ML19289E373
| ML19289E373 | |
| Person / Time | |
|---|---|
| Site: | Fort Saint Vrain  |
| Issue date: | 04/09/1979 |
| From: | Wessman G GENERAL ATOMICS (FORMERLY GA TECHNOLOGIES, INC./GENER |
| To: | Gammill W Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 7904170253 | |
| Download: ML19289E373 (59) | |
Text
N*
A GENERAL ATOMIC COMPANY S
E, CALIFOANIA 92138 (714) 455-3000 April 9, 1979 Mr. William Gammill Assistant Director for Advanced Reactors Division of Project Management U. S. Nuclear Regulatory Commission Washington, D.C.
20555
Dear Mr. Ganinill:
Enclosed are fifty (50) copies of General Atomic Company's response to Lead Items I-IV, VI and VII contained in the February 16, 1979 letter,
" Review of Thorium-0xide and Uranium Carbide Fuels". The lead items for which responses are enclosed pertain to the review of thorium oxide.
The response to Lead Item VIII, which pertains to the review of uranium carbide, is being transmitted under separate cover to enable GAC and NRC to maintain separate documentation of these reviews. We note that the February 16 letter contained no Lead Item V.
Very recent developments at General Atomic Company have led to a decision by Fort St. Vrain project management to defer initial use of thorium oxide from Reload Segment 9 to a later segment.
GAC wishes to pursue the review of thorium oxide to completion; however, we are with-drawing our request that the thorium oxide review be completed by your previously committed date of June 29, 1979.
Plans to use uranium carbide fissile fuel in FSV Segment 9 are still in effect, and we look forward to completion of the UC review by June 29.
2 Please let us know if you have any questions regarding these responses or our schedule requirements.
Sincerely, NW G. L. Wessman, Director Plant Licensing Division GLW:mk Encl.
7904170#53
e t
STATE OF CALIFORNIA ss.
COUNTY OF SAN DIEGO )
After being duly sworn, the person known to me to be G. L. Wessman of General Atomic Company, signed the within document this 9A day of April 1979.
WITNESS my hand and official seal.
,, s,
G m :~ :-
~
l~ ~ $
dd.$ N C.4 L. C A vV ? C
- 1 NCTAR, Pu8cc CV l r., a <',
PR;%.
^rFr- 'w
,- % :t SAN DsiaJ.,. "-
y My Commission Escares May 29, lia2
- e..vwwvwww.w-aw.*.wwv.i Notary Public S
RESPONSE TO NRC LEAD ITEM I FOR Th09 REVIEW I.
Th0p Fuel Performance Data Base The discussion of (a) irradiation test results and (b) fuel performance during a potential core heatup event (sections 4.2.3 and 4.4, respectively, of GLP-5640) is, in each case, rather brief.
Perhaps the several references listed in GLP-5640 contain all that is needed as supplementary information.
On the other hand, it might be possible to expedite the review if separate and more detailed discussion and analysis were provided of the overall data base for Th02 TRISO particle performance under normal and off-normal conditions.
Thus, it would be helpful to have a summary presentation of the details of the irradiation tests and performance analyses, which would in::ade the following:
~.
Fuel test particle material parameters, such as kernel diameter, coating thicknesses, densities, anisotropy factors, etc.; differences between test material parameters and potential reload material para-meters should be highlighted and discussed in terms of possible effects on performance.
2.
Fabrication parameters should be discussed in terms of the precursors, methods, etc. that were used to produce th'e test particles as compared with the future reload particles and the effects of these fabrication parameters on fuel particle component structure and performance.
3.
Test conditions, including numbers of particles tested versus number failed, method of failure detection, statistical significance of the test results, etc.
4.
The analysis of the test results should include a discussion of the significance of the apparent lack of data on high exposure Th02 TRISO fuel particles (taking into account that the fraction of fissions 1.
the fertile particles increases with time as more Th-232 is converted to U-233).
5.
An attempt should be made to indicate the expected failure fraction for each potential failure mechanism, as was attempted for UC2 TRISO and Th02 BISC particles in GA report GA-12971.
The supporting data should be correlated with the predictions.
6.
The relationship of the sic tensile stress model to the irradiation test data base should be discussed.
For example, for a given test sample, what was the calculated distribution of sic tensile stresses, how many specimens (particles) failed in the sample, and what were the calculated tensile stresses in the failed particles? As a tie-in to the questions on fabrication parameters and particle coating structure and properties, what effect does sic microstructure (grain size) and density have on strength, how are these parameters affected by fabrication, and what quality assurance procedures are used to ensure that the desired sic structure and strength are obtained?
RESPONSE to Lead Item I-1 Ref. 1 is a detailed review of the irradiation experience for coated Th0 f"'I 2
particles, which is summarized in Figure 1.
Figure 1 is a cumulative distribution which defines the irradiation test experience in terms of the burnup (% FIMA) exposure for coated Th0 fuel.
The cumulative burnup distribution applies to a 2
7 6
total of -10 BISO coated and -10 TRISO coated particles tested in the temperature 25 2 (E>29fj) HTGR.
range 600 - 1500.C and the fluence range 0.4 - 15.3 x 10 n/m The broken and solid lines in Figure 1 illustrate the BISO and TRISO coated particle data base respectively. The BISO and TRISO coated Th0 data bases are presented 2
togetter to provide a large body of collective empirical evidence supporting the chemi al stability of Th0 kernels during irradiation.
2 Reference 1 documents the fact that Th0 shows no evidence of kernei-coating 2
interactions and is chemically stable under the entire range of normal operating conditions.
In addition, a large body of generic TRISO coated particle design information on other fuel systems, e.g., TRISO coated (Th/U)C and ThC, can be 2
2 applied to gain added confidenca in the TRISO coated Th0 fuel design. This 2
.information, combined with irradiction test experience on reference type TRISO cocted Th0 fuel, pr vides high confidence in the expected performance.
Referring 2
4 to Figure 1, -2.5x10 TRISO coated Th02 particles have been irradiated to burnups 3
greater than 4.5% FIMA and '6.5x10 particles to burnups greater than 7.0% FIMA.
These burnup limits represent the mean and peak values for 6 year old fuel in FSV.
Consequently, confirmation of TRISO coated fuel performance for these relatively high burnup exposures represents a very conservative appraisal of the expected core average fuel performance.
Irradiation capsule tests GF-4, HT-31, HT-33, and HT-34 represent the majority of the high burnup (>4.5% FIMA) TRISO coated Th0 test data.
The results from these capsule tests are discussed in more 2
detail in response to Lead Item I-3.
The range of TRISO coated Th0 kernel and coating properties tested in the 2
irradiation capsule program (Figure 1) is shown in Tables 1 and 2.
The first row of Tables 1 and 2 presents the nominal properties for the reference TRIS 0 Th0 design.
The second row shows the range of irradiation test experience for 2
each of the properties. The last row presents the nominal property values for the core heatup simulation test (CHST) samples used in the safety analysis test program, which is described in the response to Lead Item II.
Two important points are evident from Tables 1 and 2:
1.
The range of irradiation test experience brackets the nominal design values for all critical properties.
2.
CliST sample properties are representative of the nominal reference design properties.
Consequently, the irradiation capsule test and'CHST results can be used as a representative basis for judging TRISO coated Th02 fuel perf~ormance under normal and accident related conditions.
RESPONSE to Lead Item I-2 The effects of fabrication parameters on TRISO coated Th02 performance are con-trolled by specifying property and process requirements. Table 3 lists fuel particle components and the properties specified to assure accept-able performance.
In addition, a brief technical justification is provided in the last column concerning the :pecification requirements for each property.
These property requirements are generic to any TRISO coated particle design and are independent of whether the kernel is an oxide or carbide.
The chemical stability of Th02 and lack of kernel coating interaction during normal operation has been clearly established (refer to response to Lead Item I-1); consequently, the predominant consideration in the TRISO coated Th02 design is to assure the structural integrity during irradiation of the load-bearing layers.
Referring to Table 3, this is accomplished for TRISO coated Th02 by specifying kernel and buffer dimensions and densities which account for the pressure of fission gases produced under peak exposure conditions.
This procedure assures that sufficient heavy metal loading is achieved without subjecting the sic layer to excessive stresses. The methodology used to achieve acceptable kernel and buffer properties is discussed in more detail in response to Lead Item I-6.
Th02 kernels are fabricated by a gel-supported-precipitate (GSP) process.
Briefly stated, the process includes producing a spherical Th0 -Gel particle 2
which' is subsequently sintered to densities typically 98% of the theoretical Th02 density.
The process lends itself to producing a high quality product which is uniform and spherical.
In addition, specification requirements
control the impurity levels within the kernel. All of the irracliation test data on reference type TRISO coated Th02 fuel is based on kernels fabricated with the GSP process; consequently, the irradiation test results are directly applicable to the reference TRISO coated-Th02 design.
Coatings in the TRISO design are all produced using well established coating technology so that comparable fabrication procedures are exercised for all TRISO coated particles regardless of the kernel substrata.
Property and process specifications for the buffer, IPyC, sic, and OPyC layers assure the structural integrity of the TRISO coated particle design.
It should be emphasized that most of the property specifications have been empirically defined through the irradiation capsule test program.
Perfonnance of the sic is guaranteed through a series of property and process specifications which ensure the strength and fission product retention of this layer during irradiation.
Critical sic properties which are specified are thickness and density; specifically, thickness coni.rols the maximum stress level in this layer (refer to the response to i.ead Item I-6) and the density controls the inherent strength.
Figure 2 is a series of plots of the fracture strength of sic versus test temperature for different sic densities and grain sizes (Ref. 2).
Two important points to be noted from Figure 2 are:
1.
The fracture strength of sic increases with temperature above 1000*C for 6 nsities greater than 3.18Mg/m3, while for the low density sic (3.17Mg/m3) the fracture strength decreases above 1000*C.
2.
The trend between sic fracture strength, temperature, and density is independent of sic grain sizes between 1 pm and 15 pm in diameter.
Property specification limits on sic require that the mean density be 23.18Mg/m3, 3
and typically densities are 3.20 Mg/m.
Consequently, the sic deposition process is controlled to produce a high density product which in turn assures optimum mechanical properties during irradiation.
In addition, process control on the sic microstructure is maintained by specify-ing an allowable range of deposition temperature.
Figure 3 is a plot of sic
coating density versus deposition temperature and coating gas parameters (Ref. 3) 1.e., the flux and volumetric ratio of active coating gases.
Two general regions are~ shaded in Figure 3 which relate to the sic microstructure:
1.
The first region at deposition temperatures less than 1450 C defines a tendency for circumferential banding in the sic which is attributed to the formation of elemental Si and/or microporosity in the structure.
2.
The second region at deposition temperatures greater than 1700 C defines processing conditions which tend to result in large columnar grains ( >15 m in diameter) and excess microporosity.
These trends indicate that deposition temperature is a process parameter affect-ing microporosity, elemental silicon, and grain size which in turn determine the structural integrity and impermeability of the sic layer.
Coating temperature specification limits on sic require that the process operate in the range 1450'-1700*C.
This process restriction, combined with the sic density require-ments, assures optimum mechanical and structural properties for sic. Also, it should be emphasized that the majority of irradiation test experience is on TRISO coated particles containing sic layers with these optimum structural properti es.
Consequently, high reliability of sic performance has been success-fully demonstrated during irradiation and is assured through adequate control of property and process conditions.
RESPONSE to Lead Item I-3 The most recent irradiation test results on reference type TRISO coated Th0 2 fuel particles are divided into two groups:
1.
Loose (unbonded) particle tests conducted in capsules HT-31, 33, and 34.
2.
Fuel rod tests conducted in cell 2 of capsule GF-4.
Table 4 summarizes the range of irradiation test conditions and number of part-icles tested in capsules HT-31, 33, and 34.
In addition, the table lists the range of kernel diameters and buffer thicknesses tested. The particles tested in the HT series are intended to test the reference type TRISO coated particle and provide added confirmation of the particle stress calculation models. (Refer
to the response to Lead Item I-6.) This feature of the capsule tests is evident in HT-34, which had nominal buffer thicknesses as low as 27pm on an otherwise reference type TRISO coated Th0.
The results of the HT-31, 33, 2
and 34 capsule tests are sumarized in Figure 4, which presents a comparative evaluation between measured and predicted pressure vessel failure.
The figure is a plot of pressure vessel failure fraction measured visually (open circles) along with an adjacent predicted value (closed circle) versus burnup exposure.
The predicted value is based on TRISO coated particle stress model calculations discussed in the response to Lead Item I-6.
A failure fraction of 10-3 on the ordinant in Figure 4 is taken to represent zero observed or predicted failures.
In addition, the two batches with predicted failure fractions of ~.55 had nominal buffer thickness of 27pm and 35pm.
Two important conclusions from this figure are:
1.
TRISO coated particle stress calculations result in predicted pres-sure vessel failure fractions which are equal or greater than observed failure fractions.
Consequently, the reference design, which is based on the stress models, is extremely conservative.
2.
The upper 95% confidence bound on pressure vessel failure for TRISO coated Th02 particles irradiated in capsules HT-31, -33, and -34 to burnups between 4.9% and 10.3% FIMA at a temperature of 1200 C is <0.3% (zero failures out of 1090 reference type TRISO coated Th0 2
fuel particles).
When one considers these conclusions jointly, there is strong analytic and empirical evidence which supports satisfactory performance of the reference TRISO coated Th02 fuel irradiated to peak exposure conditions.
Fuel rod tests conducted in cell 2 of capsule GF-4 provide added confirmation of the expected performance of reference type TRISO coated Th02 fuel.
Table 5 lists the fuel types, nominal particle dimensions, and irradiation conditions for fuel tested in cell 2 of capsule GF-4.
The nominal Th02 kernel diameters tested in this capsule are approximately 75pm larger than the reference (450 m)
~
diameter, while the nominal buffer thickness is less than the reference (70 pm) thickness; conseouently, the fuel oerformance of these particles represents a conservative appraisal of the reference design.
Figure 5 is a plot of Kr85m R/B
(rate of release / rate of birth) for GF-4 (cell 2) versus fast fluence. As a basis of comparison, Figure 5 contains both measured (solid line) and predicted (dashed line) Kr85m R/B results. The measured values were obtained in-pile, and the predicted values were based on TRISO coated particle stress calculations for Th0 and UC fuel, using nominal particle dimensions and standard deviations 2
2 and GF-4 exposure conditions.
Figure 5 shows that the predicted values are in agreement with observations.
No Th0 failure was predicted under these exposure 2
conditions.
This prediction is supported by the decrease in R/B (Kr85m) for 25 2 (E>29fj)HTGR. The reduction in R/B at high fast fluence greater than 6x10 n/m fluences is attributed to a continued decrease in the number of fissions in the fissile particles with no accompanying increase in fertile failure as the number of fissions in the fertile particles increases.
RESPONSE to Lead Item I-4 The analysis of test results on TRISO coated Th0 fuel particles includes fuel 2
subjected to high exposure conditions.
Specifically, the HT-31, HT-33, HT-34, and GF-4 test results discussed in response to Lead Item I-3 were obtained for relatively high burnup exposure conditions. As a point of comparison, the mean and peak fertile particle burnups for 6 year old FSV fuel are 4.5% FIMA and 7.4% FIMA, respectively, The HT results included burnups between 4.9%
and 10.3% FIMA with approximately 50% of the test results greater than 7.4%
FIMA.
In addition, the Th0 tested in GF-4 achieved a burnup of 4.8% FIMA.
2 The HT series and GF-4 capsule tests provided verification of the adequate performance of Th0 fuel exposed to high burnup conditions.
2 RESPONSE to Lead Item I-5 The following response is presented with the understanding that the models and methodology described in GA-A12971 have not been approved by the NRC for use in Fort St. 'vrain licensing submittals.
TRISO Th02 particle failure via kernel migration is discussed in detail in the response to Lead Item VI.
No failure resulting from kernel migration has been observed to date in TRISO coated Th02 particles, and failure via this mechanism is not expected under FSV normal operating conditions.
sic-fission product interactions have not been observed in TRISO coated Th0 2 particles irradiated at temperatures up to 1500 C, as discussed in the response
to Lead Item VII. Accordingly, failure via this mechanism is not a'nticipated during normal operation.
Failure by means of pressure vessel effects are discussed in the response to Lead Item I-6, below.
Projected failures via this mechanism are significantly less than 1%.
As noted in GLP-5640, significant reductions in coated particle manufacturing defects are anticipated with TRISO coated Th02 particles as compared to TRISO coated ThC. Accordingly, failure of defective particles is also expected to 2
be small.
When these four potential failure mechanisms are combined, and projections of core performance are made using the most recent HTGR core design methods at General Atomic Company, total fertile particle failures of less than 1% are obtained for an equilibrium FSV core fueled with TRISO coated Th02 fertile fuel.
RESPONSE to Lead Item I-6 The TRISO coated Th02 particle design is developed on the basis of calculated sic stress distributions which assure that the expected particle failure from internal fission gas pressure is less than or equal to that calculated for FSV TRISO coated ThC2 fuel.
In order to adopt a conservative desi.
stress 9
calculations are done for peak fertile fuel exposure conditions, i.e.,1250 C, 8x1025n/m2 (E>29fd)HTGR, and 7% FIMA.
Figure 6 is a schematic diagram which illustrates the methodology used to calculate the sic stress distribution in a TRISO coated particle population. The explicit functions which define the maximum tensile stress in a TRISO coated particle are based on the Kaae perform-ance model (Ref. 4).
These functions are used to calculate the maximum sic stress for individual particles based on a random selection of critical particle properties, i.e., kernel and coating dimensions and densities.
This part of the calculation is performed by a Monte-Carlo type routine which accounts for random permutations of critical properties in TRISO coated particles.
The output of the calculational routine defines a histogram which describes the frequency of occurrence of a stress value, f(o).
This distribution is then used to define a failure criterion based on the assumption that TRISO coated fuel particles with the highest stresses fail.
The failure criterion is defined by:
7 I
f=
2 f(o) do
~
FC where f (a) = calculated sic stress density distribution f = experimentally observed failure fraction FC = Failure Criterion The failure criterion is determined by comparing measured failure fractions for TRISO coated Th02 particles in a number of different test populations with calculated stress distributions.
The underlying strength of this technique is that the failure criterion is determined empirically from irradiation test results on reference type TRISO coated fuel particles.
Table 6 lists em-pirically determined failure criterion for TRISO coated ThC, (Th/U)C, and 2
2 Th0 fuel particles.
The failure criterion for ThC and (Th/U)C is -2800 2
2 2
psi (Ref. 5) and it k=:ed on 12 separate batches irradiated in capsule F-30, i.e.,
the proof test for FSV fuel. The failure criterion for TRISO coated Th0 is 2
-3400 psi and is based on 24 particle batches irradiated predominantly in the HT capsule series.
Another point worth noting is that the empirically determined failure criterion for Th0 is essentially the same as the failure criterion 2
for ThC and (Th/U)C, that is, there is no statistical difference at the 90%
2 2
significance level.
The expected failure fractions for TRISO coated Th0 and ThC are defined in 2
g Table 7 and are 0.001 and 0.004, respectively.
Thee failure projections are based on nominal fuel properties, expected property distributions, peak ir-radiation exposure conditions, and the empirically deduced failure criteria defined in Table 6.
The expected failure fraction for TRISO coated Th0 is 2
less than that of the ThC2 particle. Consequently, ample performance margin exists in the TRISO coated Th02 particle design.
The impact of sic microstructure (grain size) and density on strength is dis-cussed in detail in response to Lead Item I-2.
References 1.
O. M. Stansfield, et.al, " Performance of Th02 in HTGR Fuel Particles",
GA-A14745, March, 1978.
2.
T. D. Gulden, " Mechanical Properties of 8-sic", Jou. Amer. Cer. Soc., 52,,
11, p. 585 (1969).
3.
Gas-Cooled Reactor Programs, HTGR Base-Technology Program Progress Report, January 1,1974-June 30,1975, ORNL-5108, p. 246.
4.
J. L. Kaae, "A Mathemetical Model for Calculating Stresses in a Four-Layer Carbon-Silicon Carbide Coated Fuel Particle", Jou. Nucl. Mat., 32,,
- p. 322 (1969).
5.
HTGR Fuels and Core Development Program, Quarterly Progress Report for the Period Ending February 28, 1978, GA-A14863, March 1978, pp. 9-59.
FIGURE 1 CUMULATIVE DISTRIBUTION DEFINING IRRADIATION TEST EXPERIENCE FOR TRISO AND BISO COATED Th02 FUEL 1
1 PARTICLE TOTAL NUMBER OF l TYPE PARTICLESTESTEDl s
6 TRISO Th02 s 10 tt) i,
(
)
l
.8.
y n
E! N i
BISO Th02 s 10 l
- -) p 53 4 i
(.............. )
=ha i
3 m a Q
.6.
i' sS i
FERTILE BUR"UP IN 6 YR.i C/2 g3
=
i.
OLD FSV FUEL N Z ss 4 z i
i 1
W $
MEAN PEAK
$P 4-i.
me i
ow z"
o H
eo i.
i i
o
.2_
i.
- s 2. 5 x 104 i
x i.
~6.5x103 PA C
......... N 0
0 4
6 8
10 BURNUP (% FIMA)
Th02 FUEL TESTED BETWEEN 600
- 1500 C AND 0.4 - 15.5 25 9
x 10 n/m- (E > 29 fJ)HTGR
FIGURE 2 BEND STRENGTH VS. TEMPERATURE FOR CVD sic OF SEVERAL GRAIN SIZES AND DENSITIES 200 - 4032-99 GRAlta DI AMETER,15 m
3 DE tlS I T Y, 3.206 g/cm 150 -
100 -
i 7
200 - 3662-73 c
GRAlti OIAMETER, 4 m
~
3 DENSITY, 3.208 g/cm E 150 s
1 u
5 100
$200 - 3662-111 C
GRAlti OI AMETER, < 1 pm h
3 D E tlS I TY, 3.189 g/cm e 150 J
te
{100 i
200 - 3742-89 GRAIN DI AMETER, < l m 3
DENSITY, 3.171 g/cm 150 T
100 j
O 500 1000 1500 TEMPERATURE ('C) 6
FIGURE 3 sic DENSITY AND STRUCTURAI. VARIATIONS VS. COATING CONDITIONS H
DATA 2
FLUX 3
2 (cm / min.cm )
SYi/ BO L MTS*
O 20 0.2 A
25 0.2G NOT STATED A
~30
- 0.63
'METHYLTRICH LO ROSIL/stJ E 3.22 3.20
\\
s
\\
/
\\
.g \\.,.;.,,, g:;:...,
.. /
3.18 i,.b. :::\\x. '
1;.
? L. < v. :.. r..
g,.
~ w... /..<
W; B' : :;;.;..;. - '..
^
- -l
..g.
g?'S9:h.' ;;h
. 0* ?.
s-
~.~:.:...;\\. W,. '.. >. *,
3.1G 1:h :.?. ?.'
.., c;.. p, -
~,.... %.
. v.
".E
',[ kj'.;',l N0TOhf: l l;, ',
'h P's.'1, C%
F
. :$h, '..
REG 10t10EFli!!ilG
%'. ' j.,MNTENDENCY FOR
,.( 6.' N' ' C
'E p
3.14
QJ 1:),
{.' T[l. f d.,
' CIRCU.iFER EilTI AL y
.'.,.d,.,..-... n.,
- ...,.... -. ' BAilCli!G !?! sic
.E. &
- S,6-tj f;, ' '- r.:. STRU CTU RE S'd-T'. '.)\\.:
~
w
- c.,,,.
o
~
f ATTR1CUTED TO
- c. ',,'.if
. " ~ " '
e 3.12 N {'.a ",
W '.
. ELEM ENTA L Si
. '..' j;{d
- 9. :.2..: \\.. L ~,',' *
'f.2. '. W '
-.i Ai10/08 MICR0PO ROSITY.,
^'.
- .,,p:._.. -
o
.:.:.'.'M)9.:
..1 o
9 3.10
- ;.. *.n.
1.t. :...>.;.
.:,;)n.
d-3,,. " -
G;I.'f
..}* ' '
.p..
v7:.
- REGIOil DEFit!!i!G 55 c.
3.08 m.._.
>TE?lDEi:CY FC R
,' ~
- lj.,
7 1 ":
.,. : LARGE CO LU:/.. AR
-(.b ? -
,. ) ~.
W :. '. GRAINS ( >15 gm IN O!AMETER): :j -
M~R.7 -.'
AND EXCESS MICR0POROSIT,Y..
3
.g.,.
A
-['
~ -*~~ c[,.
~
3.06
.~
y
.py::
. c. }..
- v..:.: s
- s.,
- p. m.,...
.Y:.~....n..,..
i:.
l l
3.04 1300 1400 1500 1600 1700 1800 1900 CDATING TE.'PERATURE ( C)
FIGURE 4 COMPARATIVE EVALUATION BETWEEN MEASURED AND PREDICTED TRISO COATED Th02 PRESSURE VESSEL PERFORMANCE IN CAPSULES HT-31, HT-33, AND HT-34 IRRADIATED AT 1200 C AND BETIEEN 4.1 TO 8,3 n/m2 (E > 29 fJ)HTGR-
.. -............ - - -... - ~. - -. -
a 1
i z
1 O
T O - VISUAL,LY MEASURED nv O
O g - PREDICTED.
f
'm a
_a
_1 Q
10 g
u.
i hJ m
m m>
g nc
-2 3
10 m
i m
a y
e
~3 10
($dg,$7$'$$$-GDv{$
'[O T'(S '
O ( O '- ( O O ~~ O ~ -
,it 5'
6 7
8 9
10 11 BURNUP h FIMA)!
I I
-4 j0 gT (IEASURED G-
, PREDICTED 4-2-
~~
~5
~~ O~
'O 10 p,
f a-n E
G-
. A-EOL. -
to m
4..
s M
?. CI' lis N
-G a: 10 p
G-2-
-7 10 i
i i
i e
i i
4 Q
l 2
3 4
5 6
7 8
9' IO I
2 FAST FLUENCE,10 n/cm (E > 0.18 MeV)g7ag FIGURE 5 MEASURED AND PREDICTED Kr-85M R/B VS. FAST FLUENCE FOR CAPSULE GF-4, CELL 2 CONTAINING VSM UC2 TRISO/Th02 TRISO FUEL.
FIGURE 6 SCHEMATIC DIAGPM DEFINING METHODOLOGY USED IN PRESSURE VESSEL PERFORMANCE CALCULATIONS EXPLICIT FUNCTIONS
, TRISO = gf t,. p;,
, TE!.'PER ATURE. BURNUP, FLUENCE) 9
- ARBITRARY FUNCTION INDEX DEFINING i* COMPONENT PROPERTIES i
=
t;
= DIMENSION DISTRIBUTIONS FOR CRITICAL COMFCNFNTS: KERNEL, BU F FE R, lPyC. sic p; = OENSITY DISTRIBUTIONS FOR CRITICAL COMPONENTS: KE RfiE L, BU F F E R, lPyC, $;C e = MAXIMUM sic TENSILE STRESS MONTE-CARLO TYFE CALCULAil0NS FOR VARIOUS FERMUTATIONS OF RANDOM VARIABLES t, AND p;. e.g.
z 3
3 3
4 3 4
4 4
g i
h 1
I 2
2 2
2 h:
23 KERNEL BUFFER IPyC sic h
DIAMETER THICr,N ESS THICKNESS THICKNESS HTGR FUEL PARTICLE POPULATION STRESS DENSITY DISTT.lSUTION 3
FAILURE CRITERION
- =5 t
8 t-r7 NU AREA DEFINES EXPECTE]
{-}
-} {
PRESSURE VESSEL FAILURE E#
cf2) l l e(!!!
MAXIMUM TENSILE STRESS, e
TABLE 1 PROPERTY VALUES FOR TRISO COATED Th02 1_-
-a
.21 m
N-r L1 Ys.;--
--.i
__-l, - _, _ _
_u'__
BUFFER PROPERTIES l IPyC PROPERTIES l
KERNEL PROPERTIES Y,
TIIICK-TIIICK-l ANISO-I DIAMETER DENSITY NESS DENSITY NESS DENSITY TROPY
- f 3
3 3
CATECORY (um)
(Mg/m )
(um)
(Mg/m )
(pm)
(Mg/m )
(BAFo)
)
4
' NOMINAL PARTICLES 450 9.8 -
70 0.80 -
35 1.90 1.06 i
i PROPERTIES FOR 9.9 1.10
' REFERENCE DESIGN I --
_,m_,
- - _, - wm-n- - nu my f RANGE OF IRRA-275 - 530 9.7 -
55 -
0.90 -
25 - 40 1.80 -
1.03 -
DIATION TEST 10.0 105 1.30 1.95 1.08
- EXPERIENCE I
(MEAN VALUES)
~
i,,, n u m
_1 1
=rn, 2r.-
~ - - - -
l
, =. m m J
CIIST SAMPLES
- 1 (MEAN VALUES) i I
l 6252-00-025 512 9.91 64 1.16 30 1.87 1.06 4
6155-00-010 404 9.70 61 1.22 32 1.88 1.06
{
id 6155-01-010 491 9.77 87 1.22 30 1.96 1.06 l
6155-01-020 496 9.77 90 1.24 32 1.93 1.05 I
\\
\\
. =
=.
.~
-, =
PRESENTED FOR COMPARATIVE PURPOSES TO REFERENCE DESIGN (CHST REFERS TO CORE HEATUP SIMULATION TEST).
TABLE 2 PROPERTY VALUES FOR TRISO COATED Th02 7 - - _- -
~
l OPyC PROPERTIES sic PROPERTIES TIIICK-DEPOSITION TilICK-NESS DENSITY TEMPERATURE NESS DENSITY ANISOTROPY 3
CATEGORY (um)
(Mg/m )
( C)
(pm)
(Mg/m3)
(BAF )
o NOMINAL PARTICLE 35 3.20 1450 - 1700 45 1,87 1.04 h
h r
a PROPERTIES FOR f REFERENCE DESIGN l
L RANGE OF IRRADIA-20 - 40 3.18 - 3.22 1400 - 1750 30 - 60 1.75 - 1.95 1.02 - 1,06 ]
TION TEST k
EXPERIENCE (MEAN VALUES)
__ m 4 h CilST SAMPLES *
(MEAN VALUES)
I s
g 6252-00-025 31 3.21 1640 42 1.81 1.06 6155-00-010 29 3.22 1650 35 1.78 1.04 l
1 6155-01-010 28 3,22 1650 41 1.78 1.04 5
i I
6155-01-020 28 3.22 1650 39 1.81 1.02 i,
L;1_
m_-
.u m
PRESENTED FOR COMPARATIVE PURPOSES TO REFERENCE DESIGN (CilST REFERS TO CORE IIEATUP SIMULATION TEST).
TABLE 3 TRISO COATED PARTICLE GENERIC DATA BASE
.--,w-.-
- = - =. _
- ~..
-n.-
1 PARTICLE j
COMPONENT PROPERTY PROPERTY JUSTIFICATION
--1
--__,... _ _- ~.;.= m m...:y - =
c.~.
~ - - -
KERNEL / BUFFER DIMENSIONS, HEAVY METAL LOADING, 4
L DENSITY PRESSUPI VESSEL DESIGN
.. ~ ~
,__.=m IPyC THICKNESS, IMPERMEABILITY TO l
r
- DENSITY, CHLORINE DURING sic a
?
ANISOTROPY DEPOSITION AND STRUCTURAL STABILITY DURING IRRADIA-TION
,r-.
- c.. m:nmaammxx:mmune-a l
sic THICKNESS, ENSURE STRENGTH AND j
- DENSITY, FISSION PRODUCT RETENTION L
DEPOSITION DURING IRRADIATION 9
TEMPERATURE
-~ m -..
OPyC THICKNESS, MAINTAIN STRUCTURAL L
- DENSITY, INTEGRITY DURING IRRADIA-
[
ANISOTROPY TION P
.-_____3 e_
.c_.--
-.__ --. _ -..~. -.
TABLE 4
SUMMARY
OF IRRADIATION TEST RESULTS ON TRISO COATED Th02 LOOSE PARTICLES TESTED IN CAPSULES HT-31, 33, AND 34 (IRRADIATED AT 1200 C)
=....__._
RANGE OF NUMBER IRRADIATION CONDITIONS RANGE OF BUFFER OF rs: = m,. =. m m :.- w _: - z,
KERNEL THICK-PARTICLES FLUENCE 25 2
BURNUP
[
CAPSULE DIAMETERS NESSES TESTED 10 n/m
)
TEST (um)
(pm)
(E > 29 fJ)MTGR
(
FM)
A j
- +~
w.-;
- samrn=ammmma m-
. -w
- m.,..
HT-31 447 - 511 57 - 63
/> 03 4.1 - 6.5 4,9 - 7.2 t
4 l
- b HT-33 446 - 509 58 - 65 406 5.2 - 8.1 7,2 - 10,3 HT-34 448 - 455 27 - 83 451 5.3 - 8.3 5,6 - 9.1 Murt'=x7.m'g, M M. M M.L.)
TABLE 5
SUMMARY
OF NOMINAL PARTICLE DIMENSIONS FOR TRISO C0ATED UCp AND Th0p FUEL TESTED IN CAPSULE GF-4(CELL 2)
Irradiation Conditions Flyb'n/m2 Number of Nominal Particle Dimensions Design nce Particles Particle Type (pm)
Temperature Burnup 10c Tested (Batch Designationi)
Kernel Buffer IPyC sic OPyC
( C)
(% FIMA)
(E>29fJ)HTGR 7,000 UC2 (6151-17-020) 204 100 35 30 33 1100 75.5 9.5
'6252-03-010 529 63 30 33 37 16,900 Th02 d 6252-04-010 528 63 30 33 39 1100 4.8 9.5 (6252-05-010 519 63 30 33 49
e TABLE 6 EMPIRICALLY DETERMINED PRESSURE VESSEL FAILURE CRITERIA FOR TRISO COATED FUEL PARTICLES
~m
~
RANGE OF OBSERVED
)
PRESSURE VESSEL l
NUMBER FAILURE (%)
f IRRADIA-0F i._. _ _ m._.. a. 1....
TION IRRADIA-METALLO-FISSION FAILURE PARTICLE CAPSULE TION GRAPHIC /
GAS (a) CRITERION q
q TYPE TESTS BATCHES VISUAL ! RELEASE (PSI)
- , _ w. __=. __,_m L
(b)
ThC / (Th,U)C2 F-30 12 0 - 9.1
-2800 2
J o
J l
- c r
- =.:m w=--
-. = :.,.m x,
f Th02 HT-17, 18, 24 0 - 81.0 3.4
-3400(b)
J i
19, 28, 31, 33, 5
AND P13S 2
4
.;17 - - -
4
__u__m
..w
-.-w
_--..r*->_
.1.--
,- e Me ir- _ _ _ - -- m -- _ _
-s-
-m,__,__i (a) DETERMINED ON BASIS OF Kr-85M R/B (RATE OF RELEASE / RATE OF BIRTH).
(b) NO STATISTICAL DIFFERENCE BETWEEN CARBIDE AND OXIDE FAILURE CRITERIA AT 90% CONFIDENCE LEVEL.
r TABLE 7 COMPARATIVE EVALUATION OF PRESSURE VESSEL PERFORMANCE FOR TRISO COATED Th02 AND ThC2
=
.m s
....... -... -. w.-
u.:
NOMINAL PARTICLE DESIGN EXPECTED PRESSURE
[ KERNEL DIMI./ BUFFER h VESSEL PERFORMANCE (b ;
1 i
PARTICLE TYPE (THICK / PARTICLE DIAM.)
(FAILURE FRACTION)'
-. M L E' TT' A Vini 21 4
. M :i M C L' M a.i.
T
.< M AM-~ -
i TRISO ThC2 446/53/748
- 0. 004(c)
(
(DS 184B)(a)
'. "3E""Er
.EZ' ' l.2EEN '
"2...'.7 T M d.M """" nMZ ZfEC a 'hv'N.?
l
'i TRISO Th02 450/70/820
- 0. 001(c)
(DS 157G)(8) 1 l
\\
ra.zm
....n.......,.,.-...:....
(a) DESIGN CALCULATION NUMBER (b) sic STRESS DISTRIBUTIONS CALCULATED FOR FUEL EXPOSED TO 25 2 (E > 29 fJ)HTGR, 1250 C, AND 77. FIMA 8 x 10 n/m (c) BASED ON TRISO C0ATED PARTICLE STRESS CALCULATIONS DESCRIBED IN FIGURE 6 AND FAILUilE CRITERIA IN TABLE 6.
RESPONSE TO NRC LEAD ITEM II FOR Th02 REVIEW II.
Gaseous Fission Product Release From Th02 TRIS 0 Particles The statement in GLP-5640 (p.4-9) that gaseous fission product release from TRISO coated Th02 during core heatup is less than FSAR predicted values for Th02 particles seems inconsistent with the respective release-to-birth ratios (R/B) on page 4-7.
Further information should be pra-sented on the data base for the R/B values for each type of fuel particle.
The way these values are used in the dose calculations should be dis-cussed also.
RESPONSE
The release to birth rate ratio (R/B) is a parameter used to describe the re-lease of short-lived fission gases or gas-like species from the fuel during normal core operation.
During a core heatup accident, the reactor is shutdown, the birth r e is zero, and R/B is not a useful or meaningful parameter.
Fission product release during a core heatup is determined primarily by the effects of time and temperature on fuel particle coating failure and on dif-fusive release of fission products from the kernels of failed fuel particles.
As shown below, fuel performance data collected under simulated core heatup conditions support the conclusion (p. 4-9 of GLP-5640) that gaseous fusion product release from TRISO Th0 during a core heatup event would be less than 2
suggested by predictions made using FSV FSAR models and methods. The data also show that release from TRIS 0 Th0 w uld be less than from TRISO ThC '
2 2
Doses are ca.culated using the methods described in the FSAR.
Core Heatup Simulation Test Program The performance of HTG't fuels under hypothetical accident conditions is being evaluated in an on-going core heatup simulation test (CHST) program.
Groups of 50 to 200 particles are heated, in the CHSTs, from 1100 to 2500 C over periods of 8, 30, or 80 hours9.259259e-4 days <br />0.0222 hours <br />1.322751e-4 weeks <br />3.044e-5 months <br />. Temperatures are increased linearly with time.
Fuel performance is monitored by measuring fission product release fractions as a function of time and temperature.
e
The key fission product that is monitored is Kr85, since gaseous fission product release is related to total coating failure. At present, the release data are compared with predictions made using applicable fuel failure and fission product release models to evaluate the degree of' conservatism associated with the models.
The CHST conditions are chosen to simulate the range of conditions predicted for those hypothetical core heatup events that are considered in reactor licens-ing and siting applications.
In the case of FSV, a key event is Design Basis Accident No. 1 (DBA #1).
The FSV FSAR shows, in Appendix 0, that the minimum time required for a small fraction of the fuel to reach -2500*C during DBA #1 would be 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br />, while the average fuel temperature would not reach 2500 C.
A schematic of the CHST system is shown on Figure 1.
Testing is conducted in resistance heated graphite tube (King) furnaces.
Each furnace is penetrated by four tantalum tubes.
One tube extends about half way through the furnace, is sealed on the hot end, and contains a temperature control thermocouple.
Three of the tubes are open ended and extend through the furnace.
These tantalum tubes are used to house test samples and mullite traps for cesium collection.
Each group of particles used in a sir.gle test is separated into three samples.
Each sample is then loaded into a type H-451 graphite crucible. One crucible (test sample) is then placed in each open ended tantalum tube.
Sample test temperatures are monitored optically during heating.
The tests are conducted in flowing helium (50cc/sec/ tube) to assure that re-leasea fission products are quickly transported to their respective traps. As the helium and fission gases exit the furnace, they are combined into a single flow line that passes through (i) a bed of hot (-500 C) cuprous oxide to convert tritium to water, (ii) a desiccant to remove the water, (iii) a room temperature activated charcoal bed to remove radon, (iv) two ionization chambers and (v) a liquid nitrogen cold trap.
The cold trap, which is used to collect Kr85, is changed periodically and gamma counted to obtain quantitative Kr85 release data.
The response of one ionization chamber is integrated electronically and then related to cold trap Kr85 data to obtain a continuous measure of Kr85 release.
The second ionization chamber is retained as a backup.
Upon completing each test, the Kr85 release data are compared with the pre-test Kr85 inventory to obtain the release fraction as a function of time and temper-ature.
TRIS 0 Th0p CHST Samples and Heating Conditions Kernel and coating dimensions and densities for the TRISO Th02 samples tested to date are summarized in Table 1.
Nominal kernel and coating properties speci-fied for the reference FSV TRIS 0 Th02 design are also shown.
Kernel and coating dimensions and densities of the test samples are consistent with the FSV refer-ence design.
Two of the three test samples contain (8Th/U)02 kernels.
These were used so that test samples with kernel burnups exceeding the maximum ex-pected Th02 burnup in FSV would be included in the test program.
Irradiation conditions of the three test samples are compared with expected FSV conditions in Table 2.
Three sets of conditions are shown for FSV. The first (core average) represents average conditions expected for an equilibrium core.
The second (avg. 6 yr. fuel), represents the average conditions expected for fuel removed from FSV after 6 years of operation.
The third set (peak) gives maximum conditions to be experienced by fuel after 6 years of operation in FSV.
Three parameters are included in Table 2.
The first two (fast neutron exposure and kernel burnup) are self explanatory.
The third (fission density) is the 3
ratio of the a/erage number of fissions per particle to the volume (m ) per parti-cle inside the sic layer.
Fission density is proportional to the average fission gas pressure or metallic fission product concentration within individual test samples.
It provides a simple method for normalizing irradiation exposure to kernel / particle dimensions when comparing the behavior of various test samples.
As shown in Table 2, the irradation conditions for the TRIS 0 Th02 test samples range from the average to peak values expected for 6 year old fuel.
Seven CHSTs containing a total of 566 TRIS 0 Th02 particles have been conducted (Ref. 1). A summary of samples used, number of particles tested, and CHST con-ditions is given in Table 3.*
Results and Discussion Krypton 85 release fractions observed in tests conducted for 8.3-10.-5 hours and 67.0-71.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> are shown as a function of tcmperature on Figure 2.
Release fractions observed in the 28 hour3.240741e-4 days <br />0.00778 hours <br />4.62963e-5 weeks <br />1.0654e-5 months <br /> test fell on the boundary between the short term (<10.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />) and long term (>67.0 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />) test. data.
The results show
- Results from CHST numbers 78IHR-A-2-1 and 78IHR-C-3-1 have not yet been docu-mented.
Results are contained in GA notebooks 7287 and 7775.
that fission gas release fractions during a core heatup would 'be less than 5%
at temperatures as high as 2000*C.
They also show that Kr85 release at any given temperature will increase as the time required to reach that temperature increases.
Since the CHST conditions are consistent with conditions projected for FSV DBA#1 and the irradiation conditions of the test samples are consistent with those expected for 6 year old fuel, the data developed to date on TRIS 0 Th02 provide a very conservative demonstration of the behavior of oxide fertile fuel during a hypothetical core heatup in FSV. The data collected from the tests summarized in Table 3 were combined to detennine the expected Kr85 release fraction and the range, at a 90% confidence level, for the expected Kr85 release fraction from TRIS 0 Th02 as a function of temperature.
The range for expected Kr85 release fractions is shown as a function of temperature in Figure 3.
Results obtained from a 26.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> CHST conducted on 101 FSV TRIS 0 ThC2 fertile particles (Ref.1) are also shown on Figure 3.
The ThC2 fuel was obtained from a FSV initial core production batch and irradiated to a burnup of 4.7%
FIMA in the FSV fuel proof test, capsule F-30 (Ref. 2).
The results lead to the conclusion that fission gas release from TRISO Th02 during a core heatup event would be less than or equal to release from TRIS 0 ThC2-Predictions of Kr85 release fractions were made as a function of temperature for CHST conditions using FSV FSAR fuel failure and fission product release assumptions.
It was assumed for these calculations that temperatures increased linearly with time from 1100 to 2500 C in 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> or 80 hours9.259259e-4 days <br />0.0222 hours <br />1.322751e-4 weeks <br />3.044e-5 months <br />.
The predictions are compared with experimental results for TRia0 n C2 and TRIS 0 Th02 on Figure 4.
Predi.cted release fractions are greater than observed from both types of fuel, which leads to the conclusion that FSV FSAR fuel failure and fission product release assumptions are conservative.
The data on Figure 4 clea.ly support the statement in GLP-5640 (p. 4-9) that gaseous fission product release from TRISO coated Th02 is less than values predicted for TRISO ThC2 using FSV FSAR models and methods.
Sumary Experimental result', obtained from tests conducted under simulated core heatup conditions have shown that:
- TRIS 0 Th02 fission product release is less than TRIS 0 ThC2 fission product release.
- Kr85 release fractions from both TRIS 0 Th02 and TRIS 0 ThC2 are much lower than' predicted using FSV FSAR fuel failure and fission product release m)dels.
References 1.
"HTGR Generic Program Technology Program, Fuels and Core Development Program, Quarterly Progress Report for the Pe-fod Ending February 28, 1978", GA-A14863, March 1978.
2.
C. B. Scott and D. P. Harmon, " Post Irradiation Examination of Capsule F-30", GA-A13208, April 1, 1975.
GRAPHITE HEATING ELEMENT.
TA TUBE
.g MULLITE TEST HE IN
(
TUBE SAMPLE HE OUT DESICCANT ACTIVATED CHARCOAL CUPROUS OXIDE (500*C)
(20*C)
(20*C)
LIQUID N2
/
IONIZATION IONIZATION COLD TRAP CHAMBER CHAMBER d
=
FIGURE 1 SCHEMATIC DRAWING 0F THE CHST SYSTEM
~
e 1.0 y
a y
LENGTH CHRS) OF CHST 0.8 8.3 - 10.5 Q 0
o-5 67.0 - 71.5 0 p
0 E 0.6 Q-5 0
b 0.4 O
a 0
0 0.2 0
0
'O O'^^
bd D
^
^^'
0 1100 1300 1500 1700 1900 2100 2300 2500 TEMPERATURE (*C)
FIGURE 2 Kr-85 RELEASE DATA COLLECTED DURING 8.3-10.5 HR.
AND 67.0-71.5 HR CilSTs 0F TRIS 0 Th02 FUEL
t 1.0 y
s\\\\\\\\\\\\\\' RANGE AT 90%
CONFIDENCE, 0.8 TRISO THO 2
2 r;
g 0.6 na 0.4 N\\
a 2
\\
,k 0.2 TRIso THO M
DATks 0
t
"ss4ss
--ss 1100 1300 1500 1700 1900 2100 2300 2500 TEMPERATURE (*C)
FIGURE 3:
COMPARIS0N OF THE RANGE AT A 90% CONFIDENCE LEVEL FOR Kr-85 RELEASE FRACTIONS EXPECTED FROM_TRISO Th02 FUEL WITH DATA OBTAINED FROM CHSTs 0F TRISO ThC2 FUEL.
a F
1.0 y,
a
~
PREDICTED KR-85 0.8 RELEASE FSV FSAR Q *.
zo METHODS 8-80 HR CHST TRISO U
$ 0.6 DATk k
~
g
-s;.,
-[
g 0.4
- y 4
5 0.2
~
TRISO 2
THo
- g DATA /
'mmHMN N 0
1100 1300 1500 1700 1900 2100 2300 2500 l
TEMPERATURE (*C)
FIGURE 4: COMPARIS0N OF Kr-85 RELEASE FRACTIONS PREDICTED USING FSV FSAR FUEL FAILURE AND FISSION PRODUCT RELEASE ASSUMPTIONS WITH EXPERIMENTAL DATA OBTAINED FROM CHSTs CONDUCTED ON TRIS 0 Th02 AND TRIS 0 ThC2 FERTILE FUEL.
TABLE 1 KERNEL AND COATING PROPERTIES ") OF TRISO Th0 CORE HEATUP SIMULATION TEST SAMPLES I
2 KERNEL BUFFER IPyC sic OPyC DEN.3 DIA.
DEN.3 THICK.
DEN TIIICK.
DEN THICK.
DEN.3 HICK.
3 (Mg/m )
(pm)
(M g/m )
(pm)
(Mg/m )
(pm)
(Mg/m3)
(pm)
(Mg-/m )
(um) 6252-00-025 9.91 512.
1.16 64 1.87 30 3.21 31 1.81 42 6155-00-010( )
0.70 404 1.22 61 1.88 32 3.22 29 1.78 35 6155-01-010/020(b) 0.77 494 1.23 89 1.95 31 3.22 28 1.80 40 FSV(c) 9.9 450 1.0 70 1.90 35 3.20 35 1.87 45 (a) average properties (b)
(8Th/U)O kernels used to simulate high burnup Th0 fertile fuel 2
2 (c) reference design (for comparison with CllST samples) 9 O
TABLE 2 IRRADIATION CONDITIONS OF TRISO Th0 CORE HEATUP SIMULATION TEST SAMPLES 2
IRRADIATION CONDITIONS AST NEUTRON DATA KERNEL E
OSURE I
RETilIEVAL BURNUP FISSION ")
NUMBER (10 n/m")
(% FIMA)
DENSITY 26 6252-00-025 11.5 4.1 3.6 x 10 6
6155-00-010 5.0 8.0 5.7 x 10 26 6155-01-010/020 5.0 8.0 5.4'x 10 FSV( )
6 Cbre Avg.
2.8 1.9 1.4 x 10 6
Avg., 6 yr fuel 4.9 4.5 3.4 x 10 26 Pedt 8.2
- 7. 0 5.0 x 10 3
(a) fissions /m inside the sic layer (b) reference design
TABLE 3 TRISO Th0 CORE IIEATUP SIMULATION TEST CONDITIONS 2
TEST CONDITIONS "I
"A I
)
RE R EVAL CIIST NUf.! DER OF LENGTII NUMBER NO.
PARTICLES (hrs)
INITIAL PEAK 6252-00-025 78 IIIR-A-1-1 100 0.7 1100 2425 6252-00-025 78IIIR-B-2-1 100 28.0 1150 2395 6252-00-025 78IllR-Cc2-1 100 70.7 1125 2435 6155-00-010 78 IIIP.- A-2-1 75 8.3 1000 2580 6155-00-010 78IIIR-C-3-1 75 71.5 1100~
2390 6155-01-010/020 77 IIIR-A-6-1 55 10.5 050 2310 6155-01/010/020 78IIIR-C-1-1 61 67.1 1100 2350
RESPONSE TO NRC LEAD ITEM III ON THORIUM 0XIDE III. Effect of Kernel Type on Steam-Graphite Reactions It is indicated in GLP-5640 that, since the retention of barium and strontium (both known to catalyze steam-graphite reaction) is expected to be improved with Th02 kernels, the reaction rate of the core graphite will be reduced.
There are no numerical values provided in the report discussion, however, and so it is not possible to determine the full ramification of this potential change in reactivity.
For example, if less core graphite is reacted per unit time, does this result in greater oxidation of the core supports? If so, what is the potential safety significance? Moreover, within the core itself, a reduction in reactivity of the core graphite will not necessarily result in a reduction in total oxidation, but may merely shift the extent of oxida-tion from the top (colder) regions of the core to the bottom (hotter) regions.
In addition, if the reactivity of the core graphite is, in fact, reduced significantly, there is a greater potential for hydrolysis of the carbide fissile kernels. These concerns should be addressed in sufficient detail to permit an assessment of their safety significance.
RESPONSE
A bounding, order-of-magnitude estimate of the potential change in core graphite reaction rates due to a change in metallic fission product concentration sorbed in the graphite is presented here.
This estimate is based on information pre-viously presented to NRC in the review of GA-LTR-7 (Ref.1).
It was then demon-strated that water availability is not affected in such a manner as to signifi-Cantly Change Consequences of water ingress events, since there is an abundance of water available in any case.
During NRC review of LTR-7, several second round questions (Ref. 2) pertained to quantifying the effects of barium and strontium reaction rate.
The response (Ref. 3) showed that, for a typical large HTGR fuel element fueled with UC2-TRISO/Th0 -BISO particles:
2 1.
The barium concentration builds up near-linearly to about 429 per fuel element at end of life.
(This value is about the same for FSV elements at E0L.
For the FSV equilibrium core, the average barium concentration in the core is about 0.25mg/g of fuel element carbon.
Most of this barium is retained in the fuel particles, but a small fraction, roughly on the order of 10-2, is released to and retained in the fuel element moderator graphite.)
. 2.
Using the complex equations built in the 0XIDE-3 Code to calculate the catalyst concentration distribution in the graphite, the reaction rate increase factor due to barium and strontium was estimated to be from 1.04 to -h22 locally, depending on fuel age, for a typical graphite temperature of 1460 F.
The core average rate increase was 1.10 (see response to Question 9d in Ref. 3).
Equations for the effect of catalyst on the reaction rate were taken from page 3-32 of LTR-7:
Fc = 1+Cc exp (12.153 - 4.264 x 10-3T F) c = C a + 0.2 C r in mg/g of graphite.
The strontium con-where C B
S centration was assumed to be similar to that for barium (Cc= 1.2C a)-
B 3.
As a check, it was noted in the response to Question 9c in Ref. 3 that the overall barium concentration in the core (including within the particles) is calculated by OXIDE-3 to be about 44g per fuel element.
In the FSV FSAR accident analysis of steam ingress events, a typical condition for graphite oxidation is given on page 14.5-3 to be:
(1) temperature of 1600*F, and (2) barium concentration of 0.00lmg/g of C in the moderator graphite.
For these values, the above equations yield Fc = 1.2, which is roughly consistent with the above estimate.
During a steam ingress accident, only a small percentage of steam would react before core cooldown. Table 1 presents a summary (taken from Table 14.5-1 of the FSV FSAR) of the steam reacting during the six ingress accident cases analyzed in Section 14 of the FSAR. The amount of unreacted steam at the end of the accident is from 60 to 99 percent of the total ingress.
Thus, there is abundant water for reaction with the core support structure or for fuel hydrcly-sis. The question remaining is: will decreased reaction rates with the fuel element graphite cause higher reaction rates elsewhere (by higher H 0 pressure)?
2 It should be noted that for the high steam pressures characteristic of these accidents, the steam-graphite reaction becomes saturated and is no longer first order, but is approaching zero order (becoming independent of steam concentra-tion).
However, even if the reaction were first order, one can bound the effect on reaction rate in the core support structure as follows:
~
. 1.
The change in the amount of water reacting is proportional to the square root of the change in fuel element graphite reaction rate.
2.
According to the above evaluation, the total effect of catalytic fission metals is to change the overall reaction rate of the active core graphite by less than a factor of 1.2 (water reacting by factor of 1.1).
3.
Thus, if all catalysis were eliminated, the percent of water unreacted in Table 1 would increase by 7% in accident case 4 and by less than 2%
in all other cases.
4.
The corresponding change in burnoff of other graphite components, such as core support structure, due to Item 3 is about 7% for accident Case 4 and less than 2% for other cases for a first order reaction.
These are relatively small changes which should be masked by conservatively assumed steam-graphite reaction rates.
It is noted that these calculations are bounding estimates to put the catalytic effect into perspective. The effect of a change in the catalyst concentration due to use of Th02 fuel would be some fraction of the above values.
The results also pertain to the question of the effect of catalysis on the hydrolysis of the fuel, since the hydrolysis reaction is first order.
It should be noted further that most hydrolysis occurs over a long time period after the steam-graphite reactions, which are more temperature sensitive, stop.
Therefore, any temporary change in the amount of water consumption as it diffuses through and reacts with the fuel element graphite is negligible.
Based on the information present above, the conclusions are as follows:
- Better retention of barium and strontium in Th02 kernels changes local steam-graphite reaction rate only a few percent at most.
- Changes in reaction rate due to F. P. catalyst variation do not signifi-cantly affect water concentratien for reactions in other parts of the core or core support structure.
.
- Hydrolysis is essentially unchanged since water availability is un-affected.
(Most hydrolysis occurs after steam-graphite reactions stop.)
References 1.
M. B. Peroomian, A. W. Barsell, and J. C. Saeger, "0XIDE-3, A Computer Code for Analysis of HTGR Steam or Air Ingress Accidents", GA-Al2493 (GA-LTR-7), January 15, 1974.
2.
R. A. Clark to D. S. Duncan, "Round 2 Questions on GA-LTR-7", letter dated December 12, 1974.
3.
L. D. Johnson to R. A. Clark, "GA-LTR-7, 0XIDE-3", letter dated May 20, 1975.
_ TABLE 1 STEAM LEAK ACCIDENTS
- CASE NO.
TOTAL H O PERCENT H O 2
2 (CHAP. 14)
INLEAKAGE, LB UNREACTED 1
340 88 2
39,100 99 3
6,240 97 4
2,160 60 5
15,740 94 6
8,080 89
- FROM TABLE 14.5-1 0F FSAR
RESPONSE TO NRC LEAD ITEM IV FOR Th0p REVIEW,
IV. Carbothermic Reduction of Th02 Although it is asserted in GLP-5640 that test rt alts indicate that "the carbonaceous reduction of Th02 is not a performance consideration up to 2000 C (sic.), tests conducted at Brookhaven National Laboratory (reported in BNL-NUREG-50785) indicate that the " rate of reaction of UO2 and Th02 with graphite becomes appreciable at 800 C to 1000 C and is very rapid above 2000 C."
Some analysis and discussion of these apparently contradictory data or interpretation of data is desirable, particularly regarding its potential safety significance for the six events identified in Section 3 of GLP-5640.
RESPONSE
Data on the kinetics and thermodynamic equilibrium of the reaction of thorium oxide and carbon to form thorium carbide are presented.
The results show that the reaction does not take place in intact particles and is highly unlikely in failed fuel particles at temperatures below about 1400 C due to the expected and measured concentration of C0 in the Fort St. Vrain primary circuit.
Even in the absence of CO, the rate of the reaction at 1300 C or below is inrneasur-ably slow, taking about 4 years to completion at 1200 C.
This result refutes the conclusion of BNL workers (l), who stated that the reaction initiates at 800 to 1000 C.
The reaction does occur at high temperatures such as would be attained during Design Basis Accident #1, but only after the particle coatings fail.
The other five events discussed in Section 3 of GLP-5640 would not be affected.
Based on the assumed FSAR models for fuel performance during DBA #1 (FSAR Figures D.1-13 and D.1-14) the fission product release rate from failed fuel is greater than the Th02-C reaction rate.
Thus, there would be no effect on fission product release during DBA #1.
Recent data are presented that show particle failure at 2100 C and above atten-dant to which is inmediate and total release of fission gases.
Carbothermic reduction would occur within a few minutes (in agreement with BNL data) but would not increase the rate of fission gas release.
This is the basis for the statement in GLP-5640 that the reaction would not affect fuel performance up to 2500 C.
. Introduction A chemical reaction which may occur in the FSV core with Th02 fuel is the con-version of fertile thorium oxide particles to thorium carbide by carbothermic reduction:
Th02 + 4C + ThC2 + 2C0 The reaction is controlled by both thermodynamic and kinetic considerations.
For example, the reaction cannot proceed at a given temperature as written if the surrounding C0 pressure is equal to or greater than the equilibrium C0 pressure which is dictated by the laws of thermodynamics.
It is this fact that prevents the reaction from occurring in intact Th02 fuel particles, where the C0 pressure will be higher than the equilibrium pressure due to reactions of oxygen release.i (during fissioning) with carbon.
The reaction can occur, how-ever, in failed particles at very high temperatures during Design Basis Acci-dent #1, where the reaction equilibrium C0 pressure would be higher than the C0 pressure in the primary circuit. Should the reaction occur, it could even-tually result in complete conversion of the oxide to the carbide, attendant to which is a ^ increase in kernel volume (based on lattice parameters of oxide and carbide).
This increase in kernel volume would easily be accommodated by the porosity of the surrounding buffer carbon, particularly since a volume of carbon equal to 1/2 the kernel volume is consumed in the reaction, thereby further increasing the void volume of the surrouncing carbon.
Thermqdynamic Equilibrium Using the thermodynamic functions provided by Stull and Prophet (Ref. 2),
Holley and Storms (Ref. 3), and Leitnaker (Ref. 4), the pressure of C0 in equilibrium with the Th-0-C system can be expressed as a function of the fuel temperature as follows:
logpy(atm)=-2.28x
+ 9.63 Figure 1 is a plot of the C0 equilibrium pressure calculated by the above equation versus reciprocal temperature, At 1673K, the equilibrium C0 pressure is 10-2 kPa (10-4 atm), which is equal to 2 ppmv at a reactor pressure of 47 atm.
This means that if the reactor C0 pressure is 2 ppmy, the carbothermic
. reaction can only occur at temperatures above 1673K (1400 C).
Because C0 is a persistent impurity in the HTGR coolant, it is clear that this reaction will not occur at normal operating temperatures.
Plotted alongside the calculated line in Figure 1 are experimentally determined values obtained at General Atomic (Ref. 5) and from the literature (Refs. 6 through 8).
The General Atomic exper-imental procedure utilized a small loop in which helium at 0.25 MPa (2.5 atm) was continually circulated.
The gas pass.ed through a resistance-heated graphite tube furnace which contained the oxide-carbon specimen and then through a sampl-ing valve of a Loenco gas chromatograph, after which it returned to the furnace.
In this way, selected samples were heated and the reaction product gases were quantitatively analyzed directly at various times throughout the test.
The samples used in the equilibrium studies were mixtures of thoria and graphite powders (Th0 /C = 1/7 molar ratio) encased in a graphite crucible.
Temperatures 2
from 1813 to 2013K have been investigated.
The reaction was considered "at equilibrium" when no change in C0 pressure was noticed after 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Some of the runs were carried well past this point.
Figure 2 shows the C0 concentration data plotted versus time for five runs using two different powdered specimens. The final C0 pressures obtained at each temperature are plotted in the Arrhenios plot of Figure 1.
The data in Figure 2 indicate that the carbothermic reaction rate is a function of C0 pres-sure, which is consistent with observations by Imai (Ref.10).
Reaction Rate Studies The reaction rate of powdered samples as a function of C0 pressure can be ob-tained from the data in Figure 2.
More important, however, is the reaction rate of Th02 microspheres at conditions where C0 pressure is not inhibiting.
Reaction rate data for 500 pm Th02 fuel particles are shown in Figure 3, where rate (frac /
hr) is plotted vs 1/T.
Two types of tests were performed at General Atomic, both in the absence of CO, and as represented by the open symbols in Figure 2.
The closed symbol data is that of Hamner, Ref. 9, shown for comparison.
In the General Atomic tests, fuel particle samples (either laser drilled BISO coated Th02 particles or bare kernel Th02 mixed with graphite powder) were annealed for various times in either vacuum or flowing He conditions.
Reaction rates were determined by a weight loss (aw) method in one test series.
In the second
. test' series, rates were determined by analysis of carbide using controlled hy-drolysis at 100 C in which the amount of hydrolysis product ethane was measured.
The fraction of conversion to carbide (either by weight loss or carbide methods)
-I wa5 divided by the anneal time to obtain average reaction rates, hr This method of determining reaction rate is obviously conservative, because it ne-glects the slow down of the reaction due to diffusion of carbon through the ThC2 product which forms around the Th02 microspheres.
In Figure 3, the apparent activation energy of these data is 92 kcal/mol, which can be compared with the activation energies obtained for mixed powders of 90 kcal/mol (Imai, Ref.10) and 75 kcal/mol (Kanno et at., Ref.11).
The data shown in Figure 3 indicate that at high temperatures the reaction is quite rapid, with complete conversion occurring in about 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> at 2000 C.
At temperatures below 1300 C the rate of reaction is immeasurably low.
For example, at 1265 C, the time for complete conversion is 104 hours0.0012 days <br />0.0289 hours <br />1.719577e-4 weeks <br />3.9572e-5 months <br /> or 1.3 years. At 1200 C the time would be about 4 years.
It is clear that the rate of reaction at 1000 C or below would be virtually zero.
This conclusion is in agreement with the work of Kanno, et al. (Ref.11), who routinely heated mixed thoria and carbon powders at 1000 C (with no measurable reaction occurring) in preparation for higher temperature tests.
The results obtained at high temperature are in agreement with tne results ob-tained at BNL (Ref.1), where it was concluded that the reaction is very rapid above 2000 C.
The BNL workers further concluded that the reaction become appreciable at 800 to 1000 C; obviously, neither GA test data nor the data of Kanno (Ref.11) agree with this part of BNL's conclusion.
In the BNL tests dense cylinders of Th02 or U02 were heated in graphite crucibles in air to 2500 C in 1/2 hour or less.
(For DBA #1, no air ingress occurs.)
In the case of Th02 a flame and a loud pop were observed at 2500 C.
In the case of U02 the phenomena were observed at 2300 C.
It was explained that the loud noise and flame were caused by a rapid reaction of C0 and air.
Subsequent examina-tion of the samples revealed some melting and carbide formation.
It is GA's opinion that the BNL tests qualitatively showed that the reaction is, indeed, rapid at high temperatures.
It is impossible to extrapolate these results to lower temperatures, however, since no quantitative rate data vs temperature were obtained to permit such an extrapolation.
The GA data base, on the other
. hand, is quantitative and extends over a temperature range (see Figure 2) which allows reasonable extrapolations on either side of the measured range.
Fuel Performance As previously stated, carbothermic reduction does not affect fuel performance except during DBA #1 in the small fraction of fuel that is initially failed.
The expected average failure fraction of Th02 particles is less than 1%.
This small amount of fuel could release fission gases at an enhanced rate due to the conversion reaction. The overall effect would be negligible, however, because the total inventory of fission gases in the failed fuel is small compared to the total released during the accident.
Fission product release from fuel failing during the heatup is not affected whether the FSAR models for failure and release are assumed or whether more recent experimental data is considered.
Figure D.1-13 of the FSAR shows fuel failure occurring in the range of 1585 C-1725 C.
Examination of Figure D.1-12 at these temperatures indicates that fission product release rates from failed fuel are higher than the Th0 -C reaction rates shown in Figure 3.
Hence, the 2
fission products will be essentially entirely released before carburization occurs.
The results of core heatup simulation tests indicate that the bulk of fission gas release occurs as a result of particle failure which occurs at 32100 C.
This is illustrated in Figure 4, which is a composite of 7 temperature ramp experiments on TRISO coated Th02 particles having 4-8% FIMA burnup and is similar to Figure 2 in the response to Lead Item II.
In each test, 50 to 100 particles were used.
In Figure 4, Kr-85 fractional release vs. temperature for two heat-ing times is plotted. The fractional Kr-85 release can be reasonably assumed to be equal to fraction of particles failed, because previous work has shown that at 2000 C virtually all noble gas is released from dense kernel material (Ref. 12).
Coating failures occurring at 2100 or higher would cause immediate and total release of fission gas.
Since carbothermic reduction would occur some time af ter particle failure (at 2l00 C the time to complete reduction is approximately 1/2 hour, see Figure 3), it is concluded that virtually no appreciable change in the rate of fission gas release would occur.
References 1.
P. Soo, C. A. Sastre and D. G. Schweitzer, " Experimental Studies of Core Heatup Phe'nomena," BNL Reactor Safety Quarterly Report.
2.
Stull, D. R., and H. Prophet, eds. JANEF Themochemical Tables, 2nd Ed., NSRDS-NBS 37, June 1971.
3.
Holley, C. E., and E. K. Stoms, " Actinide Carbides, A Review of Thermodynamic Properties," in Proceedings of the International Atomic Energy Association Symposium on Themodynamics,1967.
4.
,dfrey, T. G., J. A. Wooley, and J. M. Wooley, and J. M. Leitnaker, "Themodynamic Functions of Nuclear Materials: UC, UC, U0, Th0 '
2 2
2 and UN, " ERDA Report ORNL-TM-1596 (Revised), Oak Ridge National Laboratory, December 1966.
5.
HTGR Fuels and Core Development Program Quarterly Progress Report for the Period Ending Nov. 30, 1977, US DOE Report GA-A14744.
6.
Prescott, C. H., Jr., and W. B. Hincke, "The High-Temperature Equilibrium Between Thorium 0xide and Carbon," J. Amer. Chem. Soc.
49, 2744 (1927).
7.
Pialoux, A., and Mme. J. Zaug, " Etude de la Carboreduction de la Thorine a l' Aide de la Diffractometrie de Rayons X a Haute Temperature sous Pression Controlee, Puis Sous Vide," J. Nucl. Mater. 6]l,131 (1976).
8.
Lorenz, R., H. L. Scherff, and N. Toussaint, "The Carbothemic Reduction of Protactinium Pentoxide and First Results on Protactinium Carbide," J. Inorg. Nucl. Chem. 31,2381,(1969).
9.
Hamner, R.
L., R. L. Pilloton, and T. M. Kegley, "A Method for Pre-paring Dense Spherical Particles of Thorium and Thorium-Uranium Dicarbides," Nucl. Appl. 3, 287 (1967).
10.
Imai, H., S. Hosaka, and K. Naito, " Preparation of Thorium Carbides,"
J. Amer. Ceram. Soc., 50_, 308 (1967).
11.
Kanno, M., et al., "Carbothermic Reduction of Thorium 0xide," J. Nucl.
Sci. Tech. 9_,97(1972).
12.
Myers, B. F., et al., "The Behavior of Fission Product Gases in HTGR Fuel Material," ERDA Report GA-A13723, General Atomic Company, October 1977.
i t
i
'O f.
.}
.I i
.g
._j
..y.
I
- O-E I
i
-2 10
~~~i-~~~[F.~2U T T.T
.O'
~:
- - I - -.
100
..t p
t i.. a. _ - _
CO (atm)
--- - O
-i --
j
-. p _. - _.
PPM CO @
j 700 psia
-3 10
_2 : _. 3__...
2
...._...-._..f._-. !~._.-__.
g ;...._..
__.._._i 10
..._.4
_.g
.{
.-j.___.-
._.._i_...
._...w.
i t
t 10
-._O__GA Da_ta l
l i
2
~-4 Pialoux
--- E Loren z
' ' ~ ~ ~~
y
_._______;._.._. 7
~'- $ Prescott-
- --- j - --- ;f ~
~ ~~ ~ j ~ ~
Calculated
~
.g!..
1,
-5 1700 1600.
1500 1400 C 10 i
i I,
4.8 5.0 5.2 5.4 5.6 5.8 6.0 6.2 10 /T, K Fig. 1.
Carbon Monoxide Equilibrium Partial Pressure Versus Reciprocal Temperature for the Th0 -C System 2 2
I 0
10
@ 2013 K 4
O O1973K O
f O 1908 K O
1853 K ha
^
^
^
2 6
b E
10~I f
P 1813 K O
V He PRESSURE = 0.253 MPa (2.5 atm)
SYSTEM VOLUME = 7.0 liters j
GA FLOW RATE = 500 cm / min 3
i 10-2 I
I I
I t
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 TIME (h)
Fig.
2.
Carbon Monoxide Partial Pressure Versus Time for the Th0 -C 2
System
i
}__._ t_.
i_.
i 1'
l
...--t-
._1_.
L
.4 i
r_. __.
l
- r- -- - -
l----
i i
-____.l._
1 1.0
- rj.T.
- d,;. ~
l..
...;. :J_ [
.'~ i.n_ i 2 1
.l__
._l
._._J..____._&.__..,
.. _._ 4....
.._.._.. f.
_ p..
.. _ _.. -._ ___. & l _.
.__...4___-_hm.._.
l l
I i
i
-l 10
-1
- - - t--
-.}~
r_ -..e. _ _ T _. - _ _ _ p.
- r - ~ --
Hr r----
..-_ r..- _.
4-
...._.p.__._
o
-t 7 -< -
i j
i
-2 i
.___i__._~~._
... j~ E. _..
_____ ___[
.__3__.._'~'. _7.* o,.
' ' ~
___-.._3,_.____ _ _.. _
_..___7.__.._..
. 7 7.... _ _
_. 7. _ _.
p__.
- L____
_. p.
l
'.4._.
i l
i
. _ t. _.
L.. _..
O: GA Data, Aw method
_ _.0 GA Data, carbide metOd -
-3 l
10 Z &j._ Hanner._
.,....._. _ ~ _., _.... _ _ 'o j
____._.j.._
.. j
_..q._._..___._..-_q._-__,
__,_.__.i...__
.___+_...t_-_.
.p._
l j
6 i
I !
I t
.__.._.t.
L.... :
8
-4 2400,C 2000 1800- 1600 1400 1200 10 3
4 5
6 7
10 /T K Fig. 3.
Rate of Carburization of Th0 Micr spheres 2
sem iT g
n i
0 t
0 a
/
I 5
e l
/
2 l
o
\\
w T
r N
0 o
0 f
3 2
eru t
)
a C
r 0
ep 0
(
1 m
I 2
e E
T R
U s
T v
A n
0 R
o A"
0 E
i 9
1 P
tc M
a E
r
\\
T F
\\
e
)
0 s
C 0
I a
7 e
0 1
l 0
e 5
R 5
e 2 m
i o 1
5 7
8 T
t 0
0 r
g C 1
5 K
n i 0 7
1 t 0 6
a 1
4 e
1 H
(
.g 0
i 0
F 3
0 8
4 2
01 1
0 0
0
RESPONSE TO NRC LEAD ITEM VI FOR Th0 REVIEW 2
VI. Th0 Kernel Migration 3
The most recent work on Th0,; kernel migration (C. L. Smith, Nucl.
Tech. 35, pp. 403-412) indicates that the migration is characterized by an incubation period that decreases with increasing temperature and burnup. What assurances are there then that the out-of-pile Th0 kernel migration measurements are not influenced (to an unknown p
degree) by removal from an irradiation out-of-pile and that the measure-ments out-of-pile do not provide false (low) indications of the rate in-reactor? Are there any decay processes that might affect the out-of-pile measurements, considering the period of time between the irradiation of the particles and the testing of the particles in an out-of-pile facility?
RESPONSE
Summary There are no known decay processes that would affect out-of-pile Th0 e
2 kernel migration data, e Predicted Th0 kernel migration distances, based on out-of-pile, post-2 irradiation test data (including the incubation period), are greater than or equal to migration distances observed during irradiation.
e Th0 kernel migration rates are independent of coating type (BISO or 2
TRIS 0).
Importance of Decay Processes Analyses of in-pile and out-of-pile data (Ref. 1, 2) suggest that the rate of Th0 kernel migration is independent of kernel burnup for burnups 2
exceeding 0.24% FIMA and not, therefore, related to fission product concen-trations. Although a great mahrity of the available data have been obtained from irradiated test samples (temperature range 1150 to 1700 C),
Th0 kernel migration has been observed in unirradiated samples heated out 2
of pile at temperatures exceeding 1750 C (Ref. 2). These observations suggest that the initiation of Th0 kernel migration is influenced more by 2
the presence of C0 than of fission products.
Since the C0 content of BISO or TRISO coated Th0 is not influenced in any significant way by decay 2
processes, it is concluded that there are no decay processes that would in-fluence Th02 kernel migration data collected in post-irradiation, out-of-pile heating tests.
Comparison of In-Pile and Out-of-Pile Th0p Kernel Migration Data A total of 107 BISO Th02 and 106 TRIS 0 Th02 particles have been tested in accelerated and real-time irradiation capsules (Ref. 3).
Kernel migration rates indicated by the in-pile data have been less than or equal to rates sug-gested by out-of-pile test results.
Specific examples are given below.
Th02 kernel migtation distances observed during post-irradiation examination (PIE) of FTE (Fuel Test Element) and RTE (Recycle Test Element) experiments conducted in the Peach Bottom HTGR (Refs. 4, 5, 6, 7) are given in Table 1.
Approximately 100 fertile particles were examined in metallographic cross sec-
. tions of each fuel rod. Although no migration was observed, in-pile migration distances are shown to be less than 5 pm to be consistent with the limit of resolution for metallographic examinations.
Two values for predicted kernel migration distances are shown.
Both are based on out-of-pile migration rate data.
The first assumes that migration began at time zero; the second assumes that kernel migration did not begin until the incubation period discussed in Ref. I was exceeded. No migration (<5pm) was observed even in those examples where predicted migration distances were much greater than the metallographic detection limit of Sum.
RTE and FTE data show, therefore, that predicted mi-gration distances which are based on out-of-pile kernel migration data and account for the incubation period are greater than or equal to observed in-pile migration distances.
Predicted and observed Th02 migration results from accelerated irradiation capsules P13R and P13S (Ref. 8) are shown in Table 2.
Approximately 20 parti-cles were observed in metallographic cross sections of each fuel rod.
No kernel migration was observed; however, values of less than 5 pm are shown for reasons discussed above. Although predicted migration distances were less than could be observed, the example of predicted migration 2 to 3 times that observed shows that predicted migration distance is conservative.
The range of'Th02 kernel migration coefficients (KMC) determined as a function of temperature from out-of-pile testing is compared with in-pile KMC values
determined from H HRB, and fit capsule tests (Refs.1, 3)- in Figure 1.
The HRB and HT data were not corrected for the incubation period, since the correction would be small (14 to 24%).
The H capsule data were not corrected because of uncertainties in operating temperatures.
The data in Figure 1 show that in the range of nomal FSV core operating temperatures, in-pile and out-of-pile KMC values are similar.
The data in Figure 1 also show that Th0 KMC values are independent of coating type (BISO vs. TRIS 0), which is 2
consistent with earlier observations (Ref. 2).
References 1)
C. L. Smith, " Migration of Th0 Kernels under Influence of a Temperature 2
Gradient," flucl. Tech. 3_5, pp 403-412,1977.
2)
T. B. Lindemer and R. L. Pearson, "Keynel Migration for HTGR Fuels from the System Th-U-Pu-C-0-N," J. Am. Cer. Soc 60_, pp 5-14,1977.
3)
- 0. M. Stansfield, et. al., " Performance of Th0 in HTGR Fuel Particles,"
2 GA-A14745, March 1978.
4)
T. N. Tugs and E. L. Long, " Post-Irradiation Examination of RTEs from the Peach Bottom Reactor," 0RNL-5422, December 1978.
5)
J. F. Holzgraf, et. al., " Post-Irradiation Examination and Evaluation of Peach Bottom Fuel Test Elements FTE i4 and FTE-15," GA-A13944, February 1979.
6)
C. F. Wallroth, et. al., " Post-Irradiation Examination and Evaluation of Peach Bottom Fuel Test Element FTE-4," GA-A13452, July 1977.
7)
C. F. Wallroth, et. al., " Post-Irradiation Examination and Evaluation of Peach Bottom Fuel Test Element FTE-6," GA-A13943, September 1977.
8)
C. B. Scott, D. P. Harmon, and J. F. Holzgraf, " Post-Irradiation Examination of Capsules P13R and P13S, " GA-A13827, October 8,1976.
TEMPERATURE (O )
C 1800 1700 1600 1500 1400 1300 1200
-8
.i 6
4 4
i i
i 1
~
g RANGE FOR KMC ATTHE N
90?5 CONFIDENCE LEVEL (out-of-pile data) g N
\\
10-8 O
\\
N
\\
O N
2 S
\\g
[
g
- 0 i
s
~
O d
6 0m 8
\\
z O
N O
8
\\
E!
O A O*
N O
$ gg-10 s
N O
%\\
h
\\
g g
O m
h D
Sg g
s 10"11 u
oa h
IN-PILE DATA g
97-U$
BURNUP
\\
g CAPSULE
(% FIMA)
COATING O H-1-10 0.37 BISO 10-12 O H-2-10 0.32 BISO O H-2-11 0.24 BISO T
A HRB3-3A 14.0 BISO V HRB4-3A 14.5 BISO G HT-31 8.2 TRISO 1
I f
f 4.5 5.0 5.5 6.0 6.5 7.0 4
10 /T(K)
FIGURE 1 Co=parision of in-pile and out-of-pile rh0 IO!C data g
e 6
TABLE 1 COMPARISON OF OBSERVED AND PREDICTED Th02 KERNFL MIGRATION DISTANCES FOR FTE AND RTE TEST SAMPLES MIGRATION (JJM)
FUEL MAXIMUM SURFACE PREDICTED ROD CENTERLINE TEMP.
5 CAPSULE NO.
TEMP. (OC)
( C)
FIMA UNCORRECTED (")
CORRECTED OBSERVED (
4-2-1-7 1174 994 0.7 1.0 0
<5 4-2-7 1174 994 0.7 1.0 0
<5 6-2-1-7 1210 1030 1.5 14.6 0
<5 FTE 6-2-2-7 1210 1030 1.5 14.6 0
<5 14-2-3-6 1383 1203 0.4 28.7 26.6
<5 15-2-3-9 1264 1084 1.0 8.4 4.0
<5 5-3-3-6 1183 1003 2.7 4.4 1.5
<5 1-4-3-6 1399 1210 1.5 24.7 23.4
<5 RTE l-5-3-1 1292 1112 1.4 14.9 13.1
<5 1-4-8-6 1455 1275 1.5 49.4 47.1
<5 (a)
COllitECTED FOR INCUBATION PERIOD 011 flME TO lEAC11 2.0% FIMA, WIIICIIEVER 03HES FIRST, (b)
LIMIT OF METALLOGRAPIIIC RESOLUTION = 5 FM e
TABLE 2 COMPARISON OF OBSERVED AND PREDICTED Th0p KERNEL MIGRATION DISTANCES FOR P13R and P13S TEST SAMPLES TiiO KERNEL ffIGRATION 2
h! AX. PI(EDICTED (pf t)
FUEL ROD UNColtitECTED COltitECTEDla )
DATA CENTERLINF.
SURFACE FOR FOR RETRIEVAL FIntA Tg!!P.
TE!!P.
INCUBATION INCUBATION OBSERVED (b)
CAPSULE NUh1BER (FERTILE)
( C)
( C)
PERIOD PERIOD (p.\\f )
P13R 19-7 2.4 1225 1120 2.1 0.3
<5 P13S 18-5 2.4 1335 1235 0.9 4.3
<5 P13S 19-5 2.2 1325 1230 8.4 3.1
<5 P13S 28-13 0.8 1050 995 0.1 0
<5 P13R 1G-5 3.1 1285 1180 5.1 1.8
<5 P13S 16-G 2.8 1375 1285 15.6 10.2
<5 (a)
CORRECTED FOR INCUBATION PERIOD OR TIME TO 2% FIMA, WHICIIEVER COMES FIRST (b)
LIMIT OF METALLOGRAPIIIC RESOLUTION
RESPONSE TO NRC LEAD ITEM VII FOR Th02 REVIEW VII.
sic-Fission Product Interaction in TRIS 0-coated Th02 Particles A faiely extensive data base exists which shows that the pressure vessel, sic-tensile stress correlation breaks down at temperatures >1500 - 1600 C.
This is believed to be due to a chemical interaction between the sic layer
,and fission products that have migrated out of the kernel to the inner surface of the sic layer. The data that address this phenomenon in TRIS 0 Th02 particles should be discussed, and 15e fraction of failures due to this mechanism should be predicted as a function of burnup and temperature.
RESPONSE
Post irradiation exan.ination of TRIS 0-coated Th02 particles irradiated at average temperatures up to 1500 C has shown no evidence of sic-fission pro-duct interaction under normal operating conditions.
The bulk of the data un TRIS 0 Th02 Performance at temperature: above 1600"C has been obtained in the core heatup simulation testing program which, for Th02,is described in the response to Lead Item II.