ML20112B446
| ML20112B446 | |
| Person / Time | |
|---|---|
| Site: | Calvert Cliffs  |
| Issue date: | 08/31/1984 |
| From: | Adams W, Fisher H, Litke H ABB COMBUSTION ENGINEERING NUCLEAR FUEL (FORMERLY |
| To: | |
| Shared Package | |
| ML20112B406 | List: |
| References | |
| NUDOCS 8501100399 | |
| Download: ML20112B446 (120) | |
Text
y ATTACHMENT 5 EVALUATION OF INTERPELLET-GAP FORMATION AND CLA0 COLLAPSE IN MODERN PWR FUEL-ROOS EPRI Contract RP 2061-6 Task A2 Final Report, August 1984 -
Prepared by COMBUSTION ENGINEERING, INC.
. C-E Power Systems 1000 Prospect Hill Road Windsor, Connecticut n6095 Principal Authors -
W. M. Adams H. D. Fisher H. J. Litka W. J. Mordarski, -
Contributors -
M. A. Book L. V. Corsetti I. 9. Fiero ,
R. A. Matzie F. W. Sliz Prepared for Electric Power Research Institute 3412 Hillview Avenue .
Palo Alto, California c4304 EPRI Project Manager D. Franklin Materials and Corrosion Program Nuclear Power Division kDRA 0 b7 P PDM
w 4
NOTICE This report was prepared by the organization (s) named below as an account of work sponsored by the Electric Power Research Institute, Inc. (EPRI). Neither EPRI, members of EPRI, the organization (s) named below, nor any person acting on behalf of any of them: (a) makes any warranty, express or implied, with respect to the use of any information, apparatus, method, or process disclosed in this report or that such use may not infringe privately owned rights; or (b) assumes any liabilities with respect to the use of, or for damages resulting from the use of, any information, apparatus, method, or process disclosed in this report.
Prepared by Combustion Engineering, Inc.
Windsor, Connecticut
1 s
ARSTRACT This report presents the results from a review of interpellet-gap fonnation, ovality, creepdown and clad collapse data in modern PWR fuel-rods. Conclusions are reached regarding the propensity of modern PWR fuel to form such gaps and to undergo clad collapse. CEPAN, a creep-collapse predictor approved by the' NRC in 1976, has been reformulated to include the creep analysis of claddtpge with finite interpellet-gaps. The basis for this reformulation is discussed 1in detail. The model previously used in the calculation of the augmentation ,
factor, a peak ifnear heat rate penalty due to the presence of interpellet-dios within the fuel-red, has been modified to incorporate gap-formation statistf'cs from modern fuel. Finally, the benefits of the limited gap-formation and the CEPAN reformulation for the licensing of modern PWR fuel. cods are evale,arad.
, s. ~
, s k
, . (,'
e 1' %
s i L % t
(
v.
e N
N N
\
j l
V l h 1 k
-)
ACXNOWl.EDGEMENTS A special thanks is given to T. Dynak for her careful interpretation and typing of this report. -
t 9
O r
I
CONTENTS Section Page 1 INTRODUCTION 1-1 1.1 History 11 1.2 Purpose of Report 14 1.3 Scope of Effort 14 1.3.1 Assess Extent of Gap Formation and 1-4 Clad Collapse in Modern PWR Fuel Rods 1.3.2 Reformulation of CEPAN for Finite 1-5 Interpellet, Gap e
- 1.3.3 Evaluation of 'the Benefits du'e to CEPAN 1-5 and Gap-Model Reformulation for Modern PWR Fuel-Rod Licensing 1.4 Application and Benefits 1-6 2 GAP FORMATION DATA BASE 2-1
- 2.1 Introduction 2-1 2.2 Results 2-3 2.2.1 Gap-Formation Mechar,1cs 2-3 2.2.2 Pellet-Gap Data from the Palisades Reactor 2-6 2.2.3 Pellet-Gap Data from the Maine Yankee 2-6 Reactor 2.2.4 Pellet-Gap Data from the Fort Calhoun 2-6 Reactor
.2.2.5 Pellet-Gap Data from the Calvert Cliffs 2-14 Unit 1 Reactor
! 2.2.5 Pellet-Gap Data from the Oconee 2-14
! Unit 2 Reactor 2.2.6.1 Poolside Examinations 2-14 2.2.6.2 Hot. Cell Examinations 2-25 m
f&.= --
e q , --,-,- - - - . - + - -mq.r - --* + -- '
Section Page 2.2.7 Pellet-Gap Data from the Arkansas Nuclear 2-26 One Unit 1 2.2.8 Pellet-Gap Data from the Zorita Reactor 2-25 2.2.9 Pellet-Gap Data from the Zion Unit 1 2-32 Reactor 2.2.10 Pellet-Gap Data from the Surry Unit 2 2-32 Reactor 2.3 Interpellet-Gap Gamma Scanning Techniques 2-32
'2.4
Conclusion:
2-37 3 CLN1-COLLAPSE DATA BASE 3-1 3.1 Introduction 3-1 3.2, Profilometer Measurements of Ovality and 3-1 Rod Diameter 3.3 Fael-Rod Ovality and Creepdown Data 3-2 Evaluation
, 3.4 Conclusions . 3-3
~
4' CEPAN WITH FINITE LENGTH CORRECTION FACTOR 4-1 4.1 Introduction 4-1 4.2 Derivation and Validation of the Small-Gap , 4- 1 Finite-Length Correction Factor for the CEPAN Code 4.2.1 Model of the Finite-Gap Influence on 4- 2 the Evaluation of Collapse Time 4 4.2.2 Validation of the Saall-Gap Finite- 4-6 Length Correctios Factor for CEPAN s 4.3 Application to Pellet Supported C-E Fuel Rods '
a-17 5 AUGMENTATION FACTORS 5-1 5.1 Background and Introduction 5-1 5.2 Model.Parametars 5-2 5.2.1 Gap Distribution Characteristics 5-2 5.2.2 Gap Peaking Characteristics 54 5.2.3 Maximum- Gap Size 5-14 x
Section Page
- 5.2.4 Radial-Pin-Power Distribution 5-14
~
5.3 Model Description 5-14 5.4 Comparison of Model to Data 5-26 j 5.4.1 Comparison Between Calculations and 5-26 Measurements 5.4.2 Calculation of Typical Augmentation 5-28 Factors "
5.4.3 Comparison Between Models 5-28 5.5 Impact of Changing the Augmentation Factor 5-28 5.5.1 Setpoint Analysis for Analog System 5-32 Protected Plants 4
s
- 5.5.2 Setpoint Analysis for Digital Plants 5-35 5.5.3 Fuel and Rod Performance 5-35 5.5.4 ECCS Performance Analysis 5-36 5.E.5 Safety Analysis '
5-36 6 CONCLUSIONS AND RECOMMENDATIONS 6-1 6.1 Conclusions -
6-1
~
6.1.1 Evaluation of Data Base
- 6- 1 6.1.2 Revised CEPAN Analysis Results 6-1 6.1.3 Augmentation Factor Analysis Results 6-1 6.2 Recommendations 6-2 7 REFERENCES 7-1 I
i
)
S.
1 l
ILLUSTRATIONS Figure Page 4-1 Ovality Versus Time 4-3
, 4-2 Pressure Versus Time 4-3 4-3 Finite Length Correction Factor Versus 1/a for A- 7 v = 0.259 44 Collaase Time Versus Flux 4-9 4-5 Collapse Time Versus Pressure 4-10 4-6 Collapse Time Versus Tencerature 11
, 4-7 Collapse Time Versus Ovality 4-12 4-8 Collapse' Time Versus Outer Radius 4-13
~
4-9 Collapse Time Versus Wall Thickness 4-14
~
. 5-1 Relative-Gap-Size Distribution Using a Maximum 5-5
- Gap of 0.7 inches in Augmentation Factor Model
- 5-2 Rod Location Assignment to Radial Groups and 5-6 Limiting Single Gap Power Peaking 5-3 Effect of Burnup on Single-Gap Peaking Factors 5-7 5-4 Estimation of DOT Results by Combination of Single- . 5-10 Gap Peaking Effects '
i 5-5 Relative Effect of Gap size on Single-Gap Power 5-11 Peaking 5-6 Relative Power Versus Axial Distance From Gap 5-12 Interface 5-7 Relative Effect of Gap size on End Gap Power 5-13 Peaking 5-3 Axial Of stribution of Gans Over 0.1 inches in 5-15 l Prepressurized, Modern PWR Fuel Rods.
l
Figure Page 5-9 Maximum Gap Size vs. Core Height for Modern 5-16 Prepressurized PWR Fuel from Calvert Cliffs and Fort Calhoun 5-10 Maximum Gap Size vs. Core Height for Non- 5-17 pressurized, Densifying PWR Fuel From Palisades 5-11 Maximum Gap Size vs. Core Height for Non- 5-18 pressurized, Densifying PWR Fuel from Maine Yankee 5-12 Integral Radial Pin Power Distribution for a 5-19 Typical Reactor 5-13 Augmentation Factor Model Diagram 5-25 5-14 Gap Size Distribution for Modern Fuel 5-27 5-15 Augmentation Factors for a Typical Core 5-30 5-16 Augmentation Factor as a Function of Maximum 5-31 Gap Size 5-17 PFDL Plot - 90% Power CEA Insertion at 80C 5-34 e
G 9
0 L.
~ ~
TABLES Table Pace 2-1 Ginna Fuel Characteristics for Cycles IA and IB 2-2 2-2 Sumary of PWR Interpellet-Gap, Ovality and 24 Creep Data 2-3 Fuel-Rod Interpellet-Gap Measurements for the 2-7 Palisades Reactor 24 Fuel-Rod Interpellet-Gao, Ovality and Creep 2-11 Measurements for the Maine Yankee Creator 2-5 Fuel-Rod Interpellet-Gap, ovality and Creep 2-15 Measurements for the Fort Calhoun Reactor
, 2-6 Fuel-Rod Interpellet-Gap, Ovality and Creep 2-16 n .
fieasurements for the Calvert Cliffs-1 Reactor 2-7 Fuel-Rod Interpellet-Gap, Ovality and Creep 2-19 Measurements for the Oconee-2 Reactor 2-8 Fuel-Rod Interpellet-Gap, Ovality and Creep 2-27 Measurements for the Arkansas Nuclear One-1 Reactor j 29 Fuel-Rod Interpellet-Gap, Ovality and Creep 2-28 Measurements for the Zorita Reactor 2-10 Fuel-Rod Interpellet-Gap, Ovality and Creep 2-33 Measurements for the Zicn-1 Reactor 2-11 Fuel-Rod Interpellet-Gao, ovality and Creep 2-35 l
Measurements for the Surry-2 Reactor l 3-1 Oependence of Maximum Ovality and Average 3-2 01ametral Creep on Pressure and Fuel-Pellet Densification i 4-1 Standard Case Parameters a-a 4-2 Actual and Predicted Collapse Times 1-16 5-1 Sumary of Gama Scan Data 5-3 l 5-2 Effect of Multiple Gaps 5-9
. . _ _ _ . . -. ... .. .. - .. .~. . - . -. - . - - .
Table Page 5-3 Conceptual Basis for the Augmentation Factor Model 5-20 5-4 Random Number Decision Intervals 5-22 5-5 Input Data for Typical Augmentation Factor 5-29 Calculation O
f e
/
9 9
9 9
EXECUTIVE
SUMMARY
INTRODUCTION
- Fuel used in the initial cycles of early PWRs (i.e., Ginna, Point Beach, H. S. ~
Robinson, Palisades, Maine Yankee, Beznau, Stade, etc.) consisted of densifying fuel pellets in unpressurized fuel rods. Such fuel rods exhibited relatively large interpellet-gaps which in some cases led to clad collapse. More modern fuel, consisting of nondensifying fuel pellets in pressurized fuel-rods, has been loaded into later cycles of these early reactors and almost every cycle of ,
more modern reactors (e.g., Fort Calhoun, Calvert Cliffs, Arkansas Nuclear One, Zion, Surry, etc.). This latter fuel has exhibited only smail interpellet-gaps and no clad collapse.
This report reviews the measured interpellet-gap formation and clad collapse -
[ data to derive conclusions about the propensity of (nodern PWR fuel to form such l- gaps and to undergo clad collapse. A fannulation is presented to $alculate the i
time at which cladding collapse is predicted when finite sized gaps are .
l present. Finally, reconsnendations are made regarding the continued imposition of power-peaking penalties to accommodate these effects.
BACXGROUNO Three types of fuel-rods, similar to those that have been manufactured by C-E, provide the primary basis for the analyses reported here. De first type, I
called old fuel, was constructed from densifying 002 pellets in initially unpressurized fuel-rods. The second type, called intermediate fuel, was
[ ~ constructed from densifying 002 pellets in prepressurized fuel-rods. The l third type, called modern fuel, is constructed of nondensifying 002 fuel pellets in prepressurized fuel-rods. As used by C-E in describing its fuel,
!. densifying fuel pellets are those that, upon resintering, densify by I approximately 3f. theoretical ~ density (T.O.), while nondensifying fuel pellets l are those that densified by less than 0.5'. of T.O.
When the fuel-rods of Beznau, Stade, Ginna and Point Beach, which are non-C-E manufactured old fuel, were examined in 1972, regions of clad collapse up to four f aches long were observed. Old fuel from the first cycles of C-E's Palisades and Maine Yankee reactors were also examined. Although interpellet-gaps as large as 0.68 inches were observed, there was no sign of clad collapse. Examination of modern C-E fuel-rods after multiple cycles in the Fort Culhoun and Calvert Cliffs Unit i reactors revealed much smaller interpellet-gaps and, again no cases of clad collapse. Modern Babcock and
! Wilcox (B&W) fuel from Oconee-2 and ANO-1, along with Westinghouse (W) fuel ,
from Zorita, Zion and Surrey also show these latter characteristics.
PURPOSES The-purposes of this report are to:
e llevelop the data base for and evaluate the propensity of modern fuel to form interpellet-gaps,
- Determine a conservative representation of the statistical ,
distributions of the size and locations of interpellet-gaps, e Develop a formulation for incorporating the statistical representation of the gaps in modern fuel into a prediction of the time to clad collapse, *
- e Calculate the propensity of modern fuel to collapse, , ,
e Revise the formulation for the calculation of augmented-power-peaking factors based upon the statistical data for pondensifying-fuel to incorporate the statistical distribution of gaps in modern fue1, l
e Evaluate the impact of the revised augmentation factors l formulated on safety analyses and licensing concerns, and e Recommend actions for the inclusion of the densification effects -
of modern fuel in safety and licensing issues.
INTERPELLET GAPPING IN MODERN PWR FUELS Using a gamma scanning device, C-E examined seventy-nine rods with old fuel from the first cycle of Palisades, ten rods with old fuel from the first cycle of Maine Yankee, six rods with intermediate and nine rods with modern fuel from Cycles 1.through 5 of Calvert Cliffs Unit 1, and five rods with modern fuel
- from Cycles a and 5.of Fort Calhoun. The size and position distributions of the interpellet-gaps found in these fuel-rods show the old C-E fuels had a total of 1560 !nterpellet-gaps. Only twenty-one of these gaps were greater
than 0.10 inches, and none were larger than 0.68 inches. The intermediate and new C-E fuels had a total of 31 gaps with none greater than 0.022 inches.
Furthermore, C-E has noted that the net pellet column shrinkage of this latter fuel is on the order of 0.5 inches. All sets of data show an essentially random axial spatial distribution of the number and size of gaps.
C-E reviewed the reported interpellet-gap distributions from six reactors fueled by W and B&W. Their reports show that the old fuel that they manufactured had a high probability of developing interpellet-gaps, many of which were large, including some greater than A.0 inches in length. However, their intermediate and modern fuel-rods had smaller numbers of gaps per rod examined, as well as smaller gaps. The largest gap seen in R&W's modern fuel was 0.30 inches. Most of their reported gaps were smaller than 0.10 inches.
In some instances, W reported cumulative gap size rather than specific gao sizes. The largest cumulative gap size they reported for modern fuel was 0.161 inches. Most of their reported cumulative gaps and all of their specific gaps were smaller than 0.1 inches. C-E has concluded that for any vendor's modern fuel, i.e., prepressurized fuel-rods containing nondensifying UO2 pellets, a ,
conservative upper limit in size for interpellet-gaps probably can be set at
, 0.30 inches. '
. CLAD COLLAPSE IN MODERN PWR FUELS C-E has examined 2600 fuel-rods from its Palisades, Maine Yankee, Fort Calhoun and Calvert Cliffs Unit 1 plants and reviewed the clad-collapse reports on Ginna, Point. Beach, Zorita, Zion, Surrey, Oconee-2 and ANO-1, fueled by other vendors. While some'old fuels from W and B&W have shown clad collapse, no modern fuel manufactured by any vendor has exhibited this failure.
THE CEPAN CLAD COLLAPSE PREDICTOR The C-E computer code CEPAN was originally programed to conservatively predict the earliest time for clad collapse in PWR fuel-rods based on infinite interpellet-gaps. The newer version nf CEPAN incorporates the effects of finite gaps, and is applied to worst case calculations. The effects of these small gaps are incorporated into the fuel cladding collapse model by assuming that the critical variable in clad collapse is the clad-ovality and then correcting estimates based on infinite gaos to account for end support from small gaps.
The new method defines the estimated time to clad collapse as the time for achieving the critical clad-ovality. It then determines the time of critical clad-ovality for creepdown clad failure as the infinite gap CEPAN time to clad collapse multplied by the ratio of collapse pressure for pressure-induced clad failure with finite gaps to that collapse pressure with infinite gaps. From these theoretical considerations, C-E developed the new clad-collapse time predictor equation:
' CSS Tg = T yt for 0 < t/a 1 15 P
C15 Tg =
T;t for 15 < t/a where Tg = time to collapse with finite gaps T yg = time to collapse with infinite gaps ,
P = P CSS for a simply supported tube with an t/a ratio of 15 C15 P
=
Pressure required to reach the critical ovality for a simoly ,
supported tube, which is equal to:
P = 2
[Eh/a(1-v)3.[(1.y2)/(3(1+4(1/va)2))
CSS ,
2 2
+ (h /12a ) (3+(7-v)/(1+4(1/va)2)7 where E = clad tube modulus of elasticity 1
h = clad tube thickness
! v = Poisson's ratio 1 = interpellet-gap length a = midradius of clad wall Based upon these formulae and the specific gap formation data for modern fuel, C-E concludes that Zircaloy-clad modern fuel-rods, with an initial fill gas
! pressure greater than 300 psia, and a clad thickness to midsurface radius ratio (h/a) between 0.10 and 0.15, will not experience clad collapse within 10 years of operation.
, , . , , - ~ , - - .,,-v--n-- - - , , - - , , , , --e --m, - - . . - ,
AUGMENTATION FACTORS The increased power peaking due to interpellet-gap formation has been accommodated by a penalty on linear-heat-generation rates called the augmentation factor. While this factor is generally considered dependent, upon an appropriately determined set of near-neighbor infinite-gap radial-power-peaking factors and a given cycle's expected pin by pin power distributions, its calculated values have been dominated by the impact of the gap-size distributions considered. C-E plant specific augmentation factors have in the past been based on the very conservative interpellet-gap-distribution size statistics obtained from the ganina scans of old C-E fuel. These augmentation factors had ranged in peak value between 1.045 to 1.07, depending upon the particular plant being considered.
~
I
'In this investigation, augmentation factors are calculated for a given plant l power distribution and set of infinite-gap radial penalty factors. Different i size-and-position-distribution gap statistics were used to evaluate the imoact
! of the new-fuel gap-distribution effects. The variations considered maximum gap sizes from 0.4 inches down to 0.050 inches, and a random axial distribution of gaps in both number and size. These analyses resulted in an axially constant augmentation factor whose values varied from 1.012 for a maxi:5um gap size of 0.4 inches to 1.002 for a maximum gap size of 0.050 inches.
SUMMARY
AND CONC 1.USIONS
, C-E has examined its own fuei-rods containing old, intermediate and new fuel
! and has also reviewed the measurements reported for similar fuel-rods
- manufactured by W and 88W. It is concluded that, if C-E's modern fuel-rods l contain intarpellet-gaps smaller than 0.5 inches in length, which is 607, larger l than the maximum gap size observed in any vendor's modern fuel, the clad will not collapse within 10 years of ooeration. Furthermore if the gaa is smaller than 0.A inches, the augmentation factors will remain smaller than 1.012. More particularly, in those C-E modern fuel-rods that contain interpellet-gaps, the gaps are typically less than 0.05 inches. Therefore their augmentation factor is so small (<1.002) that it is insignificant relative to the uncertainty of power-peaking calculations or measurements and can be s.et e.;ual to 1.0 for all !
analyses and licensing applications. Similar non-collapse lifetimes and low augmentation factors should prevail for other vendor's modern fuel, but specific evaluations would be required to quantify these effects.
l
Section 1 INTRODUCTION 1.1 HISTORY Early pressurized water reactors (PWRs) used unpressurized fuel-rods made with fuel pellets subject to densification upon irradiation. Among the potential mechanical. consequences of fuel pellet densification are fuel-stack shrinkage ,
, and the formation of interpellet-gaps. Such gaps, theoretically, could be large enough to permit the fuel-rod cladding to collapse. Interpellet-gaps would also result in increases in the nuclear power peaking in adjacent pellets or adjacent ungapped fuel-rods.
During refuelings at Reznau, Ginna, H. B. Robinson, and point Beach in 1072, clad flattening in unpressurized fuel-rods was observed. The collapsed fuel -
, rod sections varied from 0.5 to 4.0 inches long and were distributed over the -
top 414 of the fuel stack (J,). This phenomenon was caused by shrinkage of l *the pellet column due to densification, followed by pellet hang up which l
i prevented settling of the colunn thereby leaving interpellet-gaps, and led to the subsequent collapse of the cladding into these unsupported regions. The i
pellet shrinkage occurred in a relatively short time at low burnups (f,4 GWd/T) and at temperatures well below those required to cause densification in l an ex-reactor sintering test. In a later report (2), it was concluded that
! under typical PWR operating conditions, cladding collapse in an unpressurized fuel-rod could occur in the first cycle due to this mechanism.
These observed clad collapses led to the imposition of design criteria that
- would assure that specific effects would not occur (i.e., no clad collapse was
-permitted) or were accounted for (i.e., through imposition of augmentation-factor power-peaking penalties). Each NSSS or fuel vendor was allowed to n develop his own analysis methods to, meet these criteria, subject to NRC approval.
l m
l Until now, Combustion Engineering's (C-E's) methods have been based principally upon the data for the Palisades Q) and Maine Yankee Q) reactors. These data were derived from measurement of irradiated, unpressurized fuel-rods made with densifying fuel pellets. The data were the basis for the representation of an interpellet-gap distribution that had monotonically increasing sizes and decreasing numbers with core height. C-E developed the clad-collapse time-predictor, CEPAN (4), which assumed that the interpellet-gaps would be essentially infinite and that the fuel pellet ends would not support the clad.
C-E's fuel manufactured during this time satisfied the no-clad-collapse require-ments, but was required to have peak augmentation-factor power-peaking penalties ranging from 4.5% to 7.0%.
In August,1972, C-E initiated a series of experimental and analytical programs to: (a) determine the mechanism responsible for in-pile densification; (b) develop fuel fabrication processes and fuel-rod designs to reduce or eliminate the occurrence of densification; and (c) develop imoroved methods for predicting densification and its consequencas Q,7). The exoerimental programs which incipded fuels with different microstructures shcwed that the occurrence of in-reactor densification was related to the as-fabricated .
porosity distribution in a given fuel type Q). Fuel types prone to densification were shown to (1) contain a significant fraction of pore-volume made up of small, closed pores (< 4 um in diameter) and (2) to have a grain size less than 5 um. Densification was observed to decrease sharply as tnis small porosity was eliminated. The mechanism identified as causing in-reactor densification was annihilation of the small porosity through fission-spike overlapping and enhanced diffusion.
As a result of the experimental program, several design and fuel manufacturing changes were identified to minimize fuel densification and its effect on performance @). The most important of these included:
- e. Fuel-rod prepressurization to forestall cladding creepdcwn. The use of thicker cladding is another but less effective means to accompitsh the same objective.
e Changes in the fuel fabrication process that result in fuel with ',
very few pores smaller than 4 um and an average minimum grain size of 5 um.
e Development of a resintering test that could be used as a basis for predicting the extent of in-reactor density change due to densification.
I JL-2
,e The resinter test was benchmarked extensively for a variety of fuel types and microstructures using the results from the in-reactor experimental programs (9). This thermal treatment is now performed routinely on a statistical sampling of pellets from each fabrication lot of fuel to confirm adequate fuel
, stability with respect to densification.
Fuel fabricated by C-E since 1974 has been stable with respect to densification.
For example, fuel fabricated prior to 1974 showed typical density changes
- during resintering of about 3', of theoretical density (T.O.). In contrast, fuel fabricated since that time has shown density changes of 10.57. T.O.
Since 1973, C-E has conducted over thirty different inspection programs as part of fuel performance surveillance activities at several commercial power reactors. Approximately 600 different fuel assemblies with average burnups ranging'from s1 GWd/T to 52 GWd/T have been visually examined. Approximately i
seventy of these assemblies were of the pre-1974 design (i.e., nonpressurized fuel-rods containing densifying fuel). Slightly more than 60% of the remaining assemblies contained fuel-rods of modern design (prepressurized with non-densifying fuel) while the balance contained fuel-rods of a transitional t;ype .
. that included prepressurization with densifying fuel. No regions of collapsed 1
i cladding were observed during any of these visual examinations (10,).
During these examinations, approximately 2600 individual fuel-rods in reactor spent-fuel pools were inspected using a variety of other techniques as well.
Approximately 500 of these rods were of the pre-1974 design with the remainder being either of the transitional or modern pWR design. Gama scanning of s100 rods of the pre-1474 design showed the largest fuel column gap to be M.7 inches but showed no evidence of clad collapse (a,,9). profilometry measurements of the outer diameter of s100 fuel rods, with average burnuos from 10 to 56 GWd/T, have shown no evidence of clad collapse in prepressurized fuel-l rods even when densifying fuel was used (11 - 14),
l l
l Extensive hot-cell examinations have been completed on twenty-seven pre-pressurized test fuel-rods frcm Calvert Cliffs-1, as part of a joint C-E/Ep;1I I
, Fuel performan'ce Evaluation program, and on twelve rods from Fort Calhoun, as part of a joint CE/0 ppd /00E program ( M , 20 - 23). These rods l included different fuel types with varying propensities for in-reactor densi fication. The programs included gamma scanning of twenty five fuel-rods
_ _ , . . . _ _ _ . . . _ . W
with average burnups as high as 54 GWd/T. Only a few cases of gap formation were observed in rods containing densifying fuel (maximum gap of 0.2 inches),
and no gaps larger than 0.025 inches were found in rods containing modern PWR fuel.
As a result of the many fuel performance evaluation programs conducted by C-E over the last several years, the following conclusions regarding densification and its effect on fuel rod performance can be made:
e Fuel-rod prepressurization reduces the extent of clad creepdown, thereby allowing restacking of the fuel column in the event of densification, e Prepressurizaticn and increased clad thickness are effective in forestalling creep collapse in all rods including those containing densifying fuel.
s . Fuel can now be fabricated with practically no propensity for in-reactor densification thereby reducing the probability of gap formation in the fuel column during irradiation.
e Modern PWR fuel pellets used in conjunction with prepressurized fuel rods, are resistent to clad collapse.
1.2 PilRPOSE OF REPORT ,
The overall purposes of this report are to:
- 1. Sumarize data on the extent of gap formation and collaps- in modern ,
PWR fuel-rods (i.e., fuel rods fabricated with nondensi.fing fuel and prepressurization). This information will be used to support an update of CEPAN and as input for the evaluation of augmentation factors as they apply to modern PWR fuel.
- 2. Describe-the reformulation of the CEPAN code to include the creep analysis of cladding with finite interpellet-gaps.
- 3. EvaluatethebenefitsoftheCEPANreformulationan$limitedgap formation on modern PWR fuel-rod licensing.
1.3 SCOPE OF EFFORT 1
1.3.1 Assess Extent of Gao Formation and Cladding Collaose in Modern pWR '
Fuel-Rods .
The tworkscope for this subtask included:
L
- 1. Review, reduction and interpretation of data applicable to modern pWR l fuel-rod designs, using pertinent results from both in-reactor and ex- ;
reactor experiments.
k'l
- 2. Statistical analysis of gap-size distributions and propensity for gap formation as a function of rod-design parameters (e.g., internal pressure, fuel densification).
- 3. Characterization of creep /ovalization behavior as a function of gap size and rod-design parameters.
The results will provide a data base for use in the CEPAN reformulation and the licensing impact subtask.
1.3.2 Reformulation of CEPAN for Finite Intercellet-Gao '
The CEPAN code is a finite-difference code for evaluating the creep deformation of initially unsupported fuel cladding. Because a hypothesis of generalized plane strain is used in the CEPAN theoretical formulation, the axial dependency ,
4 in the governing system of equations was eliminated, resulting in a two-dimensional formulation involving the radial and angular coordinates of the tube's cross section. Thus, the previous model was appropriate for cladding with large interpellet-gaps. The analysis of smaller gaps using a three-dimensional mesh would require the retention of the axially-dependent terms in the equilibrium equations for a thin circular shell. The addition of terms and mesh in the axial direction would result in a code capable of ealculating creep deformation in PWR fuel-rods containing interpellet-gaps of any size.
The workscope of this subtask consisted of:
- 1. Performance of an initial evaluation to determine the most desirable method of modifying the CEPAN code for finite interpellet-gaps, i.e.,
using (1) an empirical correction factor that would account for varying finite gap sizes or (2) using the modified theoretical three-dimensional formulations.
- 2. Development of appropriate corrections or theoretical formulatisns and inclusion of their corresponding logic in CEPAN. Parametric studies were performed using.the modified CEPAN for various geometric, loading, and material values to demonstrate proper code functioning.
I Comparisons to data were included.
1.3.3 Evaluation of the Benefits due to CEPAN and Gao Model Reformulation for Modern PWR Fuel-Rod Licensing i
, The workscope for this subtask consisted of: '
- 1. Documentation of the methods used to calculate the power peaking augmentation factor and a review of how this factor affects core design.
, ,. -. --- - .-.- .?----. . - -
- 2. Development of an appropriate relationship of the size, frequency and distribution of gaps in the core for use in a revised augmentation factor calculation. New gap statistics were incorporated into the design methodology and appropriate modeling changes were made to properly reflect improved characteristics.
- 3. Evaluation of changes in single gap peaking factors and end gap peaking factors to encompass extended burnup enrichments (up to 4.1 wt%) and exposures (up to 51.0 GWd/T).
4 Performance of generic analyses for a typical PWR (based on C-E design) using the improved gap statistics; i.e., the determination of new augmentation factors.
S. Evaluation of the effects that the new augmentation factors have on the safety and setpoint analysis for a generic PWR (based on C-E design) to determine the overall benefit of the improved gap statistics.
1.4 APPLICATION AND BENEFITS The analysis of interpellet-gap size and clad collapse data for modern (non-densifying) fuel pellets in prepressurized fuel-rods leads to the conclusion that interpellet-gaps will be less than 0.5 inches in length and that clad collapse will not occur for the useful life of the fuel. As a consequence, the probability distribution of gap sizes used-to calcula*te augmentation factors has been redefined so that most gaps are small with an upper limit of 0.5 inches and these gaps occur randomly with respect to thef e position in the core. This changes the augmentation factor spatial distribution from monotonically increasing with core height to a constant, independent of core height, and decreases the peak value from approximately 1.05 to less than 1.012. This change will affect both the evaluation of the LOCA limit and the monitoring of core performance to avoid violating the LOCA limit.
The net result of applying augmentation factors based upon the revised data base and procedures will be an operational margin gain.
M ~
Section 2 GAP FORMATION DATA BASE
2.1 INTRODUCTION
Early PWR's were either partially or fully loaded with unstable, densifying, fuel pellets contained in nonpressurized cladding tubes. The shortcomings of
-this design were recognized (1) at the refueling of the Ginna Reactor in early 1972, when some of its nonpressurized, irradiated fuel-rods were found to 4
have flattened or collapsed clad sections in regions where interpellet-gaps ranging from 0.5 to 4.0 inches had formed. Fuel densification and pellet l
relocation accompanied by clad ovalization had created these interpellet-gaps.
The high primary-system coolant pressure, 2200 psi, caused the fuel cladding to creepdown until the movement was arrested by pellet clad contact. The lack of pellet support in gap. regions longer than n.5 inches allowed creepdown to continue until clad-collapse over the voided fuel column regions occurred. No -
- fuel rods with collapsed sections were found among the prepressurized rods examined at that refueling. A summary of the Ginna fuel characteristics and the observed fuel-rod collapse statistics are presented in Table 2-1. ,
- Fuel behavior similar to that of the Ginna reactor was also observed (1) at i the Point Beach Unit One and the H.R. Robinson reactors. For example, in the Point Reach reactor sixty-six fuel-rods in one region of the core had collapsed fuel-rod sections. None of these collapsed sections were longer than 4 inches in 1ength and no rod contained more than one collapsed section. All of the l failed rods were nonpressurized. No collapsed sections were observed in those
! regions of the core containing pressurized fuel-rods, i
C-E's Palisades and Maine Yankee reactors, which also were initially loaded ,
with densifying fuel pellets in nonpressurized fuel cladding tubes, also had fuel rods exhibiting fuel-column gaps during gamma scanning (1, 3,).
However, these rods did not show signs of clad collapse.
l l
l l
~ _ - . . - - , , , , ~ . - - . . . - . . - . . - . . . - - - . . - - - - - - -
TABLE 2-1 GINNA FUEL CHARACTERISTICS, for CYCLES 1A AND 1B Region l 1 2 3 4 Number of Assemolies Cycle 1A (Initial fuel loadings) 41 40 40 --
Cycle 18 (Loading after Feb.1972 shutdown) 41 40 28 12 Rods per Assembly 179 179 179 179 Enrichment, 5 of U-235 2.44 2.78 3.48 3.20 Clad 00,'(In) 0.422 0.422 0.422 0.422 Diametral Fuel-to-Clad Gap Thickness, mils 6.5 6.5 6.5 6.5 Pellet Density, % Theoretical 94 92 90 92 Helium Prepressurized No No No Yes
SUMMARY
OF OBSERVATIONS OF GINNA FUEL AT APRIL,1972 SHUTCOWN Region 1 2 3 4 Number of Assemblies 20 22 11 1 Examined Number of Fuel Rods 1040 1144 572 52 Examined
-Number of Rods with 21 83 20 0 Collapsed Sections Percent cf Rods with 2.0 7.3 3.5 0 Collapsed Sections
.i B
4 Subsequent to these evaluations, C-E, Westinghouse (W) and Babcock and Wilcox (88W) determined the size and numerical distributions of interpellet-gaps for
, different fuel designs at seven of their other operating plants. A summary of the measurements at the ten plants is given in Table 2-2.
All of the gap size measurements for noncollapsed fuel rods were made with gama scanners. Each vendor's scanners had a different aperture and scanning rate, but all scanners were considered adequate for detecting gaps larger than 0.10 inches.
This section of the report sumarizes and evaluates the published data on interpellet-gap formation in old and modern PWR fuels to provide a basis for assessing the impact of gap formation on fuel-cladding creep-collapse behavior and on local power peaking caused by the gaps in the fuel-pellet columns of modern fuel-rods.
2.2 RESULTS 2.2.1 Gao Formation Mechanisms Fuel-pellet. dimensional stability under irradiation and fuel-rod internai -
pressure are the two dominant fuel-rod design parameters that affect inter-pellet gap formation, fuel-cladding ovalization and creep-collapse behavior.
~
Dimensionally stable fuel pellets are defined as those that undergo insignificant further densification, f.e., less than 0.5% T.D., during
- feradiation. They are identified as "nondensifying fuel". Dimensionally unstable pellets undergo a further increase in density, aften up to, and occasionally beyond, 3% of T.D., under irradiation and hence shrink in length and diameter. They are identified as " densifying fuel".
f Axial shrinkage would create interpellet-gaps in a column of fuel pellets even if the pellet did not move. Radial shrinkage increases the pellet to clad diametral gap and subsequently permits greater clad ovalization. Since pellet clad contact due to ovalization does not take place uniformly, pellets became immobilized randomly as firm clad contact is established. Pellets located between immobi}ized pellets tend to restack, producing a larger gap below an immobilized pellet.
2-3 - - .
M 4
6 1ABLE 2-2 i SteelAlly 0F PWR INlERPELLET-CAP, OVAlliY, AND EREEP DATA Initial , Average Fuel fuelgod Burnup Number Cap Slae Diametral Number of Namlegg Number of Rods Type Pressure Range of Range Ovality Creep
- Vendor Reactor Rods Examined with No Caps (e) (psig) Caps (CWJ/MIU) (Inches) (loches) (Inches)
- 8. Nonpressurland - Densifyine Fuel C-E Palisades 79 4 D 15 3.77-8.42 1499 0-0.7 Not Measured Not Neesured
- N4ine Yankee 10 0 D 15 9.7-13.5 59 0-0.4 .0128 .0034 Wettinghouse Glnna 2756 Not Reported D 5.6-18.0 Not Reported 15 0-4.0 Not Reported Not Reported fl. Nonpressurland - Nondensifylog Westinghouso Zorita 15 12 ND 15 13-52.5 12 0 .86 Not Reported .003 III. Pressurfacd - Densifying C-E SC&E 6 450 1 D 18.7-49.7 12 .005 .022 .013 .0046 D43 Oconee 2 12
- 5 ,0 360 12.0-25.4 113 .000 .792 .019 .0033 (a) D = Densifying NO = Hon-densifying (b) Dwality = 0.25 (man. OD - min. 00)
(c) Caps on two of the rods were measured at poolside at the end of cycles 1 and 2 and also in a hot cell f acility at the esul of cycle 2
2 5
TAatE 2-2 (Continued) -
- SIMHARY Of PNR lNifflPELLET-C P, OVALITY, Ate CREEP DATA Initial Average fuel fuel Rod Burnup Number Cap Size Number of Number of Rods Type Pressure Range of Manig Diametral Range Ovality Creep i Vendor Reactor Rods Emanined with No Caps ~ (a) (psig) (CWJ/NTU) Caps (Inches) (laches) (Inches) i
{
IV. Pressurized - Nondensifyine .
C-E BG&E 9 4 ND 300/450 18.7-54.1 11 .004 .014 0043 .0046 I fort Calhoun 5 3 ND 400 37.2-45.5 f
4 006 .018 Hot Reported .004 Westinghouse Zorita 19 17 ND 500 29.6-54.0' 7-8 0.161
- Not Reported Not Reported Zion Unit 1 15 3 ND 450 12.4-31.0 32 .020 .150 001 .0031 Surry Unit 2 17 15 ND 450 6.3-9.1 2 .054 .097 .0075 .001 00 Oconee Unit 2 Il
- I le 360 11-25.2 29 .01 .300 .0023 .0027 ANO-1 8 8 Nu 360 18.0 0 No Caps Not Reported Hot Reported (a) 0 = Consifying ,
t2 = Nondensifying ,
(b) Ovality = 0.25 (man. 00 - min. 00)
(c) Sum of several gaps (d) Sum of two gaps (s) Caps of two of the rods were measured at poolside at the end of cycles 1 and 2 and also in a hot cell facility at the end of cycle 2 -
i I
.- ____ . ~.
The second dominant parameter, fuel-rod pressurization, reduces cladding stresses due to the pressure differential between the external primary system coolant pressure, and the fuel-rod internal pressure. This stress reduction tends to diminish tube ovality and cladding creepdown thereby allowing densification to be completed prior to firm clad-pellet contact. This, in turn, allows pellet restacking over the entire fuel column, resulting in the minimization of the sizes of the residual interpellet-gaps.
Thus, the use of nondensifying fuel and fuel-rod pressurization wuuld be expected to be effective in preventing the occurrence of large interpellet-gaps.
2.2.2 pellet Gao Data from the Palisades Reactor (Combustion Engineering) (,3)
A total,of seventy-nine corner rods fecm twenty fuel assemblies were gamma scannci at the end of the first cycle of irradiation. These rods were not
. pressurized and contained densifying fuel. Of the gaps reported (1499), only 17, were greater than 0.10 inches. The rods in which they appeared are listed
~
in Table 2-3. Because of its location, the 2.160 inch gap reported in rod A38 NW has long been considered a loading or handling anomaly and has been excluded -
from all analyses that use the palisades data. The largest remaining gap is 0.68 inches long.
2.2.3 Pellet Gao Data from the Maine Yankee Reactor j (Combustion Engineering) (,4)
Ten fuel-rods from Maine Yankee Core 1 fuel were gamma scanned at the end of one cycle of irradiation. All of the rods were nonpressurized and contained densifying fuel. The burnup levels for these rods ranged from 9.7 to 13.6 GWd/T. The data are presented in Table 2-4 The largest gap found measured 0.40 inches, while only seven of the fifty-nine listed gaps were greater than 0.10 inches in length.
2.2.4 Pellet Gao nata frem the Fort Calhoun Reactor (Ccmbustion Engineering) Q6) 6 Two fuel-rods from assembly 0038 and three reds frem assembly 0005 were gamma scanned at the end of four and five cycles of irradiation, respectively. These !
rods were prepressurized to 400 psi, contained nondensifying fuel and I accumulated burnups in the range frmn 37.2 to 48.3 G'Jd/T. The data are
Table 2-3 fuel-Rod Interpellet-Gap Heasurements for the Palisades Reactor VENDOR: C-E PeIIet dia. (inchesl . 417 fuel Batch: A, B, and C length (inchesh . 600 N
IDENTIFICATION Fuel Fuel U Pellet Rod M Average Internal End Gap B- Maximus Diametral fuel Fuel-Rod Density Pressure of Burnup Sizes E Creep Assy No. (% TD) Type (psig) Cycle Ovality)
(mwd /MTU) (inches) R (inches (inches)
A22 NE 93.5 0 15 6.94 .125 Not Reported 1 1 Not Reported
<.060 23 SE 93.5 0 15 1 6.94 <.060 6 " "
SW 93.5 D 15 1 6.94 <.060 39 " "
Nil 93.5 0 15 1 6.94 <.060 12 " "
A25 NE 93.5 D 15 1 6.94 .680 1
, <.060 1 6
NW 93.5 0 15 1 6.94 <.060 7 " "
SE 93.5 i 0 15 1 6.94 <.060 18 " " '
l SW 93.5 D 15 1 6.94 <.060 3 " "
O E
4
Table 2-3 (Continued)'
Fuel-Rod Interpellet-Gap Heasurements for the Palisades Reactor i
VENDOR: C-E Pellet dia. finchesl .417 fuel Batch: A. B. and'C len9thLinchesh .600 i
N IDENTIFICATION Fuel fuel , U Pellet Rod M Average Internal End Gap 8 Maxianus Diametral fuel Fuel-Rod Density Pressure of Burnup Sizes E Assy Creep ;
No. (1 TO) Type (psig) Cycle (HWd/MTU) (inches) R Ovality)
(inches (inches)
A12 NE 93.5 0 15 6.38 1 .125 1 Not Reported Not Reported
<.060 36 NW 93.5 D 15 1 6.38 <.060 7 " "
SE 93.5 0 15 1 6.38 <.060 22 " "
SW 93.5 D 15 1 6.38 <.060 14 " "
A35 SE 93.5 D 15 1 6.1$ .140 1
<.060 2 NE 93.5 0 15 1 6.18 <.060 16 " "
HH 93.5 0 15 1 6.18 < 060 11 SW 93.5 D 15 1 61.8 <.060 3 " "
O e
6
M Table 2-3 (Continued) fuel-Rod Interpellet-Gap Heasurements for the Palisades Reactor VENDOR: C-E Pellet dia.
~
(inches) ,417 '
fuel Batch: A, 8, and C length (inches) .600 N -
IDENTIFICATION fuel fue1~ U
- Pellet Rod M ' Average Internal End -
Gap 8 Maxima Diametral fuel fuel-Rod Density Pressure of Sizes E Ovality ;
Creep Assy No. (1 TD) Type (psig) Cycle (HWd/HTUBurnup (inches)) R (inches)
(inches) -l A38 HW 93.5 0 15 6.32 2.160*I I Not Reported 1 1 Not Reported ,
<.060 3
'NE 93.5 D 15 1 6.32 <.060 4 SW 93.5 D 15 '
1 6.32' 0 " "
SE 93.5 0 15 1 6.32 - 0 " "
l 801 NW 93.5 0 15 1 7.0 .130 1
<.060 12
~
SW 93.5 D 15 1. 7.0 .330 1 SE 93.5 D 15 1 7.0 <.060 12 " "
~
- IdI Considered to be anomalous *
~
t ,
P f
" 'I
, ,. , /
1 e
i Table 2-3 (Co,ntinued)
Fuel-Rod Interpellet-Gap Measurements for the Palisades Reactor VENDOR: C-E Pellet dia. (inches) .417 '
Fuel Batch: A, B and C len9th (inches) .600 N
IDENTIFICATION fuel fuel U >
Pellet Rod . H Avera9e i internal End Gap B Haximum Diametral Fuel fuel-Rod Density Pressure of Burnup Sizes E Creep
} Assy No. (% TD) ' Type (psig) Cycle Ovality)
(PWd/NTU) (inches) R (Inches (inches) 850 SE 93.5 0 15 81.2* .545 1 1 Not Reported Not Reported
.370 1
<.060 13 i SW 93.5 D 15' 1 81.2' .490 1
<.060 13 NE 93.5 0 15 1 81.2 <.060 2 " "
NW 93.5 l} 15 1 81.2 <.060 17 " "
! B67 SW 93.5 D' 15 1 7.7 .120 1 1
<.060 38 -
SE 93.5 0 15 1 7.7 <.060 20 " "
NW 93.5 D 15 1 7.7 <.060 19 " "
i NE 93.5 0 15 1 7. 7 - <.060 50 " "
j forty e19ht additional rods examined in nineteen, fuel assenh11es did not contain any gaps lar9er than 0.060 inches.
i 1
l .
Table 2-4 fuel-Rod Interpellet-Gap. Ovality, and Creep.h asurements for the m ine Yankee Reactor VENDOR: C-E Pellet dia. (inches) .3795 fuel Batch: Core 1 (ABC) length (inches) . .650 TliENT11TffiT6N fuel Fuel Average Pellet Rod Diametral Creep Internal End Gap hximum at h ximum fuel fuel-Rod Density Pressure of Burnup Sizes Ovality Assy No. Type Ovality
(% TO) (psig) Cycle (mwd /HTU) (inches) (inches) (inches)
A047 IIBV-007 93 0 15 11.9 .005 1 .016 .0029
.007
.006
.013
.005
.013
.009
.125
.011 -
.007
- .007-
.005
.004
.005
.005 '
.004
.004
.007 A047 Il8V-198 93 D 15 11.5 .280 1 .010 .0026
.014 C231 KCA-109 93 0 15 9.7
, 1 . .300 .010 Perforated
.014
Table 2-4 (Continued)
' fuel-Rod Interpellet-Gap. Ovality, and Creep Measurements for the Maine Yankee Reactor ~
VENDOR: C-E .
PeIIet dia. (inches) , '.3795 fuel Batch: Core 1 (A,B,C) length (inches) .650 lh(HTTFTEXT10N Fuel Fuel Average Pellet Rod Diametral Creep Internal End Gap Maximum at Maximum fuel Fuel-Rod Density Pressure of Burnup Sizes Ovality Assy No. Type (psig) Cycle-Ovality
(% TO) (mwd /NTU) (inches) (inches) (incl.es i 8042 JBP-003' 93 0 15 13.6 .009 1 .0052. .0021 8042 JRP-004 93 1 0 15 1 13.2 .400 .0128 .0034
.007 ,
B042 JHP-005 93 D 15
- 13.2 .220 .0116 1 .0026 B042 J8P-027 93 D 15 12.8 .023 .0064 1
.0015
.023
.004 B069 JCH-199 93 D 15 12.8 .071 1 .0064 .0021
.013
.018 8069 JHY-157 93 D 15 12.9 .004 1 .0106 .003
.007
.004
.009
.009
.009
.022
, .009 4
i
Table 2-4 (Continued)
Fuel-Rod Interpellet-Gap. Ovality, and Creep Measurements for the Maine Yankee Reactor VENDOR: C-E Pellet dia. (inches) .3795 Fuel Batch: Core 1 (A,8, and C) length'(inches) .650 TriERYTFICITION . Fuel fuel Average -
Pellet Rod Diametral Creep Internal End Gap Maximum at Maximum fuel fuel-Rod Density Pressure of Burnup Sizes Ovality Assy Ovality No. (1 TD) Type (psig) Cycle (Md/MTU) (inches) (inches)
, (inches)
~ 4
.006
.005
.006
.007 C231 KCA185 93 D 15- ! 9.8 .150 .0055 .0029
.120
, .021
' .007
. .008 i .008
.010
.101
.007
.010
.021
.012
.015
.008
.008 b
S
y _ _ __.__ n _ _ _.._ _ . __ _ _ . _ _ _ _ _
. -.y. . . _ . - .
, presented in Table 2-5. Four small gaps were observed in each of the two rods examined at the end of the fourth cycle. The largest gap measured 0.018 inches. The three rods gamma scanned at the end of their fifth cycle contained no measurable gaps.
2.2.5 Pellet Gap Data from the Calvert Cliffs !! nit 1 Reactor (Combustion Engineering) (g - g, 3)
Fifteen fuel rods from Batch R fuel assemblies were gama scanned after one to five cycles of irradiation with individual rod burnups ranging from 18.7 to 54.1 GWd/T. All of the rods were prepressurized to 450 psi with one exception, a rod which was pressurized to 300 psi. Six of the rods contained densifying fuel. The data are presented in Table 2-6. The largest reported gaps for the nondensifying and the densifying fuel were 0.014 and 0.022 inches, respectively. Five of the rods, one of which contained densifying fuel had no measura'ble gaps. The behavior of rod AHS-085, which was pressurized to 300 psi, was no different from the rest, which were pressurized to 450 psi.
Interpellet-gap differences between rods containing densifying and non-densifying fuel were insignificant. The gap sizes and their spatial '
distribution appear to be about the same irrespective of the number of cycles.
of irradiation.
2.2.6 pellet Gao Data from the Oconee Ifnit 2 Reactor (Babcock & Wilcox) (25 - 27) 2.2.6.1 poolcide Examinations Eight fuel-reds were gamme scanned at the end of Cycle 1 and then again at the end of Cycle 2 for interpellet-gap evaluation. Four of the rods contained non-densifying fuel and four contained densifying fuel. All of the rods were prepressurized to 360 psi and had fuel burnups that averaged 12 and 25 GWd/T at the end of Cycles 1 and :.3 respectively. The data are presented in Table 2-7 For the four rods contaim.7g nondensifying pellets, the largest gap found at the end of the first cycle Ws 0.3 inches. The largest gap at the end of the second cycle was 0.2 inches. One rod had no. measurable gaps in either the first or second cycle. For the four rods containing densifying fuel, the largest gap found at the end of the first cycle was 0.2 inches. However, at the end of the second cycle a 1.0 inch gap was found in one of the rods and a 0.5 inch gap in a second. Since both the nondensifying and the densifying 4
i Table 2,-5 fuel-Rod Interpellet-Gap, Ovality,' and Creep Measurements for the Fort Calhoun Reactor VENDOR: C-E Pellet dia. (inches) .380 l Fuel Batch: 0 length (inchesh .450' ThENTTFTCITION fuel fuel Average Pellet Rod Diametral Creep Internal End Gap Maximum at Maximum i Funi Fuel-Rod Density Pressure of Burnup Sizes Assy No. (1TO) Type (psig) Cycle Ova 11tp Ovality (mwd /NTU) (inches) (inches I (inches)
D038 KHN090 95 NOI *I 400 4 37.2 .009 .0024 .0044
. .006
.008
.006 038 KMN098 95 ND 400 4 37.8 .005 0024 .0035
- .004
.018
.006 0005 KJ0125 95 ND 400 5 47.4 No Gaps .001 .0035 0005 KJE076 95 ND 400 5 48.3 No Gaps .001 .0037 0005 KJE085 95 ND 400 5 , 47.0 No Gaps .001 .0037 I*I ND = Nondensifying e
4
Table 2-6 -
fuel-Rod Interpellet-Gap. Ovality, and Creep Heasurements for the Calvert Cliffs-1 Reactor VENDOR: C-E Pelletdia.(inches) .3795 fuel Batch: B length (inches) .450 TDENTITTCATION Fuel Fuel Average Pellet Rod Diametral Creep Internal End -
Gap Maximum at Maximum Fual fuel-Rod- Density Pressure of Burnup Sizes Ovality Ovality Assy No. (1 TD) Type (psig) Cycle (HWd/HTU) (inches) (inches) (inches) ;
BT01 NBJ-024 93 DI ") 450 1 18.7 .010 n.m.(b) ,,,,
. .015
.011 05 BT02 Alls-009 93 D 450 2 25.8 .009 .0097 .0038
. .017 BT03 11 Alls-Oll 95 D 450 2 ' 33.0 n.m. .0118 .004 3 .007
.022 .013 .0045
.007 BT03 09 Alls-016 13 D 450 2 25.8 n.m. .0114 .004 3 33.0 n.m. .0095 .0045 4 41.4 No Gaps n.m. n.m.
NBD-096 0 450 4 41.4 .006 .006 n.m.
.005 D = Densifying -
n.m. = Not Heasured ~
______m_ _ _ _ _ _ - _ _ _ __
i Table 2-6 (Continued)
Fuel-Rod Interpellet-Gap, Ovality. and Creep Measurements for the Calvert Cliffs-1 Reactor.
VENDOR: C-E Pellet dia. (inches) .3795 Fuel Batch: B length (inches) . .450 THENTTFTfKTION Fuel Fuel Average Pellet Rod Olanetral Creep Internal End Gap Maximum at Maximum Fuel Fuel-P.ad Density Pressure of 'Burnup Sizes ovality Assy No. (% TO) Type (psig) Cycle Ovality (mwd /MTU) (inches) (inches) (inches)
BT01 50 Alls-020 95 ND 450 18.7 No Gaps .0043 1 .003 BT01 47 93 ND 450 21.6 1 No Gaps n.m. .003 Alls-079 2 n.m. .0039 .0027 BT02 51 Alls-021 95 ND 450 2 25.8 No Gaps n.m.
.0039 BT02 46
- Alls-080 93 ND 450 n.m. n.m. .003 1 .003 2 29.2 No Gaps n.m. .0027 BT03 53 Alls-023 95 ND 450 2 n.m. n.m. .003 .0038 3 33.0 .008 0028 .0033 BT03 42 Alls-074 93 N0 450 2 n.m. n.m. .0046 .0035 3 37.0 .004 .004 .0033
.005
.005 l
l l
l l
1
^
-~. ,
t v= 1
{ iable2-6(Continued) '
Fuel-Rod Interpellet-Gap, Ovality, and Creep Heasurements for the Calvert Cliffs-1 Reactor i VENDOR: C-E Pellet dia. (inches) .3795 Fuel Batch: B len9th (inches) .450 ThENY1TTTATION Fuel Fuel _ Average Pellet Rod i
~~~ Diametral Creep Internal Fnd '
Gap Maximum at Maximum i Fus) Fuel-Rod Density Pressure df Burnup Sizes Ovality Ovality Assy No. (1 TO) Type (psig) Cycle (HWd/MTU) (inches) (inches) (inches)
BT03 54 4
Alls-024 95 ND 450 2 n.m. n.m. .004 .0039 3 n.m. n.m. .0032 .0033 4
41.4 .005 n.m. n.m.
i
' .008 0 ,
.008 '
.014 BT03 48 Alls-081 93 N0 450 2 n.m. n.m. .0036 .003 3 n.m. n.m. .0034 .0025 4 45.8 .004 n.m. n.m.
D047 10
- Alls-010 93 0 450 2 25.8 n.m. .0101 .004 4
3 33.0 n.m. .0108 .0046
, 5 49.7 .022 n.m. n.m.
.012 1 D047 45 j Alls-085 93 ND 300 2 n.m. n.m. .004 i
.0041 3 n.m. n.m. .0042 .0034 5 54.1 .005 n.m. n.m.
I i
Table 2-7 fuel-Rod Interpellet-Gap, ovality, and Creep
- Measurements for the OCONEE-2 Reactor VENDOR: 81W Pellet dia. (inches) .370 l, length (inches)- .445 nondensifying :
length (inches) .700 densi fying '
fGENTTFffXTTbN Fuel fuel '
Pellet . Red Average Internal End Gap Maximum Olanetral' Fuel. Fuel-Rod Density Pressure Burnup Assy of Sizes Ovality Creep No. (1 TO) Type (psig) Cycle' (mwd /MTU) (inches) (inches) (inches) 2815 15 92.5 ND 360 12.0 No Gaps 1 .0011 .0013 i 2 No Gaps .0011 .0024 2015 23 92.5 ND 360 12.0 0.10 1 .0012 .0016
[ 0.20 0.10 0.10 2 25.0 0.2 .0012 .0025 2B40 39 92.5 ND 360 12.0 1 No Gaps .0011 .0013 2 25.0 0.10 .0011 .0024 2840 47 92.5 ND 360 12.0 0.30 1 .0012 .0016 2 25.0 No Gaps 0012 .0025 2815 52 92.5 0 360 12.0 0.10 1 .0077 .0031 0.10 0.10 0.15
- 0.20 0.10
. 2 25.0 No Gaps .0082 .0045 I
9 Table 2-7 (Continued)
- fuel-Rod Interpellet-Gap. Ovality, and Creep Measurements for the OCONEE-2 Reactor VENDOR
- B&W Pellet dia. (inches) .370 length (inches .445 nondensifying length (inches)i .70 denst fying i
TUENTITTCATION Fuel fuel Pellet . Rod Ave.' age Internal End Gap Maximum Diametral Fuel Fuel-Rod Density Pressure of Burnup Sizes Ovality Creep Assy No. (1 TO) Type (psig) Cycle (mwd /MTU) (inches) (inches) (inches) 2B40 56 92.5 0 360 1 12.0 .10(7)" .0077 .0031
.20
.15 2 25.0 .10(3) .0082 .0045
.20 1.0 2B40 28 92.5 0 360 12.0 .10 41 .0026 1 .002
.15 4'l 2 25.0 .10 4) .0035 .003 -
.20 2J l
2B15 4 92.5 D 360 1 12.0 .10 .0026 .002
.15(2) 2 25.0 .1 .0035 .003
.2 (2)
.5 (a) Number in parenthesis indicates the number of gaps of the s,1ze 5
Table 2-7 (Continued)
- fuel-Rod Interpellet-Gap. Ovailty, and Creep hasurements for the OCONEE-2 Reactor (Hot. Cell) .
VENDOR: B&W Pellet dia. (inches) .370 length (laches) .445 nondensifying length (inches) .70 densifying TOTkTITICATION Fuel Fuel Pellet Rod Average Internal End Gap h ximum Olanetral fuel Fuel-Rod Density Pressure Assy of Burnup Sizes Ovality Crcep No. (% TO) Type (psig) Cycle (Wd/MTU) (inches) (inches) (inches) 2840 41 75041E 92.5 ND 360 2 24.6 .010 .002 .0027 43 '
75043E 92.5 NO 360 2 24.7 .010(3) .0023 .0027 29 75029E 92.5 ND 360 2 23.8 .010(3) .0023 .0012
.020 (3)
.040 33 75033E 92.5 NO 360 2 24.0 .030 (2) .0020 .0012
. .185 25 75025E 92.5 ND 360 2 24.7 .020 .002; .0019 47 75047E 92.5 ND 360 2 25.2 .010(3) .0020 .0019
, .020 39 75039E 92.5 NO 360 2 25.2 .020 (3) .0020 .0013 O __
. i Table 2-7 (Continued) '
Fuel-Rod Interpellet-Gap,_ Ovality, and Creep Heasurements for the OCONEE-2 Reactor (HotCell) ,
VENDOR: B&W Pellet dia. (inches) .370 :
.445 nondensifying length length ((lnches) inches) .70 densifying a
TDENTIFffATION Fue1 Fuel Pellet Rod Average Internal End Ga ',. Maximum Diametral fuel fuel-Rod Density Pressure of Burnup si 2es Ovality '
Rssy Creep No. (1 TD) Type (psig) Cycle (mwd /MTU) ( ;.eches) (inches)
(inches) 2840 53 13931 92.5 0 360 2 24.5 .010(3) .017 .0033 *
.020 .
1
.040 '
.060
.070
.110
.180
.320 '
.599 2840 55 13897 92.5 0 360 2 24.1 .010 .0198 .0033
.020 f
.040 0
.060 (
.070
.162
.228
.331 N
Table 2-7 (Continued)
Fuel-Rod Interpellet-Gap. Ovality. and Creep Heasurements for the OCONEE-2 Reactor (flotCell) .
4 VENDOR: B&W Pellet dia. (inches) .370 lengthqinches) .445 nondensifying '
length Linches) .70 densifying i
TiWW1TFTfATIB Fuel Fuei 4
Pellet Rod Average I
! Internal End Gap Maximum Diametral ?
! Fuel Fuel-Rod Density Pressure of 'Burnup Sizes Ovality Creep ,
Assy 'No. (1 TO) Type (psig) Cycle i (Wd/HTU) (inches) (inches) (inches) l 2840 52 i 13960 92.5 0 360' 2 25.4 .010(21 .0157 .0033
.020(2h
.040
- .060
.070 i
.142
.030 l 2840 42 75042E 92.5 0 360 2 25.0 .010 .0046 .0012 4
.030
.240 i 2840 42 75044E 92.5 D 360 2 25.0 .010 (2) .0043 .0012
.020
.040
.080 (2)
.221
.960 1
i l
Table 2-7 (Continued) i Fuel-Rod Interpellet-Gap, Ovality, and Creep Measurements for the DCONEE-2 Reactor t
(HotCell) , j VEN00R: 88W Pellet dia. (inches) .370 i length (inches .445 nondensifyin9 length (inches)j .70 densifying TOTNITFRATION fuel fuel ii Pellet Rod Average Internal End Gap Maximus Diametral fuel fuel-Rod Density Pressure of Burnup Sizes Ovality Creep Assy No. (% TO) Type (psig) Cycle (mwd /MTU) (inches) (inches) (inches) 2840 36 75036E 92.5 0 360 2 25.4 .010(2) 0089 .0019 .
.060 (2)
.070 s
f .140
.228
. 260
.930 2840 38 75038E 92.5 0 360 2 25.3 .020 .0091 .0019
.040 (3)
.120
. .165
.252
.292
.307
.336
.792 w
e
Table 2-7 (Continued) -
Fuel-Rod Interpellet-Gap. Ovality, and Creep Heasurements for the OCONEE-2 Reactor (lletCell) ,
VENDOR: 88W Pellet dia. (inches) .370 ,
length (inches'l .445 nondensifying !
length (inchesh .70 densi fying TiffWTITINTION Fuel Fuel s
- Pellet Rod Average
Internal End Gap Maximum Diametral fuel fuel-Rod Density Pressure of Burrup Sizes Ovality Creep Assy No. (1TD) Type (psig) Cycle ,
(NWd/MTU) (luches) (inches) (inches) .
2840 28 75028E 92.5 0 360 2 24.8 .010(2) .0071 .0016 .
.030(2) I
.070 !
.091 ,'
.098
.118 .
.221 l
. .236 i 4
i I
l
fuel-rods were pressurized to 360 psi, the larger and more numerous gaps in the densifying fuel rods are related to fuel stability.
Gamma scanning conducted on the corner rods containing densifying fuel from five fuel assemblies at the end of Cycles 3 and 4 showed no gaps larger than 0.1 inches, while ten of the rods contained no measurable gaps at the end of either cycle.
2.2.6.2 Hot Cell Examinations -
A hot cell evaluation of interpellet-gaps was conducted on fifteen fuel-rods:
seven containing nondensifying and eight containing densifying fuel. These data are also presented in Table 2-7. Gaps were found in all of the fuel rods. Among the nondensifying fuel-rods, one gap of 0.185 inches was found while all of the remaining gaps were less than 0.040 inches. In rods that contained densifying fuel, the gaps were more numerous and larger. Out of a total of seventy-two gaps in these rods, four gaps were measured between 0.5 and 0.96 inches. These gaps were due to the large and variable increase in pellet densification during irradiatfog. These density increases were between 1.7% T.D. and 5.1% T.D., which were consistent with the large variable pre.
irradiation resinter density changes that ranged from 1.5". T.O. to 3.??. T.D.
2.2.7 pellet Gao Data from the Aricansas Nuclear one Unit 1 Reactor (Babcock & Wilcox) (,2f,)
Eight fuel-rods were gamma scanned at the end of one cycle of irradiation (la GWd/T). Four of the rods contained solid nondensifying pellets and four contained annular nondensifying pellets. No measurable gaps were reported.
The data are presented in Table 2-8.
2.2.8 pellet Gao Data from the Zorita Reactor (Westinghouse) (19,- 3)
Gamma scan data from measurements made after one, two and three cycles of irradiation are presented in Table 2 0 At the and of one cycle of irradiation of rods containing nondensifying fuel pellets, the one pressuri:ed rod and two of the nonpressurized rods showed no gaps. The three remaining nonpressurized rods contained at least seven gaos, the largest being 0.86 inches. Of the four rods examined at the end of 2 cycles of irradiation, only one rod, number 266, contained a significant gap which was 0.769 inches in w _
Table 258 Fuel-Rod Interpellet-Gap. Ovality, and Creep Measurements for the Arkansas Nuclear One-1 Reactor VENDOR: B&W Pellet dia. (inches) .3635 I length (inches) .418 '
THEliTITIGTION Fuel Fuel j Pellet Rod Average '
Internal End Gap. Maximum Diametral Fuel Fuel-Rod Density Pressure of Burnup Sizes Ovality Creep Assy No. (% TO) Type (psig) Cycle .(Wd/MTU) (inches) (inches) (inches) '
NJ023Q 1-I*I I 95 '
NO 415 1 18.0 No Gaps Not Reported Not Reported 1-5(b) 95 NO 415 1 18.0 No Gaps Not Reported Not Reported 15-15(b) 95 ND 415 1 18.0 No Gaps Not Reported Not Reported .
15-1(b) 95 N0 415 1 18.0 No Gaps Not Reported Not Reported NJ023S 1-1(a) 95 N0 415 1 -
18.0 No Gaps Not Reported Not Reported 1-15(b) 95 NO 415 1 18.0 No Gaps Not Reported Not Reported 15-15(b) 95 NO 415 1 '18.0 No Gaps Not Reported Not Reported 15-II 'I 95 ND 415 1 18.0 No Gaps Not Reported Not Reported IOI Solid mliets i
(b)Annularpellets s
O
.I Table 2-9 Fuel-Rod Interpellet-Gap, Ovality, and Creep Heasurements for the Zorita Reactor VENDOR: Westinghouse Pe11et dia. .3669 length (Linches) inches)
.600 '
TlWi1TfTCITION Fuel . Fuel .
Pellet Rod Average Internal End Gap Maximum Diametral '
Fuel fuel-Rod Density Pressure of Burnup Sizes ovality Assy Creep No. (% 10) Type (psig) Cycle (HWd/MTU) (inches) (inches) (inches) 283 93.5-94.2 .340I "I 1
NO 15 1 29.4 nr nr i
285 ND 500 29.6 1 No Gaps nr nr R23 "
ND 15 1 13.4
<.8%)
.175 nr nr ,
l R24 ND 15 13.0 1 No Gaps nr nr '
l R27 ND. 15 13.0 l
1 .260 nr nr l "
R35 HD 15 13.1 1 No Gaps nr nr 266 "
ND 500 d 2 31.2 nr .0025
.769fC
.161 284 94.07 ND 15 2 41.4 No Gaps nr .0035 (d) Considered to be formed by handling procedures ,
ICI Total of several gaps (b) Total of several gaps w = W re ed I*I Total for 3 gaps .
w-
Table 2-9 (Continued) ,
fuel-Rod Interpellet-Gap, ovality, and Creep Heasurements for the Zorita Reactor VENDOR: Westinghouse Pellet dia. qinches) .3669 Iength tinches) .600 TDDITTFTCAYf0N fuel Fuel Pellet Rod Average Internal End Gap Haximum Olanetral fuel fuel-Rod Density Pressure of Burnup Sizbs ovality Assy Creep No. (1 TO) Type (psig) Cycle (HWd/HTU) (inches) (inches) (inches) 331 94.2 ND 500 2 41.8 .34I *I nr nr 333 94.2 0(b) 15 2 40.0 No Gaps nr nr
- 213 93.5-94.2 ND 500 3 39.6 ""
nr nr
- 214 ND 500 39.7 ""
3 nr nr 215 ND 500 3 39.7 ""
nr nr 216 ND 500 3 39.7 ""
nr nr 362 ND 36.5 ""
15 3 nr nr 368 NO 500 3 36.2 ""
nr nr 391 NO 500 3 38.0 ""
.0003 nr 293 ND 500 3 47.7 ""
+.0005 nr 235 ND 500 3 47.7 ""
nr nr 294 ND 500 3 47.6 ""
+.0002 nr IOI Sum of four gaps (b) Based on data in Reference 30
i i
Table 2-9 (Continued) f fuel-Rod Interpellet-Gap. Ovality, and Creep Heasurements for' the Zorita Reactor VENDOR: Westinghouse Pellet dia. (inches) .3669 j length (inches) .600 ;
TKHTITINTION fuel fuel -
Pellet Rod Average !
Internal End Gap Haximum Ofametral '
Fuel fuel-Rod Density Pressure of Burnup Sizes Ovality Creep Assy No. (1 TO) Type (psig) Cycle (HWd/HTU) (inches) (laches) (inches) 230 93.5-94.2 ND 15 49.1 3 No Gaps' nr nr ..
227 ND 500 3 48.5 " "
nr nr 228 ND 500 3 48.0 " "
nr nr 232 " " "
ND 500 3 49.0 nr nr 343 ND 15 3 . 51.6 " "
nr nr 379 " " "
ND 15 3 51.4 nr +.008 383 ND 15 52.2 " "
3 nr +.0075 344 ND 500 3 51.2 " "
nr nr 334 ND 15 3 50.8 " "
nr nr
~
384 ND 15 3 51.0 " "
nr nr 336 ND 500 3 50.8 " "
386 ND 500 ,
3 51.0 " "
Table 2-9 (Continued)
Fuel -Rod Interpellet-Gap. Ovality, and Creep Measurements for the Zorita Reactor VENDOR: Westinghouse Pellet dia. (inches)- .3669 ['
length (laches) .600 '
TEiTTFTCITION Fuel fuel Pellet Rod Average Internal End Gap Maximum Diametral Fuel Fuel-Rod Density Pressure of Assy Burnup Sizes Ovality Creep i No. (1 TD) Type (psig) Cycle (mid/MTU) (inches) (inches) (inches) 369 93.5-94.2 ND 15 3 36.8 .275I *I nr +.0009
<.86 e
330 "
N0 15 3 55.2 No Gaps- nr
+.0008 332 "
NO 500 3 54.0 No Gaps nr +.0014 e
b 4
O
i s.
l length and was located four-pellets from the top of the column. This was a pressurized rod containing nondensifying fuel that showed a large net column
. shrinkage and no evidence of local high-ovality on its profilometry trace. The investigators reported that they believed that the fuel was loose inside the cladding and shifted during handling. It is doubtful that this gap existed during operation. Examination of the three cycle rod data with fuel burnup in the range of 36 to 55 GWd/T showed that gaps were present in only one out of twenty-three rods. This rod, number 369, was not pressurized, contained one large gap of 0.86 inches and several smaller gaps that totalled 0.275 inches. -
2.2.9 Pellet-Gao Data form the Zion Unit 1 Reactor (Westinghouse)(3)
Four one-cycle fuel-rods and eleven two-cycle fuel rods frem assemblies C-64 and C-63, respectively, were scanned during the end-of-Cycle 2 fuel examinations. These rods consisted of pressurized nondensifying fuel. The data are presented in Table 2-10. No gaps of significance were recorted. The largest measured value of 0.15 inches was recorted as the sum of two gaps.
2.2.10 Pellet Gao nata from' the Surry Unit 2' Reactor .
-(Westinghouse)(32)
Four fuel-rods from assembly RD-1 and thirteen fuel rods from assembly RD-2 were gamma scanned. The rods consisted of pressurized nondensifying fuel. .
The data are presented in Table 2-11. No gaps were reported for fiftaten of the rods while one gap measuring less than 0.10 inches was reported for each of the remaining two rods.
2.3 INTERPELLET-GAP-GAMMA-SCANNING TECHN!0tlES Interpellet-gap-size determinations were made using a gamma scanner. The method is based on the assumption that the measured gama radiation intensity along the length of a fuel-rod is directly proportional to the local fuel density in the rod, provided that the gama detector is properly shielded from radiation emitted by other fuel-rods and structural materials. The measured gamma intensity is also a function of the specific collimator slit geometry and the scanning speed for a given level of burnup. 'A calibration factor is obtained for a specific collimator geometry and scanning speed by measurement of some known gap widths.
t 4
Table 2-10 fuel-Rod Interpellet-Gap, Ovality, and Creep Measurements for the Zion-1 Reactor VENDOR: Westinghouse Pellet dia. 3659 l length inches inches 1.600 :
t' TliEilTITTCITION fuel fuel
- Pellet Rod Average .I internal End Gap Maximus Diametral !
Fuel fuel-Rod Density Pressure of Burnup Sizes Ovality Assy Creep -
No. (1 TO) Type (psig) Cycle (Wd/MTU) (inches) (inches) i (inches) i I
C64 605 95 NO 450 12.42 No Gaps 1 .001 .0024 C63 616 95 NO 450 1 - 12.42 .07I *) .001 .0024 C63 652 95 NO 450 12.42 .06(a) 1 .001 .0024 :
C64 687 95 No 450 1 12.42 .15 IAI .001 .0024 601 95 ND 450 2 31.0 No Gaps
]
.0007 .0031 j
C64 610 95 NO 450 2 31.0 .04I *I .0007 .0031 j C64 618 95 ND 450 2 31.0 No Gaps .0007 .0031 i
640 95 NO 450 2 31.0 .06(b) .0007 .0031
- C64 642 95 NO 450 2 31.0 .04 ICI .0007 .0031 i
1 ,
j (b) 3,,of ggy,g,p3 (c) Sum of four gaps (a) Sum of two gaps (d) Sus of three gaps
veW, Table 2-10 (Continued)
Fuel-Rod Interpellet-Gap. Ovality, and Creep hasurements for the Zion-1 Reactor
~
t VENDOR: Westinghouse Pellet dia. 3659 1ength(inches) l (inches) . 600 l
WHTITirATFON Fuel Fuel Pellet Rod
- Average Internal End !
Gap Maximum Diametral Fuel Fuel-Rod Densit Pressure Surnup Ovality ,
of Sizes Creep '
Assy No. -(1 TD)y Type (psig) Cycle (feld/MTU) (inches) (inches) (inches)
C64 648 95 N0 450 2 31.0 .03 .0007 .0031 C64 657 95 ND 450 2 31.0 .04I *I .0007 .0031 C63 663 95 ND 450 -
2 -
31.0 I
.03*I 0007 .0031 064 667 I 95 ND 450 2 - 31.0 .02") .0007 .0031 C65 685 95 'ND 450 31.0 2 .09(b) .0007 .0031 063 691 95 ND 450 2 31.0 .04(d) .0007 .0031 (d) Sum of three caps (b) 3,,og ggy,g,p3 (a) 3,,og g,,g,p3 e
w_ _ _ --
Table 2-11 Fuel-Rod Interpellet-Gap, Ovality, and Creep Measurements for the Surry-2 Reactor VENDOR: Westinghouse Pellet dia. (laches) . 323 I length (laches) . 530 TlWNTTFITATION fuel fuel Pellet Rod Average Internal End Gap Maximus Diametral Fuel Fuel-Rod Density Pressure of Surnup Sizes Ovality Creep Assy No. (1 TO) Type (psig) Cycle (Wd/NYU) (inches) (inches) (inches)
NO-2 105 95 ND 450 8.65 No Gaps 1
$ .003 .001 NO-2 106 95 ND 450 8.72 No Gaps 1
$ .003 .001 NO-2 108 95 ND 450 8.72 .097 1
$ .003 .001 NO-2 113 95 ND 450 8.50 .065 1
$ .003 .001 !
R0-2 500 95 ND 450 7.47 No Gaps 1
1 003 .001 RS-2 501 95 N0 450 1 8.08 No Gaps $ .003 .001 RS-2 502 95 ND 450 9.09 No Gaps 1
1 003 .001 R0-2 503 95 ND 450 6.31 No Gaps 1
5 003 .001 R0-2 505 95 No 450 1 7.47 No Gaps $ .003 .001 RS-2 506 95 NO 450 7.62 1 No Gaps 1 003 .001 RS-2 508 95 ND 450 I 8.29 No Gaps 1 003 .001 RS-2 509 95 ND 450 8.80 No Gaps
, 1 .0075 .001 t
. .w - ~--
e -
7 Table 2-11 (Continued) .
Fuel-Rod laterpellet-Gap. Ovality, and C' eepr Measurements for the Surry-2 Reactor VENDOR: Westinghouse Pellet dia. (inches) .323 l 1ength (inches) .530 .
TOTNTlYTCATION Fuel Fuel Pellet Rod Avera9e Internal End Gap Maximum Diametral fuel fuel-Rod Density Pressure of Burnup Sizes Ovality Creep Assy No. (% 10) Type (psig) Cycle (mwd /MTU) (inches) (inches) (inches)
RD-2 510 95 NO 450 9.13 No Gaps 1
1 003 .001 RD-2 511 95 ND 450 7.99 No Gaps 1
1 003 .001 RD-2 512 95 NO 450 9.09 No Gaps
, , 1 5 003 .001 IN RD-2 514 95 ND 450 1 7.99 No Gaps 1 003 .001 '
HD-2 515 95 N0 450 6.31 No Gaps 1
5 003 .001 e
s
Measurement sensitivity varied widely for the data reported. Rabcock & Wilcox used a collimator slit opening of 0.103 inches. The presence of pellet gaps as small as 0.10 inches could easily be detected against a background noise level equivalent to a gap of approximately 0.025 inches. Westinghouse defined the sensitivity of the system it used as capable of detecting changes in gamma activity equivalent to that produced by a 0.050 inch gap in the fuel column.
Combustion Engineering utilized a collimator slit opening of 0.015 inches for the Palisades and Maine Yankee reactor measurements and a 0.010 inch collimator slit for the Calvert Cliffs and the Fort Calhoun reactor measurements.
Conseqently, gaps as small as 0.005 to 0.010 inches could be detected. The systems of all three fuel suppliers were entirely adequate for detection of gaps that could be of significance with respect to fuel-rod creep collapse conditions and augmentation factor calculations.
2.4 CONC 1.USIONS Gaps that formed in unpressurized fuel-ecds containing densifying fuel pellets have been found to be as large as 4 inches. For pressurized fuel-rods containing nondensifying fuel pellets, all gaps formed in the examined reactors were less than 0.3 inches R&W's modern fuel had a maximum observed gap -cf .
0.30 indhes; W's modern fuel had a maximum cumulative gap of 0.161 Jnches' and C-E's modern fuel had a maximum observed gap of 0.022 inches. In all the modern f,uel-rods, the total pellet axial column shrinkage is small, on the order of 0.5 inches. Therefore, the possibility of forming large individua}
axial gaps does not exist in modern fuels. -The axial distribution of gaps and gap sizes was also found to be random in modern fuel.
7 L
t 2-37
s thickness to midsurface radius ratio (h/a) for this study was (h/a) max "
0.12911 and (h/a) min = 0.11681.
J The curve of Figure A-9 has a maximum wall thickness value (h) of 0.027 inches, a minimum h of 0.022 inches and was based on the constant outer radius value of 0.21225 inches. Their maximum and minimum midsurface radii were therefore a
max = 0.20125 inches and am in = 0.1987 inches. These data yield maximum and minimum h/a values of 0.027 0.022 -
h/a max = = 0.1459 =
= 0.10932.
h/am in 0.19875 0.20125 i All h/a values are bounded by the upper and lower limits used in Figure 4-1.
I Because.the Fl.CF is a geometric effect, its applicability is not constrained by the requirement of a specific creep law. Thus, if the only variable within a data set is the axial gap length, any experimental data concerned with the creep collapse of cladding can be used to validate the model. Pertinent data was found in . References 36. and 37 .and is stimarized in Table A-2. In the data from Reference 37, the actual collapse times for the caps characterized by L/a '
= 6.08 and 6.27 are the average of the observed values. The experimental ovality versus time curve for the 1/a = 3.69 analysis of Reference 37 indicates that the actual collapse of this rod will not occur in the first 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> The predicted collapse times given in Table 4-2 were determined by multiplying the actual collapse time for the largest gap in the test series by the ratio of the RCF associated with the gap in question to the RCF corresponding to the largest gap in the series. When the 1/a for the largest gap is greater than 15, this procedure results in Tg = (RCF) Tyt, which is the formula derived ear 11er. The data in Table 4-2 illustrates the conservatism of this procedure since all predicted collapse times are less' than or equal to the observed collapse times if the collapse occurred during the duration of the experiment. For the three examples where actual collapse did not occur, sufficient data is available for the two cases characterized by I/a = 5.17 to demonstrate the adequacy of the theoretical prediction. For 1/a = 3.69, as discussed earlier, collapse during the first 1849 hours0.0214 days <br />0.514 hours <br />0.00306 weeks <br />7.035445e-4 months <br /> Is highly unlikely.
I I I I I i Q 28,000 -
C c CEPAN (CE c) 26,000 -
24,900 -
22,000 -
20,000 -
<n m
2 18,000 -
8 -
E ^
w 16,000 -
e d
@ 14,000 -
12,000 -
i 10,000 -
8,000 -
6,000 I I I I I f 0.020 0.021 0.022 0.023 0.024 0.025 0.026 0.027 CLAD WALL THICKNESS, IN.
Figure 4 9: Collapse Time vs. Wall Thickness, Remaining Parameters Standard Case I
W
l 24,000 4 1 o
C 3 CEPAN (CE e) '
22,000 - _
20,000 - .
us e
3 o 18,000 _ _
m I
w
- 2 16,000 =
a
.J O
u 14,000 = _
12,000 - _
10,000 f I l
0.200 0.210 0.220 OUTER RADIUS,IN.
1 l
Figure 48: Collapse Time vs Outer Radius, Remaining Parameters Standard Case I
.___.___(-13_____, __. _ _ _ _ _ _ _ . _ , _ _ , _ _ _ _ _ , _ _ _ _
i I
26,000 g i .
s
- C C CEPAN (CE d 24,000 = =
22,000 - -
. , 20,000 - -
c 3
9 m
3 18,000 = - .
N a.
3 a
16,000 - . -
C -
u 14,000 =
3 12,000 -
10,000 1 I i 0 0.5 1.0 1.5 2.0 OVALITY, MILS Figure 4 7: Collapse Time vs. Ovality, Remaining Parameters Standard Case
_w - . _ _ .
26,000 g
- g C 0 CEPAN (CE e) 24,000 =
=
22,000 -
20,000 -
m E
i--
w E 18,000 =
- 3:
o O -
18,000 -
14,000 "
12,000 =
=
l 10,000 I I I I I I
! 520 540 560 580 600 620 640 660 TEMPERATURE,0F l
1 Figure 4-6: Collapse Time vs. Temperature, Remaining Parameters Standard Case t r - -- , , -
i l
l l
I l
70,M g ; 3 , g 3 C C CEPAN (CE e) 60,00d ' -
50,000 -
m C
- s .
= 40,000
- ~
m -
E m .
% 30,000 --
-i O
u 20,000 -
f 10,000 -
I f f f f 0 f 800 1000 1200 1400 1600 1800 2000 2200 PR ESSURE, PSI 1:
', Figure 4 5: Collapse Time vs. Pressure, Remaining Parameters Standard Case
-_. . - - -. ~.
M,000-
{ g g g g J l
l
- ~
C O CEPAN (CE e) 40,000 == -
as e 30,000 - -
3 b
ui -
3 . .
7 .
g . .
a.
a d
C
- u .
~
20,000 == -
l l
r .
10,000 -
6,000- ' f f ' f
'O 0.2 0.4 0.6 0.8 1.0 1.2
! FLUX (n/cm 2.sec) x 1014 Figure 4-4: Collapse Time vs. Flux, Remaining Parameters Standard Casa
._ __. ._ . . . . . . . . . _ . _ _ _ s . _
h TABLE 4-1 STANDARD CASE PARAMETERS Outer Radius (OR) = 0.21225 inches Wall Thickness ,(h) = 0.0245 inches Initial Ovality (0.0. max-0.D. min)
- I 1 *il8 Net Pressure (External) = 1450 psi Temperature = 5900F Young's Modulus = 10.84 x 106 psi Poisson's Ratio = 0.259 Fast Neutron Flux = 0.7 x 1014 n/cm2 -sec Solution Time Step = 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> Number of points through thickness of shell = 10
-Number of circumferintial nodal points per quadrant = 25 Linear terms only retained .in equilibrium equations Creep Law:
, -6000 c, = 4 0.85e '
-
- 1.5,g ( RT ) Ak exp(-kt) + C
+- 1.3 x 1017 sinh (0.09e,)exp ( }
RT where e, =
effective strain rate (hr-1) e, =
effective stress (ksi)
= 0
. R 1.987 (ca1 X)
T = temperature (0X) 4 = fast neutron flux (n/cm 2sec, E>l Mev)
=
t time (hours)
A = 7.4'x 10-15 C = 3.87 x 10-18' k = 3.6 x 10-3 l
- (Equation 27 of Reference 5) applies to Zircaloy a under biaxial stress.and in-core conditions. i 1
I
55 . . . s a a 4 s e a a a a u = 0.259 50 1/a h/a-c1000 h/a = 0.1225 h/a = 0.1500 -
o 53.77 4A78 38.04 45 1 38.39 32.02 25.32 2 20.77 17.41 1A13
> 3 11.90 10.04 8.24 .
4 7.55 6.43 5.35
$ 40 w
5 5.21 4.50 3.80 -
6 3.34 3.: 8 2.39 2 7 2.98 2.54 2.31
- 9. 35 a 2.40 2.16 1.93 u.s 9 2.00 1.33 1.se 10 1.71 1.59 1.47 5 30 o
- 11 1.4e 1.40 1.32 .
Q 12 1.32 1.27 1.21 2 13 1.19 1.18 1.13 m 14 g 25 -.
h/a = 0.1000 15, 1.08 1.00 1.07 1.00 1.06 2 -
1.00
.$ 20 h/a = 0.1255 .
2
. h/a = 0.1500 15 =
10 -
5 =
9 9 9 9 i f f f f f f f .
'O 1 2 3 4 5 6 7 8 9 to 11 12 13 14 15
\
1 2/a I
l- Figure 4 3: Finita Length Correction Factor Vervis 9/a for u = 0.259 l
l l
Figure 4-3 illustrates the variation of Fl.CF with 1/a for 1/a<15 for three representative values of the wall thickness (h) to the midsurface radius (a) ratio (h/a).
4.2.2 Validation of the Small-Gao Finite-length Correction Factor for CEpAN A series of constant load problems were run to obtain the variation of collapse time s computed by CEPAN with the conditions representative of C-E design applications reported in Reference 1. The results of these analyses are described below and demonstrate that the FLCF's exhibited in Figure 4-3 cover the range of interest for C-E designed fuel.
Within each set of results, one parameter was varied over a range encompassing anticipated reactor operating values, while all other parameters were held 4
constant at their base case values. The base case parameters, shewn in Table 4-1, are based on an average of existing clad geometry and loading conditions.
The midsurface radius (a), calculated from Table A-l's data, has a value of 0.2 inches. This ensures that gaps greater than 3 inches are modeled as infinite
- when the formula of Equation 4-6 is used. This relationship is compatible with
~
the modeling used in Reference 35.
The effects of flux, pressure and temperature variation on cladding 1ife are shown in Figures 4 4 through 4-6; whereas, geometric factors that affect collapse time (ovality, outer radius and wall thickness) are indicated in Figures 4-7 through 4-9. Figures 4-4 through 4-3 show that the collapse time decreases with increasing flux, pressure, temperature, ovality and outer l radius. Figure 4-9 indicates that the effect of increasing wall thickness is
! to strengthen the cladding against collapse.
i-l ..-
Figure 4-3 exhib'its the variation of FLCF with 1/a for three values of h/a (0.1, 0.1225, and 0.15). The h/a value 0.1225 corresponds to the base case of Table 4-1.
The values of 0.1 and 0.15 are shown below to bound the values of h/a used ir. the studies of Reference 5.
ThecurvesofFigure4-8weregeneratedusinhhequalto0.0245incheswitha range of outer radii (OR) bounded by ORmin = 0.202 inches and ORmax
- 0*222 i
inches. This results in a maximum midsurface radius of a max = 0.2097 inches l and a minimum radius of ma in = 0.18975 inches. The range of the wall i
I- = gap length n = the collapse mode (representing the number of half waves in the circular surface collapsing) is set to its minimum I value, 7 for all evaluations.
According to Reference 34, P CI is given as 3
Eh 3
g (g_y2)
Since PC15, which is PCSS evaluated at a gap length to midsurface-radius ratio (1/a) of 15, is larger then PCI, a still more conservative value of Tyt TFL,C, is given by P
C53 T FL,C = T yt -
P (L5)
C15 ,
where ,
P
> 1.0 for 0<1 < 15 P
C15 s
This formulation has the advantage that at an 1/a ratio of 15, PCSS
= 1.0 and Tpt = T gt r
l Since Tpt 1 Tgt for all gaps, a conservative design formula for Tpt can be writt'en as T FL,C =
(FLCF) T yt (4_6)
P
', where i
P CSS /PC15 (11.0) for 0 < 1/a i 15 1.0 for 15 1 1/a c asc --
It follows from curve R pp of Figure 4-2 that Tp, = [(PCF-E)/(P A eg-PA )] x [Tyt+Tg7 (4-1) where: T U
=
The unknown difference between the time needed for the finite gap rod to reach the critical pressure for an infinite rod (PCI) and Tgt P
A
= Initial external differential pressure on rods R gp and R pp Extending curve R gp to its intersection with PCF would yield a smaller estimate of Tyt, T(, while setting both Pg and Tg to zero to define the curve labeled RFP.C, would yield the smallest and, therefore, most conservative estimate of Tyt, T(. Thus, a conservative estimate of the time to collapse for fuel with finite gaps (Tk) is given as:
T = F .T yt (4-2)
CI
~
To incorporate finite gap size information into the estimate of Tk requires P
evaluation of the ratio CF in terms of geometric and material properties o'f the cladding. CI For a given finite length, the collapse pressure for a simply supported tube
(,34,) , PCSS, is the minimum required for any support condition. Therefore, P
CF in Equation 4-1 can be replaced by PCSS, for a conservative estimation
{ of T pt. P CSS is given by 2
[Eh/a(1-v )7 [(1-v )/(n 2.1)(t#n 2(1/na)2))
P = 2 CSS
+ (h2 /12a2 )(n2 -1+(2n 2
.1 y)f(t.n 2 )(gf,,)2)),
(4 3) where: ,
E =
modulus of elasticity for cladding h = clad wall thickness l a = clad wall midradius v = Poisson's ratio for the cladding
d 3
2 A '
. O CRITICAL
- OVALITY I T I I
c(se I ed [
I Se ,
++ I I
I I
I I [
' i TIME Tjl Tg Fig'ure 41: Ovality Versus Time ,
z -
3 EA s
PCF s,
/ l PCI __
+< e- i i I
g iI I l
PA I lI I
- I II I- I R FP,C I I Tu q I II g f f I t g TIME Tgt Tyt ThL Tpt i
! Figure 42: Pressure Versus Time
. . . _ _ - _ ~ . . . _ . . _ . . . . _ _ . . _ . . _ . .
l l
l analysis, the analogous effect is an increase in the fuel-rod life due to the greater clad support provided by the fuel pellets adjacent to the gap.
Inclusion of tha three-dimensional modeling of the stress-strain equation for collapse with end support is, potentially, the best theoretical basis for predicting clad collapse. A simple alternative would be the development and validation of a correction factor method based upon the theory of collapse in the presence of end support. This section of the report describes the development of a correction factor that yields demonstrably conservative predictions of clad collapse.
4.2.1 Develooment of a Model of the Finite-Gao Influence on the Evaluation of Collapse Time It is assumed that the collapse time for a fuel-rod of a specified geometry containing an axial gao of designated length can be ascertained with the aid of a simplified model. The required time is the product of the collapse time for the fuel-rod with an infinite gap, as determined from the CEPAN code, and a finite-length correction factor (Fl.CF), 'which is solely a function of the axial gap length and the geometry and material properties of the cladding. The model -
, . considers the deformation of four rods. Two rods, R IC and RFC, containing infinite and finite gaps, respectively, are subjected to creep deformation.
Two other rods, Ryp and Rpp, also containing infinite and finite gaps, .
respectively, are subaected to quasi-static external pressure loading. All rods have the same wall thickness-to-radius ratio and the finite gaps in rods RFC and Rpp are identical in the development of this model. It is further asumed that, irrespective of the deformation mechanism acting on the clad, collapse will occur at a unique critical ovality.
l Figure 4-1 shows linear approximations of the ovality versus time relationship in rods RIC and RFC, respectively. The time to collapse with infinite gaps, Tyt can be determined from a CEPAN analysis; however, the time to j
collapse with finite gaps, Tyt, which is~ greater than tit., is unknown.
Figure 4-2 shows the quasi-static external pressure-loading histories of rods ;
Rgp and RFP that are requiredito produce the same ovality histories for the pressure driven deformation as cccurred for the creen deformation rods (see Figure A-1). The collapse pressures, PCF and PCI, can be determined from a i
static buckling analysis,
{
t
l I
section 4 CEPAN WITH FINITE LENGTH CORRECTION FACTORS
4.1 INTRODUCTION
Since 1976, Combustion Engineering has been using the CEPAN (5) code to predict the earliest time at which clad collapse could occur in C-E designed fuel-rods. This code is :onservatively based upon the theoretical model for creep deformation of unsupported fuel cladding. As such, the code is appropriate for estimating the time to clad collapse for fuel-rods that could have large gaps in the fuel pellet column. This is an acceptable formulation for the analysis of unpressurized fuel-rods containing densifying pellets.
The fuel-rods supplied to U.S. vendors are now pressurized and contain nondensifying fuel pellets, making the basic CEPAN' assumption of large gaps overly conservative. Therefore, a modification of CEPAN that more accurately -
accounts for the small interpellet-gaps observed in pressurized, nondensifying fuel-rods is appropriate. This new version of CEPAN bases its prediction of the minimum time to clad collapse on the assumption that clad collapse occurs at a critical ovality independent of the mechanism causing the. ovality change.
As a consequence, it relies on a conservatively low prediction of the time to clad collapse driven by a quasi-static pressure loading to estimate the time to clad collapse due to creep deformation. The small-gap formulation used to
. evaluate the collapse time for fuel containing finite gaps significantly increases the minimum time to clad collapse calculated by CEPAN but.still conservatively predicts the time of observed clad collapse. Evaluations of the i m. -imum time to clad collapse for modern fuel-cods show that they would not be j expected to suffer clad collapse during their useful life.
I 4.7 DERIVATION AND VALIDATION OF THE SMALL-gap FINITE-LENGTH CORRECTION FACTOR FOR THE CEPAN CODE ,
It has been demonstrated analytically (34) that for the elastic buckling of cy11ndrical shells, the critical buckling pressure increases markedly as the distance between the axial supports of the shell is decreased. In clad creep
.--_ -. - - _ _ .p-
Measured ovalities for pressurized rods ranged from 0.2 to 19.8 mils as compared to 5.2 to 16 mils for nonpressurized fuel-rods. Nondensifying fuel, had ovality values ranging from 0.2 to 7.5 mils, while densifying fuel had !
maximum ovality values ranging from 2.6 to 19.8 mils. Thus, fuel densification has significant adverse impact on fuel rod ovality; while prepressurization has no indicated benefit.
Measured creepdown values ranged from 1 to 4.5 mils for the pressurized rods as compared to 0.8 to 8.3 mils for nonpressurized rods. This supports the -
i assertion that prepressurization tends to limit clad creepdown. l l
3.4 CONCLUSION
S
% l Modern fuel, consisting of prepressurized fuel-rods containing nondensifying
~
fuel-pellets has been shown to have greater resistence to fuel-rod ovalization
~
and clad creepdown than do intermediate or old fuel-rods. Furthermore, no creep collapse failures have occurred in the modern fuel-rods. These rods varied in initial rod prepressurization from 300 to 500 psi and covered a burnup range from 6.0 to 54.0 G'Jd/T. This performance is in. sharp contrast to the 124 rods with collapsed sections observed in the nonpressurized, densifying fuel loadings of Cycle 1A and 1R of the Ginna reactor and similar failures in the Point Beach and the H.R. Robinson reactors.
It is concluded th'at prepressurized fuel-rods containing nondensifying fuel
- show no evidence that fuel failure due to clad creep collapse will occur.
i I
u
The Combustion Engineering profilometer system includes: two LVOTs placed 1800 apart (which move axially along the rod), a rod rotation mechanism, and a strip chart recorder. The rod is located by guide rollers positioned above 1 and below the LVDT. The LVDTs respond to changes in diameter as they move
.along the rod and the resultant signals are converted electronically and register on a strip chart recorder. The system is calibrated using a stepped diameter standard. The diameters of each section of the standard are precise to 0.1 nils. The system was also calibrated for ovality by using a well-characterized ovality standard with induced ovalities ranging from 0.01 to 0.016 inches. The indicated accuracy of the profile diameter measurements is
+
_0.2 mils.
'1 3.3 FUEL ROD OVALITY AND CREEPDOWN DATA EVALUATION Fuel-rod prepressurization-reduces the cladding hoop stress caused by the 2250 psi ext'ernal primary system coolant pressure. Therefore, there is expected to be a corresponding reduction in fuel cladding tube ovalization and creepdown.
i Secause nondensifying fuel-pellets will have very limited radial shrinkage, they would be expected to offer resistance to clad ovalization. These effects can be evaluated by considering the maximum ovality values and the average .
creepdown values for each of the fuel-rods examined. These are reported in ,
detail in Tables 2-4 through 2-11 and summarized here in Table 3-1.
TABLE 3-1 Dependence of Maximum Ovality and Average Diametral Creep r
on Pressure and Fuel-Pellet Densification l
Range of Average i Range of Maximum Ovality 01ametral Creep j (inches) ~
(inches)
Minimum Maximum Minimum Maximum
(
Pressurized Fuel-Rods 0002 .0198 .001 .0046 Nonpressurized Fuel-Rods .on52 .0160 .nong .00A0 Densifying Fuel-Pellets 0n20 .0198 ----- -----
Nondensi fy Fuel-Pellets .00n2 .0075' ----- -----
l 3~7- ---- - - - - - ~ '
' ~~
Section 3 CLAD COLLAPSE DATA RASE
3.1 INTRODUCTION
As discussed in Section 2.1, clad collapses were observed at the end of Cycle 1 in the Ginna reacter; the collapsed fuel consisted of initially unpressurized fuel-rods containing densifying fuel pellets. Similar clad collapses were also observed in the first cycle fuel of the Point '.each Unit One and the H. B.
Robinson reactors. Table 2-2 provides ovality and creepdown data indicative of the total lack of clad collapse seen in all examinations of C-E fuel and in all
. modern fuel from any other vendor.
3.2 PROFIL0 METER MEASUREMENTS OF OVALITY AND R00 OIAMETER .
The ovality of a' cladding tube is a measure of the difference between the maximum and minimum measured mean outer diameters of the tube. The measuring
~
~
device, a profilometer, determines the diameter at any height for the angle of orientation of the diameter by determirying the separation between the fingers riding on the tube surface and the center of the tube.
- The Westinghouse profilameter system scanned each red twice (up and down) at two rod orientations (00 and 1350 ). Hence, with two 900 opposed transducers, two) diameter traces of each rod were obtained at no and 900 and at 1350 and 2250 , respectively. The system was checked against calibration standards before each run. The data yielded a precision of plus or minus 0.04 mils at a le level for mean diameters.
The Babcock and Wilcox profilometer system consisted of a pair of sapphire rod fingers that were installed to contact the cladding across the diameter of a peripheral rod. The differential separation batween the two fingers was monitored by a linear variable transducer (LVOT). With the fingers _in contact with ?.he rod, the measuring head is translated along the'-length of the rod to obtain a diameter profile. The. probable error of a data point is reported to be approximately 0.1 mils.
TARLE 4-2 Actual and Predicted Collapse Times Predicted Collapse 13/ Actual Collaose Time (hrs) Time (Hrs) 9.78(a) 1171(a) g371, 8.55(a) 2174(a) 1442 6.10(a) 2520(a) 2481 6.08(b) g3nn(b) 13nn, 5.17(b) Collapse Not nbserved in 2501 Hrs (D) 1717 6.27(C)~ 750(C) 750*
6.17(C) Collapse Not Observed in 1500 Hrs (C) 1035 3.69(C) Collapse Not Observed in 1500 Hrs (C) 1R49**
(a) Data from Reference 36 (b) Data from Reference 37
+
(c) Data from Reference 37 h/a is approximately 0.10 for all specimens. In any data set (denoted by the same reference letter) the only variable is the axial length.
This value is equal to the actual collapse time since the FLCF associated with this gap is equal to the FLCF corresponding to the largest gap in the test series (see Page 4-16)
The discussion in the text clarifys the relationship between the predicticn of collapse time and the lack of collapse observed.
4.3 APPLICATION TO PELLET SUPPORTED C-E FUEL RODS A survey of fuel-rods with modern fuel pellets (see Table 2-2) has indicated that no axial gap approaches a size of 0.5 inches. For the base case of Table 4-1, the 1/a ratio associated with this upper limit of axial gap size is 0.5/0.2 = 2.5. Interpolating from the tabular data in Figure 4-3, the minimum FLCF (for h/a = 0.1500) associated with an 1/a value of 1.5 is $11. Since the base case collapse time of Figures 4 4 through 4-9 is N17000 hours, a conservative estimate of clad lifetime for modern C-E fuel is 11 x 17000 =
187000 hours. Clearly, collapse of C-E fuel rods is not a practical consideration when the suppcrt provided by the fuel pellets is included in the structural modeling.
O
- t. .
l l
l l
l l
l 4-R7 - -- - - - - - - - - - -
1 1
Section 5 AUGMENTATInN FACTORS 5.1 BACXGROUND AND INTRODUCTION The densification of UO2 results in the shortening of the active fuel pellet f stack height. When this reduction in stack length results in the formation of axial gaps within the fuel column, additional power peaking will occur adjacent to these fuel gaps. This arises because the decreased neutron absorption due to fuel removat more than compensates for the fission loss. Further additional peaking would result if the cladding were to collapse into the gapoed fuel column. This latter effect would occur because the voided region (i.e., gap) would be replaced to a large extent by water and the increase in moderation of fast neutrons by the water outweighs the increase in absorption in the water.
However, since the time collapsed cladding first was observed, the NRC has required fuel vendors to design their fuel so that collapse into the gap was .
avoided. Furthermore, as described in Sections 3 and 4, the collapse of clad i
is not to be expected during the design life of modern (f.e., prepressurized fuel-rods containing nondensifying fuel pellets) pWR fuel. .
Since the position, size, and number of gaps within a given volume of core cannot be defined explicitly, a statistical approach is used to determine the l potential increased peaking resulting from the presence of gaps. The
[ additional peaking due to axial gaps betwetn fuel pellets is called the augmentation factor. It is defined as the ratio of peak augmented power specified to some confidence level, to the peak unaugmented power. The augmented power distribution is correspondingly the power distribution with the statistical distribution of gaps preserr, while the unaugmented power i
distribution is the base power distribution with no gaps present.
l In the past, C-E's calculation of. augmentation factors used a computer program
.which incorporated conservative representations of the gap size distribution data from the palisades and Maine Yankee reactors. This model also assumed gao sizes that could be as long as 2.5 inches (essentially infinite in length) near
TARLE 5-1 S'JMMARY OF GAMMA SCAN DATA A. Gap Size by Plant Number of Gaps in Each Interval 0 .025 .025 .n5 ,05 .1 .1 .2 . 2 .4 4 .6 .6 .8 .8 1.0 Plant (in.) (in.) (in.) (in.) (in.) (in.) (in.) (in.)
Fort Calhoun 8 0 0 0 0 0 0 0 Calvert Cliffs 22 0 0 0 0 0 .0 0 Palisades 1433 51 5 4 2 2 2 0 Maine. Yankee 44 0 6 6 5 0 0 0 J
B., Sample Size and Fuel Type Number of Rods Number of Rods' -
Plant Gamma Scanned With No Gaos Fuel Tyne
~
Fort Calhoun 5 3 Pres 3urized, nondensifying Calvert Cliffs 16 5 Pressurized, both densify-ing and nondensifying Palisades 79 4 No1 pressurized, 'densi fying Maine Yankee 10 0 Nonpressurized, densifying I
)
l 1
6
/
,- , *e, , - - - . , - , - - . , , - - * - . - - . - --w. - - - - - - - , , , - - - , - ,---4,----w - -,. - + - - - - - -
The Palisades and Maine Yankee gamma scan data were incorporated into the data base to ensure conservative gap frequency characteristics. The fuel rods from these plants were not pressurized, contained densifying fuel, and were found to have more gaps of all sizes. They also contained gaps more than thirty times as large as any gap found in the Calvert C1tffs or Fort Calhoun fuel. These combined data, which yield overestimates of both the number and the size of gaps in modern fuel,' contribute toward more conservative gap distribution characteristics.
Eighty-nine percent of the rods sampled were found to have at least one gap, giving a value of 0.89 to f .g Based on the observed data, a uniform axial frequency distribution, f g, has been assumed. The relative size distribution function, f ,eclosely models the data, showing many small gaps and fewer large gaps, as shown in Figure 5-1.
5.2.2 Gao Peaking Characteristics '
f I
Two types of gaps may cause a local power peak in an axial region: aligned gaps, which are gaps ,that occur in neighboring rods about the rod of interest, and end gaps, which are gaps that occur within the rod of interest. The relative power peaking in the rod of interest due to aligned gaps is dependent upon the physical characteristics of the core under consideration. Local-power-peaking factors resulting from single gaps located in various lattice posit, ions relative to an ungapped rod are obtained from two-dimensional X-Y 00T calculations. Only the nearest 26 fuel rods surrounding the rod of interest have a significant power-peaking influence. As used in the augmentation factor computations, these 36 pins are divided into 5 radial groups, with the value j for each group taking on the average of the rods within that group. The l assignment of the 36 rods to the various radial groups is illustrated in. Figure j 5-2. Figure 5-2 also illustrates the percent increases in power in an ungapped rod due to a single gap at each of the five surrounding radial group locations for a typical C-E core.
If the augmentation factor is computed as a function of exposure, it is assumed I
that the reactor has operated to the burnuo of interest without gaps. The single-gap peaking factors frcm the X-Y 00T are calculated as though the gaos had formed at the time of the calculation. An examole of the variation with burnup is 41ustrated in Figure 5-3. As can be seen by Figure 51, the single-gap peaking factors are a weak . function of burnup for all five radial groups.
1 J L 0.95 .
0.94 . , ,
,/l \
,,/ -
i e 0.10 - {
z g . .
E 3 '
o -
o o
I O
u.
O D
3 g 0.05 -
m O
E 0 ! - - m u a u I u. 3 I I I I 0 60 100 S,- RELATIVE GAP SIZE (100 = 0.7 IN.)
I Figure 6-1: Relative Gap Size Distritmtion Using a Maximum Gap of 0.7 in.
In Auynentation Factor Malal 1 _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ . _ . _ _ _ _ _ _ _ _ _ _ _ _ . _ _
UNGAPPED (1) (3) (5) l RCD- 5.44 1.84 , 0.79 1 (1) (2) (4) (5) 5.44 3.48 1.56 0.79 (5) (4) (5) 1.84 1.56 0.79 i (5) (5) 0.79 0.79 LEGEND (n) RADIAL GROUP NUMBER x.xxx % INCREASE IN PEAKING IN UNGAPPED RCD DUE TO A SINGLE GAP AT INDICATED LOCATION Figure 5 2: Rod Location Assignment to Radial Groups and Limiting Single Gap Power Pesking 1
6
, , . . , - - . . , ..w-- ,v- ,a.,y.e--e,--..p v w- , - . - - - s. -r-x e r r~
l l
a
(
- N M *
- 50
<8 f zC f t j '
n
-3
- R h M
w
.I 6
- 3
>= g-lE s d i 4 c &
B. i o2 c.
- N z g ll3 .
m 3 m
o E
uh N
o A
8 s a e i ,
o E T M N e g
%'dVD 013n0 DNINV3d A-7 _ _ _ _ _ _ _ _ _ . - _ _ _ _ _ _ _ _ . _ _ _ _ _ __.
._ . ~ . - . -. --
l These peaking factors exhibit a maximum for each radial group somewhere between 20 and 40 GWd/T burnup. A generic set of these peaking factors can therefore be specified, which will be conservative for all burnups.
To obtain the peaking due to combinations of multiple gaps distributed among the various groups, the single-gap power-peaks are combined multiplicative1y.
This approach is based on a comparison (Table 5-2) of power-peaking obtained from X-Y 00T calculations for various combinations of multiple gaps with the estimated peaking obtained from both multiplicative and additive combinations -
of the single-gap peaks shown in Figure 5-i. This comparison shows that the multiplicative combination procedure provides a close, and generally conservative, estimate of the DOT results for power peaking due to multiple gaps.
Since the X-Y 00T calculations correspond to an infinite gap height, corrections for the finite length of the assumed gap must be made. These are obtained fecm R-Z 00T cases containing either a central gap (to model an end-gap) or an annular gap (to model an aligned gap). Figure 5-5 illustrates the
, shape of the curve of relative power in the rod of interest as a' function of gap length in a single adjacent red containing the gap. For a given region the -
magnitude of g7 is obtained from Figure 5-5 for the appropriate gap size in the region. This factor is applied to the infinite gap peaking values, by ,
in each of the radial groups.
The relative power peaking caused by_ an end gap is determined from R-Z 00T calculations where the gap of interest is introduced into the radial middle:'of an essentially infinite cylinder whose center is at the radial origin. These analyses result in the Ifmiting end gap size power peaking effects shown in Figures 5-6 and 5-7.
The power peak calculated by R-Z 00T for a 2.5 inch end gap that is axially adjacent to the fuel region of interest is illustrated in Figure 5-6. In view of the axial conductio*. and radiative transfer effects due to the presence of the gap in the rod of interest, an effective value of 1.01 has been employed for this end gap peaking. This value is quite conservative, since, as discussed in the next section, the maximum gap size is much smaller than 7. 5 inches. The effect of end-gap size on the end-gap peaking has also been 5-8
TABLE 5-2 EFFECT OF MULTIPLE GAPS l
Number and DOT Combination of Single Gao peaks -
Location of Gaos Calculation Multolicative Additive 2A 1.0941 1.0954 1.0932 4,A 1.1981 1.1999 1.1864 4,8 1.1235 1.1267 1.1211 2,C 1.0320 1.0324 1.0321 4,C 1.0642 1.0658 1.0643 4,0 1.0545 1.0558 1.0541 4,G 1.0249 1.0252 1.0249 2,C+4,0 1.0874 1.0894 1.0862
~
2,3+ 2,F 1.1406 1.1421 1.1347 1,A+1,8 1.0785 1.0783 1.0769 i
l i
i
. _ .. b 9?_ _ _ _ . . _ ... _ _ . _ - _ _ _ _ , _ _ _ _ _ . . _ _ _ _ . . _ . - - - _ , . _
l l
l 4 I I I 4 l l
1.06 -
1.05 -
1.04 -
e w I E !
C a.
m 1.03, - *
. 2 .
i- l W
z
=,
1.02 -
1.01 .
Mr M.
1.00- I I = 1 1 0 1 2 3 4 5 AXIAL DISTANCE FROM MIDPLANE OF GAP, INCHES Figure 5 6: Relative Power vs. Axial Distance from Gap Interfaca c%Rs-...-.-.,--.--.---.----------------
l- ,
10 i i i i i 'V
.i 0.9 -
0.8 .
l 0.7 -
e ui N
E 0.6 -
- a. -
c u.
O g 0.5 -
m
- u. - ,
- u. . -
W .
m -
h 0.4 -
a w
e 0.3 -
. 'g i
0.2 -
0.1 ,
' I I I I O'
O 1 2 3 4 5 N =
GAP SIZE, IN.
Figure 5 5: Relative Effect of Gap Size on Single-Gap Power Peaking 9 11
1.0 , , , , V 0.9 -
0.8 = -
47 - =
c ul N
E g 0.6 - -
c C
2 m
4 o 0.5 - -
o w
w w
m W E4 - -
E d
t w
0.3 - =
l- 0.2 - ,
t l
l 0.1 (- -
0 1 I I I I O 0.5 1 1.5 2 2.5 ==
END GAP SIZE,IN.
Figure 5 7: Relative Effect of Gap Size on End Gap Power Peaking l
- - - . . - - _ _ . -.-.---5.11
l calculated with R-Z DOT for various gap sizes. The magnitude for the relative end-gap-size coefficient, gkr, is illustrated in Figure 5-7.
5.2.3 Maximum Gao Size The maximum size of a gas has been selected on the basis of the same gamma scan data used to determine the gap size distribution functions. This data, as can be seen in Figures 5-A through 5-11, shows that the maximum gap size is independent of axial location in the fuel rod. Table 5-1 indicated that the largest gap occurred in a fuel-rod from the Palisades plant. This gap measured 0.68 inches in length. (This value is more than twice that of any gap
. found in modern fuel-rods.) A maximum gap size of 0.7 inches was chosen as a conservative upper bound of observed data for the initial model.
5.2.4 Radial Pin Power Distribution The radial pin power distribution is obtained in histogram form as a function of burnup from two-dimensional fine mesh (either P00 or MC) cal ulations.
These distributions are reactor type dependent. This differential information is processed to produce the integral function, R(X), which, fn turn, is used in ,
the convolution procedure described below. A typical integral-radial-pin-cower ~
distribution plot is presented in Figure 5-12 for pin powers of 1.1 and greater.
5.3 Mn0EL DESCRIPTION .
Both the conventional and the modern fuel calculational models are described in this section. A random combination of gap sizes and locations in rods surrounding a typical fuel-rod of interest is selected. Power peaking in the rod of interest due to these random gaps is computed and saved as an event.
Numerous such events are then used to construct a power-peaking frequency distribution which, along with the unaugmented radial power distribution, is used to obtain the peak augmented power. The augmentation factor is then l computed by dividing the peak augmented power by the peak unaugmented power.
l This sequence is described in more detail below, and shown in Table 5-3.
t The reactor is first divided into uniform axial regions for computational purposes. These regions are chosen to have a height of approximately four (a) f rches so that interactive peaking effects between any two adjacent regions are minimized. This region size is consistent with the R-Z 00T calculations as shown in Figure 5-6. The model, therefore, assumes that any such effects, l
w
h 15
-- - CALCULATED BY MODEL OBSERVED 0
=
u 3
e '10-d (D
E C
w w
U #
X w '
E C
u.
O e
w a .
2 3
Z~ 5-0 '
1 0 20 40 60 80 100 i
% OF CORE HEIGHT Figure 5 8: Axial Distribution of Gaps over 0.01 in for Prepressurized, Modern Fuel Rods
JL 1.00 CALCULATED BY MODEL OBSERVED 0.75 -
E
-m m
o=
A$
- u. J
=<
w>
e5 m-
- n. z 0. 50 C>e 3
su -
N2 lE 0.25 -
0 ' I I I I ' i i 0 20 40 60 80 100
% OF CORE HEIGHT Figure 5 9: Maximum Gap Size vs. Core Height for Modem Prepressurized P'NR Fuel from Calvert C:lffs and Fort Calhoun 5-16
l 2.4 -
PALISADES -
2.0 ' -
m OBSERVED
- 1.6
. z .
u ua z Cf ~. -
ma
- o. <
<3 c- 1.2 -
~
3 iz$
X ~
e jc= 0.8 -
0,7 -
e
~
- 0.68 i
0,4 -
I I I i 0 '
l 0 20 40 60 80 100 l
% OF CORE HEIGHT Figure 510: Maximum Gap Size vs. Core Height for Nonpressurized, Densifying PNR Fuel from Palisades
l u i 1.00 - l l
==---- CALCULATED BY MODEL OBSERVED 0.75 -
I us m o=
-ciy A-
- u. s
=
w ><
N
-WE
,. [y 0.50 - ,
og ' -
35 3-2W .
X
<E 2 .
0.25 =
l- 0 I i i
! 0 20 40 60 80 100 i
% OF COPE HEIGHT 1
Figure 511: Maximum Gap Size vs. Core Height for Nonpressurized, Densifying PWR Fuel from Maine Yankee l
l
- A JLO ______ _._ _ _ _ _ _ _ _ _ _ _ _
1.600 g
1.400 X 1.200 2
h ~
e
$ 1.000 w
^
g- -
z
- ;$ 0.800 gE a.
5 e
~
E \ -
us 0.600
'E a.
1 =
c d
- 2 0.400 0.200
\ l I
0.000 1.100 1.150 1.200 1.250 1.300 1.350 1.400 RADIAL PIN POWER, X l
Figure 512: Integral Radial Pin Power Distribution for a Typical Reactor i
(_ ,, - - - - - - - " - - - - - * ' ~ ~ ' ' ' ~ ~ ' ~ ~ ' " " ' ~ ~ ' ' " ~ '
~
TABLE 5-3 CONCEPTUAL BASIS FOR THE AUGMENTATION FACTOR MonEL
- 1. Rod of interest is surrounded by 36 rods in 5 radial groups
- 2. Create a set of 37 random numbers (usually by using a random number generator) and, based on built-in gap statistics, use random numbers to determine a gap distribution in the 37 rods
- 3. Calculate local-peaking at rod of interest due to gap distribution 4 Repeat for several thousand sets of random numbers to determine probability, p(A), of obtaining a given local peaking, A
- 5. From integral pin census, R(X), calculate expected number of rods exceeding a given augmented power, Y, due to the local peaking A EA (Y)
= p(A)
- R(Y/A) *
- 6. Sum over range of local peaking values to get augmented pin power distribution "
E(Y) =
r Eg(Y)
A
- 7. Interpolate at E = .355 to determine peak augnented power, Y 95
- 8. Calculate the augmentation factor as ratio of peak augmented oower to peak unaugmented power FA Ygg / Xmae
- 9. Repeat entire calculation for several regions to obtain axial variation
- 10. Apply model at several burnups throughout cycle.. Maximum augmentation factor is applied over entire cycle.
j l
between gaps in one region ind peaking in an adjacent region, are negligible.
Any gap that has its center in a given axial region is modeled to occur at the center of that region and to be contained entirely within that region, so that it will have a maximum effect on peaking.
Thirty-six rods in three rows around a rod of interest are modeled as contributing to the power peaking in the red of interest. These rods are split into five groups according to their distance from the rod of interest. For an identical gap in each rod, the members of a group will have identical effect on the power peaking.
Based upon the statistical models for the older types of fuel, the calculation "
would begin by choosing an axial region. The program would then assume that there is a maximum possible gap size that had been empirically selected on the basis of available gamma scan data. If a distribution of gap sizes up to the maximum is allowed, the probability that a gap with relative size pS will nccur in axial region i of any individual fuel rod is given by:
Pf = fp fj f, g (5-1)
~
'4here gf is the fraction of fuel rods having gaps, f g is the axial gap frequency distribution at regior.1 (i.e., the fraction of gaps located 'within region 1), and fp is the fraction of gaps that have relative gap size Sp .
- The relative size distribution, f , pis assumed to be independent of axial position. The probability that a given fuel rod does not have a gap in region i, is therefore given by:
Pf=1.0- Pf = 1.0 - fjfg (5-2)
To determine the gap distribution within an axial region, random numoer decisionintervalsbasedonPfandPfareconstructed(asshownin Tabl e 5 .1) . For the rod of interest and each of the .16 surrounding rods, a random number between 0.0 and 1.0 is then chosen. If the random number is less thanPfthentherodisassignednogap. If the random number has a value that falls within one of the other decision intervals, the rod is said to have a gap of the corresponding relative size. The actual gap size in a given rod within region i is then the product of the randomly chosen relative size and the maximum gap size possible in region i, m
l TABLE 5-4 RANDOM NUMRER DECISION INTERVALS -
- 1. Random numbers are uniformly distributed between 0.0 and 1.0
- 2. Decision intervals are chosen withi.1 this range for each gap size so that the width of the interval is just the probability of having a gap of that size E Pg P E E P 0 2 3 4 5 E
6 E 7
I i 1 I I I I I I 0.0 1.0 Pg = Probability of not having a gap
- Pp = Probability of having a gap in size interval R
'i
- 3. The value of the random number selected for a given rod then determines whether that rod has a gap, and if so, of what size 9.99
The model for modern fuel addresses these parameters slightly different. This latter model assumes thatj f is uniform. Therefore, all axial regions will have the same Pf. It also assumes that the maximum gap size is position i independent.
Having calculated the gap distribution around the rod of interest, the increase in power peaking in this rod due to the specific arrangement of gaps is computed. As previously mentioned, a series of two-dimensional 00T calculations that evaluate the effect of single and multiple gaps on rods at 1 various locations has shown that the total peaking in the rod of interest due to a given combination of gaps in the neighboring rods is best computed by multiplicative1y superimposing all the single-gap contributiens. This resulted in a more conservative peaking factor (higher total peakt ig) than if the single gap effects where combined in an additive manner.
The power increase in peaking in the rod of interest can therefore be expressed as:
35 (1 + skr k) g (1
- 9rb j)
A = b (5-3) neighboring rods where b y is the fractional increase in power at the red of interest due to a single gap of infinite length located in red i in ' radial group j; the relative
- effect of gap size on the power increase is accounted for by the gap size
- l dependent factor gp ; the quantity bk accounts for peaking due to an end gap in the rod of interest; and g gp is the size dependent factor to account for
, the relative effect on peaking of end-gap size. Any rod that does not have a gap does not contribute to the additional peaking since g assumes a value of I
zero when there is no gap.
The resultant power peaking is saved as one of thousands of such events that eventually are used to build a frequency distribution of the power peaking amplitude in the red of interest. A gap-size-frequency histogram is also constructed. The total peaking amplitude is normalized to yield P( A), a peaking probability function that represents the probability of the local j increase in peaking due to gaps having a magnitude of A.
l l The power peaking probability function and the integral radial power distribution are then comoined to arrive at an augmented gewer dis.tribution.
l l
Specifically, the formulation involves summing over all values of peaking to obtain:
E(Y) = I P(A) R (Y/A), ,
(5 4) all A Where E(Y) is the expected number of pins in a given axial region that exceed some augmented power Y and R(X) is the integral unaugmented radial power di stribution. R(X) represents the number of fuel pins in the core that have relative power greater than X. R(0) would equal the total number of fuel rods
, in the core since all rods have a power greater than zero.
After obtaining the augmented distribution E(Y), one can interpolate to find the augmented power corresponding to a specified expectation value. In particu-lar, one can determine the augmented power, Y95, for each axial region such that 95% of the tig.a there will be at most one rod having a worst power peak.
An expectation value of 0.355 is necessary to achieve this 95". confidence level.* The augmentation factor for the region is then simply Y /X 95 max' The augmentation factor calculational model as modified for the maximum gap
, size of modern fuel, described above, is illustrated schematically in Figure 5-
- 13. As noted later, the model is used to calculate augmentation factors at several burnup steps during the cycle. The highest values calculated are taken as the augmentation factors ta be applied throughout the cycle. This set of augmentation factors is used in determining margin to core operating limits through technical specification limits on linear heat rate, and for determining input to the DN8R calculation used to establish thermal me, sin.
- If P (M,Y e3 is the probability that at most M rras have a peaking worse than l Ye , then the cumulative Poisson distribution elves M m P (M,Ye ] = T~E(YC )1 EXP(-E(Y e ) ) I m=0 ml To obtain P (1,Y 3
- 0*90 95 requires an expectation value of E(Y95) = 0.355.
l_
theauennented Redlet Pla Peever Dieselbeselea X-Y $4aele Ga desse cose speelel P.Alag mast.p Cohadessoa Gep A"Id3 8 88880"* 88"d*88 h"3"'"**
feegsency Gap Sise Italeske PeAlag Feegaeacy hamber Gap
,,, g , $g,,
thesiliusion EEselbution Geassaalen Save '
I I I I Resideas haabee Decisiosi latervat '
I I '
I :
Di e G Aeound Red et lateseas g i i Peaklag Psutselsdity Distaduation -
t I i Eapactasion Values I (Augeneased Rads.1 Pia Powes Distadautical i
Peels Auganeased Powee (96% Comiidene.e Csisesnomi l t Augsusasa: ion tactus I
a a leshaical
- DNun Specidecations C lculeelon itiaeae maa Res.
Figure 513: Augmentdtion Factor Model Diagram
5.4 COMPARISON OF M00EL TO DATA Results from the implementation of the calculational model described above, are presented in' this section. The computations may be divided into two categories: gap distribution and augmentation factors. Gap distributions are compared with measurements from modern fuel to illustrate that the model's gap characteristics result in distributions that conservatively envelope this observed data. Augmentation factors are computed for a typical reactor using both the gap characteristics inherent in the model and, to illustrate the ~
effect on calculated augmentation factor, gap characteristics that more closely reflect the measured data for modern fuel.
5.4.1 Comoarison Between Calculations and Measurements ,
P Expected gap distributions were ccmputed using the model for comoarison with the gamma scan gap data from Fort Calhoun and Calvert Cliffs spent fuel rods.
, The model has been used as it would be in a calculation of augmentation factor, except that it has been applied to a set of rods the same size as the gamma scan sample, rather than to the whole ' core.
The comparison between calculated and measured distributions of gaps for the Fort Calhoun and Calvert Cliffs fuel is illustrated in Figures 5-8 through 5- '
- 11. Figure 5-8 illustrates the number of gaps larger than O.010 inches in
- axial intervals of 10% of core height. There were no gaps in this modern fuel greater than 0.025 inches, as can be seen in Figure 5-9, which shows the maximum gap size (if greater than 0.01 in.) in each of the axial regions. The largest gcp in each axial interval for the older Palisades and Maine Yankee fuel is shown in Figures 5-10 and 5-11, respectively. It is clear from these figures that the use of a unifonn-axial-frequency distribution combined with a uniform maximum-gap size results in gao distributions which very conservatively envelope the observed data for modern fuel. This is further illustrated in Figure 5-14, where the overall gap-size distribution is presented. The number of small gaps is underpredicted by the model, but the number and maximum size of large gaps is considerably overpredicted. Since the contribution to the augmentation factor results almost entirely from the larger gaps, this overprediction by the model is conservative, i.e., in a direction that results in increased augmentation factors.
AT y = 31, x=0 3 1
1 26 .
CALCULATED BY MODEL 24 - OBSERVED IN FUEL 22 .
20 =
N E 18 -= = -
c ,,'===-
E
'T U 18 - \
- \
m s*
A g 14 - \
s \
a-g 12 .
\
g w
.= \
10 - \' -
i
\
8-
\
\
8- \
\
! \
l 4=
\
\
\
2- g u _ _. ,
0 ATx=0.7
- y = 0.0 '
O 0.01 0.02 0,03 0.04 0.05 GAP SIZE (INCHES)
Figure 514: Gap Distribution for Modem Fuel (Calvert Cliffs and Fort Calhoun) l
\_ _ , . _ _ . _ - - . _ . . - . . - - . . - - - - - - - -..
5.4.2 Calculation of Typical Augmentation Factor The augmentation facter has been calculated for a typical core using the previously discussed gap distribution which is shown in Figure 5-1. The radial power distribution and gap peaking factors used for this cas~e are shown in Table 5-5. As an example of the extent to which the conservatism in the model affects the calculated results, the augmentation factor for the same typical reactor was recomputed utilizing gap characteristics more nearly like those observed in modern, stable fuels as represented by the data from Calvert Cliffs fuel. In this case, the gap-distribution parameters were selected so that the '
computed ten region gap distribution would just envelope the observed distribution. That is, in all regions except that with the largest gap, the computed gap distribution has more and larger gaps than the measured data. In the region containing the largest gap, the distribution just coincides with the data. The results of these calculations for the typical reactor are comoared inFiguie5-15. As expected, the use of this more " representative" distribution of gaps results in augmentation factors that are considerably lower than those calculated using the model parameters.
5.4.3 Comoarison Between Models .
A number of potential models were considered prior to the implementation of the
,one described above. To reduce the distribution weighting caused by the excessive number of small gaps observed in the palisades fuel (compared to ,
other plant's fuel), a model based on all data except that from Palisades das investigated. The maximum gap size was then varied within this model to show the augmentation factor as a function of the maximum gap size. As the maximum gap size was reduced, all gaps larger than the maximum gap size were included at the maximum gap size in the gap frequency distribution. This function is plotted in Figure 5-16 and illustrates that, although this model is more conservative, it still predicts augmentation factors well below 1.005 for modern, nondensifying, precressurized C-E fuel in which the largest gap observed was 0.07.5 inches.
5.5 IMPACT OF CHANGING THE AUGMENTATION FACTOR In those C-E plants using analog monitoring and protection systems, augmentation factors are used as an axially dependent penalty in the CECOR linear-heat rate (LHR) calculator performed for monitoring the LHR LCO and in calculating the LHR trip setpoints and monitoring limits. In those C-E plants
TABLE 5-5 INPUT OATA FOR TYPICAL AUGMENTATION FACTOR CALCULATION A. Radial Pin Power Distribution:
Number of Pins with Pin Power Interval Power in Interval 0.0 - 1.000 7619 .
1.001 - 1.050 4000 1.051 - 1.100 4000 1.101 - 1.150 4000 1.151 - 1.200 4000 1.201 - 1.250 anna 1.251 - 1.300 6000 ,
1.351 - L.349 569 1.400 - 1.401 -
3 ,
1.401 - 1.40101 1
- 8. Single Gap Peaking Factors:
?, Peaking in Ungapped Rod .
Radial Group due to Single Gao in Indicated Grouc 1 5.44 2 3.48 3 1.84 4 1.56 5 0.74-
JL 1.01 -
CALCULATED WITH MODEL
. -__ CALCUL'ATED WITH MORE REALISTIC GAP CHARACTERISTICS AND MAXIMUM GAP SIZE M
i e o
- U w
2 o
- p. ~
H 2
1.005_ -
IM
- 3 .
c 3
3 0 i
e 1
l
.i l 1.00 e i 4 e 0 20 40 60 80 100 i % O/ CORE HEIGHT I Figure 515: Augmentation Factors for a Typical Core
1.020 -
I j 1.016 -
l ac O
i b s 4 1.012 -
- u.
i 2 9
t-
? 4 .-
}
e 1.008 3
4 4 i
1.004 - .
i 4
} lM g i I I O 0.1 0.2 0.3 0.4 MAXIMUM GAP SIZE (INCilES) i l
Figure 5-16: Augnantation Factor as a Function of Maximum Gap Size i
^em +
d using digital monitoring and protection systems, augmentation factors are used as axailly dependent constants in the COLSS and CPC linear heat rate calculations.
Augmentation factors also impact evaluations of fuel rod performance and LOCA and non-LOCA safety analyses. In the following sections, these uses along with the effects cf the proposed changes are discussed.
5.5.1 Setpoint Analysis for Analog System protected plants C-E's analog monitoring and protection system plants use in-core detector information processed by CECOR to calculate peak ifnear heat rate and aid the operator in, avoiding violation of the LHR LCO (commonly known as the "LOCA limit"). Augmentation factors are used as an axially dependent LHR penalty in CECOR's LHR calculation. If the LHR (LOCA) limit were otherwise unaffected, a reduction in augmentation factor by 4% at a given axial location would result in a 4% gain in LHR mergin at that location.
C-E's analog plants also use ex-core detectors and preoperational analyses to determine axial shape index (ASI) monitoring limits and ASI trip setpoints, to -
protect against violation of the centerline melt limit's. The ASI monitoring ifmit setpoints are used use when the in-core detectors are inoperable. The
, quanti ty used in establishing those monitoring and protection system limits, on linear heat rate is the Power-to-Fuel-Design-Limit on linear heat rate, PFDL.
The quantity PFDL is defined as:
PFDL = (Weim . 100) / (Aug Fq 3-0 FeWavg) (5-5) where W elm =
linear heat rate SAFDL (kw/ft)
Aug F q 3-0 =
Ratio of the 3-0 peak LHR to the average LHR, including the augmentation factor for fuel densification effects; F, = Local heat flux engineering factor W
avg
= Core average value of linear heat rate, at full
- rated reactor power.
and Aug F q 3n , g,x p](g),p(),p(:)
x Z a
where F T (z) xy =
The maximum planar radial peaking-factor at height z Fg (z) = The axial peaking-factor at height z Fg (a) =
axially dependent fuel densification augmentation factor.
The quantity PFDL, or kW/ft overpower, is the percent of rated power to which the reactce could hypothetically be raised, for the given Aug Fq3 '0, before the fuel centerline melt limit of Wej,would be reached.
In the axial-power-distribution setpoint analysis, many axial pcwer distributions are generated as a function of cycle life and Control Element Assembly (CEA) insertion. As a part of this' analysis, for each core condition examined, a value of PFDL and an Axial Shape Index (ASI) are generated and stored as an ordered pair. As ghown in the above equations, the PFDL is generated using axially dependent augmentation factors. These factors -
- represent power spikes caused by gap formations in neighboring fuel rods. The ASI is a measure of the power generated in the bottom half of the core, minus the power generated in the top half of the core, divided by the total power.
Plots of PFDL versus ASI are then generated for use in the determination of excore detector linear heat rate monitoring and protection system limits. A typical PFDL plot, generated for a 90% power CEA insertion at the beginning of a cycle, is shown in Figure 5-17.
f Hypothetically, if the augmentation factors were reduced completely to 1.0 when generating the PFDL data, there would be an apprdximate 47, gain in kw/ft overpower margin within the limiting ASI range of -0.3 to 0.0 In the setcoint analysis, this gain could be used by raising the protection system limits by 4".. In addition, if the kW/ft LOCA Ifmit did not change, then the monitoring system limits could also be raised by 44.
Anaxiallyindependentaugmentation; factor would result in a margin gain over most of the ASI space, and an -
insignificant margin loss in the extreme positive ASI region. Therefore, the new augmentation factor model is expected to result in an overall gain in kW/ft-related operating margin.
170 i
160
=
b =
- 150 a
- as ..
m -
3 .
u .
8140 ,
c e
.' =
z -
i A
P
.4 w
2 130 w
2 .
.a - ,
e
- 8 *
- 120 z ,
m ._
u **
m - .
m +
C .
b110
- m x
= -
8 a.100
- c w
4 m
w z 90 j c 5
a.
80' i
70 I
60L I f f I ; ;
4.300 0.600 0,400 0.200 0.000 0.200 0,400 0.600 0,300 AXIAL SHAPE INDEX Figure 517: PFDL Plot . 90". Power CEA Insertion at SCC
. _ . . ~ . . _ - . _ . _ _ . _ . _ _ _ . _ _ _ . . . . _ _ _ . _ _ _ _ _ _ _ _ . . _ _ _ . . _ .
_ ~ . - - . - . . - . . . . . -
l 5.5.2 Setpoint Analysis for Digital plants In C-E's digital monitoring and protection system plants, the Core Operating Limit Supervisory System (COLSS) calculates linear heat rate on-line using in-core detectors and compares it to the LHR LCO (LOCA limit). At the same time, the Core Protection Calculators (CPC) calculates LHR on-line using ex-core detectors to provide a trip to prevent violation of the centerline melt limit.
In both cases, augmentation factors are used directly as axial dependent constants which are multipliers of the LHR axial distribution. If the LHR (LOCA) Ifmit and the centerline melt limit were otherwise unaffected, a ~
reduction in augmentation factor by 47, at a given axial location would result in a 47, gain in LHR margin at that location.
5.5.3 Fuel-Rod performance
! The fuel performance code FATES is used to generata the initial fuel-red conditions for input to the LOCA analysis and for input to certain safety analyses (e.g., loss of flow, and CEA ejection). The FATES-seneratad data characterizes the fuel-rod as a function of burnup. During the FATES analysis, a nominal rod depletion is performed, and the bounding LHR is representative of the peak kW/ft value allowed within the LCO's.
l Augmentation factors address the linear-heat-rate spikes that occur over only a small axial region corresponding to that of the gaps in neighboring rods. ,
Realistically, these small spikes, when considered over the length of the fuel-
[ rod have an insignificant effect on fission gas release, gap conductance, or internal pressure. Therefore, the peak linear-heat-rate assumed in the FATES analysis is the "LOCA limit" divided by the augmentation factors included in the monitoring system (stored energy, in the form of temperature distributions, for LOCA analysis reflects this factor as discussed belew). Effectively, then, augmentation factors are credited in the FATES analysis. If augmentation factors are removed or reduced in the LHR monitoring calculation, it is necessary to also remove or reduce the credit in the FATES analysis. If the LHR (LOCA) limit remains the same, the lower augmentation factor credit in FATES will result in increased fission gas release and reduced gao conductance (h g ,). The overall impact will be slightly more adverse FATES results. A small reduction in the LHR (LOCA) Ifmit could be used to account for the removal or reduction of augmentation factors in the FATES analysis.
. ._ - . . - rv m
__.__._....s . . . . . _ _ _ . ._ _ - . .
5.5.4 ECCS Performance Analysis Augmentation factors are used in two areas of the ECCS performance analysis.
First, they are used in generating the fuel performance data that are used as input to the ECCS performance analysis. Secondly, they are used in establishing the hot rod peak linear heat generation rate (PLHGR) used in the ECCS performance analysis. This PLHGR value is consonly referred to as the "LOCA W/ft Limit". ' The impact that reducing the augmentation factor has on the fuel performance data was discussed in the preceding subsection. It is expected that that effect will, in turn, conservatively affect the results of the ECCS performance analysis. The impact of reducing the augmentation factor on the PLHGR is discussed in the following paragraphs.
The ECCS performance analysis determines the PLHGR that is used in the setpoint analysis to establish the monitoring system limit for linear heat rate. The ECCS performance analysis and the setpoint analysis together model the augmentation factor in a conservative manner in determining the monitoring system limit for linear heat rate. In the ECCS performance analysis the hot cod is analyzed at a PLHGR that includes the heat generation rate effects of the augmentation factor while the fuel roda surrounding the hot rod have heat generation rates that do not include the augmentation factor. Thus, the radiation heat loss of the hot pin is enhanced.
Reducing the augmentation factor to 1.0 in the ECCS performance analysis tends to result in a reduction of the allow:d linear heat rate. This PLHGR reduction will always be less than the corresponding increase in the monitoring system linear heat rate when decreasing or eliminating the augmentation factor. In some cases the ECCS performance analysis PLHGR will not be reduced at all if the augmentation factor is decreased. The net effect of the removal of ' augmentation factors from the LOCA Ifmit analyses and from the monitoring and setpoint analyses is an expected gain in the monitoring system linear heat rate operating margin of up to at 5.5.5 Safety Analysis The FATES data are used as inout to the code used to perform the loss of ficw and Control Element Assembly (control rod) CEA ejection analyses. If the augmentation factors are reduced and the kW/ft LOCA limit remains the same, then the FATES data would produce lower hgg, coefficients. However, the rusn
impact of the lower hga, coefficients on the analyses of the above events will be minimal. To bound these events, the worst parameters from beginning, middle and end of cycle are selected as input for C-E's analysis procedures.
All use a minimem beginning of cycle gh ,p to minimize the rate of heat-flux decay after trip. Since the beginning of cycle h gap at the start of the reload fuel cycle is small to begin with, the change in the hga, coefficients due to reducing the augmentation factor would have a negligible impact on the results of the three events.
If the augnentation factor reduction also leads to a lower k'4/ft LOCA limit, the margin available to ride out a given transient may be reduced. This would affect the setpoints as dascribed above and result in an initially narrower ASI operating space. In turn this would change the starting point of the non-LOCA safety analyses, but would not have any other inherent effect an the safety analyse's.
e e
0 I
e
. _ , , _ . _ _ _ ,,-,,,,-,-.r - - - - - - - - - ~ ~ " ' ' - ' " " ~ " - " " " ' ' '
Section 6 CONCLUSIONS AND RECOMMEN0ATIONS
6.1 CONCLUSION
S 6.1.1 Evaluation of Data Rase Combustion Engineering has carefully examined fuel rods of both an old and modern design that were irradiated in four C-E reactors: palisades, Maine Yankee, Fort Calhoun, and Calvert Cliffs. None of the rods examined showed evidence of clad collapse, although some showed signs of interpellet-gap formation. C-E also examined Westinghcuse or Babcock & Wilcox reports on both old and modern fuel. Some showed signs of gap formation but no modern fuel showed signs of clad collapse. It is concluded that while modern fuels, (i.e.,
nondensifying UO g pellets in prepressurized fuel rods) undergo interpellet-gap formation, the gaps in C-E's modern fuels (and probably in all modern fuels
~ from U.S vendors) are conservatively represented as no larger than 0.05 inches, 4
and are randomly distributed along the axial extent of the fuel rods.
6.1.2 Revised CEDAN Analysis Results '
The CEPAN code, a predictor of the time to clad collapse, was modified to account for the existence of randomly distributed interpellet-gaps. It is concluded that modern fuels, similar in design to that now being fabricated at C-E, have a conservative estimate of time of clad collapse to be 18700n 4
effective full power hours. Therefore, modern fuel can be expected to be resistant to clad collapse significantly longer than its intended useful life.
6.1.3 Augmentation Factor Analysis Results Combustion Engineering's initial method for evaluating the augmented power peaking due to interpellet-gaps was based upon a conservative representation of the size, axial position, anit probability distributions for the gaps discovered in palisades fuel (i.e., densifying fuel pellets in unpressuri:ed fuel rods).
When this method and its associated statistics were applied to an analysis for a specific pin census and set of single gap peaking factors, it produced an u
axially increasing set of augmentation factors with a maximum value at the top of the core as high as 1.06. However, measurements frem both old and modern fuels show an essentially random axial distribution of gaps. Modern fuels exhibit an upper limit on gap size of less than 0.n5 inches, which is significantly less than the 0.7 inches upper limit observed in earlier fuel (i.e., densifying and nonpressurized).
The size distribution statistics based upon different combinations of Palisades, Maine Yankee, Fort Calhoun and Calvert Cliffs data in conjunction with different upper limits on maximum gap size (e.g., 0.7 inches measured at Palisades, 0.5 measured at Maine Yankee, and 0.05 inches, a conservative representation of the 0.025 maximum measured at Fort Calhoun and Calvert Cliffs) were used to determine the augmentation factor for the same pin census and set of single gap peaking factors that were used to calculate the old
- augmentation peaking factor of 1.06. These calculatiens produced an axially independent augmentation factor of less than 1.01. The augmentation factor for a maximum gap size of 0.2 inches was 1.008, while that for a maximum gap size of 0.025 inches was 1.001. .
6.2 RECOMMEN0ATIONS Based upon these analyses, C-E recommends that modern, prepressurized fuel-rods loaded with nondensifying U0 2 fuel pellets, be considered to be resistant to interpellet-gap formation and clad collapse. Therefore, tha penalty for densification caused augmented power peaking shculd be removed from the analysis and 11ca.1 sing requirements of any reactor loaded exclusively with fuel-rods of such modern desige.
i e
- ________ --__.O _ - - - _ . - _ _ . - - - - - - - - - - - - - - -
Secticn 7 REFERENCES
- 1. " Technical Report on Densification of Ligh Water Reactor Fuels,"
(SAEC Regulatory Staff Report, WASH-1236, Novenbar 14, 1972.
- 2. R. N. Duncan, H. Xnaab; et al., " Dimensional Stability of Water Reactor Fuels," Proceedings of the Joint Topical Meeting on Commercial Nuclear Fuel Technology Today, April 28-30, 1975, Toronto, Canada.
- 3. " Fuel Evaluation Model," CENPD-139-NP-A, July 1974 4 N. Funrman, et al., " Evaluation of Eual Rod Performanca in Maine Yankee Core 1 - TASX C," EPRI NP-212, Novencer,1976.
- 5. "CEPAN - Method of Analyzing Creep Co11pase of Oval Cledding,*
, combustion Engineering, Windsor, Connecticut, CENPD-137 (March, 1976). - - -
- 6. R. O. Meyer, "The Analysis of Fuel Densification," US Nuclear Regulatory Commission Report, NLREG 0085, July,1976.
- 7. D. W. Brite, et al., "EEI/EPRI Fuel Densification Project Final Report," EPRI-131, March,1975. ,
- 8. S. D. Harkness, S. R. Pati, et al ., "In-Pile Censification of LPanium Dioxide," submitted for the proceedings of tne European Nuclear Conference, Paris, France, April 21-25, 1975.
- 9. M. G. Andrews, et al ., Combustion Engineering Report, "Densification of Combustion Engineering Fuel," CENPD-118, Rev.
01, May, 1974.
- 10. M. G. Andrews, et al ., "The Perfonnanca of Combustion Engineerins Fuel in Operaticq PWR's," Presented at ANS Topical Meeting on LWR Fuel Performance, April 30-May 2,1972, Portland, Oregon.
- 11. O. E. Bessette, et al., "C-E/EPRI Fuel Performance Evaluation Program, RP 586-1, TASK A, Examination of Calvert Cliffs I Test Fuel Assemolies at End of Cycles 1.and 2," Septemoer,1978.
12.' E. J. Ruzauskas, et al ., "C-E/EPRI Fuel Perfannanca Evaluation Program, RP 586-1 TASX A, Examination of Calvert Cliffs I Test
~ Fuel Assembly after Cycle 3," NPSD-87, Septamber,1979.
- 13. E. J. Ruzauskas, et al., "C-E/EPRI Fuel Perfonnance Evaluation Program, RP 586-1 TASK A, Examination of Calvert Cliffs I Test Fuel Assembly after Cycle 4," C-E NPSD-146, October,1981.
- 14. R. G. Weber, et al., "C-E/EPRI Fuel Performance Evaluation Program, RP 586-1. TASK 3, Examination of Arkansas Nuclear une Unit 2 Characterized Fuel Assemblies aftar Cycle 1," CENPSD-174, July 1982.
- 15. E. J. Ruzaukas, et al., "C-E/EPRI Fuel Performance Evaluation Program, RP '586-1, TASK A, Examination of Calvert Cliffs I Test Fuel Assembly After Cycle 5," CENPSD-241, to be issued.
- 16. G. P. Smith, "The Evaluation and Demonstration of Methods for Improved Fuel Ltilization at Fort Calhoun, End of Cycles 6 and 7 Fuel Exmainations," prepared for the Department of Energy, DOE /ET/34010-10, CENO-414, October,1983.
- 17. R. G. Weber, et al., "EPRI/C-E Fuel Performance Evaluation Program, RP 586-1, TASK 8, Examination of Arkansas Nuclear One thit 2 Characterized Fuel Assemolies after Cycle 3," to be issued.
- 18. J. C. LaVake, G. P. Smith, "The Evaluation and Demonstration of Methods for Improved Fuel Leilization at Fort Calhoun, End of Cycles 4 and 5 Poolside Inspection Programs," prepared for the Department of Energy, CEND-383, August,1980.
- 19. Latter from J. R. Marshall (AP&L) to Robert A. . Clark (NRC), "AN0- -
2 End of Cycle 3, Fuel Examination Results, Docket No. 50-368,"
Letter No. 2CAN948403, April 18,1984.
- 20. S. R. Pati, "C-E/EPRI Fuel Performance Evaluation Program, RP 586-1, TASK A, Gas Release and Microstructural Evaluation of One and Two-Cycle Fuel Rods from Calvert Cliffs I," CENPSD-75, March,
- 1979.
- 21. S. R. Pati, "C-E/EPRI Fuel Perfonnance Evaluation Program, RP 586-1 TASK A, Gas Release and Microstructural Evaluation of Three-Cycle Fuel Rods from Calvert Cliffs I," CENPS0-119, Decenber, 1980.
- 22. A. M. Garde, S. R. Pati, "EPRI/C-E Fuel Perfonnanca Evaluation
- Program, RP 586-1, TASK A. Gas Release, Densification, Swelling i and Microstructural Evaluation of Four-Cycle Fuel Rods fran
! Calvert Ciffs I," CENP50-211, April,1983.
- 23. A. M. Garde, S. R. Pati, "EPRI/C-E Fuel Performance Evaluation Program, RP 586-1, TASK A, Hot Call Examination of a 4-Cycle l Assemoly Cage and 5-Cycle Fuel Rods Irraciated in Calvert Citffs-l I," CENPSD-234, Feoruary,1984
- 24. V. Pasnpathi, " Joint C-E/EPRI Fuei Perfonnanca Evaluation Program TASK A. Fabrication and Characterization of 14x14 Experimental Assemblies," Novemoer 1975.
. _ . . _ . . . _ _ . _ _ _ _ ~ _ . _ _ - . _ . _ _ _ _ _ _ _ _ _ - _ _ _ , . _ _ _ - -
_...y.._ -
ge . - "
- 25. T. P. Papazoglu, " Pre-Irradiation Characterization of Test Specimens in the PWR Demonstration Irradiations Program," Section 7 LRC-4733-1 (NP], November 1976.
- 26. H. H. Dan's, et al., "EPRI/ Babcock & Wilcox cooperative Program on PWR Fuel Rod Performance," RP-711-1, LRC-4733-5, April 6,1977.
- 27. D. K. Throme, et al., " Hot Cell Examinations of PWR Demonstration Fuel Assembly 2340," Sction 2.1.7, LRC-4733-6, February 1981.
- 28. T. A. Coleman and T. n. Pyecha, " Development of an Extended Burnup Mark 8 Design Elght Progress Report," 00E/ET/34213-11, BAW 1532-8, July 1982 - June 1983. -
- 29. WCAP-A2n5, Rev.1, "Zorita Research and nevelopment Program on Special Assembly Test Rods," Semi-Annual Progress Report for the Period Ending June 10, 1973.
- 30. WCAP-A272, Rev.1, "Zorita Research and Development Program on Special Assembly Test Rods," Semi-Annual Progress Report for the Period Ending December 1, 1973.
- 31. WCAP-84nt, Rev.1, "Zorita Research and nevelopment Program on Special Assembly Test Rods," Semi-Annual Progress Report for the period Ending June 10, 1974
- 32. WCAP-9255, " Interim Report Zion 'Jnit 1 Cycle 2 Fuel Performance,"
- E. J. Tarby, et al., January 1979. * ' '
- 33. WCAP-8873, " Interim Report, Surry Unit 2 End of Cycle ?. Onsite Fuel Examination of 17x17 Demonstration Assemolies After One Cycle of Exposure," J. Destefan, et al., Janucry 1978.
- 34. S. P. Timoshenko, J. M. Gere, " Theory of Elastic Stability," .
Second Edition, McGraw-Hill Book Co., New York, 1961.
- 35. " Revised Clad Flattening Model," Westinghouse Electric Corporation, Pittsburgh, PA, WCA18-3381 (July,1974).
- 36. "In-Reactor Tests of Externally Pressurized, Short, Unsupported Lengths of Zircaloy Tubing," Bettis Atomic Power Laboratory, West Mifflin, PA, WAPD-TM-1529. .-
- 37. P. L. Pfennigwerth, O. A. Gorscak, I. A. Selsley, " Deformation and Collapse of Zircaloy Fuel Rod Cladding into plenum Axial Gaps," Paper C5/5, preceedings of the Seventh SMiRT Congress, Chicago, IL, 1983.
L 79
the top of the core. Such calculations resulted in monotonically increasing augmentation factors as a function of core height and maximum values as high as 1.07 for particular cycles of specific C-E reactors.
New models, based on more realistic representation of the gap size distribution functions for modern fuel-rods would result in significantly lower values of the maximum augmentation factor. This could result in increased operating margin for reactors currently constrained by linear heat generation rate limits. This section of the report describes C-E's augmentation factor calculational procedures and the effects of the modern fuel data on the resultant augmentation factors.
5.2 MonEl. PARAMETERS Calculation of the augmentation factor requires that a number of parameters describ'ing the size and distribution of gaps within the core be specified. The impcrtant parameters used in this determination are the axial gap frequency distribution, f , tthe maximum gap size, the relative-gap-size distribution, f p, the fraction of rods having gaps, fg, the relative gap-size-effect coefficients, g p and gg e, the end gap contributio'n to peakihg, b.4, the single aligned gap contribution to peaking, bj, and the radial power distribution, R(X). These parameters are discussed in this section.
5.2.1 Gao Distribution Characteristics -
Three model parameters,g f , gf , and f rare needed to quantify the gap
- distribution characteristics. Spent fuel rods from the Fort Calhoun, Calvert Cliffs, palisades and Maine Yankee plants were gamma scanned to provide data.
All of the fuel rods were taken from assemblies scattered throughout the core.
The characteristics used in the application of this rr.odel have been empirically selected on the basis of this data, which are sumarized in Table 51. The burnup of these rods is tabulated in Tables 2-3 through 2-7 Because all data from Palisades were taken from corner rods, the fuel rod aumber in Table 2-3 refer to a specific assembly corner. This combined data shows: (1) a distribution of gap sizes peaked strongly towards very small gaps, (2) very few gaps larger than 0.025 inches (Tess than six percent for combined Palisades and Maine Yankee data and none for Calvert Cliffs and Fort Calhoun data), and (3) r
- an essentially uniform axial distribution for the larger gaps '(f.e., larger than 0.025 inches) and for gaps of ~all sizes.
L 5-2
1 PERCENT INCREASE IN POWER AT x DUE TO SINGLE GAP AT LOCATIONS A-I A C F x x 4.661 1.607 0.678 1
i B D G 3.028 1.352 0.623 H -
E 0.801 0.434 1
0.269 Figure 5 4: Estimation of DOT Results by Combination of Single Gap Peaking Effects 5-10
- - . - - - - - . . _ _ . _ . . ._- . _ . _ . .- ._. _ _ _ , _ -. ,_