ML20009B419
| ML20009B419 | |
| Person / Time | |
|---|---|
| Issue date: | 11/16/1979 |
| From: | Levine S NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Harold Denton, Minogue R Office of Nuclear Reactor Regulation, NRC OFFICE OF STANDARDS DEVELOPMENT |
| References | |
| RIL-074, RIL-74, NUDOCS 8107150409 | |
| Download: ML20009B419 (16) | |
Text
I t
(
poe g'[.
- I
UNITED STATES NUCLEAR REGULATORY COMMISSION WASHING TON. D. C. 20555
[ f,.
m, MEMORANDuti FOR: liarold R. Denton, Director Office of Nuclear Reactor Regulation Robert B. Minogue, Director Office of Standards Development j.
FROM:
Saul Levine, Director
=
Office of Nuclear Regulatory Research
SUBJECT:
RESEARCH INFORMATION LETTER # 74THE STEADY-STATE FUEL R00 BEHAVIOR CODE: FRAPCON-1
References:
1.
D. F. Ross/L. S. Tong letter to H. E. Ranson (DOE-RL),
and R. E. Wood (D0E-ID), June 6, 1979
.?
2.
J. A.
Dearien,
et.al., "FRAP-S3: AComputdrcodefor the Steady-State Analysis of 0xide Fuel Rods," TFBP-TR-164. Revision 2, March 1978 3.
D. D. Lanning, et.al., "GAPCON THERMAL-III Code Description,"
PNL-2434,JanuarP1978 4.
D. L. Hagnnan and G. A. Reymann, "MATPRO-VERSION 10, A Handbook of Material Properties for Use in the Analysis of Light Water Reactor Fuel Rod Behavior," TREE-NUREG-1180, February 1978 5.
J. D. Kerrigan, D. L. Hagrman, and S. O. Pe'ck, " FRAIL-4:
A Fuel Rod Failure Subcode," v0AP-TR-012, April 1978 6.
J. D. Kerrigan and D. R. Coltman, CVAP-TR-78-24, 19/8 This Research Information Letter transmits the description and assessment documentation of the latest version of the steady-r tate fuel rod behavior code - FRAPCON-1.
1.
INTRODUCTION y
FRAPCON-1 is a computer code that calculates the thermal and mechanical response characteristics of a nuclear fuel rod operating under steady-state conditions.
It was developed to provide both best-estimate (BE) and evaluatior model (EM) calculational ability to NRC for various uses within the Office of Research and the Office of Nt. clear Reactor Regula-tion.
The need for such a code was established by NRR/RES discussions in the summer of 1977.
These actions were initiated by a joint NRR/RES sQ 7g409791116 G107130409 CDR 4
(
Harold R. Denton Robert B. Minogue letter (Reference 1) and included responses and feedback from the partici-pating contractors - EG&G and BNWL. As a result of many discussions between RES and NRR technical staff, it was agreed that a new steady-state code would be developed by RES by merging the models of the RES-sponsored FRAP best-estimate code series (Reference 2) with the models of the GAPCON licensing audit code series (Reference 3).
It was also agreed that RES would be responsible for overall code development and for the BE models; whereas, NRR would be responsible for the development and review of the EM models. The new code was to be optimized for ease of use, running time, and corc-space usage, as well as be structured to allow for easy interchange of EM and BE models. The resulting code (FRAPCON-1) is:
- 1) used as a BE code to initialize the current RES Lest-estimate tran-sient code; 2) used as a stand-alone, best-estimate steady-state code; or
As of the time of this writing, the EM models to be used in FRAPCON-1 are being reviewed by cognizant licensing personnel.
Consequently, the contents of this letter will address only the code's capability with respect to BE models.
However, sine the models are interchangeable, the report will provide the user with t.n overall assessment of the code's abilities and enable NRR users to evaluate those models in the code which can or should be interchangeable with EM models.
It will also provide a base to which comparisons can be made between the results obtained in a BE mode and those obtained when using EM options.
2.
RESULTS AND EVALUATION Code Features and Models FRAPCON-1 is a FORTRAN IV computer code which considers the coupled effects of fuel and cladding deformation, temperature, and inte ' " qas pressure on the overall response characteristics of a fuel roo's.c
'ng under normal conditions. The cladding deformation model includes, Gci-axial, elasto-plastic analysis, and accounts for both primary and sec.ond-ary creep.
The fuel deformation includes the effects of thermal expan-sion, densification, and swelling on pellet dimensions.
-The fuel temperature model considers the effects of pellet cracking, relocation, and gas composition in the fuel-clad gap region.
Internal
~
, gas pressures are computed as a function of burnup (gas release) and average gap temperature. Material properties are supplied to the code via the MATPRO-10 subcode (Reference 4).
The code is also linked to the FRAIL subcode (Reference 5) which supplies rod failure probabilities at any point in time requested by the user.
l
-. -.-.- m
g
(
Harold R. Denton Robert B. Minogue As stated earlier, the base codes used in the development of FRAPC0ft-1 were the FRAP-S and GAPC0fl codes. The new code was modified to allow for dynamic dimensioning, and was completely modularized to facilitate easy interchange of BE and EM models. Most of the BE models in the code were taken from the FRAP-S code series; whereas, most of the EM models will be based upon GAPCON models.
Finally, the fuel pellet temperature calcula-
. tion subroutine was changed to utilize the more efficient Method of Weighted Residuals calculational scheme from GAPCON-THERMAL-III. The above changes are summarized in Table I of Appendix A, and an in-depth description of the code and its models is presented in Enclosure 2.
Assessment of Code Capabilities The independent assessment of the code was accomplished by a group other than the model developers (see Enclosure 3).
The objectives of this work were to demonstrate the best-estimate capabilities of the code and to provide nuidance to the model developers where improvements seem warranted.
During the assessment, FRAPCOM-'i results were compared with in-pile measurements and post-irradiation examination data for approximately 700 test rods. The results of these comparisons are summarized below for those response va.
Ns important to fuel rod behavior safety analysis.
Detailed results of these and other variables are available in Enclosure 3.
(A) Fuel Centerline Temperature Figures 1 and 2 of Appendix A illustrate the comparison of FRAPCON-1 predicted ard measured centerline temperatures for unpressurized and pressurized rods, respectively.
Error analysis of the plots show that the stardard deviation between measurements and predictions yield corresponding values of 170*K and 294*K, res acetively.
The smaller deviation of the unpressurized rods is reo) ably)due to the fact that a much larger data sample (61 rods vs 32 rods is avail-able for these rods which would tend to mitigate the effect of any systematic data errors present.
In both cases the code tended to slightly overpredict the data for low density fuel (<95 percent TD),
and slightly underpredict the data for high density fuel (>95 percent TD).
However, the general overall agreement is considered to be quite good since the deviations are very close to the 20-25 percent uncertainties predicted by response surface techniques which account for such input uncertainties as operating conditions, fabrication dimensions, and material property data.
(
(
i Harold R. Denton Robert B. Minogue l (B) Fission Gas Release The data comparison for steady-state gas release predictions over a ra.ge of burnups to 46000 MWD /t is shown in Figure 3 of Appendix A.
tbte that for releases less than 20 percent, the code generally uverpredicts the release, and that above 20 percent, the data become more evenly distributed.
An overall standard deviation of 16 percent was computed from the plot. This relatively large error is due to the large scatter in experimental measurements, the operating history uncertainty, and model deficiencies.
It is believed that the NRt'/ANL-developed gas release code (GRASS), when linked to the next code version (FRAPCON-2), will substantially reduce the contri-bution of the latter.
(C) Rod Internal Pressure The standard deviation between measured and predicted rod internal pressure, for both unpressurized and propressurized rods, was only 1.6MPa. When compared with the
'sults of a response surface study on FRAP-S3 (Reference 6), which compared pressure uncertainties caused by fabrication variables, operating variables: and material property input, the agreement is quite good.
The Reference 6 study yielded standard deviations of 0.9,1.59, and 2.34MPa for beginning-of-life, middle-of-life, and end-of-life conditions.
Figures 4 and 5 illustrate the data comparisons (both pressurized and unpressurized rods) of FRAPCON-1 for low burnup conditions and high burnup condi-tions, respectively. The group o.f underpredicted points in Figure 4, between 7 and 12MPa, correspond to startup measurements of two rods which exhibited significant pressure transducer drift and may, therefore, not be reliable.
(D) Fuel Axial Thermal Expansion Data comparisons for fuel axial thermal expansion were made for 20 rods (for both dished and flat pellet designs) under typical startup power ranges.
The results are summarized in Figure 6 of Appendix A and show excellent agreement for strains less than 0.3 percent.
Above 0.3 percent strain the code overpredicts the measured values because the fuel and Wdding are in solid contact which mitigates the fuel deformation due to cladding restraining forces.
The current fuel deformation models in the code do not account for this effect, but the next version of the code, FRAPCON-2, will - via both EG&G and BNWL optional fuel deformation models.
The standard deviation, calculated from Figure 6, was 0.23 percent of the stack length for strains less than 0.3 percent, and 0.56 percent for strains above 0.3 percent.
f
(
Harold R. Denton Robert B. Minogue (E) Permanent Fuel Axial Deformation The type of fuel deformation assessed here is that caused by permanent dimensional changes such as fuel densification, swelling, and compres-sion. The code currently accounts for only the first two mechanisms.
Assessment was based on data from about 200 rods at burnups less than '2900 mwd /t.
Therefore, the deformations reflect only that caused by densification. The results are given in Figure 7 which yielded a standard deviation of 0.45 percent of stack length.
(F) Cladding Deformation Cladding hoop strain (i.e., change in cladding diameter) was assessed from 130 rods and reflects the performance of the clad 61ng creep-down models in FRAPCON-1 since most of the rods experienced low burnup.
Figure 8 of Appendix A illustrates the results.
Note that although the agreement is quite good, (the standard deviation was 0.5 percent of clad diameter) the creep model tended to generally overpredict the amount of negative strain.
It is expected that revision of the creep model in the code, from the data obtained by the NRC-sponsored creep studies in the Petten reactor, will improve the predictions. The new model should be available within 1 year.
Data comparisons for cladding axial strain suffered from the same deficiences noted above for fuel axial strain predictions due to the lack of a fuel / clad axial interaction model. This deficiency should be alleviated in the next code version - FRAPCON-2.
3.
CONCLUSIONS AND USER RECOMMENDATIONS Standard deviations for the fuel behavior responses discussed above plus other less important responses are summarized in Table II of Appendix A.
Although some specific responses were slightly better predicted by FRAP-S3, the overall performance of FRAPCON-1 generally exhibits better calcu-lational accuracy than its predecessors.
In general, FRAPCONa predicts fuel behavior most accurately when:
1) the fuel rods are unpressurized, 2) fuel densities exceed 94 percent T.D., 3) initial gap sizes are less than 2 percent, and 4) when plenum volumes are more than half the total void volume.
Recommended options
~
for the user include using the Ross and Stoute annular gap conductance model coupled with the Coleman cracked fuel thermal conductivity model, recommended nominal and default values for all correlation multiplication factors defined in the users manual (Enclosure 2), and ncdalization and power profiles described in Section III.3 of Enclosure 3.
O
1 f
Harold R. Denton Robert B. Minogue~ The following caveats should be considered by the user when interpreting code results:
(A) Fuel temperatures may be overestimated at high burnup if the initial gap is large (>2 percent), and if the fuel density is low (<94 percentTD).
(B) Although the cladding hoop stress may be overestimated under hard gap closure conditions, positive cladding strains may be underestimated under soft gap closure conditions.
(C) At power levels up to 20kw/m (-6.1 kw/ft) internal pressure predictions are very good, however, for power levels above 35kw/m (*10.7 kw/ft),
the pressure nay be overpredicted for low burnup and underpredicted for high burnup conditions.
~
As noted above, model improvements are continually being*made to enhance data-prediction comparisons, and these improvements will be incorporated into future versions or modifications of the code.
1 W Saul Levine,
)irector Office of Nuclear Regulatory Research
Enclosures:
1.
Appendix A (Includes two tables and eight figures) 2.
"FRAPCON-1: A Computer Code for the Steady-State Analysis of 0xide Fuel Rods," CDAP-TR-78-032-R1, November 1978 3.
" Independent Assessment'of the Steady-State fuel Rod Analysis' Code FRAPCON-1," CAAP-TR-050, May 1979
.. _..., ~ _,, _. _.
APPENDIX A TABLE I DIFFERENCES BETWEEN FRAPCON AND FRAP-S3 I
~
FRAP-S3 FRAPCON Heat Conduction Stacked 1-D radial Stacked 1-D radial using /kdT, effective using Method of Weighted -
fuel thermal conductivity Residuals, effective fuel thermal conductivity Fuel, cladding MATPRO-9 MATPRO-10A and gas properties FRAP-T Links
-FRAP-S3/T2,T3,T4 FRAPCON/T5 Programming features-Modular coding Modular coding, Dynamically dimensioned with respect to nodelization and power-time steps, 4
modified code iteration structure to optimize efficiency
/"N I
G 4
~
c
(
I 3
5 3
3000 g
y a
5 aa00 I_
4#a u
Run 0
$e</
s m36 e
a 43-15
= 53.54 C
S g 1500 o g.60.632.613
+ 65-69 7 219-221 5
4 319-320 g
,o(
4 "=
es 321 324 6 370-37I 7,,,,
- -426.428.422
. See
.2 aw b 588.589 4
O 9
6 11.
e 592-594 8in
~
$00 614A.6118.615R
=
,r' Predicted =Weasured i
i i
0 0
$00 3000 IS00 2000 3500 3000 seco Measured fuel centerline temperature (K)
Fig 1 FRAPC0ti-1 predicted against measured centerline temperatures for unpressurized rods.
53
(
(
seco aooo o
o 5 aoco u
g.
E
~
a st..s*
~
.aco
' W.r.
- 4 c
3 maa aun G
4 2
g m;;pg.ve.m.c4.4v.c.
g.
v6;2g g a s.454.455.
a 7
,,,o 8 415.416
.3
> 441-441.553 ts 9 450-453 6
& 596.597 5
9 soo
.f-- Predicted = Measured a
o soo scoo isoo sono moo sooo seco Measured fuel centerline temperature (K) l Fig. 2 FRAPCON-l predicted against measured centerline temperatures for pressurized rods 54
c
(.
t00
.-r,,,
ao
.6y a
C e
n' a
1.apr 4.
- u
~
A
- n.
- y m
s o
T n to :
.m e l
E 8
E.
o g
.+g*
- o.~.
2 l
+
2
=",
E Run g n.7 76.n.e0.es.es.e7 e
1 99.100.103.10s-i07
.e s.
'is:ia:istis:i!!:ix.im-i46i g
0 213.214.217 20s-207.211 w
? 273-239 T 2st-2s3
}
n 286-297.298
=
+ 299.302-343 T
E 4 386-318 e 319-323.326.327
'O e 131-334 O
6 g 336.339
, 434 1 01 7
a 196-199.s02-s07,s09-sl3.si9 e s43-s4s
> sss ass 7.ss9.s6i-s63.s66-s6e.s70.s7ej a see.sa9 4 62s.627 e 633.632 a 6T3.636.639
Predicted = Measured
- 648-644 7 61s.616 0 01
' '.1 oD1 a
I to 100 Measured fission gas release (T.)
Fig. 3 FRAPCON-1 predicted against measured fission gas release f raction.
l l
69
(
(
i E
i i
i e
i i
i Run E
O 33.34.50-52
- 273.274,278.133,437 139,446 118 e 279.290,463A S 314,315 O 367-389 IS S 117-123 m
e 5
14 4*
2
.* 4 '*
g ;* *.s' e
lla r..-
a n.
e 5 io
?
c
. t &e
'8 s R,,.. *.
~
s
- ~
e..-
7 e,-
% e
+--- Predicted = Measured 4
2
.5 0
2 4
6 8
10 la 14 IS IS E
Measured rod internal pressure (MPa)
Fig. 4 FRAP CON-1 predicted against measured rod internal presspre 77
-.-w
/
(
N.
10
~
g g
e n-2 w
7
~
=
o b
w 6 C.
o'
/
/
,h Cu s o
Run
,C e'
O 33.31.50-52 e
+ 70 4
~
e 319*.
,f
$ 321-324 6
o 331-334
'O e
a 336-333,386 o
e <si:423
% 3
--- Predici ad= Measured T
u n.
2
," ' a sa a
,. o*
+
+
+
O I
2 3
4 5
6 7
8 9
10 Measured rod internal pressure (MPa) 1
.. Fig. 5.
FRAPCON-l predicted against measured rod internal pressure at high burnup conditions.
81
^*
2A 14 22 2
e M
w C
13 2
a C
Run iA Q
$ST 4
C 3l,32 M
a 40-12
- 58.60.81 L4 O 318
+ 367 g 9 4
0 o TSI
'M"
?
12 7
v 152,3s3 83 a
e' e 386 x
A
= 105.106 e 701,702 o
8 g
g 1
=
a a4 e
s e
a T
Qa Predicted = Measured -
.a
.9 g *9 2
Os y..-
w Q.
o
.a Q4 Y
x 7
7 NA Y
G2 o
0
,a2
-02 0
Q2 Q4 OA QA I
L2 L4 la la 2
22 14 2A Measured fuel axial expansion (7.)
l 1
l Fig. 6 FRAPC0t!-l predicted against measured fuel axial expansion I
during heatup.
1 88
(
l t
as n
tt w
o 3
0 Y
A e -05 O
U O
. gg
+
g T
0,.* 8 0 Run
, 63 O
Og o 359 362
.O a 363-E
-14 0
O 391-400 C
0 0
= 405.406 0,'.'
e 5% 5%.5M c
0 e 600-602 ec 9 603-605 w
-2 b
p *0 e 606-610 e 340-345 g,
.B 19s.499.501-503.507.509-Si!.513
$16.517.519 x 528-542
]
-U e 543-548 o 621-626.628 o
7 647-656
& 88.89 Predicted =Wensured
~
C-
-3 i
_33
-35
-3
-2.5
-2
-15
-l
-05 0
05 1
Measured permanent fuel axial deformation (P.)
Fig. 7 FRAPCON-1 predicted against measured permanent fuel axial deformation.
92
4 2
i e
i Run a sg.:o0,103,io5-tor a 106 800.182-114.118.120 122
,124.125.127-132.134.835.137 143 146.148-152 7 211.217 IO m 273-239 e
Lt o 253 253 w
E 270.290
- 316.317.348 e'
E
- 418 414 e'
g
, e 460-166.468-478.490-409 L.
8494.4 % -499.502-507 g
509-518.513.515-517.519 w
e 543-546 g
a 5%-559.562.563.566.567 %9, 570.571.574
,8
- 0. ee..649. 65 8 -6%
647
- e. 525-542 g3 554.552.555 h6 0
v
--o
.* - y g' C
- c al v,,
e _o3 le%=en E
[
flab 5m
.ca e
T
.g z
,6 s
2
.- ^de
~
d-Predicted = Measure 4l
-tm 444 '
e e
i i
e i
,g
-2
-13
-l
-0S 0
05 1
13 2
Measured permanent cladding hoop strain (".)
Fig. 8 FRAPCON-1 predicted against measured permanent cladding hoop strain.
96
TABLE II FRAPCON-1 MODEL ASSESSMENT -
SUMMARY
OF STANDARD DEVIATIONS BETWEEN MEASUREMENTS AND PREDICTIONS Sampie Size Standard Deviation Output Parameter
(# of Rods /# of Points)
FRAPCON-1 Fuel Centerline Temperature 32/274 (Pressurized Rods) 294K 61/472 (Unpressurized Rods) 170K Released Fission G as 145/145 15.9 %
Rod Internal Pressure 20/330 (Unpressurized Rods) 1.38 MPa t
28/285 (Pressurized Rods) 1.93 MPa Gap Closure Heat Rating 88/88 11.4. KW/M Axial Fuel Thermai Expansion 18/160 0.37 %
Permanent Fuel Axial Deformation 97/354 0.45 %
Permanent Cladding Hoop Strain 154/358 0.47 %
Permanent Cladding Axial i
Strain 96/119 0.15 %
Cladding Surface Corrosion Layer 40/69 5.8 micron Cladding Hydrogen t
Concentration 33/46 37.2 ppm 2
Gap Conductance 17/112 (Unpressurized Rods) 10821 W/m K 2
20/115 (Pressurized Rods) 21200 W/m K Fuel Off-Centerline Temperature
'20/111 208K 4
(
l Harold R. Denton Robert B. Minogue 1
\\
e:
The following caveats should be consy' red by the user en interpreting code results:
(A) Fuel temperatures may be overestimated at high burnup if the initial gapislarge(>2percen, and if the fuel nsity is low (<94 percentTD).
(B) The cladding hoop stress may be overes$1 mated under hard gap closure conditions, whereas, claddi g strains y be underestimated under soft gap closure conditions.
(C) At power levels up to 20kw/m 56,1 kw/ft) internal pressure predictions are very good; however, for power levels above 35kw/m ( 10.7 kw/ft) the pressure may be overpredi tsd for low burnup and underpredicted for high burnup conditions.
As noted above, model improvements are continually being made to enhance data-prediction comparisons, and these i provements will be incorporated into future versions or modifications of the code.
/
Saul Lev;'ne, Director Office of Nuclear Regulatory Research
Enclosures:
1.
Appendix A (I:.cludes two tables and eight figures) 2.
"FRAPC0ff-1: A Computer Code for the Steady-State Analysis of 0xide Fuel Rods," CDAp-TR-78-052-R1, November 1978 3.
" Independent Assessment of the Steady-State Fuel Rod Analysis Code FRAPC0ff-1," CAAP-TR-050, May 1979 DISTRIBUTION SUBJ RB CY CIRC JL CY CliR0ti SL CY BRANCli RF GM RF Gtt CY Dli CY RSR:W:Fb CEJ CY GPMarino/pr LST CY 10/4779 TEM CY
,,R 1 FB,
JLarkins SLevine
, D!g
...$ o n..._..,.,Jo hnson/ Ton.g _.TEHur,l ey RBadnitz sunuauc -
_1.9.8 79...........! LL/ 79.....
10/..!79 10/_l79........
10/.... / 79
.10/.../79..
aan -
(
g Harold R. Denton
~ Robert B. Minogue s The following caveats should be considered by the user when interpreting code results:
(A) Fuel temperatures may be overestimated at high burnup if the initial gap is large (>2 percent), and if the fuel density is low (<94 percent TD).
(B) Although the cladding hoop stress may be everestimated under hard gap closure conditions, positive cladding strains may be underestimated under soft gap closure conditions.
(C) At power levels up to 20kw/m (-6.1 kw/ft) internal pressure predictions are very good; however, for power levels above 35kw/m ("10.7 kw/ft),
the pressure may be overpredicted for low-burnup and underpredicted for high burnup conditions.
As noted above, model improvements are continually being made to enhance data-prediction comparisons, and these improvements will be incorporated into future versions or modifications of the code.
Orif: 01 Of-.:d 0-1 Sault.i;i.m,
Saul Levine, Director Office of Nuclear Regulatory Research
Enclosures:
1.
Appendix A (Includes two tables and DISTRIBU1 ION eight figures)
SUBJ 2.
"FRAPCON-1: A Computer Code for the CIRC CHRON Steady-State Analysis of 0xide Fuel Rods," CDAP-TR-78-032-R1, November BRANCH RF 1978 GM RF GM CY 3.
" Independent Assessment of the DH CY Steady-State Fuel Rod Analysis Code C
FRAPCON-1," CAAP-TR-050, May 1979 TEM CY RB CY JL CY SL CY RSR:W:F i
GPliarino/pr i
10////79
- See previous yellow for concurrence g
T
- omu, g
Sg RES R
RSR:W:FB* -
RR DAHoatson Johnson / Tong TEikrley k
Steige w o,,,
.19L26/79
_10p//79 j/J/79 _,, _l,
_1 /C/79 _,,1l/MH9, _
79 i
om -
-.