ML17261A010
| ML17261A010 | |
| Person / Time | |
|---|---|
| Issue date: | 06/07/1978 |
| From: | Levine S Office of Nuclear Regulatory Research |
| To: | Case E, Minogue R Office of Nuclear Reactor Regulation, NRC/OSD |
| References | |
| RIL-0029 | |
| Download: ML17261A010 (16) | |
Text
-
UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20556 JUN 71978 MEMORANDUM POR:
E. G. Case, Acting Director Office of Nuclear Reactor Regulation R. B. Minogue, Director Office of Standards Development FROM:-
Saul Levine, Director Office of Nuclear Regulatory Research
SUBJECT:
/
RESEARCH INFORMATION LEITER - #29 FUEL ROD ANALYSIS COMPUTER CODE:
FRAP-T3 Introduction and Summary This memorandum transmits the results of completed research to prepare and test the third modification of the computer code FRAP~T (Fuel Rod*
Analysis Program - Transient).
FRAP*T is a FORTRAN IV computer code*
b~ing developed to predict the transient response of a LWR fuel rod*
_during postulated Jccidents such as Loss-of~Coolant_Accidents (LOCA),
Power Cooling Mismatch Accidents (PCM), Reactivity Initiated Accidents_
(RIA), or Inlet Flow Blockage Accidents (IFB).
FRAP-Tis also being developed to perform the calculations needed for planning and analyzing Power Burst Facility (PBF) and Loss of Fluid Test (LOFT) experiments.
Although the code calculations are made on a best estimate (BE) basis, substitution of alternate models and correlations could be easily made to make evaluation model (EM) calculations.
FRAP-T3 is the third annual
_update of the code and as such provides a relatively mature analytical capability.
Improvements upon FRAP-T2 are primarily in the area of cladding behavior. Aspects of various versions of the code are shown in Table I.
The importance of improving our fuel behavior codes.is recognized in a series of user requests:
REG*:RSR-88, "Fuel Pin Analysis Development, 11 dated March 14, 1973; REG:RSR-118, "Regulatory Need for Additional Safety Research on Reactivity Initiated Accidents-," dated November 21, 1973; Section 6.8 of the "Regulatory Assessment of the AEC Water Reactor Safety Research Program, 11 dated August 12, 1974; "Review of Fuel Behavior Project Description," dated May 6, 1975; 11NRC/NRR Technical Safety Activities Report, 11 dated September 11, 1975.
E. R. B. Minogue These user requests are for analytical models, tested against data, which will predict fuel failure and failure propagation thresholds in power reactors. A calculational tool is also needed to interpret PBF, LOFT and Halden experiments, to provide audit capability for vendor codes such as STRIKIN-II, FACTRAN, LOCTA-IV and THETA, and to support specific SD and NRR activities. This memorandum and its enclosures describe the FRAP-T3 code, its testing and our evaluation of its applicability and capability.
Results - Code Features In FRAP-T3 the coupled effects of mechanical, thermal, internal gas and material property response on the behavior of the fuel rod are considered.
Given appropriate coolant condition and power histories, FRAP-T3 can calculate rod behavior for a wide variety of off-normal situations and postulated accident conditions (e.g., BWR or PWR power transients, flow coastdown, load loss or coolant depressurization).
Further details of code features (e.g., models, input requirements, output parameters) are given in Appendix B.
Results - Code Qualification An essential part of producing an operational computer code, which can be used with a known degree of confidence, is the independent testing process (described on pages 257-267 of Appendix C).
The results of such testing of FRAP-T3 are as follows.
Figure 1 compares measured and pre-dicted centerline fuel temperatures for unpressurized rods.
The good agreement, generally within 10%, suggests that heat transfer is well represented by the MacDonald-Broughton (
11cracked pellet") gap conductance option which was used for these calculations.
Figure 2 indicates a similar comparison for rods prepressurized with helium, showing less satisfactory agreement.
However, a second FRAP-T3 gap conductance model is available, following Ross-Stoute, and this option provides good thermal predictions for prepressurized rods. 1. Figure 3 shows predictions of plenum gas pressure.
Most of the high pressure results fall within 10%
of the measured values. Accurate prediction of this pressure is important to the ballooning behavior of fuel rods in a hypothetical LOCA.
Figure 4 compares single rod PBF (annulus geometry) test data with calculations using two of the Critical Heat Flux (CHF) options which are available to FRAP-T3.
Lack of a better fit may be accounted for by peculiarities of the PBF test train configuration (e.g, standoff screws and flow area).
Figure 5 shows fuel temperature response following scram.
An adequate 1TFBP-TR-186, "Gap Conductance Test Series, Test GCl-3, Test Results Report and Summary of Piggyback Tests, 11 March 1977.
\\...
E. R. B. Minogue calculation of the dissipation of stored energy and decay hea.t immediately after.scram ~s especially important for analyzing actident sit~ations.
- Finally, Figure 6 compares results of a standard problem run with FRAP-T arid with the German code SSYST.
An agreement between these two independent codes implies some validity of the code predictions.
Evaluation In the context of LWR system transients, FRAP is well suited to be used as a component code to descri~e fine details of fuel rod behavior.
Furthermore, sensitivity studies with FRAP will facilitate definition of the simplest acceptable fu~l description in systems codes.
Substantial effort has gone into FRAP-TJ to make it a mechanistic and sophisticated code.
The independent testing process has shown that several new models and subcodes, some of which are unique to FRAP-T, are important to making realistic calculations. These include the material properties package2,
the failure subcode3, three dimensional cladding ballooning, a complete heat transfer correlation package, a transient plenum temperature model and an axial gas flow model.
The material properties pack~ge and the failure subcode have been well received by the Fuel Code Review Group.
Quantitative characterization of the uncertainty associated with parameters predicted by FRAP-T3 (e.g., plenum pressure, fuel centerline temperature, cladding ballooning or burnout power) has been made, and representative sa~ples are shown in the figures.
Developments are continuously underway to remove some of the present limitations of applicability of FRAP-T3 and these developments will be incorporated in future versions of FRAP as new research data and modeling permits.
FRAP-T3 has been transmitted to the Argonne Code Center and is programmed and running on the CDC 7600 computers at INEL (Idaho), Berkeley (California) and Brookhaven (New York).
We would be happy to assist yo~r staff in running any of the FRAP standard problems listed in Table II in order to directly demonstrate the code's capability.
2TREE-NUREG-1005, 11A Handbook of Materials Properties for Use in the Analysis of Light Water Reactor Fuel Rod Behavior, 11 December 1976.
3TFBP-TR-1B9, 11 FRAIL 3: A Fuel Rod Failure Subcode, 11 April 1977.
E. R. B. Minogue Appendices Appendix A contains the six figures and two tables referred to in the text. Appendix B provides a succinct description of code features and some co1TTT1ents concerning use of the code.
Appendix C, report TFBP-TR-
- 194, 11 FRAP-T3-A Computer Code for the Transient Analysis of Oxide Fuel Rods, 11 provides detailed descriptions of the code afld its testing.
~n~or Office of Nuclear Regulatory Research
Enclosures:
As stated
APPENDIX A
C3 0
f E..
~
- c.
E Cll I-Cll c Ill c:
Cll L)
~
- u.
"O Cll u j
~
C"l I-
~
c: a:
~
e Symbol Run Reference e
2 1 HPR-80 2000 i
3
- 377 IFA-226
- 424 1
D 425 IFA-431
- 426 0
42".'
8 431 PBF 1500 1000 500 500 1000 1500 2000 Measured Fuel Centerline Temperature (° C)
Predicted versus measured fuel centerline temperature for unpressurized rods.
FIGURE 1
'\\
~*
~
u 0
~
Q)
- i Cl)
Q) c.
E Q) r-Q) c -..._
Q) c Q) u C1l
- i
~
"'O Ill u
"'O Q) 0...
("')
r-0...
q:
a:
~
2500 Symbol Run Reference
- c.
273} PBF 274
- n5 PBF 2000
- :?77) 0 278 PBF 280 PBF it.
415 ) IFA-429,
[..,
416 430 PBF
- 433) PBF c
434
- 438) PBF 1500
~
439 c
441 HBWR 442}
- 443 HBWR e
444 445 I) 1000 1
Q 500 Measured Fuel Centerline Temperature (°C)
Predicted versus measured fuel centerline temperature for pressurized rods.
FIGURE 2
~
4
- c.
1800 Symbol Run Reference 1600 0
"'l 274 PBF 278 280 1 PBF 0
1400 D
279 6
418
'419 v
420
"' "' 1200 4
421 IFA-429 8:
~
<I 422
- i
'i' 423 Cll 417 c: 1000 iii c:
~
i:
'O 800 0 a:
'O Cll u
-0 3721 IFA-142 Cll 600 c:
373
~
387 \\
t-IFA-225
'Cl.
e 389
<(
a:
400 433 PBF 437 '
438 PBF e
439 200
- Measured Rod Internal Pressure (ps1a)
Predicted versus measured rod internal pressure during heatup.
- FIGURE 3
1 1 1 0 09
~
b
~
~ 0 8 c
0
- u. :t
(.)
a; 0 7
)(
.2
- u.
Run ID-6014. 6015. 6016 Reference-P6F FRAP-T3 Prediction1 JCHF Corre1a11on 0
B&W*2 3
W-3 f,.,~*-----J7 6 4 Ill>,,.. Measured Values Syslem Pressure 2160*2230 psia Inlet Enthalpy, 644 6tullb
- Heated Dia Hydraulic Dia.* 2 5-3 O Inlet Quality. O
~ 06 05 No CHF CHF 04 03L----..l-~---...J---'c::m~.;...&..~.L.-~---_...l.---__J~---'---~...J 0 35 0.40 0 45 0 05 0 10 0 15 0 20 0 25 0 30 Thermal Load on Subchannel (kW/cm')
PBF Inlet mass flux at CHF onset vs rod* thennal output for low flow area t~st~.
FIGURE 4
0 0
~
- J -"'...
Q) a.
E Q)
I-Q) c:
~
c:
Q) 0 a;
- J u..
15001 1000 500 Predicted power adjusted to match
-~-~
initial measured temperature
- \\*.
\\*..
\\*.,
~*.,>.,...,*..
~*.
~....
~ - - -
FRAP-T3 Prediction
- Halden data Diam. Gap 2.5-2. 7%
Time after Scram Initiation (sec)
FRAP-T3 fuel temperature response after scram at 5600 MWd/t.
FIGURE 5
~ 1000 w
a:
. a:
w
- c.
- E tiJ 800 I-600 CJ z c
- c
<(
..J 0
400 6,, Experimental Data 30 20 1i)
---r-~** I
/
i
- . * *T"l"'T'........ -:: 9-....... : --r-* '"..
l 9
FRAP-T
-"' !a I
~
I 15 10 5
s.__ __...1-__ --1.~__,_;...__--=-*.::c:~--&.~---.l--~_._~=~~--~o 0
10 20 30 40 50 60 70 80 90 TIME (s)
FRAP-T3 and SSYST Predictions for Standard Problem PNS-4238 LOCA Heatup Test FIGURE
- 6 l.&J
~*
z -<
a:
I-Ul
Phenomenon Mode.led
- . Heat conduction Gap conductance Plenum gas temperature Metal-water reaction Internal pressure Cladding.defonnation Decay heat TABLE I CAPABILITIES OF VARI.GUS VERSIONS.OF FRAP-T FRAP-Tl Stacked 1-D radial Modified Ross and Stoute Coolant temperature
+ l0°F Baker-Just Compressible, laminar gas flow, constant Hagen number Uncoupled stress-strain
- equations, no fuel-cladding inter-action, no ballooning model, no creep
. No model FRAP-T2 Stacked 1-D radial Modified Ross and Stoute, Cracked pellet Six-node transient energy balance, boun-dary conditions from surface temperature subcode Baker-Just Compressible, laminar gas flow, constant Hagen number Triaxial coupled plastic stress-strain equations, fuel-cladding inter-action, intennediate balloon model, no creep No model FRAP-T3 Stacked 1-b radiaJ, 2-D. r-e Modified Ross and Stoute, Cracked pellet Six-node transient energy balance, simplified boundary
-conditions Cathcart Ideal gas law, compressible, laminar gas flow, variable Hagen number open porosity considered Triaxial coupled plastic stress-strain equations, fuel-cladding inter-action, advanced balloon model, strain-rate effects, cold-work and fast neutron.
flux effects, computation optimization, no creep; ANS model 5. 1 ( 1971 )
Phenomenon Cladding failure Fuel deformation High flow film boiling heat transfer correlations Low fl ow film boiling heat transfer correlations Critical heat flux correlations Slip ratio correlation Water properties Fuel, cladding and gas.properties TABLE I (continued)
FRAP-Tl Failure if instability strain exceeded GAPC0N-l Model Groeneveld Berenson B&W-2 Barnett Modified Barnett Homogeneous RELAPJ tables MATPR0-2 FRAP-T2 Failure if total circumferential strain exceeded GAPC0N-I Model, free thenna l expansion model Groeneveld Dougall-Rohsenow Tong-Young Condie-Bengston Groeneveld B&W-2 W-3 Barnett Modified Barnett General Electric Modified Bankoff-Jones Wagner steam tables MATPR0-6 FRAP-T3 Failure probability computed, overstress, overstrain, eutec~ic melting, and oxidation failure types modeled GAPCON-1 Model, free thermal expansion model Groeneveld Dougall-Rohsenow Tong-Young Condie-Bengston Modified Bromley (a<0.6) free convection (a~0.6)
B&W-2 W-3 Barnett Modified Barnett General Electric Marchaterre-Hoglund Wagner steam tables MATPR0-8
TABLE II
~
FRAP-T STAN DA-RD PROBLEMS.
' \\
TYPE DESCRIPTION DATE LOCA PWR Cold Leg Break Using Supplied Heat Transfer Coefficients or RE LAP Coolant Conditions LOCA TREAT Test 2, BWR Rods 1971 Slow Power Hamp Halden Reactor Project, Norway 1967 Power-Cooling-Mismatch PBF Test PCM 8-1 1976
-Reactivity I nitiatep
-Accident (RIA)
BWR Hot Standby Conditions, 250 Cal/G RIA SPERT Test GEX-692 1969 ATWS*
BWR Main Stearn Isolation Closure Valve Accident (90% Relief)
I
APPENDIX B DESCRIPTION OF CODE FEATURES The phenomena modeled by the code include:
(1) h'eat conduction; (2) elastic-plastic cladding deformation; (3) fuel-cladding mechanital inter-action; (4) transient fuel rod gas pressure; (5} heat transfer between' fuel and cladding; (6) Cladding-water chemic~l reaction; and (7) heat transfer from claddirig to coo) ant. - Consideration of the mechanical deformation of the fuel and cladding is of particular significance, since a realistic prediction of rod geometry during an accident (e.g., LOCA) is desired.
The probability of pellet cladding interaction related failures is calculated, even though the models n~eded for a true description of.
local effects are missing.
Effects of prior irradiation must be input from another stiufce (e.g., FRAP-S).
FRAP-T3 is linked to a modular material properties package, MATPR0-8, which contains correlations for all fuel, cladding, and gap gas properties needed by the code.
Each correlatio~ is contained in a separate function subprogram or subroutine.
No material properties need be specified by the code user.
FRAP-T3 is also linked to the Wagner water properties package, which was developed for.the RELAP-4 code.
This package defines
- subcooledr saturated, and superheated water properties.
FRAP-T3 requires input data (in either metric or engineering uriits) which specify cold state fuel rod geometry, transient power, transient condition of coolant surrounding fuel rod, and amount (or pressure) and type of gas in the fuel rod.
Input data are also required to specify mesh size (radial and axial incremental dimensions used in computation), time step and accuracy.
This permits the code user to have some control over the computer CPU time needed to execute a problem.
Transient coolant condi-tions can be specified in several ways.
These options have been chosen to provide maximum flexibility.
For example, card input of coolant condi~
tions or heat transfer coefficients, or magnetic tape input of coolant conditions calculated by RELAP-4 can be used.
Code printout, which occurs at input specified time intervals, includes:'.
fuel rod radial temperature distribution at an arbitrary number of axial positions, fuel diameter, gas gap thickness, gap conductance, claddinci dia~eter, axial length change, internal pressure, power, surface heat flux, and cladding hoop strain.
The code can be instructed to generate plots of the above output parameters as a function of time.
It is also possible to generate 16mm motion pictures of t~e output.
Base~ on ou~ review of FRAP-T3, we believe the following observations would be helpful to code users; (1) At hot plenum pressures above 500 psi the MacDonald-Broughton gap conductance model (so called cracked pellet model) predicts excessive values and the Ross-Stoute model option is recommended.
(2) Two model options are available for computing fuel radial displacement (free thermal expansion model or GAPCON-THERMAL-1 model).
Since FRAP-T3 was verified (and to some extent developed) using free thermal expansion, that model option is recommended.
(3) The stress-strain model in MATPRO is not applicable above 1500 F (temperature at which a metallurgical phase transformation begins in Zircaloy).
This generally causes an overprediction of cladding circumferential strain at burst.
Measured strains of 0.1 to 0.7 in/in are predicted to be 0.6 to 0.9 in/in.