnuo and the initial conditions of the fuel rods at the beginning of life. The geometry of the gap and the composition of the gas in the gap are particularly important for the calculation of heat transfer and the estimate of stored energy.
During the time between initial loading and the time of a postulated accident, the fuel experiences slow changes in state which in a steady-state code are treated as a sequence of quast-steady events. The transient fuel behavior code, on the other hand, is used to describe transients that are fast enough to merit consideration of the rate of storage or depletion of energy, be it thermal, chemical, or mechanical.
Most of the development of models or their t.orrelation development activity falls under the cognizance of the WRSR Fuel Behavior Branch because of the direct dependence on results of the tests sponsored, monitored, and interpreted by that Branch. Interfacing of these models and incorporating them into Fuel Behavior Codes falls under the cognizance of the WRSR Analysis Development Branch.
C-2
1.4 Code Verification Code verification is a necessary adjunct to all code development. Code development ceases only when sufficient data have been obtained to verify that the code describes the undergoing physical processes within an acceptable degree of accuracy. The acceptance criteria on accuracy must be tailored to the analytical sophistication the code employs. An advanced code is expected to model physical processes much more realistically than some preliminary code or even an inter-mediate level code. Hence, the acceptable uncertainty margins for an advanced code must be significantly tighter than those pemitted for simpler codes. A simpler code will be called
" verified" if its results do not exhibit gross departures from test data. If this were not the case the code verification activity would seek to identify the particular modeling deficiency that must subsequently be removed by the code developer (s). The same process would be followed in the verification of an advanced code except that, in this case, the acceptance criteria become more stringent. The code verification activity will, therefore, have two main purposes:
To reveal modeling deficiencies.
To arrive at dependable uncertainty bands for the calculated results obtained with a given code.
Subsidiary verification activities include:
Obtaining the measurement accuracy requirements for purposes of verification.
Reviewing and verifying estimates for the uncertainty bands of recorded data.
Performing code sensitivity analyses to ascertain which physical processes or which system components strongly influence the final results (e.g., the peak clad temperature), to a degree that warrants a high priority modeling (analysis) and/or experimental effort.
General coments on sensitivity analysis are given in Section 3.4.2.
j Analysis verification is performed on three levels. The first level is performed by the code developer. It involves testing the computer code, or subcodes, against closed fom solutions pertaining to some highly simplified or idealized geometries and conditions. Where these cannot be obtained, the model developer must make an effort to corroborate the basic modeling schemes using available test data. As an illustration, the momentum flux term was recently introduced in the RELAp-4 code, in an approximate manner. This particular treatment of the momentum flux can and should be verified against the available steady-state and transient data for flow of single phase and of two-phase fluids through contractior.s and expansions. Such submodel veri-fication should not be based on comparison of results of a total system code with integral system test data. Various "model development" tests need to be formulated and performed by the code developer to verify the basic modeling schemes. The code developer is also obliged to performmodelingsensitivityanalysesdescribedinSection3.4.2(a). Finally, the code developer is obligated to perfom a small amount of overall comparison of code prediction with test data, prior to handing over the program to the Verification Group. This will ensure that the code runs through all the necessary LOCA regimes and thet it produces at least physically reasonable results.
C-1
4 The second and the third level verification activities are performed by the Verification Group.
]
'In the second level the results calculated using the code arc compared with data obtained from
]
the Separate Effects Tests in which only one component or phenomenon is viewed at a time, or a small portion of the total system is studied. The third level includes comparison with integral system test data (such as Semiscale and LOFT), sensitivity analyses pertaining to the code uncertainties in input data used in calculations, and evaluation of the uncertainties in the final calculated results obtained in the use of the : odes.
1.5 Advanced Code Development 1
The present system codes (e.g., RELAP-4) consider the two-phase fluid to be a ho'nogeneous mixture of liquid and its vapor, in thennal equilibrium, so that both the liquid and the vapor are at the same temperature. The homogeneous mixture model has been found to be applicable for representation of the bubbly and of the mist flow regimes. These two regimes, however, represent only the beginning and the tall end of a flashing process. The flow structure for all inter-mediate regimes (which occupy the largest fraction of the LOCA time span) admits velocities of the separate phases which can greatly differ from each other and which, under certain conditions, can even be of opposite signs. Artificial homogenization of liquid and its vapor can lead at times to serious deficiencies in prediction of pressure drops, energy transfer, choked flow, and I
the rate and direction of transport of the individual phases of fluid. It is also well known 4
that the rate of phase change (evaporation or condensation) during fast transients can depart significantly from that predicted under the assumption of thermal equilibrium. A fast transient
{
is not confined to temporal transients; macroscropic sections of two-phase fluid undergoing a j
steady critical flow are also subjected to very fast changes as they approach the plane of j
choking. The present homogeneous mooels, as well as the present "censtant slip" models, do not adequately predict the observed discharge flow rates in short pipes, i.e., of the kind that need to be considered in LOCA analyses. Our present LOCA analyses can be (and are) made in such a
}
manner as to produce conservative results. The type of analyses to be employed in the advanced
)
code must be sufficiently realistic to provide significant improvement in the estimate of con-l servatism in present analyses for conditions and geometries for which test data are not complete.
I l
The ultimate goal for the development of an advanced system analysis code is the ability to model, as realistically as possible and in a feasible manner, the thermodynamic and flow pro-l cesses of the coolant undergoing blowdown or other transients in all flow regimes. Modeling of the fine structure of the fluid (i.e., turbulence, growth and mutual interactions of vapor bubbles, details of interfacial phenomena such as wavelet and subsequent droplet fonnation, vortices, details of the temperature profiles near phase interfaces, etc.) does not fall into a j
category describable as feasible, nor does it belong in the system code. A macroscopic view of the fluid which does not oversimplify the processes to such an extent as to lose the essential phenomena is, however, within our reach. During the subcooled portion of blowdown, structural /
}
hydraulic coupling can affect the instantaneous pressures, hence this coupling affects the hydraulic loads on the reactor vessel and/or steam generator internals which in turn are either flexible in comparison with the vessel wall, or can be momentarily displaced. In most cases of interest, it can be argued that the rigid wall essumption results in conservative loads. In certain cases such conservatism is very excessive and impcses severe design limitations. There l
I C-4
is a need, therefore, for an effort to develop improved analysis techniques that include hydro-elastic coupling in multi-dimensional flow passages resulting in a "best estimate" code appli-cable to the subcooled and transition blowdown regimes during which the blowdown loads experience maxima. This hyd m-elastic coda development is to complement the advanced code, since the advanced code places emphasis on the saturated blowdown, refill, and reflood regimes.
?
1.6 Development in Parallel ~ Paths Many aspects of the descrfption of two-phase flow and heat transfer and other physical processes occurring during a postulated accident are either not completely understood or not amenable to a precise mathematical treatment. The following list illustrates some of the cases that may be encountered:
Physical process not completely understood.
Process understood but precise mathematical formulation elusive.
Mathematical formulation possible but solvability either questionable or too demanding to
~
be practical.
Simplifications employed in physical modeling without a verified regime of applicability.
Simplifications employed in mathematical solution that may lead to inconsistency between the differential equations that are actually solved. In complex systems of equations consistency may be hard to prove.
When one or more of such conditions are anticipated or encountered in the code development, it is prudent to seek alternatives (in either the model fonnulation or the solution procedure or l'oth) and to pursue parallel development. The second path may involve the same degree of f
sophistication as the first path although more commonly it seeks a more easily tractable, hence l
a simplified approach. The parallel path developr.ent is a necessity for development of complex system codes wherein the development of many subroutines or subcodes is performed concurrently and significant delays in any one of them can greatly upset the whole development effort.
Complex codes must be organized in a modular fashion with well derined interfaces that allow the parallel path approach. Our definition of parallel path involves work on the same aspects of
~
the same code We it the existing code or the advanced code) but with different approaches. It is obviou;. therefore, that parallel development is costlier; however, it provides security for timely availability of safety assessment tools.,
2.
PRESENT STATUS 2.1 Existing System Analysis Codes 2.1.1 Overview Preliminary versions of both the RELAP-4 EM and the RELAP-4 BE codes had been developed by ANC by the end of December 1973. Sensitivity runs have shown that toth versions needed further C-5
improvements. The EM version was subsequently adjusted to comply with all the Comission's Acceptance Criteria for ECCS. The present emphasis is directed towards completion of the improvements in the BE version. The reflood regime of a LOCA (for PWR analysis) is presently analyzed with the RELAP-4 FLOOD code. Hot channel analysis is performed in a de-coupled manner wherein the plenum pressure and/or core flow are specified by the RELAP-4 or RELAP-4 FLOOD codes.
The hot channel analyses presently employed are RELAP-4 and T00DEE-2 for the PWR blowdown refill snd the reflood regimes, respectively. For BWR analyses the corresponding codes are RELAP-4 and MOXY. T00DEE-2 and M0XY are purely thermal codes.
2.1.2 Improvements Perfomed Thus Far on RELAP-4 BE In the past, application of the bubble rise model to each vertically oriented control volume resulted in the so-called " stacking" or the " layer-cake" effect which caused the fluid to separate in the alternate layers of liquid and vapor, stacked one above the other. This unrealistic situation was overcome in the first half of FY 1975 through the introduction of a simplified void drift model. This mdel uses a correlation for the relative velocity between vapor and liquid which, in conjunction with the mixture velocity, allows calculation of the individual phase velocities and flow directions. Enthalpy transport of each phase causes smooth separation of phases and allows definition of only one mixture or liquid level in any one vertical stack of the control volumes. This particular improvement was absolutely essential for application of the code to Small Break analysis. Additional improvements, also needed for Small Break analysis, such as the phase separation and draining of liquid through horizontal pipes, were also added during that time period.
The next important improvement, needed for both the large and the small break analyses, was mada in modeling of the PWR downcomer flow. This was accomplished through the use of the Wallis correlation to define the maximum allowable relative velocity between the phases during counter-current flow. This improvement was found to be applicable for the flow saturated liquid through the reactor inlet nozzles. However, recent comparisons with test data have indicated the importance of the degree of subcooling of the liquid entering the upper portion of the downcomer.
This situation cannot be handled by the RELAP-4 code in a normal fashion since its thermal equilibrium model does not allow a simultaneous presence of subcooled liquid and vapor in the same control volume. Consequently, further code modifications will be performed in which a direct l
use of test data will be employed to define the liquid penetration through the downcomer, as a i
function of liquid flow and temperature at the reactor inlet nozzles and of the magnitude of the reverse steam flow at the core bottom.
A great deal of work was also devoted to the problem of control volume " packing" which takes place when cold water is injected into a control volume carrying steam. Instantaneous condensa-tion causes a drastic local decrease in pressure which draws, into that control volume, the l
fluid from all the adjacent control volumes. Inertia of the inflowing liquid often results in pressure spikes when the control volume becomes overpacked with liquid, causing the code to fail or to become unstable unless extremely small th= steps are utilized. The latter could not be 4
tolerated in a feasible small break analysis. A method to overcome this problem was introduced very recently and so far it seems very promising.
C-6
2.2 Containment Analysis Codes In August 1973. ANC released to the Argonne Code Center the CONTEMPT-LT Code. This code super-sedes the CONTEMPT and CONTEMPT-PS Codes and is applicable to both dry containment and wet well pressure suppression concepts. In its present form CONTEMPT-LT is limited to three simple geometries, namely one dry well, one wet well and one annular compartment. The only energy addition or loss in the annulus included in the calculations is that resulting from in-or out-leakage of fluid and thennal exchange at the walls. The dry well may also contain spray systems, fans, and liquid pumps. The time histories of mass and energy flow rates at the break (primary systembarrier)arespecifiedasinput. The thermodynamic and flow processes in each region are highly simplified.
Work has also been initiated on the analysis of the short term behavior of the containment fluid (e.g., the first few seconds after initiation of a LOCA) using a modification of the RELAP-4 code (presentlynamedRELAP-4C). The current study includes consideration of two-phase-two-component (air / steam / water) mixtures and the partial dynamic storage of data in the computer core, in anticipation of multi-element representation of the containment regions.
Finally, a limited effort is underway at ANC on development of both a short and long tem ice condenser containment code, to enable the Regulatory staff to perform independent analyses.
2.3 Fuel Behavior Codes 2.3.1 Steady State Codes A number of stei y-state fuel performance codes are available including those of the reactor vendors and fuel fabricators who have steady-state codes that are used for licensing. These may be more sophisticated than the codes to be described below. However, they are proprietary and incorporate material property models based upon experimental data obtained with corporate funds, and so have not been made available to the public. In fact, each organization that performs nuclear fuel performance calculations usually maintains 3 or 4 fuel rod behavior codes. The reason for this is that the codes are basically special purpose programs that describe only certain aspects of fuel rod behavior. For example, to calculate the long tenn irradiation behavior of a fuel rod, a progiam is chosen that solves the steady-state ccnduction equation coupled with a mechanical behavior model that includes correlations and models describing irradiation effects. Because many of these programs are highly specialized to predict only certain aspects of rod behavior, they have been " finely tuned" to agree with the corresponding experimental data. This makes their results reliable over the range for which they were calibrated, but their accuracy outside that range is unverified.
In this country, the more extensive steady state fuel performance computer codes are represented by CYGRO-3, LIFE, and the GAPCON-THERMAL-1 Codes. CYGRO-3 is a thermal-mechanical fuel deforma-tion analysis program developed at the Bettis Atomic Power Laboratories. It computes one-dimensional fuel and cladding stresses and strains as a function of time through a specified f
irradiation history. Both elastic and inelastic behavior are allowed. Complex fuel and cladding C-7
__.~ _,_
=__
irradiation growth models are included in the code along witn models for cladding collapse, fuel cracking, fuel pore migration, thermal bowing and fuel-cladding sliding friction. Conceptually, j
CYGRO-3 is the most comprehensive one-dimensional fuel rod deformation computer code presently l
available. However, the known comparisons between CYGR0 and e'xperimental data are rather limited.
l The fuel and cladding material properties, gap conductance and pressure history, and coolant pressure and temperature history are not calculated by the code but must be input (the code is set up to run in conjunction with the Bettis " Environmental Data Package," most of which is j
classified). Consequently, this code has found use within the nuclear industry for parametric and sensitivity studies of, for example, the effects of fuel cracking, clad collapse, clad j
anisotropy, and the friction coefficient between clad and fuel on clad deformation during cyclic operation or on cladding strain during up-power transients, rather than as a general design tool. Both prediction and measurement of cladding dimensional changes for small deformations t
are subject to significar.t uncertainties. The unified construction of CYGR0 has made changes or alterations by other than the designers quite difficult.
The LIFE-II code is designed to predict the in-pile behavior of cylindrical fast reactor fuel j
elements. A generalized plane-strain analysis combines models for fuel restructuring, migration of fuel constituents, fuel swelling, fission gas release, hot pressing of the fuel, and clad swelling. Twenty-five categories of material property correlations are used. The code is still under development, and detailed comparisons with experimental data are not available. Changes have been found desirable in the fuel restructuring, gaseous fission product swell hg, fuel hot pressing, fuel-clad friction coefficient and clad time-to-rupture correlations to improve the code's ability to predict fast reactor fuel performance. Although many of the material property mode h are limited to fast reactor fuels, the needs described above are also common to LWRs.
The GAPCON-THERMAL-1 code calculates gap conductance, temperature profiles, and certain other thermal and mechanical conditions in an oxide fuel pin. The code in its current form was -
developed for use by the Regulatory staff in evaluating the thermal performance models supplied by the fuel vendors. The code can be used in calculating fuel temperatures for several coolant, i
cladding, and fuel material combinations. Changes in the diametral gap width are modeled, including differential pellet-cladding thermal expansion, fuel swelling, and fuel expansion induced by cracks and subsequent thermal ratchetting. Fuel and cladding creep deformation may be input. Comparisons of calculations and experimental data indicate that fuel temperatures are predicted reasonably well for short term irradiations and small gaps. Fuel temperature calculations for extended periods of irradiation and for large gaps become more uncertain because of inadequacies -
4 related principally to the kinetics of gap closure and fission gas release. Ongoing work is focused on reducing the uncertainty in these areas. However, the GAPCON-THERMAL-1 code does not attempt to model all the fuel rod mechanical deformation and material property changes that may occur as a result of irradiation.
1 Work is presently in progress, at ANC and at PNL, on the development of the FRAP-S (Fuel Rcd Analysis Program-Stcady State) code, using GAPCON-THERMAL-1 and the FUEL code as the basis. The I
work on the FRAP-S code includes: a) Extension of thermal models to cover a wider range of fuel burnup and fuel rod operating conditions, b) Incorporation of the existing data on fuel-clad mechanical interaction, clad creep and Zircaloy strength and ductility, c) Improvements in the calculation of internal pressure, fission gas release, fuel restructuring, and gap closure.
C-8
r l'
2.3.2 Transient Code The present hot channel codes solve the transient heat conduction equation in the radial direction for both the fuel pellets and the cladding, for a given number of axial space increments. Heat transfer between the fuel pellet and the cladding is governed by the gap :onductance which is correlated to gap thickness via the cladding temperature and the initial gas pressure. Temperature dependent thermal properties, material phase changes, metal-water reaction rates, and axial and radial variations of pcwer densities are considered in the solution of the conduction equation.
Burnup effects can be treated through property input parameters. Rapid effects that may occur during the post-CHF transient, such as fission product release and additional fuel cracking, are not explicitly considered. The fluid flow model and calculation of heat transfer between the fluid and the cladding are, in most cases, treated in the same manner as in the System Analysis code, except for the M0XY and T000EE codes. The M0XY code is applicable to BWR hot channel calculation and ir,cludes thermal radiation exchange between fuel rods. The T00DEE code is applicable to PWR hot channel calculation for the reflood regime. Both codes employ externally specified historics of the local fluid temperatures and convective heat transfer coefficients at the clad surface. The present RELAP-4 hot channe; code was preceded by the THETA 1-B code. The latter embodies features of the HEAT-1 fuel behavior code. The FRAP-T (Fuel Rod Analysis Program - Transient) Code which is presently in the development stage will unify the thermal, chemical, and mechanical aspects of transient fuel behavior. In its present form the FRAP-T code calculates (1) fuel rod (pellet and cladding) radial temperature distribution uo to 20 axial positions; (2) the length of fuel stack; (3) the length of cladding; (4) cladding hoop stress and cladding hoop strain at up to 20 axial positions; (5) radial deformation of the fuel rod up to 20 axial positions; and (6) the pressure of gas in the fuel rod.
Aspects of fuel rod behavior which are presently modeled include:
Coupling of temperature, internal ps pressure, and mechanical deformation.
Radiation heat transfer accross the gap.
Phase change of fuel and cladding; energy required for heat of fusion; volume increase at melting.
Effect of axial variation of gap temperature and gap thickness on gas pressure.
All the modes of heat transfer from the cladding to the coolant previously considered in the HEAT-1 code.
In the area of post-failure fuel rod behavior the existing analyses are both sparse and highly simplified. The only computer code in the public domain is NURLOC-1, which calculates slumping of molten fuel, cladding, and supporting materials in a steam environment.
2.4 Code Verification A need has existed for a more systematic approach to program planning in water reacter safety research. Work has been initiated in which needs for experimental data and analysis development C-9
for LOCA evaluation will be defined. The purpose of this task is to identify requirements for analysis capability and for experimental data meded to verify codes applicable to all accident sequences such that the required overall system prediction capability is achieved. Refinement and assessment of code capability is developed through the iterative process of reactor accident analysis and experiment pre-predictions. As a result of this process, changes in the experi-mental program may be required.
Present planning is directed to the PWR LOCA design basis accident. In addition, other postu-i lated accidents require systematic evaluation of analysis and experimental needs, e.g., small breaks and BWR LOCAs, reactivity initiated accidents, and anticipated transients without scram (ATWS).
Milestones for both experimental and analytical programs will be detailed in a Project Descrip-tion report.
The following verification activities have been performed:
a.
Requirements for LOFT Measurements were defined.
b.
Performed studies to determine the sensitivity of cladding temperature to variation in:
core pressure differential and core flow; fuel decay heat and sensible heat; fluid flow and heat transfer during reflood; refill time; blowdown heat transfer; and time to CHF. These studies employed a " Propagation of Variance" technique which utilizes Taylor series expansion and linearization concepts. The results were presented in tabular form that, for the computed change in Tclad, depicts the weight factor attributed to the effects of all of the important system components.
The disadvantage of this technique is that the non-linear effects cannot be accounted for in a flexible manner. For example, simultaneous variation of some parameters may have a different effect on T than that computed by the linear superposition. Monte Carlo and clad the Response Surface techniques are being presently investigated to remove this deficiency.
c.
Statistical analysis techniques needed for optimization of the verification program were developed. These statistical analysis techniques are also being used to obtain a " black box" correlation for the Semiscale pump by applying regression analysis to the available test data.
d.
Evaluat!on was made of the GAPCON-THERMAL-1 Code. This work included separate analyses for both PWR and BWR commercial fuel rods, model sensitivity studies at various burnup levels, and an attempt to relate analytical uncertainty to the experimental results, The input perturbations in the sensitivity analyses reflected diversity of design and differences in characteristics among different vendor plants rather than fabrication tolerances.
2.5 Advanced Code Development In fiscal year 1973 the only AEC sponsored advanced LOCA code development was being carried out by the Aerojet Nuclear Company (ANC). It consisted of the one-dimensional loop code (SLOOP),
C-10
the multi-dimensional core and plenum codes (SCORE and SPLEN, respectively), and the Executive Code which links the above codes together.
Towards the end of FY 74 the advanced code development program became more diversified in order to infuse this very difficult analytical development with thoughts and ideas belonging to a wider circle of the scientific cocinunity. By the middle of FY 75 the advanced LOCA code development could be subdivided into two main areas: the Systems Codes (or " loop" codes) and the Component Codes.
2.5.1 Systems LOCA Codes ANC's SLOOP code belongs to the system code category. It was joined by.Brookhaven National Laboratory's TH0R code and by Los Alamos Scientific Laboratory's TRAC code. SLOOP is one dimensional and based on the two-fluid model which allows relative motion of the phases and, eventually, thermal non-equilibrium. THOR will have the same features as SLOOP except that it will be based on the Drift Flux (or Diffusion) model which employes four rather than six con-servation equations. TH0R will eventually become the basis for an advanced Evaluation Model, to replace RELAP-4 EM. The TRAC code will allow mixed geometry (1-D 2-D, and 3-0).
It will utilize predominately the Drift Flux model although it will also employ the two-fluid model, where necessary.
Both TH0R and TRAC are new with development being initiated in FY 75.
2.5.2 Component Codes The Component Codes category is comprised of ANC's SCORE and SPLEN codes. LASL's KACHINA, and PacificNorthwestLaboratory's(PNL) COBRA-4 In their present form SCORE, SPLEN, and COBRA-4 are all based on a two-phase, single component, homogeneous mixture model, in thermal equilibrium.
~All are directed toward the ability to treat relative motion of the phases and thermal non-equilibrium. SCORE and SPLEN are truly three-dimensional while COBRA-4 employs only two momentum equations (axialand" lateral"). SCORE and COBRA-4 are " core" codes, able to view either the whole core or a single fuel assembly or even a single (hot) channel. SPLEN is a " plenum" code which emphasizes the effects of fluid turbulence. KACHINA is the most advanced of all the component codes. It presently employs conservation equations that allow treatment of multi-phase, multi-component transient flow with unequal velocities and in thermal non-equilibirum, in two-dimensions (axial symmetry). Its main purpose is to provide a test bed for the numerical treatment of,any complex component. Once verified, this code will serve as a yardstick against which simpler codes will be gaged, in lieu of the comparisons with test data when the latter are unobtainable.
2.5.3 Linking of Codes All advanced codes are being written in the modular fom for ease of interfacing with the modules developed elsewhere and for integration with some efficient data management code. ANC has previously been developing the Executive Code to serve this purpose. It was recently replaced by DATATRAN-2, developed jointly by KAPL and Bettis laboratories, through an effort that exceeded twenty man-years.
C-ll
. -. ~
4-2.5.4 - Correlations and Model Development Experiments Experimental and numerical work has begun on developing correlations or models, for:
Entrainment, deposition and interfacial friction for the annular dispersed flow regime.
~
Particle-fluid forces arising in accelerating two-phase flow.
Wall contact surface area for the gas and the liquid phase as a function of void fraction, pipe diameter and flow regime.
Steam / water interface surface area.
~
Interface friction coefficient.
3 Planning was also done for experiments on two-phase flow in piping tees and elbows. These experiments together with some of the others listed in Section 2.5.4 are of the "model develop-ment" variety, in contrast to the larger scale " separate effects" and " integral" tests that are the responsibility of the Systems Engineering Branch.
3.
RESEARCH PROGRAM _
Before describing plans for the analysis deve%pment to be performed in the previously listed five major areas, it will be useful to outline the general plan concerning the work priorities for system code development. Previous activities in the area of system code development (prior to FY 1974) have put a great deal of emphasis on development of the advanced code without suffi-cient attention having been given to completion of the existing code. This led to the situation, wherein the existing system code (RELAp-4) was not yet satisfactory, while the advanced code was l
still in the early development stage. In order to remedy this situation, completion of the necessary improvements in the existing code was given highest priority. In this particular I-area, special attention was given to those improvements and/or modifications that have to be made in the Evaluation Model version, to enable compliance with Regulation staff requirements.
Modification and verification envisaged for the Best Estimate version are taking longer to complete. By the end of the calendar year 1975 over 90% of the planned activities should be completed, with the work continuing thereafter being associated mainly with code maintenance f
-activities. The future work schedules for the advanced code are described in Section 3.5.
3.1 Improvements in the Existing Code 3.1.1 RELAP-4 BE (Best Estimate Version) i A number of meetings with RSR consultants and other experts in the field were held to establish the priorities in further improvements deemed necessary to arrive at the Best Estimate code version. They fall into three categories.
C-12
_m
~ The most important improvements that are now being addressed with top priority include:
a.
4 1)
Treatment of critical discharge at the break and a consistent internal choking model.
~
It was agreed that HEM (Homogeneous Equilibrium Model) should replace the Moody model for all fluid qualities (at the break) exceeding 2 to 5% (the actual value of the
[
" transition" quality will have to be found through comparisons with test data). For the lower fluid qualities a smooth transition between the momentum equation daninated flow (for subcooled or saturated flow) and HEM model will have to be developed.
2)
The liquid penetration and bypass model for the PWR downcomer will be improved, as
. mentioned in Section 2.1.2.
J 3)- Treatment of lower plenum entrainment will be introduced.
j 4)
Core reflood calculations will be based on the local conditions in the core rather than
. through the use of FLECHT correlations. This calls for the ability to define the details of the heat transfer coefficients in the vicinity of the quench front and two-dimensional heat conduction in the cladding. The latter will be attempted through the definition of a closely spaced finite difference mesh segment which moves along the fuel rod together with the quench front. A liquid entrainment correlation will be employed based on local steam velocities and a separate flow channel added to represent the fluid regions around the unheated internal boundaries such as control rods, instru-ment thimbles, etc. This flow channel will allow fallback of water from the upper i
plenum, accumulated there due to entrainment in the core.
i b.
Improvements in this category may be characterized as " fine tuning." They will involve:
1)
Addition of a transient CHF correlation. A new model was recently developed at RS9 which shows promise but needs further verification.
i 2)
Chen's correlation will be utilized for the nucleate and the flow boiling regimes.
i 3)
Improved post-CHF transition and film boiling correlations introduced.
4)
Improved treatment for voiding of the pressurizer.
r 5)
Improved and consistent treatment of the momentum equation at all flow area changes and at the ECC mixing section.
6)
Improved two-phase pressure drop treatment especially at flow area changes, c.
In order to facilitate code sensitivity and uncertainty studies, the magnitudes of the uncertainties (and their probability distributions) will be determined for all important parameters and correlations which may have a bearing on the computed values of maximum clad temperature.
C-13
l j
^
The code wi11 be modified to acconnodate definition, at input time, of the individual
" settings" of all such parameters, within their uncertainty range, thus allowing the sensitivity studies to be performed via Monte Carlo or other techniques.
After these improvements are introduced the code will be temporarily " frozen" and subjected to extensive verification and sensitivity analyses. Such studies may uncover further room for improvement. In our discussion thus far the emphasis was placed on the BE code for description of a PWR LOCA. As soon as that version is completed and " frozen." improvements will be intro-duced concerning analysis of BWR LOCAs. In this instance the first priority will be given to the description of jet pumps, especially for the "off-design
- operation conditions. Spray cooling will be studied next, followed by inprovements in the treatment of the steam separator and dryer. A parallel effort will be undertaken to introduce self-initialization, i.e., code calculation of a compatible steady-state thermal-hydraulic condition.
3.1.2 Modelling of the Reactor Core (BE Version)
The practice of decoupling the hot channel analysis from calculation of the remainder of the core has been criticized because the lateral boundary conditions between the hot channel and other core regions are not insensitive to the forces responsible for core flow redistribution; i.e., there may be a rather intimate hydraulic coupling between all regions of the core,' with recirculation paths encompassing the core and the inlet and exit plenum regions, i
The effect of tte core on system behavior is a result of both local and overall energy deposi-tion in the coolant and possible local changes in the geometry of the core channel. These, in turn, affect the enthalpy, density, flow regime, and flow resistance of the fluid. The last three effects determine the instantaneous pressure differences between the upper and lower
)
plenums. If the details of core flow redistribution and of other internal processes do r.ot significantly affect the histories and the magnitudes of the upper and the lower plenum pressures I
these two plenums will then supply the proper " insensitive" boundary conditions for a decoupled view of the themal, hydraulic and structural details of the core.
In order to verify this assumption, it will be necessary to model core thermal-hydraulics with a multi-dimensional code such as COBRA or SCORE and couple it to the RELAP-4 system code. A number of runs in which the detail of core geometry is gradually decreased will have to be performed to arrive at the simplest representation of the core which will not significantly perturb the upper and the lower plenum conditions. A single core channel model using RELAP-4 j
would be the simplest situation to treat analytically, but this condition probably cannot be achieved physically. Once the simplest adequate representation of the core has been found, all analyses of the system (including the sensitivity analyses for components or phenomena residing outside the core) should nomally be performed with that configuration. However, in some analyses, it will be necessary to perform further calculations to ascertain whether changing the problem has changed the adequacy of treatment of the core. The histories of plenum pressure and enthalpy calculated by the system analysis code would be stored on tape for the more detailed de-coupled core analysis.
i C-14 3--
e-_v.,
e, vr,
- nwv,
,c-wn,
The decoupled core analysis is to replace the decoupled hot channel analysis. It should, there-fore, contain:
a.
" Coarse" mesn multi-dimensional geometry for all core regions except the hot assembly, b.
" Fine" mesh multi-dimensional geometry for the hot assembly flow passage, j:
c.
Ability to calculate pin-to-pin thermal radiation in the hot assembly.
.d.
Fuel behaviot-transient code (FRAp-T) description of themal energy exchange between pellets, j
gap, cladding, and coolant and of mechanical deformation of the hot pin cladding.
The same decoupled core model can then be used for sensitivity analyses of effects caused i
by either changes in plenum conditions, flow redistribution, or geometry change in the core.
3.2 Improvement of Containment Analysis Codes j
3.2.1 Short Term (Initial Blowdown) Analysis L
During the very early part of a LOCA initiated by a large break, the walls or structures in the vicinity of the piping rupture may experience significant shock loads from propagation of the pressure wave, as well as significant jet impact loads. Both kinds of load could be reasonably j
well calculated using the multi-dimensional compressible flow analyses which consider a homogeneous j.
mixture of air / steam / water. -The mesh size in the approximate path of the jet would have to be small enough to define steep changes of properties. This situation is similar to the meshing of the reactor core in which the hot assembly region is fine-meshed in contrast to the remainder of i
the core. Only the compartment or region containing the rupture would have to be represented multi-dimensionally; other regions would respond sluggishly to the initial deposition of blow-down energy and could be modeled via relatively large control volumes. In this early blowdown regime the assumption of 1007, entrainment of water droplets (coming through the break) in the steam / air mixture, together with no heat transfer from walls, would be acceptable.
The model adopted for choked flow of two-phase-two-component mixtures will have to be verified experimentally. Checks for the choked-flow condition need to be performed for all flow passages
-between compartments. Data obtained from the full scale Marviken (Sweden) containment test program and from the Battelle (Frankfurt, Gemany) tests on a 1/4-scale subcompartment con-tainment systen need to be evaluated and used as a means to check out the containment codes.
3.2.2 Long-Term Analysis The long-tem containment analysis code should include the following features (among others):
General extension of capability to handle many compartments, allowing intercompartment flow
- and general application of various safeguards systems (sprays, fans, water pools).
C-15
Improved numerical techniques (as in the German IMEX Code).
~
Treatment of two-phase-two-component mixtures' and their critical flows.
Liquid droplet entrainment and separation, via criteria based on local velocities.
3.3 Improvement in Fuel Behavior Codes The following list of improvements to be made on both the steady-state and the transient codes is ordered according to disciplines and according to the work priority within each discipline.
The work order is keyed to the appropriate experimental activities which are to produce test data needed for both the formulation and verification of analytical models. Various abbrevia-tions employed in the list will now be described:
PBF - Power Burst Facility -
PCM - Power Cooling Mismatch (in PBF)
PIE - Post Irradiation Experiments LAB - Laboratory tests, location and scope to be defined LIT - Literature survey.
3.3.1 Steady-! tate Analysis (FRAP-5)
The final version of this code should include:
Creep of. fuel and clad.
Densification of fuel.
Cracking of fuel.
Restructuring of fuel.
Material properties of fuel and clad.
Elastic - plastic deformation of fuel and clad.
Gas release from fuel.
Power distribution in fuel.
Heat transfer between fuel and clad.
4 C-16
,.--#,-.m
,.,r,,
Chemical interactions of clad.
The following is a list of the planned activities for FRAP-S code development:
3.3.1.1 Thermal Analysis Restructuring, improved gap closure and gap conductivity.
Flux depression.
Continued work on improvement of gap conductance analysis using PBF and Halden data.
UO2 conductivity model for cracked pellets.
Improved gap conductance model using high burnup Halden and PBF data.
3.3.1.2 Mechanical Analysis Calculation of fuel-clad mechanical interaction and contact pressure.
Improved models for fuel and cladding plastic deformation.
Models for localized fuel-clad mechanical interaction and axial ratchetting.
UO expansion model based on data for cracked pellet cycling.
2 3.3.1.3 Pressure Response Improved models for void volume, plenum temperature and gas release.
Model for fill gas absorption.
Improved model for gas temperature in the void.
Gas absorption model based on Halden inpile data.
Improved pressure response based on Halden inpile data.
3.3.1.4 Material Properties Fuel and clad creep.
Clad strength and ductility change with irradiation, based on LIT data.
Fuel densification.
C-17
Clad corrosion.
Clad ductility and damage models using PIE data.
Improved fuel creep and swelling nodels based on Halden and PIE data.
Models for fuel swelling and clad strength, ductility, and damage.
3.3.1.5 Control Routines Axial power shape and simplified input-output routines.
Linking of FRAP-S to FRAP-T. This linkage is to be updated once a year, starting in December 1974.
3.3.2 Transient Fuel Behavior - FRAP-T Code The aspects of fuel rod behavior that are not in the present version of the FRAP-T code include:
Axial conduction of heat.
Cladding deformation caused by bending and shear stresses.
Cladding deformation at each axial node is presently considered to be uncoupled from deforma-tion at adjacent (axial) nodes.
Fuel density change caused by restructuring and grain growth.
Creep of fuel and cladding.
3 Effects of fuel porosity on thermal conductivity.
Effects of fuel cracking.
Transient release of fission gases.
Post-rupture fuel behavior.
Fuel-cladding mechanical interaction.
Cyclic plastic deformation of cladding.
Mechanical and metallurgical interactions between cladding and spacers (or grids).
Coupling of fuel rod response with the coolant thennal-hydraulic response. The present usage of the FRAP-T code is based on a decoupled analysis. The fluid flow ar.d thermal space-time histories are calculated first, and used as input in the FRAP-T calculation.
C-18
I
'The following.'ist depicts action items.
3.3.2.1 Thermal Hydraulic and Fuel Rod Behavior Interaction FRAP-T linked to RELAP-4 (completed November 30,1974)
FRAP-T linked to COBRA FRAP-T linked to SCORE 3.3.2.2 Heat Transfer Improved coolant heat transfer r.edel under clad swelling (ballooning) conditions (local flow effects).
2-D (r e) heat flow study.
Incorporate 2-D heat flow model.
Improved Gap Conductance model using Halden and PBF data.
Improved UO and gap conductivity models based on high burnup Halden Data.
2 3.3.2.3 Pressure Response Gas release model based on LIT data.
Updated gas flow model using irradiated rod LAB data.
Updated internal pressure and gas flow models using PCM and Halden data.
Improved gas release model using LAB data.
Gas flow model using Halden inpile data.
Improved gas release model using inpile Halden and PBF data.
3.3.2.4 Cladding Defomation 2-D clad swelling model.
Axial deformation and rod bowing model.
Improved clad swelling model based on PBF irradiated sample and PCM parameter tests.
C-19
3.3.2.5 Fuel-Clad Interactions 1-D fuel-clad mechanical interaction model.
2-D fuel-clad mechanical interaction and clad strain localization model.
Improved fuel-clad mechanical interaction model based on PBF parameter tests.
3.3.2.6 Material Properties Decay heat.
Fuel and clad creep models.
Improved Zirc-H O reaction rate model based on LAB data.
2 Transient swelling model based on LAB data.
3.3.2.7 Failure Criteria Cladding strength and ductility models from LIT data.
Zirc-Inconel eutectic melting mode.
Improved clad strength and ductility model based on LAB data.
3.3.2.8 Molten Fuel Behavior Study of rod melt-through and molten fuel flow blockage.
Study of molten fuel / water / steam interaction.
Study of coolant pressure pulses using SPERT and PCH-20 Data.
Issue of study on molten fuel behavior.
Incorporation of molten fuel models into FRAP-T.
3.3.3 Fission Product Transport Analysis An analysis development program will be initiated in r? 1976 in conjunction with an experimental program to model mathematically the migration of fission products issuing from a ruptured fuel pin. The analysis will be concerned with estimates of the fission product release fraction, chemical state and particle size, interaction with the coolant and with the wetted walls, escape through the rupture (s), and transport within the containment building. The final analytical models may have to be interfaced in some way with the system LOCA (thermal-hydraulics) codes.
C-20
3.4 Code Verification 3.4,1 General Aspects Work will continue on:
Measurements requirements.
Collection of test data, examination of the data range and the definition of the uncertainty bands.
Definition of standard problems and comparisons between test data and NRC code predictions.
The former activity is sponsored by the NRC Regulatory staff.
As NRC computer codes become operational the work load will shift towards the most essential tasks of the Ver.ffcation Group identified in the Introduction. Better techniques must be sought and implemer.ted for performing the sensitivity analyses identified in Section 3.4.2.
Work will be initiated on the feasibility of performing probabilistic rather than deterministic LOCA analysis.
Criteria must be established for classifying a Code as " verified." This task will become progressively more difficult as more advanced codes are developed which require more accurate and elaborate testing for their verification.
3.4.2 Sensitivity Analyses Two types of sensitivity analysis will be identified:
Sensitivity analysis performed during code development and verification for the purpose of ascertaining that the analytical models are sensitive to variation of important parameters.
Sensitivity analysis to determine effects of changes in some input parameters over the range of their uncertainties on the prediction of such in4portant system responses as the maximum cladding temperature.
The above two types of sensitivity analysis call for radically different approaches. The purpose.of these comments is to outline these approaches and to direct attention to areas that need to be explored for running meaningful and economical sensitivity analysis.
a.
Sensitivity studies performed during code development and verification should determine the effects on certain selected key information, such as cladding temperature in the hot channel and changes in the more intimately and then the successively less intimately related processes affecting changes i'n the selected key information. The procedure can be structured in a tree formation, starting from a single point (Tclad) and branching out through successively removed degrees of directness of coupling. A most sensitive or C-21
sufficiently sensitive analytical description must be found for each branch before pro-ceeding with the next. Organization of the analysis in this manner will result in a sensitive code as well as a quantitative estimate of the contribution to the error (or uncertainty) resulting from simplification in calculational methods or assumptions incorpo-rated in the model used for each process. This.information would greatly facilitate eventual estimates of the overall limits of error in predictions made using the code. The physical remoteness of components of the system is, in most respects, a good index to their remote-ness in terms of effect on the key information. For example, the start is normally at the hottest region of the cladding which is physically affected by heat from the fuel pellets (via the gap) and by the fluid flow within the hot channel of the core. This, in turn, is affected by the flow in the remainder of the core and by the upper and the lower plenum conditions. These, in turn, are affected by the conditions in the downcomer and the external piping which, in turn, are affected by the pump, the flow from the rupture, the steam generator (in the case of a PWR), and the ECC injection system. The question of the degrees of freedom required for, say, the geometric modeling, can be addressed relative to each component at a time, with assurance retained that the code describes, in a sensitive manner, other processes at higher levels of the tree.
b.
Sensitivity analyses for effects of changes in input parameters are performed with a finished code. It is important, however, that these analyses be performed with a sensitive code. Only then will the conclusions be meaningful and realistic as to the spread in the computed results caused by perturbations in input data. Some input data describe initial conditions of the system. Some set boundary conditions, which can be time dependent if they involve manual or automatic initiation of mechanisms that follow a time sequence or contain inherent time delays. Boundary conditions may also refer to prescribed changes in modeling assumptions or correlations for example, following from the Comission's ECCS Acceptance Criteria. Since sensitivity analyses are performed with a complete code and perhaps many times for each reactor type, it is important that the code be relatively simple, yet sufficiently sensitive. It is also important that statistical concepts be used in formulating the matrix of required runs. Finally, it may turn out that the most economical way of getting the information required from sensitivity analyses would be through use of the code re-cast into perturbation form. Such a form would not require a complete re-run of the total code for each case, but would make use of the solution calculated for the base case as a starting point.
3.5 Advanced Code Development 3.5.1 Role of Model Development Tests The existing codes, such as RELAp-4, are simply structured and, therefore, must resort to the use of gross correlations. The latter do not apply to some basic physical process but, rather, lump the effects of the test geometry with combined effects of may basic processes. While simple codes are adequate for safety analyses, when provided with numerous conservative assump-tions, their use for predicting the rea'.istic behavior of full scale plants (under pnstulated accident condicions) is severely impaired.
C-22
i Consequently, the advanced BE codes must consider geometrical effects through the treatment of multi-dimensional two-phase flow where such is warranted and consider the basic exchange processes occurring at the wall / fluid and at the liquid / vapor interfaces. These basic proce:;ses involve the exchange of mass, momentum, and energy, in various flow regimes and under steady-state and transient conditions. What these statements imply is the ability to define the instantaneous shape and the extent of the liquid / vapor interface. Rigorous description of some of these processes is beyond the present state-of-the-art. Besides, a very detailed microscopic view of the exchange processes is not feasible in the codes that have to deal with complex interactions between various system components. Hence, such codes must be so structured as to allow various degrees of the detail of the component's description, depending on the objective of the particular analysis and on the available state-of-the-art.
The model development tests will, therefore, encompass a two-pronged approach. One path will deal with the pure or basic research with the purpose of advancing the state-of-the-art in describing the basic transport processes. The second path will deal with processes whose analytical description is within the present state-of-the-art and which lump a number of the very basic processts together. The latter processes, however, would still involve fairly local events so that their description could safely be employed in the analyses of full scale plants.
As an example, the pure research path could involve the description of the vapor nucleation site distribution in bulk and on the wetted surfaces, for the type of fluid and surface conditions encountered in LWRs. It could also involve the establishment of vapor nucleation criteria under Ifquid depressurization and/or wall heating conditions and vapor bubble growth and interaction l
1eading to coalescence and change of the flow regime. The example of the second path (applied i
f research) would be the definition of the flow area or the computational cell averaged volumetric generation of vapor, during depressurization,that would lump the effects of bulk and surface r
nucleation, bubble interaction, etc., and utilize a relaxation model which describes how the non-equilibrium process relaxes into the equilibrium process and under what condition. The second example of the " applied research" type experiment is the determination of critical flow through pipes or apertures, study of two-phase (steady and transient) flow through orifices.
l elbows, contractions, expansions. Tee's, void migration in multi-dimensional flow fields, study of liquid entrainment during reflood, etc.
What then is the role of pure research if the advanced codes cannot afford viewing the processes on the microscopic scale? To answer that legitimate question it should be pointed out that formulation of simpler models which the code can amploy for practical calculations, can be built up from the knowledge of the basics. The alterr.itive is to use the test data directly from the
" applied" research and either guess the describing equations and adjust the coefficients according to the test data trends, or to " derive" correlations through test data. The lesser the content of the bisic analytical support to any given lumped parameter model, the more risky becomes the process of extrapolation to widely different fluid flow, heat transfer, and geometry conditions.
Work is underway in the search of literature and on the specification of particular objectives for various model development experiments. It is envisaged that the universities will provide significant support in this area.
C-23
3.5.2 System LOCA Codes These codes are being developed in the building block fashion, starting with the description of steady and transient two-phase flow in a pipe. Numerical treatment of steady-state processes is considerably simpler so that emphasis can be placed on the better definition of various exchange ter.as in the conservation equations. The contractors are presently examining the effects of various models, for these exchange terms, that currently exist in technical literature. Their paucity has amplified the need for the model development tests described above. So far the only transients that were studied involved blowdown from a single pipe. More comparisons will be made with test data involving steady-state discharge through pipes and apertures. It should be pointed out that, although such tests involve steady mass flow rates, regions of fluid experience very strong pressure and thermal transients in their passage through the aperture. The differ-ences between the two-phase, one-component and the two-phase, two-component, steady-state critical flow tests gave useful infomation about the momentum transfer and mass transfer within the fluid. The ability to ccrrectly describe the critical discharge of two-phase fluid under tran-sients is one of the most important aspects of LOCA analysis.
Alternative building block type sub-codes are being employed in this first task to evcluate advantages and disadvantages of the drift flux vs. the two-fluid model. For example, LASL is employing SOLA-DF for the former and SOLA-TIF for the latter, both steming from a very versatile two-dimensional SOLA code. Such building block sub-codes will, eventually, make up LASL's Systems LOCA code (TRAC). Simultaneously, work is underway to develop efficient coupling of one and multi-dimensional regions representing pipes and plenums, respectively.
The codes are being written in a modular fom and calculational schemes are being examined for allowing various system components to be advanced in time, through different time steps to obtain economy in the running time. Furthemore, the techniques will be examined for the-possibility of switching the calculational modes; from drift flux to the two-fluid model and back, where clear advantages are seen for proceeding in that fashion.
As previously mentioned, " executive" or data management codes will be utilized to enhance the ability to perform some of the tasks listed above and for allowing adaptation, at input time, of either simple or the more complex descriptions of any one system component.
3.5.3 Component Codes Work will commence, early in FY 1976, on introduction of the void drift (Drift Flux) model into the core codes COBRA-4 and SCORE. An attempt was made, in the first part of FY 1975, to model core reflood dynamics multi-dimensionally, using both the COBRA-4 and the SCORE codes. It soon became clear that these codes suffered from numerical diffusion which prevented them from tracking the water level in the core -- an essential part of the reflood analysis. Work has already started towards removing this deficiency.
At Los Alamos, the SOLA-TIF code is being employed to study multi-dimensional flow in a PWR downcomer. The " liquid-full" cases were completed and work is in progress on two-fluid repre-sentation. This is done in stages starting first with single leg injection then going into C-24 l
l
multiple-leg injection. In addition, the "no mass transfer between the phases" case is studied before allowing for condensation effects.
The KACHINA code was used in the study of blowdown from a single pipe. This will be followed by gradually increasing the problem complexity: insertion of an internal orifice in the pipe to effect an abrupt area change; bottom discharge from a tank initially containing steam in the upper region; top discharge from the same tank, to study the ability of tracking the " level swell"; steady flow of the initially subcooled liquid through a heated channel; co-current injection of cold liquid into a pipe carrying steam, etc. Test data for code verification exist for all these " building block" cases.
LASL has indicated that some non-equilibrium flow situations may require tighter implicitness in the numerical scheme than available in KACHINA. A new code, DIABLO, has been formulated for that purpose.
C-25
f APPENDIX D BRANCH PROGRAM PLAN FOR METALLURGY AND MATERIALS BRANCH 1.
OBJECTIVES i
The objective of safety research for the Metallurgy and Materials Branch is to generate a more confident basis for criteria and analytical procedures for design, fabrication and operation of 3'
the pressure vessel, piping and associated components of the primary system pressure boundary of LWRs so as to improve definition of the probability of their failure and failure' r:, odes, and to i
establish ways by which failure probability can be reduced if this is considered necessary.
f.
The present areas of particular interest to LWR safety include: (a) criteria for fracture-safe analysis of heavy-section pressure vessel.s and piping considering especially the relationship of fracture toughness and crack arrest in the elastic-plastic transition regime (b) the degrada-
- tion of fracture toughness and ability to arrest a running crack as affected by neutron irradia-tion', (c) the effects of thennd shock that might result from injection of cold (ECCS) water into a hot reactor vessel-following a LOCA,' (d) the load, strain, frequency, temperature and environmental facters controlling fatigue crack initiation and growth in both irradiated and unirradiated materials,'(e) ultrasonic and electrochemical techniques for pre-service inspection, and acoustic techniques for in-service flaw detection and monitoring, and (f) fundamental and
]
applied aspects of materials resistance to stress corrosion cracking and to the neutron irradia-
. tion environment.
Special attention is given to the' integrity of the primary system boundary of LWRs because of i
the need to ensure confinement of the nuclear core materials at all times, and thus the need to preclude the type of failure in the primary system that might lead to breach of this confinement.
The primary pressure boundary of current LWRs consists of a steel pressure vessel of thickness approaching 12 inches, and primary piping of thickness as much as 4 inches. The typical materials of construction of vessels, including A533-8 plate and A508 Class 2 forgings, and for piping, including A106-B and 304 stainless steel, have been studied extensively to develop information on trends for mechanical property behavior under appropriate test conditions of temperature, stress, neutron irradiation and reactor environment. These studies have necessarily been done primarily with laboratory-scale test specimens because the massive size of reactor components coupled with the section thickness and neutron irradiation considerations make testing of full-scale vessels or components either prohibitively expensive or almost technically unfeasible.
Nevertheless, behavior of the full-section-thickness materials and components must be predict-able from criteria and trends developed largely with small-scale laboratory test specimens.
Thus, despite knowledge already attained concerning properties of primary system component materials, improvements in information are still sought to round out the basis for judgments affecting continuing reactor safety.
0-1
-The approach used for the branch plan is to identify those areas of greatest importance to LWR safety and initiate high priority studies on the problems. The research programs are then formed to develop analysis criteria, or testing procedures, or new materials, with all results.
ultimately being incorporated into improved industry code rules and standards, and improved LWR safety designs.
2.
PRESENT STATUS Planning of the program identified for the Metallurgy and Materials Branch has required a back-ground knowledge of the current research status and state-of-art for each subject area, as well as knowledge of the current licensing criteria for each subject. This background information is summarized below for each of the six main subject areas of research for the branch.
2.1 Criteria for Fracture Toughness and Crack Arrest in LWR Vessels and Piping 1
a 2.1.1 Current Status i
Fracture toughness of LWR vessel and piping materials has been studied extensively by many organizations over the years, with much of the data and trends developed using relatively small, laboratory test specimens of thickness equal to or less than one inch. However, some very useful correlations have been made between the small specimen results and data from 8-to 12-inch-thick compact tension (CT) specimens and 12-inch-thick dynamic tear (DT) specimens. Pipe-rupture studies have been conducted in the ductile regime using piping generally less than 2 curve. This curve inches thick. At present, fracture toughness is governed by the reference KIR defined the most conservative limits, or the so-called lower bound, of all available valid / static i
and "K," (arrest) data on specimens of a thickness up to 12 inches.
Because Kyc, dynamic kid g
curve, additional data points only a few heats of steel were used for the specimens of this KIR from more heats and product forms of materials are needed. Crack initiation is governed by the static K curve, whereas crack arrest is presently being defined by the limits of the KIR curve Ic i tsel f.
The understanding of crack arrest has been actively pursued in recent years; although
~
the methodology for predicting when and where and under what conditions an unstable running crack will stop has yet to be established or experimentally validated, research is now underway to solve this problem.
2.1.2 Licensing Criteria Linear elastic fracture mechanics provides the analytical procedure presently used by the NRC for evaluating the stress-flaw size fracture toughness relation for onset of rapid fracture in the non-ductile range. Establishment of a specific criterion for crack arrest would enhance The currently imposed criteria for fracture toughness and crack arrest in LWR
-safety assurance.
materials appear in ASME Code Section III, Appendix G, " Protection Against Non-ductile Fracture" and in 10 CFR Part 50, " Appendix A, General Design Criterion 31 - Fracture Prevention of Reactor Coolant Pressure Boundary," " Appendix G. Fracture Toughness Requirements."
l 1
0-2 j
.These rules apply to vessels,.and to piping greater than 2.5 inches in thickness; the criterion 4
for nozzles is that their fracture toughness properties must equal or exceed those of the vessel.
2.2 Irradiation Effects on Fracture Toughness and Crack Arrest in LWR Vessels 2.2.1 Current Status The important effects of neutron bombardment upon reactor vessel beltline structural materials include the upward shift in the RT temperatere by several hundreds of degrees F, the reduc-NDT tion in shelf-level energy absorption from the high, pre-service levels, the increase in both yield and tensile strength and the reduction in tensile ductility. All of these factors combine to reduce fracture toughness and the potential for crack arrest. Because of the problems inherent with research on irradiated materials, including space limitations in reactors, shipping cask sizes, and radiation limits for hot cells, most research on irradiation effects to vessel materials has been conducted with test specimens of one inch thickness or less. However, important correlations are beginning to emerge between results from small specimens and those from 2-inch and 4-inch-thick CT specimens being tested in 1975.
2.2.2 Licensing Criteria The present licensing criteria for irradiation effects to ferritic pressure vessel materials are contained in " Appendix H. Reactor Vessel Material Surveillance Requirements" of 10 CFR Part 50.
In summary, the fracture toughness of an irradiated vessel must be assessed by establishing the neutron-induced increase in RT temperature, as measured by Charpy-V specimens, by translating NDT from pre-to post-irradiated behavior along the 50 ft-lb. energy level, and at no time permitting the upper shelf energy level to drop below 50 ft-lbs or the specimens to exhibit less than 35 mils lateral expansion.
2.3 ECCS Thermal Shock to Reactor Vessels 2.3.1 Current Status The background knowledge related to the requirement for maintaining vessel integrity if ECCS water should be injected following a LOCA is essentially the same as that indicated as discussed above under fracture toughness, crack arrest, and irradiation effects. Analysis methodologies are being developed to permit a rational integration of all pertinent factors, including crack initiation and arrest, thermal stress gradient, and the alteration of these parameters as a function of time, crack length, vessel material properties and temperature range.
The available toughness and arrest data and the analysis methodologies require additional experimental testing to permit validated conclusions. Finally, some test may be desirable to ensure continuing vessel integrity following a LOCA.
D-3
-~
2.3.2 Licensing Criteria Regulatory Guide 1.2 "Themal Shock to Reactor Pressure Vessels" describes a suitable way to implement the pertinent General Design Criterion 31 of 10 CFR Part 50. The essence of Criterion 31 and the consequent requirement for thermal shock, is that "the reactor coolant pressure boundary shall be designed with sufficient margin to assure that when stressed under operating, maintenance, testing and postulated accident conditions (1) the boundary behaves in a non-brittle manner and (2) the probability of rapidly propagating fracture is minimized."
2.4 Low Cycle Fatigue and Fatigue Crack Propagation 2.4.1 Current Status Research findings have revealed an increase in fatigue crack propagation rate for steels when cycled at I cpm rather than 60 or 600 cpm. The slower rate of I cpm approaches, but does not duplicate, the typical slow cycling rate of operating vessels. The rate of loading, the hold-time under load, and the superimposed small cyclic vibrations upon a mean load also can cause increases in tre crack growth rate of steels. Thus, the growing base of fatigue data is being continuously reviewed to ensure that it is adequately conservative in light of these findings.
2.4.2 Licensing Criteria At present, design rules are given for fatigue design in ASME Code Section III, for stresses over 35,and temperatures < 700 F.
These rules, which are invoked for licensing puru ses, apply to unirradiated material prior to service. The research needed will update the present code rules and permit code consideration for irradiated materials. However, new criteria based on crack propagation under fatigue loading may be desirable in the future.
2.5 In-Service Inspection and Flaw Detection 1
The specific areas of interest in this subject relate to (a) on-line surveillance-monitoring of flaws and flaw growth during operations, (b) pre-service and periodic ultrasonic inspection of reactor components, (c) detection and measurement of degree of sensitization of austenitic stainless steel to strass corrosion cracking, and (d) elimination of heat affected zone cracking in low alloy steel welds associated with post weld heat treatment.
2.5.1 On-Line Monitoring of Flaw Growth 2.5.1.1 Current Status Continucas on-line surveillance represents a goal, because feasibility is yet to be demonstrated.
The technique to be employed is acoustic emission, and while it has been developed greatly in the recent past, is still requires final development and proof test in operating nuclear systems.
Furthemore, it is noted that data on crack growth, recorded as acoustic emission signals must be validated by ultrasonic testirg ouring a shutdown-period inspection. Continuous acoustic D-4
emission inspection durir:q welding of nuclear components, for detection of micro-cracking on interpass cooling, has been established in non-nuclear applications, but must be carried forward in the nuclear application to demonstrate the proof-of-principle.
2.5.1.2 Licensing Criteria Acoustic emission has not yet been proven as equivalent or better than current code-accepted inspection techniques. Therefore, no criteria exists yet for application of acoustic emission to nuclear reactor component inspection.
2.5.2 In-Service Ultrasonic Inspection 2.5.2.1 Current Status inspection of nuclear reactor components by ultrasonic techniques is required both prior to service and during shutdown aeriods for periodic in-service inspections. The present inspection procedure is for evaluation of the pulse-echo amplitude and search unit position informatior, to determine discontinuity size. It is also typical practice for an operator to continuously watch an oscilloscope screen for evidence of the discontinuities; the records produced by these practices are voluminous, and could benefit from some means for convenient storage and rapid recovery and reevaluation.
Ultrasonic inspection is nonetheless well advanced and is invoked extensively in the ASME Code in Section XI (Summer 1973 Addenda) entitled " Rules for In-Service Inspection of Nuclear Reactor Coolant Systems." Thir, section now outlines the allowable flaw sizes in terms of fracture mechanics parameters for all locations within all types of reactor vessel components.
2.5.2.2 Licensing Criteria The current criteria for inspection and interpretation of discontinuities in reactor vessel components is contained in Section XI of the 1973 Sumer Addenda to the ASME Boiler and Pressure Vessel Code.
2.5.3 Stress Corrosion Cracking Susceptibility 2.E.3.1 Current Status The present procedures for detennining the degree of susceptibility to intergranular attack leading to stress corrosion cracking (SCC) are given by ASTM Standard A262-70 " Recommended Practices fcr Detecting Susceptibility to Intergranular Attack in Stainless Steels." The tests described in this standard detect susceptibility to intergranular attack in austenitic stainless steels at the purchase stage; the required improvement is for a test that will permit measurement of the degree of susceptibility to intergranular stress corrosion in the heat affected zones of austenitic stainless steel welds.
D-5
2.5.3.2 Licensing reiteria The ASME Code Section III gives rules for proper fabrication to preclude sensitization to inter-granular SCC ir. reactor vessel piping welds. These rules and the ASTM Standard have been adopted as the licensing criteria, and have been extended in Regulatory Guide 1.44 " Control of the Use of Sensitized Stainless Steel." with respect to required pre-heat and post-weld heat treatment procedures.
2.5.4 Heat Affected Zone Cracking 2.5.4.1 Current Status it has been observed that the German pressure vessel steel 22 NiMoCr 37 can suffer from cracking in the heat affected zone following post-weld heat treatment. Although no similar occurrence has been documented for USA pressure vessel steels, similar cracking was found in the weld of the HSST Program Intennediate Test Vessel V-4.
Therefore, HAZ cracking needs to be studied in an orderly way to assess the possibility of more frequent occurrence in US Pressure Vessel Steels, and the significance of HAZ cracking to vessel integrity.
2.5.4.2 Licensing Criteria Because this phenomenon has only recently come to light in German pressure vessel steel, and has not been observed in USA pressure vessel steels, it is not currently addressed in a specific Regulatory Guide. Thorough examination of welds is recommended for detection of potential cracking.
2.6 Radiation and Stress-Corrosion-Cracking Resistant Materials 2.6.1 Radiation Resistant Materials 2.6.1.1 Current Status Neutron-induced embrittlement in ferritic pressure vessel steels has been studied extensively, with the result that embrittlement can be significantly reduced simply by complying with the draft ASTM recommendation for upper limits of 0.10 wt. % copper and 0.012 wt. % phosphorus in the chemical composition of the steel. Furthermore, a mechanism by which copper affects neutron embrittlement in pressure vessel steels has been proposed. Finally, systematic studies of heat-treatment of irradiated steel at temperatures above the operating temperature have shown a significant portion of the preirradiation toughness can be recovered in this way.
2.6.1.2 Licensing Criteria The current licensing criteria for new materials simply reinvoke those stated above for fracture tough' ness requirements of vessels and pipes. This is appropriate, and provides a suitable guide for development and ultimate code acceptance of new materials. With respect to annealing (heat treatment of irradiated vessel steel) only the Army SM-1A reactor vessel was ever proposed and approved for annealing. Guidance can be taken from this action, as well as from the draft ASTM standard on pressure vessel annealing now in the final stages of the ballot-approval procedures.
D-6
2.6.2 Stress-Corrosion-Cracking Resistant Materials 2.6.2.1 Current Status Stress-assisted cracking in oxygenated reactor coolant continues to occur in seamless, small diameter, austenitic stainless steel piping. Intergranular stress-assisted cracking has not been observed in larger diameter piping made by rolling and welding of plate material. The primary factors causing this phenomenon are known. They include oxygen ir. the coolant, high stresses and a degree of sensitization of the stainless steel. The exact combination of the factors which actuali,, produce cracking has not yet been conclusively established.
2.6.2.2 Licensing Criteria Upper limits are set on allowable oxygen concentration for normal oreration and for short abnormal transient occurrences. Calculated stresses must not exceed ASME Code design levels; sensitization of stainless steel, either by furnace heat treatment or welding, is centrolled by Regulatory Guide 1.44 " Control of the Use of Sensitized Stainless Steel."
3.
RESEARCH PR0rtJ 3.1 Criteria for Fracture Toughness and Crack Arrest in LWR Vessels and Piping 3.1.1 Objec*ive The objectives of this work are:
to develop fracture-analysis procedures and design criteria for prediction of stress levels and flaw size required for crack initiation and for the subsequent propagation and/or arrest in thick-section LWR pressure vessels and primary piping for elastic-plastic and fully plastic material behavior.
to test and validate experimentally the procedures, criteria, and design curves such as the Reference K curve, using the range of thin (less than one-inch) to thick (up to 12-inch)
IR specimens and using prototype vessel tests.
3.1.2 Current Status Intensive research efforts over the past few years have resulted in development and validation of Linear Elastic Fracture Mechanics (LEfM) as a fracture-analysis methodology for pressure vetsels and piping operation under sufficient triaxial constraint and temperature below (NDT +
120*F). The criterion based on this method is embodied in ASME Code Section III, Appenoix G and has been validated for static K with 12-inch spacimens from HSST steel plate 02, and for dynamic
!c K with 8-inch-thick specimens. Criteria for fracture-safe operations under conditions re-Id sulting in elastic-plastic and fully plastic loading is now of interest. This is because arrest of a crack initiating in a local brittle region for example, is dependent upon the increased toughness of the surrounding material. To quantify the transition in properties from brittle to increasing toughness requires additional infomation on material responses such as possible D-7
a crack initiation, propagation and arrest under appropriate test parametc*s of tempuature, stress and irradiation-induced properties and their gradients. Research is underway at many laboratories and universities to develop an elastic-plastic fracture analysis criterion using j
both the J-integral and Equivalent Energy concepts, as well as to predict the elastic-plastic stress state at the tip of a crack using a 3-dimensional finite element computy code. These techniques must be studied carefully,from a theoretical standpoint for both pressure vessels and i
piping. Then the results mdst'b"e validated by experimental investigations using a spec'trum of test configurations ranging from small specimens to prototype vessels. Unirradiated naterials must be used to validate the cefterion, but it must subsequently be validated.using irradiated
~
materials as well. Vesselsandhipingmustbetestedunderbothhydraulicandpneumaticloads, f
3.1.3 Research Program 3.1.3.1 Analysis Development and Test for Vessels and Pief,ng The objective of this research program is to develop the theory and methodology for elastic-plastic and fully plastic fracture analysis of thick sdtion reactor pressure vessels and primary piping. The analysis procedure must permit quantitative prediction of allowable flaw sizes and stress levels using appropriate toughness values under operating conditions (normal, upset, emergencyandfaulted). The analysis procedure must consider onset of crack extension under slow (static) and rapid (ATWS pressure spike) loading conditions. Concepts to be used include (but are not limited to) J-integral and Equivalent Energy. The fracture toughness measurement method and specimens must be developed to lead to adoption of standards by organizations such as ASTM, ASME, etc. The specimens and test method must be evaluated for adequacy in reactor surveillance activities.
The theory and methodology must be developed for analysis of crack arrest in thick section reactor pressure vessels for situations such as thermal shock (resulting from operation of the ECCS) and embrittled region crack pop-in (ERCP-1). The analysis procedure must permit predic-tion not only of whether or not a crack will arrest, but the conditions for crack arrest (crack size in a given wall thickness, etc.), in tenns of the nature of the event, the initial operating conditions, the structural configuration and the initial crack configuration, and the appropriate toughness criterion.
As input to the structural analysis for crack arrest it is anticipated that the kid - a Mation as a function of temperature and irradiation may be needed. The analysis must consider the influence of kinetic energy of the structure (during crack extension) on crack arrest by evaluat-ing the " path dependency" of tne process including multiple events of onset of crack extension and arrest. The initial analysis will be based on linear elastic material behavior: the need for development of an elastic-plastic analysis will be evaluated from (and based on) the elastic analysis results.
The crack arrest specimens and test method and analysis must lead to the ador. tion of standards by organizations such as ASTM, ASME, etc. The specimens and test method must be evaluated for adequacy in reactor surveillance activities.
i D-8
Toughness data pertaining to both onset of rapid crack extension (K and kid) and crack arrest gg must be obtained using standard test methods from multiple heats of appropriate product fonns for vessels and piping. In the unirradiated condition this shall include plate, forging, weld i
and weld heat-affected-zone material of current and past usage in operating systems. When small specimens are employed for this purpose it must be shown that the results conform with large specimen results. Data and results must be applicable to validate or modify the reference KIR 4
curve. Trends shall be developed using small test specimens. However, validation of the theory and analytical procedures must be accomplished using large specimens or prototype vessels and piping subjected to hydraulic and pneumatic loads under limit conditions of temperature, pressure and deliberately-emplaced flaws, i
I 3.1.4 Milestones 3.1.4.1 By June 1975, the following is to be accomplished:
Further verify the criterion for elastic-plastic and fully plastic behavior for pressure vessels undergoing slow loading, considering the J-integral and Equivalent Energy concepts for toughness.
Evaluate the mechanics of the crack arrest phenomena and establish a valid crack arrest criterion in the linear elastic regime.
Further develop a three-dimensional finite element code for stress analysis around cracks
[
in pressure vessel structures undergoing slow loading elastic-plastic and fully plastic i
behavior. Cracks located in the cylindrical wall and in the nozzle corner should be given first priority.
Evaluate fracture toughness specimens and the toughness (K and kid) of representative Ic heats of vessel material in all product forms.
3.1.4.2 By June 1976, the following is to be accomplished:
Recomend procedures for testing and interpretation of fracture toughness tests with reference to the K curve and for use in surveillance applications. Through previous IR support of ASTM type activities, expedite ASTM standard test methods for kid (rapid loading),
and J-integral (slow and rapid loading).
Using appropriate crack arrest criterion, complete work to recommend specimen requirements, a standard test method and interpretation, and an ASTM recomended practice. Evaluate need for elastic-plastic crack arrest criterion.
Complete development and refinement of three dimensicnal computer code for stress analysis around cracks in pressure vessels undergoing slow and rapid loading elastic-plastic and fully plastic behavior.
D-9
Based on currently-available data, quantify the integrity of reactor piping systems under normal, upset and emergency conditions, in terms of allowable vs. probable flaw sizes and orientation, as well as potential for crack propagation and/or arrest.
3.1.4.3 By June 1977, the following is to be accomplished:
Employing the appropriate crack arrest criterion and recomended practice for specimens, test and interpretation, evaluate heats of vessel materials in all product forms. Through previous support of ASTM type activity, expedite ASTM standard test methods for crack arrest behavior.
Complete development of structural analysis employing the crack arrest criterion and toughness data for thermal shock and ERCP-1 events.
Review the Code rules in order to introduce any modifications that may be required due to improved knowledge of toughness properties of vessel materials in all product forms, particularly with regard to crack arrest phenomena.
Define the design criteria for closely-spaced nozzles in cylindrical shells, and establish limit loads for nozzles, branch connections and tees.
3.1.4.4 By June 1978, validate procedures for analyzing fracture toughness and crack arrest for LWR vessels and piping.
3.1.4.5 By June 1980, incorporate clastic-plastic fracture toughness and crack arrest criteria into code design rules.
3.1.5 Verification of Fracture Analysis Criteria by Prototype Vessel Tests Fracture toughness and crack arrest criteria that have been developed with tests of small and large specimens should further be substantiated with tests of pressure vessel configurations of prototype size, under both hydraulic and pneumatic loading at elevated and nonnal temperature levels. Such tests will provide the additional experimental data required for comparison with the pretest prediction of flaw size and stress level for failure as well as the type and extent of failure.
l Hydraulic testing of 6-inch-thick pressure vessels has been carried out with carefully sized flaws placed in the vessel walls and in the nozzle regions, under a series of temperature and stress conditions. The flaw regions were embrittled by hydrogen charging (hydrogen embrittle-ment) to produce a natural crack as a result of a dynamic crack pop-in. Testing was conducted to result in very long ($12 inch) and d'eep (one-half the thickness) pop-in cracks in order to evaluate the crack arrest criterion and methodology. The testing will aid in establishment of the relation between the fracture toughness and arrest criteria and the details of the ultimate fracture, and hence establish the design margin of safety against rapid crack extension.
D-10
Pneumatic testing of vessels and piping is to be conducted to investigate the extent of fracture which would occur if a reactor vessel or the piping were to fail in a fully pressurized operating mode. Specifically, the investigation will consider the progress and extent of crack extension that could occur as a result of the Drge amount of stored energy available in a high temperature system, (essentially pneumatic loading resulting from system blowdown). In addition, the potential damage over the range of failure, from leaks to fragments, is to be considered.
3.1.6 Milestones Redefine the goals for the final two intermediate vessel tests V8 and V10, of the HSST program to cover pneumatic loading - December 1974.
Complete individual vessel test reports: complete final report on intermediate vessel test series (V1 through V7 and V9) showing ability of present analysis criteria to predict flaw-stress level combinations for different type of fracture in heavy-walled pressure vessels
- June 1975.
Develop methodology for assessment of the consequences of reactor vessel failure under hydraulic-pneumatic loading; initiated procurement for long-lead time items - June 1975.
Prepare hydraulic-pneumatic testing facility; qualify procedures for producing simulated-embrittled vessels, and for prrducing deep, sharp flaws; conduct initial hydraulic-pneumatic vessel test - June 1976.
Complete test series to validate criteria for evaluation of vessel failure under pneumatic loading - June 1977.
3.2 Irradiation Effects on Fracture Toughness and Crack Arrest in LWR Vessels 3.2.1 Objective l
The purpose of this research is to extend the data and techniques applied in the evaluation of fracture toughness and crack arrest of unirradiated materials to neutron-irradiated material.
Such irradiations are to be as representative as possible of operating LWR service in_ temper-ature, fluence, neutron spectrum, etc. Fracture toughness and crack arrest types of test specimen and methods for use with irradiated material will be the same as those applied in the tests of unirradiated material. Static, dynamic and crack arrest test results from small irradiated laboratory-type specimens must be shown to be comparable to results from static, dynamic and crack arrest testing of irradiated thick section (> 4-inch) laboratory test and structural prototype specimens. Investigations must be planned to permit use of results to establish reference K data libraries for irradiated materials.
IR 3.2.2 Current Status At present, one static K and two dynamic K data points have been developed for irradiated gg Id material from 4-inch and 2-inch thick specimens, respectively.
D-11
- Although trends of irradiation effects are well known for many heats and fonns of vessel steels, they have been largely developed with small test specimens from which stress intensity, K.
values cannot be extracted, and they may not be fully applicable to larger specimens and structures.
3.2.3 Research Program The data base relevant to validation of irradiated material K curve behavior must be accom-IR plished with the validated specimens developed under Section 3.1.3.1, and must include multiple heats of representative pressure vessel plate, forgings, welds and weld HAZ. Valid data at high toughness levels are required for subsequent application to heavy-section operating reactor pressure vessels.
The establishment of this data base for irradiated material leads directly to a testing of the fracture-safe analysis criterion by hydraulic or pneumatic testing of a prototype test vessel, under realistic operating temperature and pressure with the ultimate modeling of toughness gradient simulating that resulting from neutron irradiation. The test must provide for initia-tion of a r' nning crack in simulated-embrittled material of an intermediate-scale pressure u
vessel and will thereby confirm the capability of the crack arrest methodology to predict arrest of a dynamic running crack before vessel failure.
3.2.4 Milestones Complete all testing of irradiated specimens from HSST programs, including correlation
. specimens removed from previously tested 4-inch thick compact tension specimens - June 1975.
Complete final report on study of combined effect of water reactor environment and irradia-tion on fracture toughness of heavy section steel from HSST program, with special emphasis.
on correlation of small specimen results to those from 4-inch thick specimens - December 1975.
Use validated small-size specimens and testing procedures to develop initial information on irradiated vessel steel in different for reference bank of dynamic fracture toughness Kid product fonns and of different heats - June 1976.
Develop K values from irradiated specimens of 4-inch thickness or greater of materials Id for establishment of irradiated material reference K curves - June 1976.
IR Complete series of tests on simulated-embrittled vessels conducted under hydraulic-pneumatic loading and prepare final report evaluating margin of safety inherent in current pressure vessel design, and guidance on materials and operations to protect against dynamic crack propagation following initiation - June 1977.
and crack arrest data on irradi-Establish reference bank of dynamic fracture toughness KIR ated specimens from a number of different steel heats and product forms representing current usage in power reactors; use this data bank as a basis for deriving reliability of criteria for fracture toughness and crack arrest - June 1978.
D-12
j Continue expansion of data bank of fracture toughness K and crack arrest data on IR
' irradiated specimens, and incorporate modified criteria into Code rules and procedures for licensing and regulatory analysis - June 1980.
. 3.3 ECCS Thermal Shock to Vessels 3.3.1 Objective The objective. of research in this area is to provide further experimental justification for the analytical methods used to predict the extent of crack propagation that would occur in a hot reactor vessel subjected to the injection of cold ECCS water following a hypothetical LOCA.
3.3.2
. Current Status
-If a LOCA were to occur, cold water ' injected by the ECCS would come in contact with the hot wall of the pressure. vessel, thus inducing thermal stresses. Tnese stresses have been calculated to j
approach the yield point for a short distance into the vessel wall. In addition, the pressure vessel would have been subjected to some amount of irradiation so that the inner vessel surface, which would be subjected to the cold temperature gradients, would have a reduced toughness. If it is pessimistically assumed that the vessel has a small flaw on its inner surface in the irradiated region subject to the thermal stresses of the ECCS injection, it may be possible to demonstrate onset of crack extension. Conversely irradiation damage decreases toward the outside of the vessel through the vessel wall, and the steel will be much tougher (being less irradiated),
and because of the ECCS injection, at a higher temperature.
Thus two sets of oppcsing conditions are present in the thermal shock situation: those tending to initiate a fracture and those tending to arrest a running crack.
j 3.3.3 Research Program The behavior of a) crack initiation and propagation and b) crack arrest is being experimentally investigated for vessels under ECCS induced stresses. Furthermore, verification of analytical 4
methods for predicting the type and extent of fracture is also underway. Verification of this analysis requires subjecting a flawed, heavy-section test cylinder to test conditions that simulate the gradients in materials properties and the thermal stresses imposed by the ECCS injection following a hypothetical LOCA.
To design this test, it was necessary to estimate the metallurgical embrittlement gradient I9 resident in an A-533-B or A-508 Class 2 pressure vessel (6-inches thick irradiated to s 3 x 10 n/cm >lMeV), the temperature gradient, the composite toughness gradient and the thermal stress when the vessel is exposed to ECCS water. Simulation of these parameters in the test cylinders i
is being accompitshed. Techniques for assuring a naturally sharp flaw to initiate onset of crack extension in a realistic manner under thermal stress have been developed and will be implemented in the test cylinder. It is desired that the crack pop-in occur near the inner surface and arrest approximately mid-way through the cylinder thickness. The analytical predictions can then 4
4 s
D-13 e
i
.-.m.
..---m,_
m_
t be better assessed. ' The suitably prepared test cylinder should then be subjected to the simulated ECCS. shock test. ' Acoustic emission detection sensors should be used in the test for both tech-nique development and diagnostic use in themal shock tests.
3.3.4
- Milestones Analysis of metallurgical _and thermal stress gradients, and testing of techniques for their duplication in a test cylinder of A-533-B or A-508 Class 2 steel; place order for test cylinder. Develop test matrix, crack sharpening technique - December 1974. Establish test variables, and predict type and extent of fracture - June 1975.
Complete preliminary test with sub-size mockup cylinder under controlled experimental simulation of ECCS injection of cold coolant into a hot vessel following a LOCA - September 1975.
Complete final test of full-scale mockup cylinder undpr controlled experimental simulation of ECCS thermal shock - June 1976. Complete report on themal shock test, including deduced margin for safety against vessel fracture subjected to ECCS thermal stresses - December 1976.
3.4 Low Cycle Fatigue and Fatique Crack Propagation 3.4.l' Objectives The objective of this task is to establish the magnitude and characteristics of crack initiation and propagation from low cycle fatigue of LWR vessels and piping. These will be based on tests with plate, forging, piping, welds and weld HAZ in both unirradiated and irradiated states (as applicable) under PWR and BWR coolant environments. This knowledge should pemit improvements in the code rules for fatigue design for unirradiated materials, and formulation of code rules for materials following in-sarvice irradiation.
3.4.2 Current Status It is recognized that small flaws, material defects and inhomogeneities will always exist to some extent in materials to be used in nuclear service. Although such irregularities will initially be below the established limits which would require repairs, they will experience growth as a result of fatigue during nomal operation. The potential for fatigue crack growth in reactor structural materials should be experimentally assessed to gain confidence that flaws cannot grow to a " critical" size. Therefore, it is important that the data base and the precision of understanding of crack growth of reactor structural materials, in plate, forging, piping, weld and weld HAZ continue to be improved especially for the environment and cyclic rates which represent realistic reactor operating service.
Fatigue design rules are published in Appendix 1 of Section III of the ASME code for unirradiated steels at nominal stresses of 35, at temperature; 1 700*F. Some crack growth rate data have 0-14
f been established for irradiated materials under reactor service, fluch of the work on unirradiated
' steels was conducted at relatively rapid cyclic frequencies, which revealed an effect of frequency and environment on fatigue crack growth rates. More recently, however, slow cyclic rates of I cpm and less have shown increases in crack growth rates from cycling in a water environment.
Because slow cycling more nearly approaches the realistic service performance of an operating reactor, crack growth rate data developed using varied loading times and long hold-times must be emphasized to a greater extent in future research.
i 3.4.3 Research Program l
4 A systematic study is to be conducted of the details of the effect of coolant environment, neutron irradiation and temperature on fatigue crack initiation and crack growth rates. Cycles for initiatfor of fatigue cracks as well as the relation between fatigue crack growth rate and
~
AK are being stcdied. The effects of interest include strain rate (especially in the low strain range), tension hold-time (especially very long 1 0.1 cpm) mean K (at small AK amplitudes), the scatterband resulting from heat to heat variations, and studies at the low range of AK to establish K more accurately. The full range of pressure vessel and piping materials and product th forms is to be studied. To be included are Types 316 and 308 stainless steel weld material with low percentages of delta ferrite. Pressure vessel materials are being irradiated and tested with environmental simulation appropriate to PWR/8WRs. Piping must be tested under PWR/BWR coolant i
environments. Results obtained with laboratory test spcimens are to be correlated with behavior of similar material under field conditions. A comparison is needed of in-reactor cyclic response j
against results obtained out-of-reactor.
3.4.4 Milestones Evaluated crack growth rates to determine potential for growth of flaws to critical size -
during anticipated normal operation - December 1974.
l l
Selected materials, completed initial neutron irradiations (as appropriate) and crack growth rate testing underway on many heats and product forms of A-533-B, A-508 Class 2 A-106 carbon steels, and 304 and 316 stainless steel under representative LWR operating service - June 1975.
Initial assessment of effect of environment and long hold-time (10.1 cpm) under tension on crack growth rate - December 1975.
Study interactions between fatigue and water environment in primary system materials - June
- 1976, Recomendations for unirradiated and irradiated-condition crack growth curves in ASME Code - June 1977.
Comprehensive data trends for fatigue crack growth for irradiated and unirradiated primary system materials in LWR environments. Evaluations of whether in-reactor fatigue is more serious than shown by ex-reactor fatigue test results - June 1978.
D-15
Update Code design rules for crack growth rates for unirradiated vessel and piping materials, and generate new rules for irradiated vessel materials - June 1979.
3.5 In-Service Inspection and Flaw Detection 3.5.1 Objective The purpose of this task is to develop testing procedures for pre-service inspections and in-service detection / monitoring of flaws, defects, etc., and of flaw growth in LWR vessels and piping, and to understand the causes of heat affected zone cracking in low alloy steel welds fol-lowing post-weld heat treatments.
3.5.2 Current Status and Research Program 3.5.2.1 Stress Corrosion Susceptibility Test A rapid and accurate test is required to determine the degree of susceptibility of material to intergranular stress corrosion cracking (SCC). Susceptibility to intergranular SCC can arise, for example, from slow cooling following overheating during welding in the field. Thus, it is mandatory that vessels and welds be inspected prior to service to ensure that areas of sensiti-zation are located before service 50 that repairs can be made promptly and easily. Detection should be as quantitative as possible, and must be non-destructive. The currently used " Copper-Copper Sulfate-Sulfuric Acid Test" and the "0xalic Acid Etch Test" require improvements in detection reliability, and quantitative measurement of degree of susceptibility to SCC. Further-more, these tests are not useful in the field. Alternate test procedures should be developed initially in the laboratory, and must be rapid and inexpensive. They should then be developed further to the point of fonnation of validated standard field techniques for quantitative elec-trochemical determination of degree of sensitization of austenitic stainless steel.
3.5.2.2 In-Service Ultrasonic Inspection l
It is necessary to improve ultrasonic inspection capability for carbon and stainless steel plate and piping to be more ccWtible with the stringent allowable flew-size requirements of Section XI of the ASME Code requirements for In-Service Inspection of reactor vessel components. The up-grading of ultrasonic inspection should focus on developing all aspects of the infonnation resulting from a pulse-echo test, including phase, frequency, amplitude and search unit position.
Sensitivity of the results to the specific operator, a specific calibration test or a specific transducer should also be reduced. Means for storage and ready retrieval of the information for meaningful reevaluation must also be developed. The importance of the ultrasonic inspection records is expected to increase, with reference being made to past records for camparison with current information. For such comparisons to be made most accurately, it is requi ed that t
sensor location, operating characteristics and changes in tne search unit wrought by aging be precisely recorded along with the ultrasonic pulse data for positive evaluation and future reevaluation of the recorded data. The attenuation of UT signals by austenitic stainless steel base metal, weld deposits and the interface must be characterized in order to permit extraction of the flaw signal from the background noise. Finally, criteria must be established to accept D-16
or reject flaws based on f. heir size, location, orientation and propensity for growth to critical size during reactor operating lifetimes.
3.5.2.3 Acoustic Emission Flaw Detection Techniques are being developed for continuous, on-line detection of flaws and defects, and mon-itoring of their growth. Detection of flaws, and the monitoring of flaw or defect growth during
- service is one of the most powerful means of preventing unexpected failure of large prir.ary system components during service. Both acoustic emission and ultrasonic testing have been shown to have the capability to detect flaws and thereby monitor growth, but it is desirable to contin-ually upgrade the capabilities of both techniques for qualitative and quantitative infonnation on flaw size, shape, location and orientation upon which positive safety judgements can be made.
Acoustic emission and ultransoaic testing are being developed as complementary tools for on-line flaw detection. Acoustic emission detectors will be placed on vessels and piping to monitor signals emitted during operation, and ultrasonic inspection conducted during shutdown periods for positive definition of size, shape, and orientation of flaws corresponding to the signals.
This task is beginning by gathering acoustic emission data from operating LWRs, to evaluate the problems still to be resolved for on-line use of acoustic emission. This step is the next logical one, as acoustic emission has already been studied quite extensively on a laboratory basis, and it now needs to be 'ried in the field. After acoustic signals have been analyzed, ultrasonic equipment should be used to verify and quantify the signals as defects. A laboratory program should be concurrently conducted on fully characterized natural flaws to validate both the detection and the quantification abilities of the techniques developed.
Flaw detection by acoustic emission during welding of piping is already underway, and will be
~
extended to vessel welding inspection. For this work, piping (or vessels) are instrumented with transducers in order to pick up the bursts of acoustic emissions resulting from cracking during post-weld cool-down. Signal patterns are then correlated to x-radiograph data to validate the l
technique, as well as to establish a library for reference in future inspections.
l 3.5.2.4 Heat Affected Zone Cracking The causes of heat affected zone cracking in low alloy steel welds following post-weld heat treatment are to be established. The applicability of the causative factors to USA pressure vessel steels is to be established and proven in tests. If applicable, the effect should be characterized as to degree and type of cracking, grades of material affected, conditions that contribute to the cracking and if practical, the mechanism. Methods and materials to preclude this phenomena would have to be developed and qualified for future use.
3.5.3 Milestones Establish the correlation between electrochemical measurements and degree of suscepti6ility to intergranular stress corrosion in austenitic stainless steel welds - June 1975.
Extend to field applications the laboratory-developed test procedure for detection of intergranalar stress corrosion and develop validated standard procedures and hardware -
June 1976.
D-17
Frove the feasibility of computer processing of ultrasonic pulse-echo data from a phase-tensitive synthetic array of sensors - June 1975.
Provide verification of ultrasonically-detected flaws and upgrade computer system from two to three dimensions - June 1976.
Complete verification of ultrasonically-detected flaws to demonstrate ability to satisfy ASME Section XI requirements, and develop " stand alone" systems for field use - June 1977.
Establish sensitivity of continuous acoustic emission surveillance of welding cracking during fabrication of nuclear components, and application to on-line inspection - April 1975.
Develop procedures for validation of acoustic signals as real defects during on-line pro-duction of nuclear components, using other NDT techniques and metallographic examination -
June 1975.
Develop data to support standardization of acoustic emission technique for on-line crack detection, and begin to draft recommendations for industry code rules and inspection pro-cedures - June 1976.
Accumulate acoustic signal data from an acoustic sensing system in an operating power reactor; initial evaluation of the physical aspects of emission and reflection of the signals phenomena - December 1975.
Incorporate into the acoustic sensing system the refinements that have been derived from operating experience; employ ultrasonic, eddy current, and other techniques to interpret acoustic signals; continue laboratory-research on the phenomena - June 1976.
Establish the cause of heat affected zone cracking in low alloy steel welds following post weld heat treatment, and determine the applicability to USA pressure vessel steels - June 1976.
Provide means of verification of acoustic detection of flaws and cracks under operating reactor conditions; incorporate results into standards - June 1978.
Gain industry code acceptance of procedures for acoustic detection techniques and crack size quantification procedures - June 1979.
3.6 Radiation and Stress-Corrosion-Cracking Resistant Materials 3.6.1 Objectives The objective of this work is to identify the mechanisms by which irradiation causes degradation of toughness in LWR vessels and piping, to verify the mechanisms for Stress Corrosion Cracking D-18
in austenitic stainless steel, and to validate development and provide code verification of new alloys resistant to these phenomena.
3.6.2 Current Status and Research Programs Increasing margins of safety can be achieved for components of the primary system not only'by improved design practice, but also by improvements in materials of fabrication. Materials having high toughness and crack arrest properties, as well as resistance to stress corrosion cracking and new processing, fabricating and welding techniques are under constant development.
These must be continually monitored and evaluated for incorpJration into designs and specific 4-tions for future primary system components.
3.6.2.1 Radiation Resistant Materials Research on radiation resistant materials is to begin with study to improve understanding of the fundamental mechanisms responsible for neutron irradiation embrittlement using transmissian and scanning electron microscopy. Parallel studies will also continue to develop alloys less sen-sitive to radiation. This will be accomplished by continued evaluation of plate, forging, weld and weld HAZ plus bolting and piping materials, both for the unirradiated state and following neutron irradiation as applicable. Typical production alloys will be used, as well as those which show promise as a result of the fundamental studies. Retention of fracture toughness and crack arrest characteristics under irradiation are to be emphasized, using evaluation techniques established in Sections 3.1 and 3.2 of this program plan; however, fatigue and crack growth will also be explored and considered in the materials improvement process. Embrittlement relief, otherwise known as annealing, will be studied in an orderly fashion so as to permit accurate predictions of response of irradiated material to various time-temperature sequences that might be encountered in operating LWR reactor systems. For such studies, fracture toughness, crack arrest and tensile data will be necessary for best interpretation of potential annealing of embrittlement.
Criteria for the acceptability of newer, high strength pressure vessel steels, and their associated weld metals, must be established. The influence of such factors as strain aging, temper embrittle-ment, neutron irradiation, chemical composition, heat treatment, upper strength level, etc.,
should be included.
Development of new radiation resistant materials must lead to code case approval for incorporation of these materials into new LWR construction. The library of annealing data is expected to lead to development of a Regulatory-approved procedure for use of post-irradiation heat treatment for recovery of fracture toughness and crack arrest properties of irradiated LWR reactor vessels.
3.6.2.2 Stress Corrosion Cracking Resistant Materials Structural materials resistant to the corrosive effects of the LWR environment have been under development for many years. This research is exceedingly complex, in that studies must be con-duc.ed over a wide range of many <ariables, such as mean and superimposed stresses, oxygen, chlo-ride and pH levels, material chemistry, fabrication and installation practices including welding.
D-19
- ~ _. _
0 I
heat treatment, grinding, machining, plus fit up and residual welding stresses. The ranges of all such variables for the potential materials of application will be reviewed, and confimatory testing of specific materials and operating conditions will be undertaken to prcvide assurance 4
I
- that proposed materials and conditions will perform without cracking in service.
3.6.3 Milestones Develop initial criteria'for selection of radiation resistant ferritic steels having high 4
capacity for retaining toughness and crack arrest capability during LWR service - June 1975.
Establish limits of embrittlement and shelf drop from conenercial heats of A-533-B steel as functions of residual element content. Develop better models to describe mechanisms of radiation damage in ferritic steels under LWR service - June 1976.
i Post-irradiation reference toughness recovery trends including effects of residual elements
- June 1977.
)
Complete development of standard procedure and data matrix for post-irradiation heat treat-ment of irradiated reactor vessels - June 1978.
a I
Verify data for code approval of high strength, high toughness radiation resistant ferritic steels for high fluence applications - June 1979.
Gain industry code approval for employment of new radiation resistant materials in LWR environments - June 1980.
Survey of pertinent factors and materials relating to Stress Corrosion Cracking in austen-itic stainless steels. Analytical method for calculating residual stresses at girth butt welds. Confim failure modes of piping in BWRs - June 1976.
Define failure conditions for stress corrosion cracking of austenitic stainless steel piping, and develop reconinendations for elimination of stress corrosion cracking problems in BWRs
- June 1977.
I e
a l
i D-20 l
i
.,...e.-
-n,--~,-,---..,-,..,..,,,_.,v
-,..n_
.,.,--.m,,,_,-.,,.__m..-.,,
,,,,, ~, -.,.
_,-.--,.n..,_-
.,.-,_n
APPENDIX E BRANCH PROGRAM PLAN FOR ENVIRONMENTAL AND SITING BRANCH 1.
OBJECTIVES The main purposes of the environmental and siting safety research program are to provide information to help ensure the safety of plants as affected by the environmental conditions at specific sites, and to ensure that future research needs for site assessment in all regions can be met on a timely basis. Accomplishing these purposes involves the understanding of, and ability to predict, the effects of severe natural phenomena such as earthquakes, tornadoes, and floods. Also involved is the development of engineering methods to help ensure the adequacy of plant structures, systems, and components to withstand imposed environmental loads safely.
Environmental and siting safety requirements are closely related to the licensing process. An objective of the research and development program in this area also is to help speed up the nuclear facility licensing process, both through development of improved site evaluation methods and by the collection of baseline information.
Protection and enhancement of the environment are important, and closely related to safety.
However, this program plan is not directed specifically at the requirements of environmental protection or enhancement, except as they may be related to the safety of the facility and the public.
l 1.1 Organization Of Program Plan The program plan is organized to provide a rational framework for describing research needs, assigning priorities, and scheduling tasks. An objective of the plan is to pemit the display and review of long-term and short-term siting needs, and to accommodate future additions or deletions without significant changes in the organization of the program plan or its utility.
Five broad program areas which encompass different asoects of siting research needs are identified.
These are:
Regional Studies Engineering Design Studies Siting Concept Safety Studies Source Phenomena Studies Evaluation Methodology Studies E-1
The order of presentation of these program areas is roughly correlated with currently assigned priorities. However, priority assignment, as discussed below, is more properly considered at detailed levels within the program organization.
The five program areas are further divided into topical study or task areas on the basis of study discipline, siting concept, or safety function. This second level of program organization is outlined in Table 1.1.
It is at. this second level that assignment of relative priorities is initially cor.sidered, and the order of presentation within each program area approximates the current assessment of crder of priority within that category. Environmental and siting safety questions enccmpass a wide variety of study disciplines, and their resolution may depend upon the extent of geotechnical data which have been collected in the many different siting regions of the country. This diversity does not necessarily impede the progress of the investigations, but it does make difficult the grouping of various research tasks for purposes of assessing current status and planning future schedules. The outline given on Table 1.1 is intended to serve as an index of the general areas to be addressed by this program plan. It is also intended to serve as a point of reference for coordinating related research efforts within the NRC and with other agencies and organizations.
1.2 Assignment Of Priorities And Scheduling The principal basis for determination of priority is assessment of the safety need to be met by performing the research task. Other important considerations are projected costs versus safety benefits, the current state-of-art in the task area, and potential impact on the licensing process. An evaluation of possible alternative solutions to the same questions may also enter into the assignment of priorities.
Guidance for assignment of priorities has come from Regulatory as expressed in discussions and memorandum reconnendations, from the ACRS as expressed in connents at hearings and in correspond-ence, and from discussions with members of the geological, geophysical, and engineering corrnunity.
The current program plan consists of tasks or projects which are currently assigned a high priority. Thus, many of the task areas shown on the iidex, Table 1.1, are not included in the present plan. Those not presently included may be added as higher priority tasks are accom-plished or as priority assessments change in response to new siting issues. Similarly, a few tasks are included in the program plan which are not scheduled for immediate support, primarily because of limitations of funds or other resources, but also because deferral is regarded as acceptable. An outline of program areas, task areas and individual tasks which are included in the current program plan is given in Table 1.2.
As can be seen in reading the task titles, the program of environmental and siting safety research mainly consists of applied or directed research. Other agencies and institutions are relied upon to accomplish most of the basic research which may be of future value to the program. The coordination of NRC programs with these organizations (e.g., USGS, NOAA, NSF, universities) helps in the concurrent resolution of siting-related questions and the pursuit of related basic research. Nonetheless, the highest priorities for NRC support will be assigned to research projects directed toward current safety questions which are important to selection and utiliza-tion of sites for nuclear facilities.
E-2
TABLE 1.1 INDEX TO SITING QUESTIONS IN ENVIRONMENTAL AND SITING SAFETY I.
Regional' Studies (Studies mainly to acquire baseline data necessary for site evaluations in regions of high sitinginterest).
1.
Geologic and Seismologic Mapping and Field Studies
' 2.
Investigations at NRC Facilities (environmental updating) 3.
Regional Meteorological Characterization'(severe storms, diffusion) 4.
Regioral Hydrology (dispersion, dilution capabilities) l
- 11. Engineerine Design Studies (Studies of engineering data input, earthquake-and tornado-resistant design, analytical l
methods, and qualification testing) f i
1.
Evaluation of Strong-Motion Records and Sites (basic input) 2.
Analytical Methods Used in Seismic Design l
3.
Structure-Media Interactions 1
4.
Earthquake Simulation Methods and Seismic Qualification of Structures and Components 5.
Seismic Requirements of Specific or Standardized Facilities i
6.
Design Studies for Shock-Isolation Systems j
7.
Design Studies for Instrument Warning Systems 8.
Combined-Loading Studies 9.
Tornado-Resistant Design Studies j
10.
Hurricane-Resistant Design Studies 11.
Tsunami-Resistant Design Studies Ill. Siting Concept Safety Studies I
(Interdisciplinary studies of siting concepts) i l.
Safety Considerations for Nuclear Parks 2.
Safety Assessment of Offshore Plants 3.
Optimum Plant Locations (regional siting guides) l 4.
Safety Considerations of Underground Plant Sites l
l IV. Source Phenomena Studies F
(Studies of specific phenomena, their recognition, effects and treatment) j 1
i 1.
Geological and Seismological Phenomena (earthquake magnitude and recurrence, surface i
rupture, liquefaction, amplification, etc.)
E-3 1
- l i
~
4..
2.
Meteorological Phenomena (tornadoes, hurricanes, floods, etc.)
3.
Sabotage 4.
Seismically induced Flooding (dam failure, tsunami)
V.
Evaluation Methodology Studies (Studies to improve methods, to enhance the state-of-art, and to provide factual bases for site evaluation criteria--development of methods as opposed to their application) 1.
Soil Mechanics and Response (in situ and lab measurement of properties and response) 2.
Seismological Methods (focal dynamics, regional attenuation, earthquake prediction, etc.)
3.
Geological Methods (including sensor applications) 4.
Geophysical Exploration Methods (including offshore) 5.
Meteorological Methods 6.
Hydrological Methods 7.
Rock Mechanics E-4
TABLE 1.2 OUTLINE OF PROGRAM PLAN FOR ENVIRONMENTAL AND SITING BRANCH i
I.
Regional Studies l.
Geologic and Seismologic Mapping and Field Studies a.
Charleston Region, South Carolina b.
Coastal California c.
Eastern Mojave Desert d.
New England and St. Lawrence " Rift" e.
Mississippi \\ alley Earthquake Zone 2.
Investigations at ERDA Facilities a.
Hanford, Washington, Flooding Assessment b.
NRTS Seismic Evaluations 3.
Regional Meteorological Characterization a.
Tornado Distribution 4.
Regional Hydrology a.
Coastal Variations of Tsunami Effects II. Engineering Design Studies 1.
Evaluation of Strong Motion Records and Sites a.
Analysis of Source Paraneters vs Distance b.
Deep Velocity Measurements at Selected Stations 2.
Analytical Methods Used in Design a.
Advanced Earthquake Modeling Techniques b.
Structural Analysis Using Three-Dimensional Motions c.
Determination of Location for Input for Seismic Motions 3.
Structure-Media Interactions a.
Experimental Comparison of Compliance-function and Finite-Element Methods 4
Earthquake Simulation and Seismic Qualification a.
Simulator Requirements and Capabilities E-5
_j
- 5. IDesign Requirements of Specific or Standardized Facilities.' Including Fuel Cycle Facilities a.
Seismic
' b.
Meteorological 6.
Design Studies for. Shock-Isolation Systems a.
Isolated Frame b.
Floating 7.
Design Studies for Instrianent Warning Systems a.' Seismic Scram Feasibility 8.
Combined Loading Studies
- a. ' Analysis of Hypothetical Combinations of Environmentally-Induced Loads III. Siting Concept Safety Studies 1.
Nuclear Park Safety Analyses a.
Safety Considerations for Shared Safety Systems b.
Combined Meteorological Effects c.
Accident Interaction Effects d.
Safety Considerations of On-Site Personnel 2.
Safety Assessments of Offshore Siting a.
Seismic Considerations in Breakwater and Mooring Design b.
Foundation Materials Properties c.
Assessment of Vulnerability to Waterspouts and Hurricanes d.
Ship Collision e.
Fault Tree Analysis of Conceivable Accidents at Offshore Plants 3.
Regional Siting Guides a.
Assessment of Concepts for Designating Sites b.
Eastern U. S. Siting Factors 4.
Safety Considerations for Underground Sites a.
Comparisons with Surface Sites b.
Availability of Acceptable Sites c.
Cost / Safety Benefit Analyses E-6
. )
IV. Source Phenomena Studies 1.
Geological and Seismological Phenomena a.
Relative Activity of Multiple Fault Strands b.
Fault Activity in the Eastern U. S.
2.
Meteorological Phenomena a.
Tornado Velocity and Pressure Measurements b.
Dispersion Under Adverse Conditions c.
Tornado Missile Behavior and Effects 3.
Sabotage a.
Threat Assessment 4.
Aircraft Impact i
a.
Vulnerability Assessment 5.
Internal Missiles a.
Generation, Effects, and Countermeasures V.
Evaluation Methodology Studies 1.
Soil Mechanics and Response a.
Production Field-Testing of Large-Strain Shear-Wave System b.
Standardized Procedures for Dynamic Triaxial Testing 2.
Seismological Methods a.
Earthquake Prediction Methods E-7
..m a
r m
- - ~,
~
t 2.
PRESENT STATUS j
Procedures for site safety evaluation of nuclear facilities have evolved through two main processes.
One is the process of site investigation conducted by utilities and their consultants to locate suitable sites. The other is the process of safety evaluation by the Regulatory staff and their consultants during the different phases of licensing. Guidance for documentation by an applicant
+
of the required _ site-related safety considerations is outlined in the appropriate chapters of the Standard Fonnat and Contents for Safety Analysis Reports published by the NRC and designated as Regulatory Guide 1.70. The general requirements for siting safety considerations are specified in Title 10 CFR Part 100 Reactor Site Criteria.
The principal concerns and main topics of past research in site safety have centered around i
geologic faulting and earthquake occurrence and effect. In this area the required investigations, seismic and geologic design bases, and applications to engineering design are specified in Appendix A to Title 10 CFR Part 100 (Seismic and Geologic Siting Criteria for Nuclear Power Plants, Federal Register, November 13,1973). These criteria were developed jointly by the NRC, the i
l Geological Survey, and the National Oceanic and Atmospheric Administration during the course of i
site safety evaluations related to licensing. Occurrence of tornadoes has been studied intensively with the goals of developing regional criteria on the intensity and the measurement and predic-tion of dynamic loads and missile effects, i
Several engineering design studies have been made to investigate the feasibility of protecting j
plants from seismic effects, including surface displacement of several feet and ground accelera-I tions up to 1.0 g.
Suggested design approaches have included isolation of containment and other Seismic Category 1 structures through embedding them in layers of inelastic materials, or j
floating the entire plant at inland or offshore sites.
r Evaluation of the potential acceptability of different siting concepts (conventional, variations j
of floating and underground, and sub-sea concepts) along the California coastline has been the subject of one study. Safety and other siting factors, including environmental impact and construction costs, were considered.
Engineering, meteorological, and seismological aspects of individual siting concepts, especially offshore and underground, have been considered in other evaluations.
Geological, seismological, and meteorological effects are the continuing subjects of research.
Although considerable progress has been made in predicting location, probability of occurrence, f
and intensity of effects related to environmental phenomena, conservative criteria are being employed in order to accommodate uncertainties and to provide safety assurance. Detennination of.the actual levels of conservatism expressed in siting criteria is a fundamental goal of f
phenomena studies.
f-l 1
i E-8
3.
RESEARCH PROGRAM Additional infonnation on the status as well as general descriptions of planned research activities in each of the five program areas is discussed below. Detailed project descriptions will be developed during the coming year and published as an annex to this program plan. These descrip-tions will outline the overall scope and objectives of individual projects. For ongoing projects, the results of past research efforts and the present and planned work activities of the partici-pating contractors will also be described.
3.1 Regional Studies Regional studies of earthquakes have been concentrated mainly in the Western U. S., principally California, because of the high level of earthquake activity. This high level of activity increases the difficulty of reactor siting in the region, but it also provides the opportunity to develop methods of recognizing and studying faulting and earthquake effects.
Geologic environmental maps in the coastal region have been developed to portray the location, recency of latest movement, and geologic relationships of known faults. Developed techniques of both land and offshore exploration have been applied to map the coast from the Mexican border to Point Arguello. The results of this study were published during FY 75 (USGS, MF585). Work is continuing northward toward Monterey Bay and should be extended to the vicinity of Point Arena during FY 76 and FY 77. Consideration will be made for extending the study north of Point Arena depending upon final decisions regarding coastal reactor siting in California and the other Pacific states.
Similar methods are being applied in the eastern Mojave Desert as a result of expressed interest in locating inland sites near available sources of cooling water (the Colorado River).
Regional studies in the Eastern U. S. are also underway. Study of the Charleston Region, South Carolina, was begun in FY 73 with the installation of a microearthquake monitoring net. This and accompanying geological and geophysical studies of the region of the very large 1886 earth-quake are continuing. Field studies in adjacent areas of the Appalachian Mountains and the Piedmont region will help to provide infonnation for better understanding the recent tectonic history of the region.
Other eastern regions where infrequent, but sometimes large earthquakes occur are scheduled for study. Because of the generally lower level of earthquake activity, and the rarity of exposed, recently active faults in the Eastern U. S., methcJs similar to those being employed near Charleston will be applied. These involve monitoring microearthquakes in an attempt to identify active structural features beneath the surface, and to study their causes, mechanisms, and distribution. Accompanying field studies of the surface and subsurface geology will be made to establish the tectonic framework or patterns within each region. Through interrelationship of the historic seismic activity, the current microearthquake activity, and the geologic and tectonic patterns, improved evaluations of seismic hazards can be accomplished. Regions scheduled for study include the New England and Anna, Ohio earthquake zones. The Mississippi Valley earth-quake zone (area of the large 1811-1812 New Madrid shocks) is being investiaged through the USGS Office of Earthquake Studies.
E-9
Completion of the eastern regional studies cannot be reliably scheduled at this time. However, three to five years probably will be tne minimum time required for useful results in each region.
A seismotectonic provinces map of the entire eastern U. S. (USGS, MF 620) was published during FY 75. It represents an experimental effort to combine the existing, but limited, seismic and tectonic information into a single format. This will be a highly useful product, relating the known equipotential seismic activity of areas to known structural and tectonic provinces.
However, since it is based on limited data and field studies its main application will be to provide background regional information needed for detailed site studies or for guiding future research.
Regional 12ation of tornado occurrence, touchdown, path length, and velocities are being studied with the goals of improving risk estimates and providing more precise information for tornado resistant design in different siting regions.
Regional evaluations of the Probable Maximum Flood (PMF) calculated for major river systems of the East and Gulf Coasts are being made under the Regulatory Technical Assistance Program.
Regional hydrologic studies include evaluation of factors which potentially may influence tsunami
(" tidal wave") draw-down and run-up in different coastal stretches. Potential effects from both locally generated and open-sea tsunamis will be considered. The open-sea study should be completed furing FY 76.
3.2 Engineering Design Studies This research is directed toward improving engineering procedures and designs used to assure resistance of plant structures, systems, and components to loads imposed by effects of severe environmental phenomena. It also includes examination and improvement of the fundamental data from earthquake strong-motion records on which earthquake design and analysis are based.
Analytical methods are widely employed in the seismic design of nuclear facilities. Characteriza-tion of ground motion at a particular site may involve the application and scaling of strong-motion earthquake data, the generation and application of synthetic seismograms, the modeling of sot.rce mechanisms and wave propagation through heterogeneous media, and the representation of complex interactions of the structures with the founding media. Dynamic response of the struc-tures and large components or systems is analytically determined from the characterized ground response and structure-media interaction.
Current practice emphasizes horizontal ground motion input which is characterized in one direction only, whereas actual earthquake motions are three-dimensional and display few consistent relation-i ships in phase or amplitude among the three measured directions of motion. Research is needed to investigate the quantitative relationship between isolated treatment of the three directions of motion and simultaneous treatment of motions as they may occur in real earthquakes.
An additional area of analytical treatment in design which may warrant research to resolve current questions on the correspondence of modeling to physical conditions is specification of E-10 1
_. ~.
the correct location for input of seismic motion. The current practice is to characterize.
ground motion at the foundation level, which is consioerably below surface grade. However, free-field response spectrum curves are generated from earthquake records of surface or near-surface motions, and foundation grades of different structures at a given site may vary cons 1derably in elevation. Investigation may be needed to clarify which questions of input treatment may derive from different analytical approaches, and which ones are related to the degree of physical correspondence of model parameters with nature. Potential results of such studies may also be relevant to determination of structure-media interactions, particularly in the case of deeply embedded and underground structures.
Seismic qualification may be accomplished either by suitable dynamic analyses or physical testing.
Verification testing of large structures and components under conditions equivalent to loads applied by large earthquake motions has been impractical to date. Research is planned which will examine alternative verification procedures, i.e., feasibility of larger and more powerful earth-quake simulators (shaking-tables), testing of physical scale models, attaining large earthquake-equivalent loadings with explosively-excited simulators, and full-scale plant excitation through point-by-point reduced dynamic loading. Specific information is needed to determine the appropriate seismic qualification method to be used for different large components or systems.
The combined loading of earthquake motion and internal pressurization as during a LOCA influences containment design. An experiment is planned to test a model section of a reinforced concrete wall under combined biaxial stress and cyclic shear to measure the " interface shear transfer" across cracks of different orientations.
Design concepts have been investigated to determine the feasibility of isolating nuclear plants from the effects of fault rupture and intense ground shaking accompanying large earthquakes.
Perhaps the most practical of these concepts is the floating inland platform. Previous design studies will be updated to determine possible benefits and costs, and the additional safety assurance which may be provided by techniques that protect against severe vibratory motions.
Nuclear plants are required by regulation to shut down following the occurrence of an earthquake whose motions exceed those of the Operating Basis Earthquake (OBE) specifitd in the design. It has been suggested that additional safety protection might be afforded by inclusion of an auto-matic seismic scram system. Possibly damaging transients caused by activation of the system, advantages of controlled shutdown, and the likely need for continuous electrical power during an earthquake emergency have been cited as reasons against installation of the automatic seismic scram systems. Furthermore, the inherent safety margins provided by the design for vibratory accelerations between the values of the Operating Basis Earthquake (OBE) and the Safe Shutdown Earthquake (SSE), may indicate that verification of the margins will provide a greater assurance of safety than inclusion of a seismic scram system. These considerations, along with alterna-ti"e sensor-net geometries, will be addressed.
3.3 Siting Concept Safety Studies Siting concepts that have been examined as possible solutions to certain site-related safety concerns include floating and underground site utilization. The potential ease of E-ll
~
site-qualification and economies of scale provided by increasing the number of generating units and associated facilities at a previously licensed site provide incentive to the concept of nuclear parks. The concept of designating pre-accepted sites will require specific consideration of the number of generating units to be accommodated and the feasibility of constructing the plants underground, or on floating inland or offshore platforms.
Evaluation of these siting concepts will consider the safety issues and practicality of implementa-tion of each concept. The safety issues involve potential uncertainties which may be introduced by the siting concepts, as well as the environmental or other effects which the concepts are intended to mitigate. Practicality of implementation of a particular concept involves considera-tion of not only costs and state-of-art, but also the availability of siting environments suitable for application.
Conventional nuclear plants with one, or a few, generating units located at surface sites and meeting current licensing requirements will be used as the basis for comparisons. Regardless of the potential acceptability which may be shown by the general studies, safety analyses are expected to continue on a site-by-site basis.
3.4 Source phenomena Studies Improvement of knowledge of potentially severe natural phenomena to the point of developing new or more reliable criteria for their treatment in engineering design is the primary goal in this research area. Detailed studies and measurement of the effects produced by severe seismic and meteorologic phenomena provide an understanding of the possible range of intensities and their geographic distribution. Statistical studies of historic and geologic records of events provide an understanding of both the geographic and temporal distribution of occurrences. Esen though the threats posed by sabotage may not be strictly environmentally related, their assessment is dependent upon the particular concepts employed in the plant design and siting.
Information for improving criteria is to be developed through studies correlating fault length and displacement with size of accompanying earthquakes; distribution of episodes of fault move-ment along strands of major faults and fault zones; correlation of field, laboratory, and analytical data on liquefaction of soils and amplification of earthquake motions; large-scale and small-scale meteorological diffusion; and tornado velocity, pressure drop, and missile behavior. Increased emphasis also will be directed toward assessment of the earthquake risks in the Eastern U. S.
These studies will cover the determination of recency of movement of faults postulated to be recently active, and the investigation of geologic structures in regions of microtremor occurrence where active faults have not been identified. Statistical evaluations of earthquake recurrence and localization will also be made.
3.5 Evaluation Methodology Studies Development of new and improved methods and standard procedures of site investigation is the main objective of this area of research. Improved procedures in soils engineering and seismology are currently under development. Additional development of fault-dating methodology is underway.
E-12
Methodology is ordinarily an inherent part of research and the progress of its development commonly is linked to the early phases of applied studies.
Specific standard procedures are being developed for conducting cyclic triaxial tests which are used to determine liquefaction potential and dynamic stress-strain properties of soils. Standardi-zation is needed because of the large number of laboratories performing the tests and the variety of equipment used. Reactor site soils investigations comprise a significant portion of the total applications of these tests. Evaluation of the applicability of laboratory test results to ift situ material properties is necessary to provide confident prediction of soil behavior under earthquake loading conditions. These studies will be completed during FY 75.
Development of a system for iftsitu measurement of the dynamic response of soil at large strains, as well as small, has been previously supported by the Division of Reactor Research and Develop-ment. Discernible and consistent changes in shear velocity (and modulus) with variation in strain levels up to 10"I% have been demonstrated. Very clean shear-wave records have been obtained. Production field testing of this equipment will be completed in FY 75 and testing will be implemented at sites from which strong motion records of earthquake have been made.
Investigation of experimental methods in earthquake prediction may be warranted in highly seismic regions because of the potential for providing additional safety assurance. Much of the research j
in this area is supported through other agencies and institutions. Promising methods whose instrumentation and operation are particularly compatible with nuclear plant operations will be considered.
2 f
d i
E-13
ENVIRONMENTAL AND SITING BRANC's LIST OF PROJECTS UNDERWAY OR TO BE INITIATED DURING FISCAL YEAR 1976 AREAS PROJECTS a.
Charleston, S. C., Earthquake Zone b.
Northeastern U. S. Seismicity c.
Anna, Ohio, Earthquake Zone d.
Faulting rad Seismicity, Southern Piedmont Coastal California Recency of Faulting 1.
Reatonal Studies e.
f.
Eastern Mojave Geology and Seismology g.
Tornado Distribution in the U.S.
h.
Design Tsunamis a.
Materials Properties at Strong Motion Record Sites 2.
Engineering b.
Seismic Qualification and verification of Design Studies Analytical Methods c.
Containment Studies a.
Underground Siting 3.
Siting Concept b.
Offshore Siting Safety Studies c.
Siting Guides a.
Recent Fault Activity 4.
Source Phenomena b.
Earthquake Statistical Analysis Studies c.
Tornado characteristics d.
Sabotage Threat Assessment a.
Atmospheric Dispersion-Verification of Models b.
In Situ Measurement of Shear Modulus c.
Procedures in Evaluation of Ground Motion at Soil Sites (complete FY 75) 5.
Evaluation d.
Liquefaction Analysis Experiments Methodology (completeFY75)
Studies e.
Procedures for Conducting Cyclic Triaxial Tests on Soils (complete FY 75) f.
Fault-Dating Methods g.
In Situ Stress Measurement in Surface Rocks (completeFY76) h.
Earthquake Prediction 4+
E-14
__