ML19208B061
| ML19208B061 | |
| Person / Time | |
|---|---|
| Site: | Fort Saint Vrain  |
| Issue date: | 09/30/1978 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19208B060 | List: |
| References | |
| NUDOCS 7909190008 | |
| Download: ML19208B061 (20) | |
Text
~ - -.
g u
J
,r ENCLOSURE 1
)
TOPIC.ALREPO{CVALUATION Report Number: GA-LTR-23 2
Report Titic:
Safety Analysis Report: Use of UC Fissile Fuel Particles
~
2 in Fort St. Vrain Fuel Elements Report Date:
September 1978 Originating OrOnization: General Atomic Company Reviewed By:
Core Performance Branch, Division of Systems Safety, March 1979 y
Sudlary of Topical Report This topical report was submitted by General Atomic' Company (GA) via letter (Ref.1)' in November 1978.
In that letter a were requested to rniew the report and concur with GA's opinion thac the substitution of uranium carbide fissile fuel particles in the Fort St. Vrain reactor for the thorium / uranium carbide particles,cerently in use is acceptable. The letter stated that GA plans to use uranium carbide fissile fuel particles for reload fuel element manufacturing as soon as practicable after receiving our approval (although the final detemination of reload and schedule for substitution is the respon-sibility of Public Service Company of Colorado (PSC)).
The topical report is divided into three major sections: (1) a perfomance analysis of' UC fuel, encompassing the nuclear, themal, and fission prod;ct 2
fuel particles on design, (2) a safety analysis of the effect of the UC2 postulated-accident sequences considered in the FSV FSAR, and (3) a summary of the UC2 particle design data base.
In general, the approach used was to compare the ana',:ical results based on the reference (Th/U)C2 fissile par-as the fissile' ticles with the results obtained with the assumption of U,C2 1liO 7909190 CO h c% 6
a'
?
material. The objective was to demonstrate that UC2 particles would perfonn comparably to the reference (Th/U)C2 particles and that the use of UC2 would result in negligi51e changes in the nuclear and themal behavior of the core and would not result in reduced safety margins or reliability compared to the reference (Th/U)C fissile particle core. Thus, the report is intended to 2
demonstrate that fission product release will be unchanged by the substitution of UC fissile particles for (Th/U)C2 particles.
2 The core nuclear perfomance analysis addresses fuel loading and excess reac-tivity, power distribution, fuel burnup and expnsure, shutdown margin, and md withdrawal accidents. The rod withdrawal accident is afforded special atten-tion in the topical report because control rod withdrawal is identified in the FSV FSAR as the worst-case reactivity initiated accident; control rod ejection accidents are not considered credible for the FSV reactor, and the consequences of a rod ejection accident are, therefore, not analyzed.
The effects that UC kernel will have on (a) fuel rod thernal conductivity, 2
and (b) kernel migration rates are addressed in the report section on the core thennal analysis. Fission product release is discussed in a separate sub-section of the report. The substitution of UC kernels for (Th/U)C kernels 2
2 in the fissile fuel particles is, according to the report, not expected to result in (1) a reduction in fuel red thennal conductivity or (2) increases in kernel migration rates or fission product release.
The topical report's " safety analysis" section is used to examine events and accidents previously analyzed in Chapter XIV of the FSV FSAR to detennine if the substitution of UC fuel kernels for those with (Th/U)C kernels in all or 2
2
!!b b lk$
, part of the FSV core could alter the consequences of postulatad accidents.
Events given a re-examination include loss of nonnal shutdown cooling (limit-ing case: cooldo*wn on one firewater-driven circulator), moisture inleakage, pennanent loss of forced circulation [" Design Basis Accident (DBA) No.1"],
rapid depressurization/ blowdown ["DBA No. 2"], and the rod withdrawal accident mentioned above.
It is concluded in the report that the existing FSAR results of accident analyses conservatively bound any perturbations resulting from the introduction of UC fuel particles.
2 Summary of Regulatory Evaluation Evaluation of Recent Test Data - In an attempt to expedite the reviews of both GA-LTR-23 and a companion topical report on Th0 fertile fuel particles 2
(GLP-5640), we generated a list of lead review items (Ref. 2) which was trans-mitted to General Atomic via letter (Ref. 3). Following the transmission of that letter, a meeting was held between the cognizant NRC topical report reviewer and GA personnel (Ref. 4).
Presentations on the lead items were made by the GA representatives. Subsequently, a fonnal response to " Lead Item VIII" (the only item that pertained to the UC2 review) was submitted by letter dated April 9,1979 (Ref. 5). Our inquiries, GA's responses, and our topical report evaluation will be incorporated into GA-LTR-23 via amendment.
Our review focussed primarily on the advances that have been made in the state-of-the-art regarding TRISO UC2 particle testing and fuel failure model develo;xnent in the period since the licensing activity on large HTGR concepts has been curtailed (from late 1975 to the present).
In late 1976, NRC review activities on HTGR fuel culminated in two documents:
(a) a NUREG report 9hb I4
., entitled " Evaluation of High Temeerature Gas Cooled Reactor Fuel Particle Coating Failure Models and Data", NUREG-0111, (Ref. 6), and (b) our input to the "I'nterim Safety Evaluation Report", or "ISER" (Ref. 7), on the General Atomic Standard Safety Analysis Report," or "GASSAR" (Ref. 8). Because the HTGR fuel review studies that we had performed during the period covered by the above two NRC documents had been comprehensive, whereas our HTGR licensing activity during the succeeding period has been virtually nil, we believed that a properly-focussed review of a safety analysis on the use of UC 2 fissile fuel particles should center on the advances made in the state-of-the-art during the latter time period.
As stated in NUREG-0111, irradiation test data on TRISO-coated
- UC was tco 2
sparse to pennit significant statistical analyses of failure probability.
We, therefore, proposed modifications to the GA fuel failure models that had been oresented in an earlier topical report (Ref. 9). These modifications were in the fann of added conservatisms that were intended to compensate for uncertainties in the data and a lack of complete understanding of the inter-relationships of coating fabrication, structure, and performance in reactor.
The major deficiency in the data base was an overall lack of statistically significant irradiation test data on fuel that could be characterized as the so-called " reference" design. Data at temperatures above normal operation were particularly sparse. " Lead Item VIII" (Ref. 2), therefore, was intended to focus on any new evidence that might exist in support of tne credicted failure rates for each failure mechanism identified in reference 9.
General
The "TRISO UC., fissile fuel particle for FSV has a dense, approximately 200 um diameter, DC kernel of 93" enriched uranium, which is coated with an 2
inner, low-density, approximately 100 um thick, buffer pyrocarbon (PYC) layer and an outer composite coating of sic (approximately 35 um thick),
sandwiched between two layers of dense oyrocarbon (35 to 40 um thick)...
7 9%
1 4)
o Atomic's formal response to Lead Item VIII (Ref. 5) was followed by our request for ' addit,i,onal information (Ref.10). That request was responded to in reference 11.
The 1ew evidence (that is, data that did not exist at the time of NUREG-0111) presented in support of the TRISO UC2 particle failure predictions consisted of (1) core heatup simulation test (CHST) results (Ref.12,13), (2) results from large-scale irradiation tests in the French Spitfire Loco Experiment SSL-2 (Ref.14), and (3) data obtained from tests conducted in cell 2 of capsule GF-4, which was irradiated in the Silee reactor at Grenoble (Ref.15).
As indicated in Fig.1 of reference 5, the irradiation test experience for 5
reference TRISO coated UC fuel includes 3.16x10 particles tested at temoera-2 25 2
tures between 885 to 1240"C and fluences of 3.5 to 12.2x10 n/m. Whereas at the time of issuance of NUREG-Olli no TRISO coated UC2 particles had been irradiated to the peak fissile burnup for 6-year-old fuel (75% FIMA*), now 3
'aporoximately 9x10 UC2 particles have been irradiated to burnups yJ7% FIMA, 5
and 2.5x10 UC2 particles have been irradiated to the mean fissile burnuo in 6-year-old fuel ('71% FIMA). Based on an emoirically determined pressure vessel failure criterion of 231 MPA (33,500 psi) for the tensile stress in the sic coating layer fr. TRISO UC2 particles (Refs,11,16), the expected failure fractions for TRISO coated UC and (Th/U)C were stated to be 0.002 and 0.005, 2
2 respectively. Since the expected failure fraction due to oressure vessel failure under nonnal steady state conditions is* less for UC than for (Th/U)C 2
2 particles, this comparison provides supoort for GA's assertion that the substitution of UC for(Th/U)C is acceptable.
2 2
- FIMA - Fissions of initial metal atoms.
p / ()
\\
ut
' The SSL-2 irradiation test contained 85,200 UC2 particles tested in 27 reference type HTGR fuel rods. The fuel rods were assembled in a graphite body which simulated the design geometry of an HTGR fuel element. Thus, this test constituted a large scale proof test of UC2 performance. Observed Kr-85m R/B (rate of release / rate of birth) values were substantially less than predicted, which implied that the in-pile failure probability of UC 2 may be even less than that determined by the TRISO coated particle stress model.
A comparison of predicted and measured Kr-85m R/Bs was also made with the data obtained fecm the cell 2, capsule GF-4 tests, but in this case the predicted Kr-85m release values were in agreement with observations. In the latter test 7000 UC2 particles were irradiated to 75.5% FIMA at 1100*C, while in the SSL-2 test the burnup and peak fuel temperatures were 72% and 1200-1300'C, respectively. These steady-state irradiations extended the data base on the reference TRISO UC2 particle, and they provice assurance that the performance of the particles will meet or exceed predictions over the range of normal operational conditions anticipated in the FSV reactor.
In addition to normal steady state operation, HTGR fuel performance must be evaluated in terms of potential transient and accident conditions. As noted earlier, at the time of issuance of NUREG-0111 data at temperatures above normal operation were particularly sparse. In an attempt to remedy this deficiency, General Atomic is performing core heatup simulation tests (CHSTs).
Groups of 50 to 200 particles, previously irradiated to burnups of 23 to 60%,
are being heated from ~1100* to '2500*C over periods of 28, 30, or 80 hours9.259259e-4 days <br />0.0222 hours <br />1.322751e-4 weeks <br />3.044e-5 months <br />.
Coating failure is detected indirectly by monitoring Kr-85 activity. The range of test conditions chosen for this program (see Refs.12,13) approximates the hypothetical core heatup events considered in the reactor licensing and 145 r/
o500
7 siting applications; for example, as shown by the thermal analysis in Appendix D of the FSV FSAR, the minimum time required for a small fraction of the fuel to reach 2500*C d,uring " Design Basis Accident No-- 1" (D8Ml), which is a permanent loss of forced circulation, would be 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br />. To date, four CHSTs, containing a total of 600 TRISO UC2 particles, have been conducted.
Based on a comparison of observed Kr-85 release fractions (on both UC and 2
(Th/U)C particles) with those predicted using FSV fuel failure and fission 2
product release assumptions, the results indicate that (a) the perfomance of TRISO UC and TRISO (Th/U)C would be similar during a core heatup and (b) 2 2
predicted release fractions are greater than observed for both types of fuel.
These results provide support for GA's assertion that replacing the present (Th/U)C fuel with UC fuel would have no effect on FSV reactor safety margins.
2 2
In sumary, our evaluation of new test data that were generated on TRISO coated UC fissile particles in the period follcwing the issuance of NUREG-0111 2
indicates that the data base on the reference UC fuel design has been extended 2
significantly and that the irradiation perfomance has been good. Fission gas releases under both normal steady state as well as simulated transient conditions have been.as good as or better than predicted releases. Although these test results have not been directly correlated with the individual fuel failure mechanisms and failure rate predictions discussed in GA-Al2971 (Ref. 9) or NUREG-0111 (Ref. 61, they do constitute evidence of the acceptable performance of TRISO UC fuel particles.
2 Evaluation of performance Analysis - As noted earlier, the " performance analysis" section of the topical report addressed the core nuclear, core thermal, and fission product release aspects of the fu 1 design. The core neutronics
$6h kN
parameters change only very slightly with the separation of the uranium and tnorium into different particles. The overall thorium to uranium ratio does not change--thus the flux spectrum, rod worths, and power distributions do not change. Core reactivity changes due to slightly different self shielding factors for uranium and thorium are less than 0.2 percent in effective multip-lication factor. Shutdown margins are similarly only very slightly affected.
The one nuclear effect that requires examination is the delay in Doppler feedback due to the time required for heat transfer between the fissile and fertile particles. Of the transients and accidents for which this is important the rod withdrawal accident is limiting. For this accident the power rise is sufficiently slow that the predicted consequences are not significantly altered. For other transients the effect is either similarly negligible or acts to reduce the consequences.
On the basis of the above considerations, we conclude that the use of uranium carbide fissile particles is acceptable with respect to the core neutronics of the Fort St. Vrain reac+wr.
The core themal analysis section of the topical report addressed the effects of UC fuel kernels on fuel rod thermal conductivity and the " amoeba effect" 2
(kernel migration). Fuel rod thennal conductivity has been the subject of earlier reviews (Ref. 7,17) and there are analyses (Ref.18) that support General Atomic's assertion that fuel rod thermal conductivity is not signifi-cantly affected by the choice of fissile kernel. Thus, we agree with GA's conclusion that the value of 4.0 Btu /h-ft *F (6.9W/m *K) used in the FSAR thernal analysis remains acceptable. Since the kernel migration coefficient (KMC) of UC is essentially the same as for (Th/U)C, we agree with GA that 2
2 the bases of the Core Thermal Safety Limit are not exceeded with UC fissile 2
fuel.
,, ~j f ^Q]
's,
.g.
Wii.h rtgard to fission product release, we agree with GA that there should be no change in gaseous fission product release to the coolant because (a) there is no expertinental, evidence of a difference in gaseous or metallic fission product release characteristics between the (Th/U)C2 and UC2 particles and (b) the R/Bs from the two particles are the same. Thus, the " Design" circulat-ing inventories (FSAR Table 3.7-1) should remain as applicable as before as source tems in accident analyses.
Evaluation of Safety Analysis - As noted earlier, the topical report's " safety analysis" section contained a review of the postulated accidents previously analyzed in Chapter 14 of the FSV FSAR.
The purpose of the review was to detemine if the substitution of TRISO fissile particles with UC kernels 2
for those with (Th/U)C kernels in all or part of the FSV core could alter the 2
consequences of postulated accidents in such a way that the worst case con-ditions, previously defined during the FSAR review, were exceeded.
Five events were given a detailed examination in the topical report: (1) rod withdrawal accidents (RWA), (2) loss of nomal shutdown cooling, (3) moisture inleakage, (4) pemanent loss of forced circulation (DBA#1), and (5) rapid de-pressurization / blowdown (DBAf2). Of these five types of events, the first one, rod withdrawal, was addressed in the preceding subsection of this report evaluation, where it was concluded that the neutronic consequences of an RWA will not be changed significantly. This is also true of the events involving loss of nomal shutdown cooling, of which the limiting case is cooling with one circulator driven by the fire-water system. Since the change in kernel type from (Th/U)C does not affect the thernal properties of the particles, 2
this type of accident is unaffected, t. (o (,
h a
/v
.. ~.. ~ -
. Of the moisture ingress cases treated in the FSAR, a steam generator subheader rupture, compounded by concurrent failure of the moisture monitor system and dumping of the wrgng (non-leaking) steam loop, was stated to have the greatest potential for graphite oxidation and fuel hydrolysis in the shortest time period following the accident.
In analyzing the potential effect of the substitution of UC for (Th/U)C2 on the consequences of this event, three phenomena were 2
addressed:
(1) steam-graphite reaction, (2) hydrolysis of failed fuel, (3) fission product release from oxidized graphite.
The steam-graphite reaction was analyzed in tenns of the potential catalytic effect of barium and strontium reaction rate. The steam-graphite reaction rate is expected to be unchanged, and hence the rate of production of CO and H2 and the amount of graphite reacted during moisture ingress events should be unchanged, because the amount of barium or strontium available to catalyze the steam-graphite reaction should be unchanged. For reasons to be discussed below, the retention of metallic fission products is expected to be the same for UC and (Th/U)C2 particles. This expection is, however, based primarily 2
on theoretical / analytical grounds rather than experimental data. Post-f rradia-tion examination of some of the lead test elements (Ref.19) recently loaded as part of reload segment 7 should provide some confinnatory data in this area.
Based on the information and analyses on hand and contingent upon the anticipated PIE results, we believe that there is reasonable assurance that the substitution of UC for (Th/U)C fissile fuel kernels will not affect significantly the 2
2 steam-graphite reaction and that the peak primary system pressure, therefore, will not exceed present values cited in the FSAR.
The second area of concern for moisture ingress events, hydrolysis of failed
- Y5
\\
i
,, ( u
,o
-~-
. fuel, occurs only when steam diffuses through the graphite block and reacts with fuel kernels whose coatings have failed prior to the moisture ingress event. Hydrolysts of failed fuel can result in the release of a fraction of the noble gas inventory from the core to the primary coolant system. The rate of hydrolysis and associated noble ges release is a function of local fuel temperatures, steam concentration, and the chemical species undergoing hydrolysis. Since, as indicated earlier, the number of UC2 particles with failed coatings is not expected to be greater than would be the case with (Th/U)C kernels, since local temperatures and steam concentrations are not 2
affected by the change to UC2 particles, and since UC2 is less subject to hydrolysis than ThC2 (see FSV FSAR page 14.5-6), the removal of thorium from the fissile particle will reduce its tendency to hydrolize. Therefore, we conclude, based on the above reasoning, that the substitution of UC #0 "
2 (Th/U)C fissile fuel kernels in the FSV reactor should not result in fission 2
product releases (due to hydrolysis of failed fissile fuel) greater than cur-rent FSAR values.
The third item of concern for moisture ingress events, the amount of activity released to the primary coolant system from oxidized graphite, is proportional to the amount of graphite reacted and the concentration of fission products within the graphite. Since both of these quantities are, for reasons dis-cussed above, expected to be unchanged with the UC2 kernel, we agree with GA that fission product release should not be expected to exceed FSAR values.
In the analysis of the consequences of a pennanent loss of forced cooling (DBA No.1), four categories of this event are considere1:
\\Y i
c; 60
..,_me,e n.
- - ~ ~ -
--.aw.
m--
~- ~-~
12 (1) thermal results wherein metallic components might fail, (2) structural results which might affect the core, reflector barrel, or core support, (3) nuclear consequences which affect shutdown capability, and (4) fission product release and offsita' doses.
With regard to category 1, the core afterheat relationship and fission product inventory and distribution are unchanged by the substitution of UC2 f0" (Th/U)C. Therefore, thermal effects are not significantly altered, and the 2
FSAR conclusions regarding thermal effects remain unaltered. Similarly, structural effects are not influenced by the nature of the fuel particle.
The nuclear consequences of a pennanent loss of forced cooling concern the potential reduction of shutdcwn margin that could occur in the overheated core as a result of compaction and melting of control rods or spatial redistribution of control poison, fission product poisons, or uranium and thorium. Control rod melting and compaction and the redistribution of fission product poisons are unaffected by the type of fissile kernel, and the distribution of uranium and thorium in the fuel elements is also unchanged. Therefore, the diffusion through and evaporation of fission products from the graphite web are unaffected.
In the analysis of the consequences of a rapid depressurization/ blowdown (DBA No. 2), the radiological release from a postulated DBA No. 2 is taken to consist of essentially all the activity in the circulating primary coolant to the event plus a fraction of the activity plated out on the surfaces of the primary circuit. Since, as discussed earlier, the use of UC kernels is not 2
expected to result in any increase in the circulating and plateout activity, we agree with GA that the radiological consequences of this accident should be unchanged.
3bb
~
' We conclude, that of the accident scenarios examined in this section, no requirement for additional analysis resulting from the introduction of UC2 particles can be identified for FSV and that reasonable assurance has been provided that worst case conditions previously defined for accident analyses and found to be acceptable during the FSAR review will not be exceeded.
Evaluation of Materials procerty Data - As noted earlier in this evaluation, the major portion of our review of the safety-related consequences of using UC fissile fuel particles in FSV was directed toward examining the data 2
generated since the issuance of NUREG-Olll. Aside from the irradiation test data, which were discussed in detail above, another area where substantial new infonnation has been generated involves the' fission product lanthanum attack on sic. Relatively recent experiments (Ref. 20) have new yielded enough data to permit determination of an equation for the rate of reduction in sic thickness for TRISO UC2 as a function of temperature and thermal gradient. At 1200*C (the peak fuel temperature in the FSV core), and assuming that a particle would experience this temperature throughout its 6-yr. core residency, the total reduction in coating thickness would be about 15 um, which is less than 50% of the original sic thickness. Since this series of experiments also indicated that metallic and gaseous fission product releases were not large even after a 50% reduction in sic thickness, we conclude that reasonable assurance has been provided that there is adequate performance margin in UC2 particles in terms of this phenomenon.
Another phenomenon that can have an effect on fission product release is kernel migration (" amoeba" effect), which involves the asymmetric movement of the kernel toward the hot side of the coated particle.
If the kernel were to migrate far enough through the buffer layer to contact the structural Q
C; h
\\'
. coating layers (i.e., the inner pyrocarbon layer), the overall structural integrity of the particle could be affected. Although the kernel migration rate of UC is essentially the same as for (Th/U)C, the coating thicknesses 2
2 (including the bu.f,fer pyrocarbon layer) are larger so the consequences of kernel migration would be reduced. The bases for the core thermal safety limit (Technical Specification SL 3.1) would, therefore, not be exceeded by the substitution of UC for (Th/U)C2 particles.
2 With regard to fission product release in general, the similar crystalline structures and melting points of (Th/U)C and UC imply that the diffusion 2
2 kinetics of different fission product species should be similar. Moreover, empirical evidence cited by GA (Ref. 21) does not indicate any difference between gaseous fission product release from different carbide kernel types.
We believe that this constitutes indirect, but significant, evidence that existing FSV FSAR st tements regarding fission product release from (Th/U)C2 particles (or a (Th/U)C ThC fuel system) apply equally to UC2 particles 2
2 (or a UC ThC fuel system).
2 2
Post Irradiation Examination (PIE) and Surveillance - Although there is a considerable body of experimental data concerning the behavior of TRISO UC2 fuel particles, surveillance, including interim and. post-irradiation examinations, is required to confinn the safety analysis of any new fuel design feature.
We have in recent months issued several position statements regarding surveillance and PIE of both test and reference fuel in FSV; eg., Refs. 22 and 23. These statements address, in part, thd insertion of eight test elements (Ref.19), some of which contain TRISO UC fissile particles, in 2
reload " segment 7" (loaded in May 1979). Although we have had no fonnal surveillance requ rement for the eight test elements as a condition for J.
9 60
this insertion, we have noted (Ref. 2L that safety analyses supported by rest its of such post-irradiation surveillance may be required for future loads of fue.1 of new designs. Thus, before final approval of future reloads of TRISO UC2 particles can be granted, we will require (a) results of surveillance examinations on those test elements that contain TRISO UC2 particles, and (b) a comitment to perfom PIE on future large-scale TRISO UC loadings. With regard to the eight test elements FTE-1 through FTE-8, 2
the currently planned DOE-funded PIE program described in amendment 2 (Appendix A) of Reference 19, should provide a considerable amount of confimatory information and should provide sufficient infomation for licensing purposes. In particular, the planned destructive examinations of FTEs 2, 4, and 6, which contain TRISO-coated UC candidate FSV reload fuel 2
particles, should provide pertinent information on potential fuel damage (including incipient failure of the particle coatings), and release of metallic (eg., cesium and strontium) fission products.
Regulatory Position We have completed our review of topical report GA-LTR-23, which is intended to serve as a referential document that provides the basis for allowing the substitution of TRISO UC fissile fuel particles for the current TRISO 2
(Th/U)C fissile particles in the Fort St. Vrain gas cooled reactor.
In 2
our evaluation we focussed primarily on the irradiation test data that have been generated on the TRISO UC fuel design since our last major review 2
effort on TRISO UC fuel (1975-76). While this required the submittal and 2
review of some supplemental infomation, we also reviewed the material supplied in the submitted report. Based on our evaluation of the infor-mation provided in (a) the topical report, (b) responses to questions, and (c) a meeting with General Atomic, we conclude that there is reasonable g h) 5
assurance that the substitution of TRISO UC f r TRISO (Th/U)C fissile 2
2 fuel particles in the FSV core will (a) result in negligible changes in the nuclear and thermal behavior of the core, and (b) will not result in reduced safety margins or reliability compared to the reference core.
Because data acquisition on TRISO UC thermal irradiation perforn'ance is 2
ongoing (via, for example, the core heatup simulation tests on irradiated particles), we will require, as part of any future application for insertion of reload TRISO UC fuel, that General Atomic Company provide 2
timely information regarding the results of any tests involving TRISO UC2 particles. Specifically, we will require that the results of (a) the ongoing CHST program and (b) the PIE and surveillance of the TRISO UC -
2 containing test elements inserted as part of reload Segment 7, and (c) results of any other irradiation test programs be provided with future TRISO UC rel ad applications.
2 Should the method of fabrication of the TRISO UC2 particles be changed in such a way as to have a potentially significant effect on their future performance, we will require GA to provide evidence to support the perfomalce predictions for the particles produced from the altered materials or process parameters. An example of a potentially signifi-cant change would be the substitution of high temperature isotropic (HTI) pyro-carbon coatings (derived from methane) for the current low temperature isotropic coatings (LTI) derived from propylene. General Atomic has acknowledged, in response to staff question 231.7 (Ref.11),
that changes to fuel specifications can be made only in accordance with quality assurance procedures that comply with the requirements of 10CFR50, Appendix a, C
\\ ~3 )
i 3b0 n
..-_-..n.
. This evaluation applies only to TRISO UC fissile fuel particles to be 2
used irt the Fort St. Vrain reactor, because the analyses were perfonned in terms of the transient analysis for FSV and the cornparative effect of TRISO UC versus the FSV reference (Th/U)C2 particles. A separate 2
analysis would be required fcr any application of TRISO UC fissile fuel 2
in a high temperature gas-cooled reactor having a design differing from Fort St. Vrain's.
e T'3D
<q ah l
REFERENCES 1.
G. L. Wessman'(GA), Letter to William Gammill (NRC), November 17, 1978.
2.
K. Kniel (NRC), Memorandum to T. Speis, " Lead Items for Review of Th02 and UC Topical Reports," Feb. 8,1979.
2 3.
William Gammill (NRC), Letter to G. L. Wessman (GA), February 16, 1979.
4.
M. Tokar (NRC), Pkmorandum to X. Kniel, " Trip Raport:
Visit by M. Tokar to General Atomic Company, Las Alamos Scientific Laboratory, and the Fort St. Vrain. HTGR," April 6,1979.
5.
G. L. Wessman (GA), Letter to William Gammill (NRC), April 9,1979.
6.
M. Tokar (NRC), " Evaluation of High Temperature Gas Cooled Reactor Fuel Particle Coating Failure Models and Data." NUREG-Olll, November 1976.
7.
D.FN. Ross (NRC), Memorandum to R. P. Denise, " Input to GASSAR ISER,"
December 15, 1976.
8.
General Atomic Standard Safety Analysis Report (GASSAR-6), GA-13200, Docket No. STN 50-535.
9.
C. L. Smith, " Fuel Particle Behavior Under Normal and Transient Conditions,"
GA-Al2971 (GA-LTR-15), October 21, 1974.
t
\\J I 7
e6D
REFERENCES (Cont)
- 10. William Gamill (NRC), Letter to Calin Fisher (GA), May 29, 1979.
11.
C. R. Fisher, (GA), Letter to William Gamill (NRC), June 13, 1979.
- 12. HTGR Fuels and Core Development Program, " Quarterly Progress Report for the Period Ending May 31, 1978," GA-A14953, June 1978.
- 13. HTGR Fuels and Core Development Program, " Quarterly Progress Report for the Period Ending May 31,1978," GA-A15093, September 1978.
14.
W. J. Kovacs and D. P. Hamon, " Spitfire Loop Experiment SSL-2: Pre-irradiation Evaluttion of Fuel," GA-A13520, September 30, 1975.
15.
W. J. Kovacs and D. F. Harmon, "Perirradiation Report's GA Fuel Materials for GF-4", GA-A13475, September 1,1975..
16.
C. B. Scott and'D. P. Hamon, " Post-Irradiation Examination of Capsules P13R and P135", GA-A13827, October 8,1976.
17.
Victor Stello, Jr. (NRC), Memorandun to R. P. Denise, " Topical Report Evaluation on Thennal Conductivity of Large HTGR Fuel Rods," November 26, 1975.
18.
W. R. Johnson, "Thennal Conductivity of Large HTGR Fuel Rods, "GA-Al2910, April 30,1974.
REFERENCES (Cont) 19.
" Safety Analy, sis Report for Fort St. Vrain Reload Test Elements FTE-1 through FTE-8," General Atomic Report GLP-5494, June 30,1977.
20.
C. L. Smith, " sic-Fission Product Reactors in TRISO UC and WAR 2
UC 0 Fissile Fuel; Part 1 Generic Technology Program:
Kinetics of xy Reactors in a Thennal Gradient," GA-A14313, September 1977.
21.
B. N. Myers, et al, "The Behavior of Fission Product Gases in HTGR Fuel Material," GA-A13723, October 1977.
22.
R. L. Tedesco (NRC), Memorandum to W. L. Gamill, " Fuel Surveillance Policy - Application to Fort St. Vrain," December 13, 1978.
23.
R. L. Tedesco (NRC), Memorandum to W. J. Gamill, " Fort St. Vrain Fuel Surveillance," Fe,bruary 16, 1979.
ENCLOSURE 2
~
,4 UNI'ED STATES NUCt. EAR REGULATORY COMM!ssION 3
o, WASHINGTCN, D. C. 20555 DEC 131978 MEMORANDUM FOR:
W. J. Gamill, Assistant Director for Standardization and Advanced Reactor, DPM FROM:
R. L. Tedesco, Assistant Director for Reactor Safety, DSS
SUBJECT:
FUEL SURVEILLANCE POLICY--APPLICATION TO FORT ST. VRAIN The lack of a commitment by Public Service of Colorado (PSC) to perform adequate surveillance, including post-irradiation examination (PIE),
of the fuel in Fort St. Vrain (FSV) has been a continuing concern for a period of at least four years (NRC memorandum, M. Tokar to D. F. Ross, December 16, 1974).
As indicated in Rev.1 of the Standard Review Plan, Section 4.2 Fuel System Design, a post-irradiation examination fuel surveillance program is expected for each plant to detect anomalies or confirm expected fuel performance. While the plan is primarily for LWR type plants, we believe that this action is applicable to FSV. Because FSV is a first-of-a-kind reactor, with a fuel system unlike that of any other, its first core, standard fuel design should be subjected to a compre-hensive surveillance program including significant PIE. The extent of an acceptable program will depend on the history of the fuel design; that is, on whether the proposed fuel design is the same as current operating fuel or incorporates new design features.
Earlier this year, we reviewed (TAR-4693) a topic &l report describing eight fuel test elements (FTEs) proposed to be loaded with " Segment 7" (first reload) in FSV. We indicated that the safety analysis was acceptable, but that a commitment to perform, and to report the results of, PIE on the eight test elements was needed. On April 6, 1978, a letter was sent to PSC (R. P. Denise to J. K. Fuller) requesting a commitment to perform, and. a description of, a PIE program on both the standard,. reference fuel, and the test fuel elements.
In response, separate letters (both dated June 20, 1978) were received on (a) the standard fuel and (b) the test elements. In both cases, references were made to planned PIE to be perfomed under 00E funding.
In neither case, however, was there a comitment to perfom PIE in the event that 00E funding were reduced or withdrawn. Through tele-comunications'and meetings with PSC, we indicated that this lack of comitment was unacceptable. The most recent meeting, which focused exclusively on the test element PIE, was held in Sethesda on Decemoer 4, 1978.
Contact:
M. Tokar, x27603
\\g
<,!?
$00
'7N1MLO3]
m
~
W. J. Gangill DEC 13 EB As a result of this recent meeting, we tried to get more insight into past NRC practice and needs for fuel surveillance. In particular we discussed the matter in depth with DOR, who have handled a number of test assemblies in LWRs during the last four years.
No clear surveillance policy existed, but a consistent pattern does exist. We found for test assemblies, that extensive PIE is always perfomed (that is the purpose of the test fuel), and that the results are usually reported to NRC, but that requirements for this have not been made for LWRs. This is contrary to the direction we were taking with FSV. With 00R's assistance, we have arrived at a recomended plan for fuel surveillance in comercial power plants. This is described in Enclosure A and is applicable to FSV.
In light of this new reccmended plan, which is consistent with past practice with LWRs, we withdraw our request for required surveillance on the 8 test elements proposed for FSV. At the same time we must insist that a fim comit:nent be secured from PSC to perfom detailed surveillance on their reference fuel, which is a first-core loading by our definition. Until such time as a fim cemitment is established for an approved first-core surveillance program, we will not forward any approvals for test element irradiation or other fuel-related requests from PSC. As we have previously noted, the proposed FSV fuel PIE programs currently planned under DOE funding are acceptable.
All we really require at this time, therefore, is a statement frca PSC that, should future funding changes require modifications to the current PIE program for standard fuel, the modifications would be submitted for NRC review and approval. PSC should also realize that, for future reloads of fuel of new designs (some components of which may be included in the eight test elements), safety analyses supported by results from post-irradiation surveillance will be required.
Robert L. Tedesco, Assistant Direc*ar for Reactor Safety Division of Systems Safety-
Enclosure:
As stated Distribution: (see next page)
\\b\\
j.
m DEC 131978 W. J. GaW11 3-cc:
R. Mattson V. Stallo P. Check T. Speis P. Williams R. Ireland G~ Kuzzyc:
Fuels Section
=
4 1
4 e
9
~
Enclosure A Recomended Plan for Fuel Surveillance in Comercial Power Plants
' Experimen 1 Test Assemblies--The irradiati5n of experimental test assemblies is encouraged. Since these assemblies will be limited to a small core fraction and since a safety analysis will have been perfonned, no extensive PIE is (generally) needed to assure safe operation of the plant containing the test assemblies.
Lead Prototype Assemblies--Lead prototype assemblies differ from experimental test assemblies inasmuch as a fo115 wing core reload is scheduled. Surveillance-of the lead prototype assembif es would thus be required in suppo'rt of the following core reload but not (generally) to assure safety of the cycle containing the lead pro-totype assembifes. An instructive example is the requirement for Surry to perform surveillance on 17x17 lead prototype assemblies in support of a first core 17x17 fuel in Trojan. The surveillance requirement exists because of the timing of the Trojan core loading; if timely results from Surry were not obtained, the assurance of safe operation of Trojan would be compromised. Westinghouse acted as a broker in that case and got Surry to make cornitments to NRC in behalf of the Trojan submittal.
First Core Loading--Detailed surveillance, including interim exam-inations, is required to confirm the safety. analysis of a new fuel design. This detailed surveillance has been required en the first two plants to use the design in order to sample a statistically large number of assemblies and also to sample effects of different
,1 3
\\b manufacturing and operating histories.
(3
-.---o..
w
Mature Core Loadings--Simplified surveillance is now required on a
' routine bal'is for mature fuel designs. This.requirem2nt is a result of activities on Regulatory Guide 1.119 (withdrawn in favor of SRP revisions) and appears in SRP-4.2, Rev.1.
The requirement is an attempt to catch anomalies that result from insidious changes in plant operation or fabrication histories. Recent problems with
~
poison red failures and guide tube wear support the need for such wide-scale surveillance.
0
~
e e
O e
6b O3
- %-mea-ih=
me,
w-
-.g41
- m egege