ML20071B809
| ML20071B809 | |
| Person / Time | |
|---|---|
| Site: | Armed Forces Radiobiology Research Institute |
| Issue date: | 02/25/1983 |
| From: | Rickard D DEFENSE, DEPT. OF, DEFENSE NUCLEAR AGENCY |
| To: | Atomic Safety and Licensing Board Panel |
| References | |
| NUDOCS 8303010163 | |
| Download: ML20071B809 (145) | |
Text
{{#Wiki_filter:b ,/ g g y couluM OW N U, i ^ UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSIdN @ 28 @' r BEFORE THE ATOMIC SAFETY AND LICENSING BOARDS ma In the Matter of ARMED FORCES RADIOBIOLOGY Docket No. 50-170 RESEARCH INSTITUTE (Renewal of Facility (TRIGA-Type Research Reactor) License No. R-84) LICENSEE'S MOTION FOR PARTIAL
SUMMARY
DISPOSITION The Licensee submits this motior pursuant to 10 C.F.R. 2.749 and urges the Atomic Safety 7.ad Licensing Board to determine that, at least as to some of the issues being litigated in this proceeding, there is no longer any genuine dispute as to the material facts involved and that as a consequence there can be no genuine issue remaining to be heard. As is indicated in the Board's Memorandum and Order of January 28, 1983, there are ten general areas in dispute. Those ten general areas consist in some instances of several factual issues each of which involve several more subissues. This motion, except as will be indicated shortly, discusses those issues in the same order (and with the same titles) as they appear in the Appendix to the Board's January ,
/ 28, 1983, Order. One exception to this procedure is the need to depart (very slightly) from the titling approach since some of the subparts of the ten general areas of dispute require treatment at some length. A second exception
'8303010163 '830225 PDR ADOCK 05000170 G PDR OD
to this approach relates to Contention 3, Emergency Plan.
.The Board has previously indicated that motions concerning a
the' Emergency Plan would not be entertained prior to publica-
, tion by the NRC Staff of its evaluation of the Emergency Plan. Hence, consideration of issues associated with the Emergency Plan must be deferred for the present.
Each of the contentions discussed will be addressed in summary fashion in the body of the motion. That is, the issue (s) raised by the contentions will be restated so as to focus the Board's attention on the precise areas of disagreement. This will be followed by a synopsis of the facts pertinent to the issue and Licensee's analysis of their significance. In addition, as to each conveniently segregable set of issues identified in the body of the brief,-there is appended to the brief a fuller discussion
- (to the extent necessary) of the logic associated with the Licensee's analysis of the significance of the issue.1 Stated another way, the body of the motion presents an executive summary of those matters treated at length in the attachments.
CONTENTION 1 - ACCIDENTS I This contention deals with the " Fuel element clad failure accident DBA." Intervenor contends that Licensee's
, analysis of this Design Basis Accident is erroneous because Licensee assumes that such clad failure would occur at 1/ Attachment 1 contains information concerning the profes-sional credentials of Licensee's principal Affiant.
2
. . ~. . . - =- -
a psak. fuel elemant temperature of less than 100 C. Inter-venor contends that such clad failure would be more likely to occur at fuel temperatures greater-than 400 C and thus
, result in a greater gap activity and fission product release than the HSR (or more correctly) Safety Analysis Report (S AR) postulates.
As the issue is now framed, there can really be no a genuine dispute about what peak fuel element temperature ' was used in the preparation of the Safety Analysis Report (SAR) . This temperature,'while not identified in the SAR in degrees, can raadily be derived from the percentage of gap activity. This derivation is articulated in detail in Attachment 2. The SAR (at page 6-12) clearly reflects that a fractional release of 0.1% was used. This gap activity is associated with a fuel temperature greater than or approximately equal to 600 C. Since Intervenor's concern.is with respect to clad failures at fuel temperatures 0 ' in excess of 400 C and with realistic gap activities and releases for a fuel element clad failure accident DBA, i all of which the Licensee has appropriately considered, l this contention must be summarily dismissed. j CONTENTION 2 - ACCIDENTS II(l) - l Fuel Element Storage Rack Failure l This contention deals with a single fuel element i i storage rack failure assumed to be fully-loaded with 12 2/ The language of the contentions refers to the HSR (Hazard Summary Report) which was filed with the license application. This document was all that was available 6 at the time the contentions were drafted. Since that time, the Safety Analysis Report has been prepared and filed. The SAR replaces in terms of significance, at least, the HSR. References in the' motion to the SAR are thus understood to be synonymous with the less comprehensive i and older HSR mentioned in the contentions. CNRS has indicated that this' point of clarification is correct. See Dr. Stillman's December 18, 1982, Deposition at pp. I 17 line 16 through 18 line 1. _,~ _
._.. _ ___. _ _ . 3_.
stainless-steel clad TRIGA fuel elements. Intervenor contends that the SAR does not provide reasonable assurance
.4 that such an accident cannot occur because: 1) no criticality , calculations for a twelve element configuration are provided and, 2) no statement of experience is cited to indicate / support the Licensee's claim that it takes approximately 67 stainless-steel clad TRIGA fuel elements to achieve criticality.
c In addi'lon, the Intervenor contends that a storage rack failure accident is of a different kind and a greater severity than accidents treated in the SAR and should De categorized as a DBA. While there may have been some doubt at the time that this contention was drafted concerning whether or not twelve elements in a worst case neutronic configuration could achieve criticality and while the Intervenor may l not have been aware at-that time of the calculations which demonstrate the number of elements of the type used at l AFRRI needed to achieve criticality, the record is now l complete. The unrebutted evidence of record (as a result of Licensee's answers to Intervenor's interrogatories) demonstrates that a fully-loaded storage rack (containing 12 fuel elements) cannot under any circumstances achieve criticality. Moreover, the record now contains ample
, data demonstrating that approximately 69 stainless-steel clad TRIGA fuel elements are necessary to achieve criticality.
The data in question is summarized in Attachment 3. ( 4 l l
In light of the facts now of. record, it.is obvious that it is incredible that an. unrestrained critical or supercritical configuration of stainless-steel clad TRIGA
+ elements in the reactor pool could be achieved as a result of storage rack (s) failure. What's more, even if criticality were somehow achieved, it would be at the bottom of the reactor pool. There would thus be no deleterious consequences i for the reactor staff or the public.
Licensee submits that Intervenor's contention expresses specific concerns that have been adequately addressed by the Licensee. Specifically, the calculations, analysis, and supporting documentation, the absence of which cites as a basis for its contention, have been presented and do, in fact, provide adequate assurances. Further, Inter-venor's claim that such incidents represent accidents of a different kind and greater severity than those treated in the Licensee's Safety Analysis Report (SAR) is totally without support in fact. Therefore, based on the undisputed evidence of record, this contention must be summarily dismissed. CONTENTION 2 - ACCIDENTS II(2) - l Failure of an Experiment This contention deals with an experiment failure
. concurrent with a malfunction of confinement safeguards.
Intervenor contends that such an accident could occur s with releases potentially in excess of regulatory limits and submits that such accidents are of a different kind and greater severity than those treated in the AFRRI SAR 5
and, therefore should be more properly designated as DBAs. There.never was any significant factual dispute regarding this contention. In preparing this contention the Intervenor
, obviously consulted NRC~ Docket 50-170 and recorded what it believed to be the essential facts surrounding past occurrences at AFRRI. Each of the eleven inconsequential malfunctions occurred. Not one of them, however, is as significant as the Intervenor would have the Board believe.
The' critical point that must be understood in connection with this contention is that the Intervenor has failed to make the connection between occurrences such as these and releases to the environment. The hurdle which the Intervenor has failed to pass is one of causation as is amply addressed in Attachment 4 (which discusses each of the malfunctions cited by the Intervenor). The Licensee has illustrated the extreme unlikelihood, and often inappropriateness, of the Intervenor's cited malfunctions with respect to their relationship or importance i to confinement isolation, source term generation, and i environmental release. Licensee submits that malfunctions i and failures can occur but that they are extremely unlikely. If one considers multiple, independent, and concurrent
- ' events--which are, even by the Intervenor's own admission, l . necessary for environmental releases to actually occur, they become incredible. Licensee further submits that w
I 6 i
t 1 - its'own h'istory of safe, reliable operation over the last 21 years is evidence more convincing;than anyone could ever present or postulate--particularly in comparison
- , with a new licensing action for a similar facility, Moreover, Licensee has demonstrated its in-place system of reviewing,-
approving, and limiting experiments and irradiations of-materials with a view towards limiting the consequences of a release, should one occur. And finally, Licensee has addressed a worst-case experiment failure with an assumed total release to the unrestricted environment from the standpoint of assessing worst-case consequences; and these consequences are insignificant by themselve,s and certainly miniscule with respect to other accidents that are also treated in the AFRRI SAR. i In short, Licensee submits that it has adequately I addressed experiment failures with an assumed total release to the unrestricted environment. Further, Licensee submits that Intervenor's claim that such accidents are of a different kind and greater severity than those accidents treated in the AFRRI SAR is totally without support in fact. ( Moreover, Licensee has demonstrated working (and proven) l l systems or mechanisms to: 1) adequately identify malfunctions l l- and failures in a timely fashion; 2) adequately provide l . backup systems to protect single malfunctions / failures from having an impact; 3) limit the probability of single l w i l-7 i l
malfunctions /f ailures to a reasonable level; 4) make multiple concurrent malfunctions / failures extremely unlikely; and
- 5) ensure adequate review and limitation of materials to be irradiated, illustrating that the consequences of -
a release, should one occur, have been adequately evaluated and have minimal impact on the unrestricted environment and the general public. Therefore, based on the undisputed evidence of record, this contention must be dismissed. CONTENTION 2 - ACCIDENTS II(3) - Negative Temperature Coefficient
.The Intervenor contends that Licensee has failed to demonstrate that the TRIGA reactor's negative temperature coefficient of reactivity remains negative when hydrogen i
is presumed lost from damaged TRIGA fuel elements. This must be so, since the Intervenor claims that Licenseo has failed to demonstrate that the negative temperature coefficient will automatically shut down the reactor; the only way in which this could occur is if the temperature l c'oefficient of reactivity somehow becomes zero or positive with a presumed loss of hydrogen from damaged TRIGA fuel elements. The Intervenor has expanded (somewhat) upon its theories in response to interrogatories and in answers to questions posed during Dr. Stillman's deposition. As is evident from the Affidavit of Joseph A. Sholtis included in Attachment 5, the evidentiary basis for this contention is only partly present. That is, there are a total of three independent 8 l
e scientific contributors (each of which is negative) associated with the TRIGA reactor's inherent negative temperature coefficient of-reactivity which acts as an effective reactor
. safeguard. Dr. Stillman has addressed only one of those three contributors. While the Licensee cannot fully agree that Dr. Stillman's analysis is correct as to the one contributor in this area that he attempts to answer (since his postulated loss of hydrogen introduces negative reactivity),
we will assume for purposes of discussion that he is. Given that assumption, there is no real disagreement concerning the scientific facts involved. That is, the Licensee has previously provided (in response to CNRS's Interrogatory
- 6) evidence which addresses all three major scientific contributors. This prior submission coupled with the affidavit in Attachment 5 provides ample basis upon which to permit the Board to decide.
In summary, Licensee has demonstrated that the TRIGA reactor's temperature coefficient of reactivity will always be negative and inherent, regardless of whether the TRIGA fuel is damaged and hydrogen is presumed lost or not. Licensee has also illustrated the extreme unlikelihood l and difficulty in removing hydrogen'from TRIGA fuel which, l even if it were presumed to occur, would still not force
. the TRIGA reactor's overall temperature coefficient of reactivity to a zero or pcsitive value. Licensee has also demonstrated that each TRIGA fuel element in the core, which is presumed damaged with an associated loss 9
of-hydrogen,-will contribute less and'less (with increasing 2 hydrogen loss) to the core's neutron. population, power level,'and fission density. Thus, each will have a suppressed s neutron population, power level, fission density, and fuel-temperature (in comparison with the'other undamaged TRIGA elements in core) , which will be suppressed more. and more with increasing hydrogen loss. Thus, the consequences-i. of accidents stemming from damaged TRIGA fuel where hydrogen is presumed lost simply cannot become more severe since 4 conditions are'not aggravated but actually reduced in 2 these damaged elements. Finally, failed TRIGA fuel accidents j are not of a different kind than those accidents that are treated in the~AFRRI SAR, since clad failures are explicitly treated in the AFRRI SAR. Therefore, Licensee
- submits that Intervenor's contention is totally without 4
support and must be dismissed. f CONTENTION 2 - ACCIDENTS II(4) - 1 Multiple Cladding Failure Accidents r i This contention deals with multiple fuel element clad failures occurring concurrently in time. Intervenor
. contends that concurrent, multiple clad failures have j not been considered in the SAR, and further contends that i -
such an accident could result from cladding material defects,
, an uncontrolled power excursion, a LOCA, sabotage, aircraft collision, or a natural "Act-of-God" accident. The Intervenor
{- also contends that a concurrent multiple clad failure accident is of a different kind and greater severity than 4 10
those accidents that are treated in the AkRP.I SAR and should be more properly designated as a DBA. In essence,
'therefore, the Intervenor contends that concurrent multiple clad failures due to their postulated causal mechanisms are in fact-credible.
This entire contention constitutes an attack on Licensee's SAR and NRC's judgment as expressed in its Safety Evaluation 4 Report (SER) without basis in fact. This contention is i built entirely upon conjecture for which supporting evidence in fact has never been provided. The Intervenor claims that concurrent multiple clad failure events can result, for example, from clad defects, yet the Intervenor has characterized such an occurrence as "very unlikely." (See page 108 lines 19 through 25, inclusive of the transcript lof Dr. Stillman's deposition in New York on 10 Dec 82.) Moreover, when the Intervenor was asked by the Licensee in its first-round interrogatory #24b, "Have there been any prior multiple fuel element cladding failure accidents ! in any TRIGA reactor?", regardless of the cause, the Inter-venor stated, "Not that we know of." The fact of the l matter is that there has never been a concurrent-multiple clad failure accident in the entire history of TRIGA reactors from any cause. Moreover, concurrent multiple clad failure b accidents are not viewed (by Licensee or the NRC) as being credible events in a TRIGA reactor and, therefore, designation l - of them as DBAs would be absurd. i 11 i l l
In short, the evidence is not, to the extent that it sts, really in dispute. -The Licensee recognizes the-remote theoretical possibility'that multiple cladding
. failures can occur (see Attachment 6) . The Intervenor i suggests but never-demonstrates that a number of mechanisms can in fact produce multiple cladding failures. What is missing is some sort of credible causal connection between a postulated initiator and multiple cladding failures.
, The Licensee submits that this causal connection is critical to this contention and in its absence the contention must be dismissed. CONTENTION 4 - ROUTINE EMISSIONS I This contention attempts to show that radioactive materials produced from TRIGA operations are released to the environment in violation of federal (10 C.F.R.
- 20) guidelines. Several examples purport to show that these guidelines are exceeded because either the equipment, methods, or reporting systems are not adequate to detect
- violations or that limits are, in spite of procedures used, exceeded.
The Intervenor alleges that AFRRI has "not demonstrated that airborne and waterborne radioactive emissions from l routine operations and disposal of solid wastes will be
, maintained within the limits of 10 C.F.R. 20." In support of this claim the Intervenor states that environmental
- monitoring is inadequate to determine radiation doses 1
l 12
to the public due to inh'lation a or'injestion because (a) film dosimetry detects only external gamma radiation, (b). the particulate radioactivity monitor for airborne effluents (a pancake - probe.GM counter) is not isokinetic and therefore cannot be used for meaningful evaluations, , (c) a beta self absorption factor was omitted from calcula-
-tions or environmental analysis and (d) a model'used to derive its dose assessments to the environment is not realistic. These statements, however, fall to show inadequate environnental monitoring. The succeeding paragraphs will delineate the errors associated with each of these allegations.
l AFRRI does not use film dosimetry for environmental monitoring. The Intervenor's answer to Licensee's interroga-tory number 27a.1 states: "The use of film to detect external gamma radiation is a technique that is much inferior to the use of thermoluminescent dosimeters, whose sensitivity i is much greater and far more reliable." AFRRI has used i thermoluminescent dosimeters for years and is in fact pleased that Intervenor's expert witness, Dr. Ernest Sternglass, agrees that the system in use is a more sensitive, reliable l system. The air particulate monitoring system is isokinetic. I The Licensee's answer to Intervenor's Interrogatory 28 l , shows that the particulate monitoring system is in fact l l l . l 13 I
isokinetic and is therefore reliable for particulate sampling; it is used even though it is not required by the license. Indeed, use of an isokinetic particulate air sampling
. system provides for " meaningful evaluation" of air being exhausted from the reactor facility and generates the "better data" which the Intervenor says should be available.
AFRRI does in fact continuously sample air downstream < from the high efficiency particulate air filters through an isokinetic sampling system and in several years of sampling has not recorded a single instance of escaped particulate isotopes exceeding 10 C.F.R. 20 limits. This sampling history includes, of course, "the small amounts of particulate material mixed with the larger amounts of Argon" that the Intervenor feels is of concern. The Intervenor cites an NRC-cited calculational omission concerning a beta self-absorption factor. In addition, the Intervenor suggests that its lack of information on how quarterly environmental samples of water, soil, and 1 vegetation are prepared and analyzed proves AFRRI's environ-mental monitoring is inadequate. The NRC did discover that measurements were made without the use of a beta self-absorption correction factor. However, even with the correction factor applied, all releases were well
. below all regulatory requirements. There was at no time a significant possibility of exceeding regulatory limits since the standard procedure requires specific radionuclide l
14
2 a analysis if gross concentrations exceed one tenth of the regulatory limits for unspecified nuclides. As far as environmental sampling methods are concerned, contrary
- to the implications in the stated contention, no items of non-compliance were found by the NRC during inspections ' o regarding the methods for preparing and analyzing samples or-the instrumentation used for these purposes.
The concentric cylinder set model, despite the statement made in this contention by the Intervenor, is not used to " derive its dose assessments to.the environment and from which it concludes its effluents are within regulatory limits." In fact, the entire statement is incorrect. The concentric cylinder set model only supplements environ-mental TLD's--it is not itself used to determine compliance with any regulatory limits. In addition, no responsible organization including the NRC, has found the model to be unrealistic and the Intervenor has yet to describe what it finds to be unrealistic about the model. (See NUREG-0851, " Nomograms for Evaluation of Doses from Finite Noble Gas Clouds," W. J. Pasciak, USNRC, January, 1983, pp. 227-240.) This contention also expresses great concern about
, a small water escape line from the orig.inal AFRRI stack.
. The line, which contained a water trap, allowed rain water entering the old stack to drain away. The line led from 15
. __ _ .,.____..,,.m, _- ._ . . , ,
the base of the stack access, across an equipment area and through a wall to the outside. When a new, larger stack was installed, the old stack was rerouted so that
. rain no longer entered this section of the stack. The line, which then became unnecessary, was removed and the exit point from the stack base was capped. Incidentally, an analysis (conducted upon discovery of this " problem")
i of the air flow in the stack and at the location of the exit (more than 3 feet below the normal air flow path) shows that it is not " highly probable" that releases in excess of 10 C.F.R. 20 Appendix B occurred. In fact, just the opposite is true. It is highly improbable that any releases occurred much less any that would have exceeded 10 C.F.R. 20. In further support of its allegation that the Licensee's environmental protections are inadequate, the Intervenor states that airborne release reports for 1962, 1963, and 1964 show that releases from the AFRRI stack exceeded the MPC concentrations for unrestricted areas. A summary report prepared from available data in 1972 failed to show any releases that could violate NRC restrictions. (This documented evidence has been reviewed during many subsequent NRC inspections and would surely have resulted in a Notice of Violation from the NRC had such a violation actually occurred.) How a letter sent (6 Oct 1961) 8 months before the AFRRI reactor first went critical can show evidence of isotopic release to the environment greater 16
than that allowed by federal guidelines is certainly curious. I The fact that the Intervenor continues to express concern with twenty year old data suggests that AFRRI is succeeding.
-in maintaining its environmental releases at commendable, not~ reprehensible, levels. Even in the " worst" of times (1963 and 1964) , the evidence of record demonstrates compli-ance with 10 C.F.R. Part 20 limits (not noncompliance.
I as suggested by the Intervenor in its statement that whole body doses in unrestricted areas exceeded 0.5 rem). In summary, the evidence before the Board clearly demonstrates that the Licensee's environmental monitoring
, methods, equipment, and (most importantly) results are fully in accord with the regulatory requirements of the Commission. Hence, the Licensee submits that this contention must be dismissed.
CONTENTION 5 - NEPA I l CONTENTION 6 - NEPA II i These two contentions are directed principally to the attention of the NRC Staff. Both contentions allege defective compliance by the Staff with the provisions [ of the National Environmental Policy Act (NEPA) found at 42 U.S.C. 4321 et seg. The statute requires an evaluation of some sort in connection with ". . . major Federal actions j . significantly affecting the quality of the human environment
- ... (42 U.S.C. 4332 (2) (c)) . The NRC has, by regulation i
! (10 C.F.R. Part 51), established a scheme by which it decides what level of evaluation is required in a particular j 17 L
a _ licensing proceeding. The NRC: Staff accomplished (in January of 1982) an " environmental impact appraisal" (EIA) in which it documented its basis for a " negative declaration."
. That is, based on the EIA,-the Staff' determined that an
" environmental impact statement" (EIS) , which is a detailed statement prepared by the NRC to comply with.the provision of NEPA cited above, is-not required prior to renewal of this license. In making this determination, the NRC Staff had before it, among other things, the " environmental report" submitted by the Licensee as part of its application 1-for license renewal as well as a twenty year historical e
record on this particular facility. The Staff then properly applied 10 C.F.R. 51.5 to the relevant environmental facts. That section identifies l eleven instances-in which an EIS is required. All of l l the specifically identified instances are clearly of far greater significance than the renewal of a license for a comparatively small research reactor. The twelfth instance is a " catch-all" instance in which actions not specifically identified may be subjected to the detailed analysis of an EIS. The Staff evidently determined that this action was not of such a magnitude to require an EIS. The intervenor asserts that, based on an inadequately
. prepared EIA, the Staff erroneously concluded that an EIS was not required. It should be recognized that these contentions were both prepared well before the Staff had prepared and published its EIA. The Licensee submits 18
W that, with the publication of-the EIA, the NRC Staff has properly discharged its duties under NEPA. Hence, Contention 6 (NEPA II) should be dismissed. Contention 7 (NEPA I)
. should likewise be dismissed'since it relies for its analysis of the requirement for an EIS (presumably under 10 C.F.R.
- 51. 5 (a) (12) ) on the totality of the Intervenor's other contentions, which as is evident from this motion, should also be dismissed.
CONTENTION 7 - SECURITY This contention suggests that, for two categories of reasons, physical security at-AFRRI is so inadequate that the reactor license should not be renewed. First, the Intervenor asserts that the Physical Security Plan is inadequate.3 Second, the Intervenor asserts (citing examples) that the Licensee's " history of security violations" demonstrates that the " controlled access areas" are ineffec-tively protected. , l As the Board noted in its Order of August 31, 1981, at page 13, the security issues were to be restricted to the building at AFRRI (Building #42) in which the reactor 3/ This portion of the contention may well have been dropped by the Intervenor. The Physical Security Plan prepared by the Licensee and subsequently approved by the NRC is protected, for obvious reasons, from disclosure
- to the general public (10 C.F.R. 2.790(d)). As of the present, despite offers by the Licensee to have a properly qualified physical security expert review the plan, no , such review has occurred. Moreover, none of the allegations of inadequacy have addressed the supposed shortcomings of the Physical Security Plan.
19
is located. None of the five examples cited by the Intervenor relate specifically to the reactor facility in Building
#42, and thus are of little consequence. Indeed, the . reactor facility and its controlled access area were not .
questioned by the Defense Audit Service (DAS) in what is still a " draft" audit report as being deficient in physical security or having been mismanaged. The physical security protection standards required of licensees who possess special nuclear material of moderate or low strategic significance (AFRRI's is low) relate principally to detection of breaches of physical security. As is indicated in 10 C.F.R. 73.67 (a) (2) (i)-(iv) a physical protection system providing for "early detection and assess-ment of unauthorized acc,ess or activities by an external adversary;" "early detection of removal of.special nuclear material;" and nctification to NRC and " appropriate response forces of its removal in order to facilitate its recovery"
- suffices for facilities such as AFRRI's. None of the l examples cited by the Intervenor demonstrate that the regulatory standards alluded to above are not being met I
at AFRRI. In other words, the Intervenor has shown nothing that would indicate that the NRC's approval of AFRRI's L Physical Security Plan was erroneous. Indeed, the first
. two of Intervenor's examples communicate the fact that mechanisms exist by which access can be " controlled" in a manner which assures AFRRI's ability to achieve early 20 l
l _ -
detection. In summary, the evidence of record demonstrates clearly that the regulatory requirements have been met and thus this contention must be dismissed.
, CONTENTION 8 - ACCIDENTS III This contention deals with multiple clad failures postulated to be caused by either a power excursion or a LOCA such that sudden elevated temperatures occur in turn causing multiple clad failures followed by either an explosive zirconium-steam interaction or an explosive zirconium-air interaction, respectively, depending on the accident initiator being either a power excursion or a LOCA. The Intervenor contends that such accidents constitute " maximum credible accidents beyond the design basis of the reactor (class 9 accidents)" but that they nevertheless can be expected to occur at the APRRI reactor.
Even though the Licensee has demonstrated that multiple clad failures due to a power excursion or a LOCA are not credible elsewhere in this Motion on contentions 9. Accidents IX, and 2. Accidents II.4, and although the Licensee has also demonstrated that it is extremely unlikely that signifi-cant amounts of hydrogen can be driven out of failed TRIGA fuel contention 2. Accidents II.3, the impossibility of explosive zirconium-steam or zirconium-air interactions
, occurring even at elevated temperatures in TRIGA fuel must still be discussed. The Intervenors have pointed this out as their primary area of concern.
21
'First, before beginning the discussionlof zirconium-steam and zirconium-air interactions in TRIGA fuel, one Very interesting point should be made. That is, the Inter-venor's own wording of this contention classifies there events.as " maximum credible accidents beyond the design basis of the reector (class 9 accidents)" and yet also submits that these postulated " maximum credible accidents beyond the design basis of the reactor (class 9 accidents)"
can nevertheless be expected to occur'at the AFRRI reactor. Licensee is at an obvious loss in trying to resolve this clear contradiction. That is, how can accidents which are admittedly "beyond the design basis" possibly be " expected to occur"? The Intervenor seems to suffer from the same problem. When the Licensee asked the Intervenor's technical consultant, Dr. Irving Stillman, "Can accidents which are, by your own admission and contention, which are beyond the design basis, actually be expected to occur?", Dr. Stillman replied "Yes." Further, when Licensee followed up on this line of questioning during the depositio'n of Dr. Stillman in New York on 18 Dec 1982, by asking Dr. 1 Stillman, "You feel that it could happen over the lifetime of the facility?", Dr. Stillman replied, "Yes, obviously." And yet when Licensee continued by asking Dr. Stillman, "Has it ever occurred at any TRIGA reactor facility that you are aware?" Dr. Stillman replied, "No, not that we know of." (Sc.e page 131 lines 9 through 25, inclusive, 22
of the transcript of Dr. Stillman's deposition taken in New York on 18 Dec 82.) Moreover, when the Licensee asked Dr. Stillman to put a qualitative estimate of likelihood on whether explosive zirconium-steam or zirconium-air interactions would occur at AFRRI if clad failures and elevated temperatures were presumed to exist, Dr. Stillman characterized such interactions as being "Unlikely." (See page 132 lines 1 through 24, inclusive, of the transcript of Dr. Stillman's deposition in New York on 18 Dec 1982.) The Licensee submits that multiple clad failures alone are not credible. Even if multiple clad failures did occur, along with elevated temperatures, the explosive j ricconium-steam and zirconium-air interactions are simply not possible. (See General Atomics Report (GA-A15384, "TRIGA Low-Enriched Uranium Fuel Quench Tests," by J.R. l Biddlecome, et al., GA Project No. 4314, July 1980. See i ! also, " Fuel Elements for Pulsed TRIGA Research Reactors," i by M. T. Simnad, et al., Nuclear Technology, Vol 28, January 1976, pp. 31-56 at page 37. Both of these documents are contained in Attachment 7. Very simply stated explosive zirconium-steam and zirconium-air interactions simply cannot occur for TRIGA fuel at AFRRI. This statement is based on actual experiments performed by General Atomics. In one of these experiments, unclad U-ZrH x fuel slugs were heated inductively up to 1200 C and then immediately quenched in water. No zirconium-steam interaction resulted even though hydrogen was driven 23
off. In fact, only minor surface slug cracking resulted. For the other experiment series, again unclad U-ZrH x fuel slugs were inductively heated up to a temperature of 850 C and then air was introduced into the chamber. Here also, no explosive zirconium-air interaction occurred. These experiments are detailed in the reports in Attachment 7. It is extremely noteworthy that during the deposition of Dr. Irving Stillman that when Licensee asked Dr. Stillman:
"Do you have any documentation that shows that explosive zirconium-steam and/or zirconium-air interactions do or can occur for TRIGA fu,el?" Dr. Stillman replied, "No, not for TRIGA fuel I don't." (See page 132 line 25 through page 133 line 3, inclusive, of the transcript of Dr. Stillman's deposition taken in New York on 18 Dec 1982.)
It is also noteworthy that the Intervenors are aware of the General Atomics experiments which refute the possibility that explosive zirconium-steam and zirconium-air interactions can occur in TRIGA fuel and yet "poo-poo" these results and try to disclaim them. (See Intervenor's response to NRC Staff request for admissions #9 where Intervenors state, " Tests of chemical reactivity (such as those described in your assertion) were performed with TRIGA fuel elements at the General Atomic Corp. Laboratories. To have complete confidence in their results would be comparable to asking the tobacco industry to determine the effect of cigarette 24
smoking on-the incidence of lung cancer.- In other words, . General ~ Atomic is the manufacturer of both the TRIGA-reactor and its fuel elements,-hardly an impartial scientific
- , study is to be expected.") -On the other hand, though, when the Intervenor supplemented.its responses to Licensee's first round interrogatories #35g and 35i, the Intervenor
- stated in both responses, "To the best of our knowledge, General Atomic Company has not attempted such' experiments even though they are ideally set up to perform them."
The Licensee submits that the Intervenor is, at best, confused or, at worst, will only accept that information which is agreeable to itself. In summary, the Licensee has demonstrated elsewhere ! in this Motion that multiple clad failures are not credible for power excursions or a LOCA since conditions for causing such clad failures cannot be attained. Licensee has also demonstrated, elsewhere in this Motion, the extreme difficulty and, thus, extreme unlikelihood of driving significant amounts of hydrogen from failed TRIGA fuel. Moreover, the Licensee has demonstrated for its reactor that even if multiple clad failures and elevated temperatures are presumed, explosive zirconium-steam or zirconium-air inter-actions will not occur. The Licensee submits that the experiments performed by General Atomics on such chemical reactions serve as proof to substantiate this claim of impossibility. The Licensee submits that the Intervenor's 25
. ._ _ .__ . __ _ . _ . _ _ . . ~ _ _ - . _ - . _ _ _ _. _ _ . . _ _. _ _ . . -
4 clains under this contention'are totally without support. Therefore, based on the discussion provided herein and the results of the cited' General ~ Atomics experiments,
- this contention must be summarily. dismissed.
CONTENTION 9 -' ACCIDENTS IV This contention deals with a presumed LOCA, where the core becomes uncovered. The Intervenor. contends that if core uncovering occurs concurrently with pulsing operation, i that multiple clad failures could result. There is no real disagreement among the parties that the multiple cladding failures postulated by the Intervenor can only be achieved after elevating the fuel temperature drastically over a substantial time interval. Part of this increase in fuel temperature, according to the Intervenor's a hypothesis, comes from the " loss of cooling and shielding ( water" (or LOCA) . The other part comes from the repeated i continuation of pulsing operation during the LOCA. Indeed, the scenario advanced by the Intervenor demands a LOCA coupled with repetitive pulsing. The heart of the disagree-ment is whether or not all of these conditions can occur simultaneously (if at all) . The Licensee has demonstrated (in Attachment 8 hereto) that numerous safeguards must fai,1 (extremely unlikely) and gross operator error must be assumed to permit the selective and fast repetitive firing out of the transient control rod during a LOCA. However, even if this incredible 26 y.- , ,- - , -- . , - , ,, _m _ war ____. . - - - . _ _ - _ , .- __ _ . - . - - _ -- - . . ,
s_ Y series of events and malfunctions were nevertheless presumed to occur, the Licensee has still demonstrated that actual repetitive pulses cannot occur at a frequency greater than about one every 10 seconds by which time fuel temperatures will have basically recovered to ambient conditions, The Licensee, moreover, has demonstrated that uncovered fuel regions cannot effectively contribute to fission, and, thus, fuel temperatures in such uncovered regions cannot become aggravated any further, i.e., beyond the conditions that are posed as a result of the LOCA by itself. The Licensee submits, therefore, that the Intervenor's claims under this contention are totally without support. Licensee further submits that its LOCA analyses within I the AFRRI SAR together with a former LOCA analysis submitted. under Docket 50-170 in 1964-65 as part of a previous license amendment are true, realistic, and, in fact, conservative and provide reasonable assurance that no clad failures , are expected in conjunction with a LOCA at AFRRI. CONTENTION 10 - ROUTINE EMISSIONS II This contention was originally part of what is now Contention 4 - Routine Emissions I. In essence, the Inter-venor alleges in Contention 10 additional examples of historical events at AFRRI that supposedly demonstrate that radiation monitoring methods are inadequate and that - prior violations of regulatory limits did result from
, routine reactor operations.
27
The Intervenor argues that annual _ doses to the general public have previously exceeded 0.5 rem. In support of this argument, the Intervenor identifies three AFRRI Environ-mental Release Data and Perimeter Monitoring Reports (dated 4 5/27/66, 9/20/66, and 12/14/71 )- and an Autumn, 1979, written communication from AFRRI to Mr. Joe Miller, a i member of CNRS. Licensee submits that there is no material issue of fact that remains as to this contention. .AFRRI's reports to'the AEC/NRC speak for themselves. Properly understood, these reports demonstrate that the' conclusion reached in other documents in Docket 50-170 are correct. For example, the NRC Staff concludes at section 12.9 of its SER-that:
-The results of the environmental radiation dosimeters
( (film or TLD) located on the NNMC grounds have l averaged less than 3 mrems/yr for the last 10 j years. The average of the highest individual i readings for the last 10 years is-less than l 15 mrems/yr. In addition, the NRC Staff dealt with and adequately explained (at SER Section 12.9) the significance of 1-5 mrems/ hour dose rate and the high environmental monitoring station reading reported on 12/14/71 as follows: During the 1960's AFRRI operated an x-ray facility in support of its research program, and a nearby , perimeter monitoring station cogsistegtly gave 4/ While this date is given as 12/14/77, Licensee suggests that 1971 is the year intended by the Intervenor. l 28
e a reading much higher than any other, or the ~ average of the others, in the perimeter monitoring set. Both because of the proximity of the x-ray lab and because,there is no credible way the reactor airborne effluents could always flow-toward the station, it is concluded that readings at that detector station were not related to the reactor or other NRC-licensed operations. In short, it is clear from an examination and proper l 1 interpretation of the evidence of record that no material disagreement of fact remains and that this contention . must be dismissed. CONCLUSION As is clear from this motion and its attachments, the vast majority of the issues before the Board in this proceeding are no longer the subject of a-genuine dispute from a factual point of view. While there was clearly l a great deal at issue at the time the contentions involved were admitted, the discovery process has provided data which narrows the focus of this dispute considerably. l The Licensee therefore requests that the Board dismiss r Contentions 1, 2, 4, 5, 6, 7, 8, 9 and 10. The Licensee further suggests that a suitable schedule for the disposition of Contention 3 be established. Respectfully submitted,
, ,e f7 * , AV D C. RI ARD Deputy General Counsel Defense Nuclear Agency Counsel for Licensee 4
29
~ ,. . __ - - --
O e ATTACHMENT 1 1 i I e B l i l l r
J Oualifications/ Resume JOSEPH A. SHOLTIS,3R., M A3OR, USAF Armed Forces Radiobiology Research Institute (AFRRI) Radiation Sciences Department (RSD) Radiation Sources Division (RSRS) Naval Medical Command, National Capital Region Bethesda, Maryland 20814 PROFESSIONAL CIVILIAN FDUCATION 1970 B.S. Nuclear Engineering, The Pennsylvania State University, University Park, PA. (nistinguished Military Graduate). GPA: 2.9 1977 M.S. Nuclear Engineering, The University of New Mexico, Albuquerque, NM. GPA: 4.0 1977 - 1980 Ph.D. Course work, Nuclear Engineering, The University of New Mexico, Albuquerque, NM. GPA:4.0 PROFESSIONAL MILITARY EDUCATION 1975 USAF Squadron Officer School, Maxwell AFB, AL. (via correspondence). 1978 USAF Air Command and Staff College, Maxwell AFB, AL. (via non-resident seminar at Kirtland AFB, NM). CIVILIAN & MILITARY COURSES /TR AINING 1968 Log',- & Event Tree Ana!ysis, California State College, California, PA. 1972 Advanced Nuclear Power Plant Technology, Georgia Institute of Technology, l_ Atlanta, G A. 1 1973 Essentials of Fluid Mechanics: Statics and Dynamics of Fluid Flow, Air Force Institute of Technology, Wright-Patterson AFB, OH. ( 1973 Nuclear Weapons Effects, Air Force Institute of Technology, Wright-Patterson AFB, OH. l 1975 USAF Laboratory Management of R&D Procurement, Kirtland AFB,' NM.
. 1976 USAF Nuclear Accident Disaster Preparedness, Kirtland AFB, NM.
l 1976 Environmental Impact Statements for the DoD, General Services Administration,
. Dallas, TX.
1977 Nuclear Criticality Safety short course and laboratory work shop, University of New Mexico and Los Alamos Scientific Laboratory, Los Alamos, NM and Taos, l NM. I
. _ . . . . - - , . _ - _ . _ _ . . _ - - . - . . - - . . _ , . _ . -_m. . _ .. - - , - - ,
F 1981 Medical Effects of Nuclear Weapons, Armed Forces Radiobiology Research Institute, Bethesda, MD. 1982 Nuclear Weapons (advanced course), Interservice Nuclear Weapons Schcol, Kirtland AFB, NM. 1983 Ballit.dc Missile Staff Course, USAF Air University, Vandenberg AFB, CA. PROFESSIONAL EXPERIENCE 1982 - Present Chief, Radiation Sources Division and Reactor Physicist-In-Charge, Armed Forces Radiobiology Research Institute, Bethesda, MD. 1981 - 1982 Reactor Branch Chief and Reactor Physicist-In-Charge, Armed Forces Radiobiology Research Institute, Bethesda, MD. 1980 - 1981 Research Reactor Operations Officer, Armed Forces Radiobiology Research Institute, Bethesda, MD. 1978 - 1980 USAF laboratory Associate and DoD Member of the Technical Staff, Advanced Reactor Safety Division, Sandia National Laboratories, Albuquerque, NM. 1974 - 1978 Chief, Space Nuclear Systems Safety Section, Air Force Weapons Laboratory, Kirtland AFB, NM. 1971 - 1974 Foreign Aerospace Nuclear Power Systems Analyst, Foreign Technology Division, Wright-Patterson AFB, OH. 1968 - l 1971 Mine Safety Analyst / Statistician, U.S. Bureau of Mines, Pittsburgh, PA. COMMITTEES, CONSULTANTSHIPS, AND SPECIFIC PROJECT EXPERIENCF 1971 - l 1974 Identified, Evaluated and Characterized the Design Performance Capabilities of Foreign Ground and Aerospace Nuclear Power Systems, including the Soviet "Romashka" and " Topaz" Space Reactors, Foreign Technology Division, Wright-Patterson AFB, OH.
~
1971 - 1974 Briefer: "The Soviet Technological Challenge," Foreign Technology Division, i Wright-Patterson AFB, OH. 1974 - 1976 Project Officer, Feasibility and Safety Analysis and Component Testing of Nuclear Propulsion and Power Systems for the USAF, Air Force Weapons Lab, Kirtland AFB, NM.
1974 - 1978 Project Officer, Nuclear Safety / Risk Assessments for the Launch of U.S. Space
. Nuclear Power Systems, Air Force Weapons Lab, Kirtland AFR, NM.
1974 -
~
1978 Technical Advisor, Interagency Nuclear Safety Review Panels for the the Viking A & B, Lincoln Experimental Satellites 8/9, Pioneer 10 & 11, and Voyager i & II Launches; Air Force Weapons Lab, Kirtland AFB, NM, HO, USAF. Washington, D.C., HO Air Force Systems Command, Andrews AFB, MD, NASA-Kennedy Space Center, and NASA-Houston Space Center. 1975 - 1976 Member, DoD Tri-Service Working Group on Nuclear Power for the Don. 1976 - 1978 Technical Advisor, Blue Ribbon Panel on Advanced Space Power Systems for the DoD in the 1980's and Beyond, HQ, USAF, Washington, D.C. 1976 - . 1978 Project Officer, AFSATCOM II/Ill Nuclear Safety / Risk Evaluation and Environmental Impact Statement, Air Force Weapons Lab, Kirtland AFB, NM. 1976 - 1978 Member, New Mexico Governor's Panel (New Mexico Erergy Resource Registry) on Energy and Scientific Manpower Resources, Santa Fe, NM. 1976 - 1977 Evaluation Team Member, Procurement of Kilowatt Isotopic Power System (KIPS) for DoD/ doe / NASA Space Use, Germantown, MD. i 1976 - 1978 Inspector, Kirtland AFB Nuclear Disaster Preparedness Inspection / Implementation Team, Kirtland AFB, Nu. 1976 - 1978 Member, Mark 12/12A Reentry Vehicle Test Launch Search and Recovery Team, Enewetak Atoll. 1976 - l 1978- Lead Project Officer, Safety Evaluation of the Space Shuttle Launch Vehicle, Air l Force Weapons Laboratory, Kirtland AFB, NM. 1978 Member, Nuclear We apons Stockpile INRAD Survey Team, Air Force Weapons Laboratory, Kirtland AFB, NM.
~
1978 - l 1980 Technical Advisor, doe /NRC Probabilistic Risk Assessment Review Group. 1978 - 1980 Technical Advisor, Reactor Safety Committee, Sandia National Laboratories,
- Albuquerque, NM.
i
1978 - 1980 Principal Investigator, . Accident Initiation and Engineered Safety Systems:
. LMFBR Accident Delineation Study, Sandia. National laboratories, Albuquerque, NM. ?
1978 - 1980 Project Officer, Characterization of Sandia Lab's Annular Core Research Reactor (ACRR) Performance Characteristics, Sandia National Laboratories, Albuquerque, NM. 1978 - 1980 Project Officer, Evaluation of LMFBR Transient Overpower (TOP) Accidents and Their Initiators & Proposal of In-Pile Experimentation to Study TOP Accident Progression and Phenomenology, Sandia National Laboratories,' Albuquerque, NM. 1979 - 1980 Lecturer, The Three-Mile Island Unit 2 Accident, Albuquerque, NM. 1981 - Present Member, AFRRI Reactor and Radiation Facility Safety Committee, Armed Forces Radiobiology Research Institute, Bethesda, MD. i 1981 - Present Instructor: " Principles of lonizing Radiation" and " Electromagnetic Pulse" units of Medical Effects of Nuclear Weapons Course, Armed Forces Radiobiology Reserach Institute, Bethesda, MD. 1981 Member, Cobalt-60 Recovery Team: Project HERMAN, Armed Forces Radiobiology Research Institute, Bethesda, MD. 1983 - Present Invited Lecturer: " Principles of Ionizing Radiation" unit of Military Applied Physiology Course, Uniformed Services University of the Health Sciences, Naval Medical Command, National Capital Region, Bethesda, MD. CERTIFICATION l 1980 USNRC Reactor Operator, License No. OPS 363 j 19S1 USNRC Senior Reactor Operator, License No. SOP 3942 i PROFESSIONAL AFFILIATIONS Member: American Nuclear Society (ANS) i Member:TRIGA Reactor Owners / Users / Operators Group Member: Test, Research, and Training Reactors (TRTR) Organization Member: American Association for the Advancement of Science (AAAS) Member: New York Academy of Sciences (NYAS) Member:The Planetary Society l i
~ - - - - -,-
Member: Americans for Rational Energy Alternatives (AREA) , Member: Scientists and Engineers for Safe Secure. Energy (SE-2) Member: Society of American Military Engineers (SAME) Associate Member (NomineetAmerican Society of Mechanical Engineers (ASME) PUBLICATIONS
-Sholtis 3 A, Jr. "'Ihe Dissociating Gas Power Cycle (U)," Foreign Technology Division Bulletin, Foreign Technology Division, Wright-Patterson AFB, OH,16 Apr 1974, (SECRET /NOFOR N).
Sholtis 3 A, 3R. Title Classified, Foreign Technology Division Bulletin, TCS-384491/74, SAO/FTD-SP-13-01/06-74, Foreign Technology Division, Wright-Patterson AFB, OH, 14 May 1974, (TOP SECRET). Sholtis 3 A, 3R. " Soviet Aerospace Nuclear Reactor Technology (U)," Contribution to Defense Intelligence Agency Task T70-02-OlB, " Soviet Nuclear Power Technology (U)," Foreign Technology Division, Wright-Patterson AFB, QH, 31 Oct 1972,
, (SECRET /NFD).
Sholtis 3 A, Jr. " Aerospace Nuclear Reactor Technology - Western Europe (U)," Contribution to Defense Intelligence Agency Task T74-02-09, " Nuclear Power Technology - Western Europe (U)," Foreign Technology Division, Wright-Patterson AFB, OH, 31 Oct 1973 (SECRET /NFD/NDA). Sholtis 3 A, Jr. " Radial and Axial Neutron Flux Profiling for Small Heterogeneous Reactor Cores by Redistribution of Fuel," AFWL-TR-75-246, Air Force Weapons Laboratory, Kirtland AFB, NM, Mar 1976. Sholtis 3 A, Jr. " Empirical Correlation Describing the Impact Pesconse of Two-Foot Diameter Spheres with Internal Energy Absorbing Material Simulating an Airborne Reactor Containment System," AFWL-TR-76-93 (Rev.), Air Force Weapons Laboratory, Kirtland AFB, NM Aug 1976. Sholtis 3 A, Jr. " Description and Analysis of Kilowatt Isotope Power Systems (KIPS) Under Developinent for Soace Application in the 1980's, "AFWL-TR-76-207, Air' Force Weapons Laboratory, Kirtland AFB, NM, Feb 1977. Holtzscheiter E W, Kelleher D, Mitchell G, Crawford M L, and Sholtis 3 A, 3r. " Safety Methodology for Space Nuclear Systems," AFWL-TR-77-104, Air Force Weapons Laboratory, Kirtland AFB, NM, Oct 1977. Sholtis 3 A, Jr. " Preliminary Safety and Environmental Assessment (PSEA) of a Nuclear-Powered Strategic Satellite System (S SS)," Internal Air Force Weapons Laboratory /DYVS Report, Air Force Weapons Laboratory /DYVS, Kirtland AFB, NM, 1 Jul 1978. Sholtis 3 A, Jr. " Synchronous Satellite / Spacecraft Collision Probabilities," Internal Air Force Wdapons Laboratory /NSO Technical Report, Air Force Weapons Laboratory /NSQ, Kirtland AFB, NM,8 Oct 1976.
Sholtis'3 A, Jr. " Economic Impact to the U.S. of No Breeder Reactor Program (LMFBR) and No Reprocessing of Spent Nuclear Fuel Over the next Thirty Years," paper presented at the First doe Nuclear Data Conference, Albuquarque, NM, Oct 1977. Sholtis 3 A, Jr. " Environmental Impact Statements for U.S. DoD Space Nuclear Systemb," USAF Nuclear Surety Information, Vol 12, No. 45, Jul-Sep 1978. Sholtis 3 A, Jr. and Crawford M L. "Preorbital Risk Assessments for the Launch of U.S. Space Nuclear Systems,". Internal Air Force Weapons laboratory /DYVS Report, Air Force Weapons Lab /0YVS, Kirtland AFB,' NM, Oct 1978. Sholtis 3 A, Jr. " Simulated Response of an Airborne Reactor Containment System to Impact," paper presented at the American Nuclear Society 1977 Western Regional Student Conference,23-2 Mar 1977, Oregon State University, Corvallis, OR. Shottis 3 A, Jr. " Mission, Design, and Safety Considerations of Aircraft Nuclear Propulsion for the DoD, " Independent M.S. Study Report, University of New Mexico, Department of Chemical and Nuclear Engineering, Albuquerque, NM, Jun 1977. Sholtis 3 A, Jr. " Impact Testing and Analysis of Airborne Reactor Containment Vessels," paper presented at the University of New Mexico, Chemical and Nuclear i Engineering Seminar, Albuquerque, NM, 22 Mar 1977. Williams D C, Sholtis 3 A, Jr., Rios M, Varela D W, Worledge D H, Conrad P W, and Pickard P 5. "LMFBR Accident Delineation Study: Approach and Preliminary Results," Sandia National Laboratories, Albuquerque, NM, paper presented at the ANS/ ENS International Meeting on First Reactor Safety Technology, Seattle, WA, 19-23 Aug 79. Varela D W, Sholtis 3 A, Jr., and Worledge D H. " Justification for Low-Ramp Transient Overpower (TOP) Experiments," Sandia National Laboratories Technical Report to 3.E. Powell and the Nuclear Regulatory Commission Office of Advanced Reactor , Safety Research, Sandia National Laboratories, Albuquerque, NM,26 Mar 1979. Williams D C, Varela D W, Worledge D H, and Sholtis 3 A, Jr. " Delineation of LMFBR In-Core Accident Phenomenology," SANn79-Oll3A, Sandia National Laboratories, Albuquerque, NM, Aug 1979. Sholtis 3 A, Jr., Rios M, Worledge D H, Conrad P W, Williams D C, Varela D W, and Pickard P 5, "LMFBR Accident Delineation Study, FY 79 Interim Report," SAND 79-0100A, Sandia National Laboratories, Albucuerque, NM, Aug 1979. Sholtis 3 A, Jr. " Analysis of the COSMOS 954 Reentry (U)," Internal Air Force Weapons Laboratory /NSCM Report, Air Force Weapons Lab /NSCM, Kirtland AFB, NM, (Cm%hl). Rios M, Sholtis 3 A, Jr., Williams D C, Conrad P W, and Pickard P 5. "LMFBR Accident Delineation Study, Phase IA Final Report, Sandia National Laboratories, l Albuquerque, NM,1 Oct 1978. i
. Williams D C, Sholtis 3 A, Jr., Conrad P W, and Pickard P S. "LMFBR Accident Delineation Study, Dbase I Final Report," NUREG/CR-1507, SAND 80-1267, Sandia National Laboratories, Albuquerque, NM,15 Nov 1980.
Sholtis 3 A, Jr. "LMFBR Accident Delineation: Development of the Methodology and its Application to Transient Overpower (TOP) Accidents, SAND 80-1413, NUREG/CR-1550, Sandia National Laboratories, Albuquerque, NY,1980. Williams D C, Sholtis 3 A, 3r., and Sciacca F W. "LMFBR Accident Delineation and the Evaluation of Research Priorities," SAND 80-1634A, TANSAO-35-1-676 (1980), Vol 35, Sandia National Laboratories, Albuquerque, NM, paper presented at the 1980 ANS/ ENS International Conference on Fast Reactor Safety, 16-21 Nov 1980, Washington, D.C. Sciacca F W, Sholtis 3 A, Jr., and Williams D C, "LMFBR Accident Delineation Study: Assessment of Post-Accident Phenomenology," TANSAO-35-1-676 (1980), Vol 35, ISSN: 0003-018X, pp. 386-387, Sandia National Laboratories, Albuquerque, NM, paper presented at the 1980 ANS/ ENS International Conference on Fast Reactor Safety,16-21 Nov 1980, Washington, D.C. Sholtis 3 A, Jr. " Nuclear Criticality Safety Analysis of Hypothetical AFRRI-TRIGA Fuel Element Storage Rack Accidents," Internal Armed Forces Radiobiology Research Institute /SSD Memorandum for Record, Armed Forces Radiobiology Research Institute /SSD, Bethesda, MD,19 Jan 1981. Sholtis 3 A, Jr. and Moore M L. " Reactor Facility, Armed Forces Radiobiology Research Institute," AFRRI TR81-2, Armed Forces Radiobiology Research Institute, Bethesda, MD, May 1981. Sholtis 3 A, Jr. " Analysis of Cocked Fuel Elements in the AFRRI-TRIGA Mark-F Ueactor," Armed Forces Radiobiology Research Institute, Bethesda, MD, paper presented at the 8th TRIGA User's Conference,8-10 Mar 1982, Idaho Falls, ID. Moore M L and Sholtis 3 A, Jr. "AFRRI Reactor Relicensing Effort," Armed Forces Radiobiology Research Institute, Bethesda, MD, paper presented at the 8th TRIGA User's Conference,8-10 Mar 82, Idaho Falls, ID. Sholtis 3 A, Jr. " Analysis of the Consequences of a Hypothetical Worst-Case Reactivity Excursion (i.e., Inadvertent Pulse) While Operating the AFD.RI-TRIGA Reactor in I the Steady-State Mode at Full Power (1.0 Mwt), " Internal Armed Forces Radiobiology Research Institute /SSD Memorandum for Record, Armed Forces Radiobiology Research Institute /SSD, Behtesda, MD,5 Feb 82. Sholtis 3 A, Jr. " Xenon Buildup and Associated Negative Reactivity Worth Determinations over Time during AFRRI Reactor Power Operations as well as after Scram," Internal Armed Forces Radiobiology Desearch Institute /RSRS-Reactor Technical / Operational Data Peport, Armed Forces Radiobiology Research Institute /RSRS-Reactor, Bethesda, MD,1982. Sholtis 3 A, Jr. "AFRRI Emergency Evacuation and Fire Plan," Armed Forces Radiobiology Research Institute Instruction 3020.2G, Armed Forces Radiobiology Research I'stitute, Bethesda, MD, 30 Sep 82.
. Sholtis 3 A, 3r. Responses to NRC Staff Ouestions concerning the "AFRRI Peactor Facility Safety Analysis Deport (SAR)," Facility License R-84, Docket No. 50-170, incorporated as an addendum to the AFRRI SAR dated Jun 81, Bethesda, MD, 9 Oct 81.
Sholtis 3 A, Jr. " Emergency Plan for the AFRRI-TRIG A Reactor Facility," License R-84, Docket No. 50-170, Armed Forces Radiobiology Research Institute, Bethesda, MD, Oct 1982. Smoker R R and Sholtis 3 A, Jr. "AFRRI Radiation Sources Division Instructions RSD 5-1 through 5-9 inclusive," Armed Forces Padiobiology Research Institute /RSRS, Bethesda, MD, dates on individual RSD Instructions vary from 1980 through 1981. Numerous responses to intervenor (CNRS) interrogatories under the contested AFRRI Reactor license renewal proceeding before USNRC. Numerous one-time and recurrent monthly, quarterly, and annual technical progress reports and project status reports over the last twelve years. HONORS, AWARDS, AND ACCOLADES 1965 Elected to National Honor Society 1965 Selected for Washington County, DA, Gifted Student Program at California State College, California, PA 1966 Honor Graduate, Monongahela High School, Monongahela, PA 1969 Vice-Commandant's Award, AFROTC Field Training, Plattsburgh AFB, NY 1970 Commissioned 2Lt, USAF, Distinguished Military Gradua e, The Pennsylvania State University, University Park, PA 1971 National Defense Service Medal 1972 Junior Officer of the Ouarter, Foreign Technology Division, Wright-Patterson AFB,OH 1972 Air Force Systems Command, Certificate of Merit 1974 Air Force Commendation Medal 1975 USAF Outstanding Unit Award 1976 Air Force Commendation Medal,1st Oak Leaf Cluster 1977 USAF Outstanding Unit Award,1st Oak Leaf Cluster 1977 American Nuclear Society, Conference Session Best Paper Award 1978 USAF Certificate of Appreciation 1978 Tendered Regular Commission, USAF
. 1980 Air Force Commendation Medal,2nd Oak Leaf Cluster 1980 Sandia National Laboratories, Honorary Staff Award 1981 US Army Reactor Shif t Superintendent's Badge
\
1981 Defense Nuclear Agency, Certificate of Achievement
. 1983 US Army Reactor Commander's Badge 1983 Charter Nominee to First (1983) Edition of "Who's Who in Frontier Science and Technology" SECURITY CLEARANCES 1971 -
Present Top Secret (DoD) 1978 - Present 0-clearance (doe) RELEVANT EXPERIENCE
SUMMARY
Twelve years experience in the design, evaluation, characterization, analysis, safety, development, procurement, use, operation, and risk assessment of thermal and fast nuclear reactor systems and radioisotopic power systems for both ground and aerospace applications. Two years experience planning, coordinating, conducting, and assessing in-pile reactor experiments in support of the U.S. Advanced Reactor Development and Safety , Analysis Program administered by USNRC. Seven years experience with the design, safety, operation, maintenance evaluation, administration, and use of research reactors; five years of which specifically involved TRIGA Reactors. USNRC-Licensed Senior Reactor Operator and Physicist-in-Charge for the AFRRI TRIGA Reactor Facility. Twelve years active commissioned service as a Nuclear Research Officer, USAF. Nine years managerial / supervisory experience involving technical nuclear projects and personnel. Chief, Radiation Sources Division at AFRRI with direct control over five radiation source facilities and nine technical staff personnel. Three years direct experience and participation on established TRIGA Reactor Facility Safety Committees. l l l l .-. ._. _ _ .- .-. . . - - -. - . , - - ,
W O e ATTACHMENT 2 1 [ I i
AFFIDAVIT O_ F_ JOSEPH A. SHOLTIS,3R. Joseph A. Sholtis, Jr., being duly sworn according to law, deposes and says: The Intervenor's centention centers around the fuel temperature assumed to exist and utilized by the Licensee in analyzing this clad failure accident DBA. The Licensee clearly states in its SAR, page 6-12, last paragraph - that, "Although the measured amount of radioactive noble gases for the operating conditions in the AFRRI reactor fuel would indicate a gap activity percentage of less than 0.01 percent, the theoretical limit of 0.1 percent gap activity for fission product gases of noble gases and iodines, as stated in reference 2, will be used in the consequence analysis for the Design Basis Accidents (Section 6.3.4)." (See also attached affidavit of Mr. Frederic D. Anderson.) Reference 2 cited above is a General Atomic Company Report titled, The U-ZrHXAlloy: Its Properties and Use in TRIGA Fuel, by M.T. Simnad, dated February 1980, and characterized i as GA Project No. 4314, GA Report No. E-Il7-833. A copy of Figure 5-1 from page 5-3 of this GA Report is provided below and graphically shows that a fractional release of 0.1% (or 10-3), which Licensee uses in its analysis of this DBA, is associated with a fuel temperature of approximately 6000C,if the theoretical maximum curve is utilized, or a fuel temperature of approximately 800-10000C, if the actual experimental data points are utilized. This figure proves that the Licensee did not assume in its analysis of the fuel element clad failure accident DBA that clad failure would occur at a fuel temperature of "less than 1000C" as the Intervenors contend. Instead, it shows that the Licensee utilizes a conservative release fraction associated with a minimum fuel temperature of approximately l e e - wm- - - - - a--__
6000C. In fact, Dr. Irving Stillman during his deposition on 18' Dec 82 tied the 0.1% release l
- fraction that the Licensee uses to a fuel temperature of approximately 500-6000C. (See P
Transcript of Deposition of Dr. Irving Stillman on 18 Dec 82, page 71 line 8 through page 72
- line 3.)
This is borne out in Licensee's answer to In'tervenor's first-round interrogatory number 1 as well as in Licensee's answers to NRC Staff's questions on the AFRRI SAR, specifically Licensee's response to NRC Staff question #67 concerning the AFRRI SAR. It should be noted that these NRC Staff questions on the AFRRI SAR together with the responses provided have been incorporated into the AFRRI SAR as an attachment or addendum. It should also be pointed out that the release fraction for accident conditions is associated with the normal operating temperature, ny the temperature during accident conditions. This is because the fission products released as a result of a fuel clad failure are those that have collected in the fuel-clad gap during normal operation. (See last paragraph, labeled 3., on page 3-4 of GA Report E-l!7-833, a copy of which is provided below.) Therefore, the Licensee uses in its analysis of the fuel element clad failure accident DBA a release fraction of 0.1% which is associated with a fuel temperature greater than or approximately equal to 6000C. Moreover, the 0.1% release fraction utilized by the Licensee + is conservative since it is characteristic of a normal operating fuel temperature greater than or approximately equal to 6000C - fuel temperatures at which the Licensee's reactor does not normally operate. t Intervenor has stated in its answers to Licensee's first-round interrogatories as well as , during the deposition of Dr. Irving Stillman that they (the Intervenors) base their contention i L 4 m_. , _ . . .. ,, - _. .-__ _ . _ _ - .,
=_. . _ - - .._. - . - .
t i statement that the Licensee utilizes a fuel temperature of "less than IOOOC" in its' analysis of the fuel element clad failure accident DBA on the 0.2% radiciodine release fraction cited in Licensee's SAR. Here also the Intervenor has misinterpreted the facts of the matter since the Intervenors believe that only 0.2% of the radioiodines contained in the gap get out- ! of the element for a c!rd failure event. (See Transcript of Deposition of Dr. Irving Stillman on 18 Dec 1982 on page 72 lines 7 through 23 inclusive.) In fact, however, what the Licensee actually assumes is that all of the radioiodine contained in the gap at the time of a clad p failure gets out and into the reactor pool -water, which has a bulk temperature of i approximately 25-300C, while conservatively 0.2% of the radioiodine that gets into the reactor pool water is assumed to come out of solution and gets into the reactor room air. This point is quite clearly stated in the AFRRI SAR on page 6-13, last paragraph, as well as in Licensee's response to NRC staff's question #43 on the AFRRI SAR. In summary, therefore, on both counts the Intervenor is incorrect and no evidence exists to indicate that Licensee uses a fuel temperature of "less than 1000C" in analyzing the fuel element clad failure accident DBA. On the contrary, Licensee utilizes a release fraction which is associated with a fuel temperature greater than or approximately equal to 6000C and which has been shown by actual experiments to be conservative. d d$ - . H A. SIMJfS, . 1 Sworn to and subscribed before
.me on this isTday of M ,1983.
I Q', Wb / 7 8'y .w .myk u M 'j l. Y WV
1: Affidavit-of Fredric D. Anderson
- # I, Fredric D. Anderson, being duly sworn, do state as follows:
.. 1. .I was employed as a nuclear safety consultant to Dames & Moore to perfonn the safety analysis for the Armed Forces: Radiobiology
'- Research, Institute's TRIGA research reactor (Docket No.- 50-170). I am currently employed by the Nuclear Regulatory Commission as a Senior Reactor Engineer in the Division of Licensing. ( 2. This affidavit addresses the contention that the source term used in the safety analysis for the design basis accident of a fuel
. element cladding failure was non-conservative.
- 3. I have reviewed Section 6.0, " Safety Analysis", of the' AFRRI React'or
. Facility Safety Analysis Report in its entirety and, subject to the s_upplemental information which follows for clarification, do hereby adopt the portion discussing fuel element cladding failure (Section 6.3.2) as true and correct to the best of my knowledge and belief.
To respond to the above-stated contention, additional information is presented in my affidavit. i i Source Term Used in AFRRI Safety Analysis of Fuel Element Cladding Failure , 'As stated in the Safety Analysis (Section 6.3.2), I used a source term of 0.1 percent of the steady-state fission product inventory for the noble L ' gases and radioiodines present in the gap of a fuel element. This source term would be available for release in the event of a fuel element cladding failure. The value of this source term for gaseous fission products was selected on ! j the basis of General Atomics experimental data and theoretical analysis given in GA Report No. E-ll7-833 (GA Project No. 4314) for various fuel temperatures l and irradiation-times. For the AFRRI reactor operating conditions of a fuel
- temperature of 600 C and assuming infinite operation, a fractional release l from the fuel material to the fuel element gap of 0.1 percent of the' gaseous -
l . fission products was theoretically possible. From a 1966 experiment under the AFRRI reactor operating conditions, a measured value of 0.01 percent of- the gaseous fission products was obtained for the release fraction to l the fuel element gap. Therefore, the source term used in the safety analysis l for a fuel element cladding failure (0.1 percent) is a factor of 10 more conservative than the measured value (0.01 percent). In order for the assumed source term of 0.1 percent for a fuel element cladding failure accident
.- in the AFRRI research reactor to be consistent with the GA experimental data, the operating fuel temperature would have to be 1000 C or a margin of 400 C from actual operating fuel temperatures.
i . l
)
i' L
Based on the above analyses and data, I firmly believe that the source term for gaseous fission products used in the AFRRI safety analysis for a fuel element cladding failure is appropriate and conservative.
&Yd <.)
Fredric 0. Anderson Subscribed and sworn to before me this o?.@ day of December,1982. b Ndtary fublic My Commission expires Od>z /, /98/o V U
\
r . l i 1 l l l I I -
[iak:n
~
fromiGA Rcport No. E-Il7-833, GA Proj;ct No. 4314 Th7 U-ZrHx Alley ~Its Properties and Use in TRIGA Fu 1, by M.T. Simnad, Fcbruary,1930] 0 10 TiiEORETl CAL MAXIMUM
' O 10'I -
O O
~
10 - O U 5a Y d g 10 C E f 8 10' - O
/
1 J! ~~ ! 10
-5 _
l f MEAN VALUES FROM: O 1966 EXPERIMENT
- 1971 EXPERIMENT, POST-lRRADI ATION ANNEAL l
O SNAP, POST-IRRADIATION ANNEAL E 1966 EXPERIMENT, POST-IRRADIATION ANNEAL I ' 10" 1600 2000
. 0 400 800 1200 TEMPERATURE (*C)
EL-1615 75LC373 Fig. 5-1. Fractional release of gaseous fission products from TRIGA fuel showing theoretical maximum, and experimental values above 400*C corrected to infinite irradiation (from Ref. 10) 5-3
aktn from:GA Report No. E-Il7-833, GA Proj:ct No. 4314 Th7 U-ZrHx Alloyz Its Propertirs and Use in TRIGA Fuel, by M.T. Simnad, February,1930.} The curve in Fig. 5-La_pplies to a fuel element which has been
- irradiated for a time sufficiently long that all fission product
. activity is at equilibrium and the release fraction is at its theoreti-
- cal maximum. Figure 5-1 shows that the measured values of fractional releases fall well below the curve. Therefore, for safety considera-tions, this curve gives very conservative values for the high-tempera ,
ture release from TRIGA fuel. Also worthy of note are the following conclusions from the TRIGA.
, fission product release experiments: .
- 1. Because the s;2ples were unciad, the high-temperature measurements were made on essentially dehydrided U-Zr.
Post-irradiation annealing musurements indicate that the dehydriding process did not significantly affect ,
- the release rate.
- 2. Part of the 1971 experiments was the measurement of the release from a post-irradiation anneal of a sample of fuel that had been irradiated to a burnup of SS.5% of
! the U-235 (or 1.1% of the total uranium atoms). The results of this part of the experiment indicated that i the effects of long-term irradiation of the fuel on fission product release-are small, at least for total burnup equivalent of the maximum that has been achieved.
- 3. The release fraction for accident conditions is character-9 , ,
istic of the normal operating temperature, not the tempera-I ture during accident conditions._ _ . _ _ This is because the fission products released as a result of a fuel clad failure are those that have collected in the fuel-clad
,f , gap J ring normal operacion.
I j 5-4 L
w ATTACHMENT 3 4
AFFIDAVIT OF JOSEPH A. SHOLTIS,3R. Joseph A. Sholtis, Jr., being duly sworn according to law, deposes and says: In a prior Memorandum for Record dated 19 January 1981, I presented conservative criticality calculations for a hypothetical AFRRI fuel storage rack accident that prove that a twelve element configuration of stainless-steel clad TRIGA fuel elements cannot achieve criticality under any circumstances. (A copy of this 19 Jan 31 AFRRI Memorandum for Record is provided as an attachment to this affidavit and was previously provided in response to Intervenor's Interrogatory Number 2.) In fact, Intervenor has stated in its supplemental response to Licensee's first-round Interrogatory #9e that the calculations provided in Licensee's 19 Jan 81 Memorandum do (in Intervenor's view) provide reasonable assurance that twelve stainless-steel clad TRIGA fuel elements cannot achieve criticality under any circumstances. This same 19 Jan 81 Memorandum also cited a source of experience for Licensee's statement that it takes approximately 69 stainless-steel clad TRIGA fuel elements to achieve criticality. A source supporting Licensee's statement of experience that it takes l approximately 69 stainless-steel clad TRIGA fuel elements to achieve criticality was also I referenced in Licensee's answer to Intervenor's first-round interrogatory #3; a copy of this response is also attached to this affidavit together with a copy of AFRRI's internal Radiation Sources Division Instruction, RSD 5-8, " Reactor Core Loading and Unloading Procedures," dated 27 March 1981. In summary, these documents identify and illustrate the actual results of the initial stainless-steel clad TRIGA fuel element core loading at AFRRI in 1965, using the standard 1/M approach-to-critical loading technique.
Licenne is at a loss in trying to understand why Intervenor contends that a fuel element storage ec!: accident, that is assumed to result in a criticality excursion,' can be considered an accident of a _"different kind" since criticality excursions are explicitly treated in the , AFRRI SAR in sections 6.2.2, 6.2.3, and 6.2.5. Licensee is also at a loss in trying to understand how a fuel element storage rack accident , that is assumed to result in a criticality excursion, regardless of its credibility or likelihood, can constitute an accident with an associated " greater severity" than accidents, like reactor power transients and clad failures, which are treated in the AFRRI SAR. This is because i i such an event, incredible as it appears, would occur at the bottom of the reactor pool under i approximately 19.5 feet of water where adequate shielding is provided. In addition, any associated inadvertent power excursion or transient would be automatically terminated by the negative temperature coefficient of reactivity together with expansion action which would render the initially critical configuration, which is unrestrained, permanently subcritical. That is, the pool was designed to accomodate criticality, regardless of whether l 11 occurs normally in the core or in an unplanned critical configuration of unrestrained l TRIGA fuel elements, and the same intrinsic mechanisms that terminate a planned pulse would also act to terminate an uplanned excursion, except that, in addition, for the unrestrained configuration, expansion of the system during the excursion would also tend to move the elements away from one another and thus render the resultant configuration l permanently subcritical. Therefore, there would be no deleterious consequences of such an event - even if it were presumed to somehow occur. In summary, Licensee 1) has provided conservative calculations which Intervenor even
, accepts as adequate assurance that a twelve element configuration of stainless-steel clad TRIGA fuel elements cannot achieve criticality,2) has provided actual historical data for an
AFRRI core loading to support its claim that it takes approximately 69 stainless-steel clad TRIGA fuel elements to achieve criticality, 3) has, in its 19 Jan 81 Memorandum, implicitly provided an indication of the extreme difficulty required, and thus the extreme unlikelihood, of establishing an unrestrained critical or supercritical configuration of stainless-steel clad TRIGA fuel elements in the reactor pool, and 4) submits reasons and justificution why a fuel element storage rack accident, that somehow manages to result in a criticality excursion at the bottom of the reactor pool, would have no deleterious consequences to the staff or the public. 0 ff
- x. sagvps, Ja 0d.
Sworn to and subscribed before me on this .7,56 day of 7A,1983. e, -
._._ -G qw OfA
% / 714 T. J 7wq FA e
'a eNl. he-r"w $ n :., SCIENTIFICSiUPORTDEPARIMENT MEMORANDUM FOR RECORD: 19 January 1981
SUBJECT:
Nuclear Criticality Safety Analysis of Hypothetical AFRRI TRIGA Fuel Element Storage Rack Accidents
. 1. An analysis was performed to substantiate that a criticality excursion would not result in the unlikely event that a fully-loaded AFRRI fuel element storage rack were to fail.
i
- 2. For the purposes of analysis, it is conservatively assumed that when the storage rack fails, all twelve fuel elements contained in the rack escape and fall to the bottom of the pool. In addition, it is conservatively assumed that the twelve fuel elements come to rest at the bottom of the pool in the I
most reactive neutronic configuration possible. Moreover, it is con-servatively assumed that the optimum configuration of fuel elements at the bottom of the reactor tank is fully reflected by water over a complete solid angle of 42 storadiana even though only 27 steradian water reflection would actually exist. l l 3 Fuel elements used in the AFRRI reactor are standard stainless-steel clad TRIGA elements containing U-ZrHy,7 with 8.5 weight percent' uranium at a nominal U 0 enrichment of 20 percent (See Figure 1). Each fuel element contains a nominal maximum 38 grams of U235,
- 4. Figure 2, reproduced from TID-7028 (1) , is based on experimental and analytical data and indicates that the minimum critical mass, merit.' ' #
- heterogeneous, 20% enriched, fully water reflected [30 system in its most reactive configuration, is 1.1 kg of U 35. Since our assumed twelve element configuration contains a total of (12 fuel elements) X (38 grams U 235 / fuel j element) = 456 grams U235, it would have a mass fraction critical, m/m crit.'
l less than or equal to 0.455 kg ' U 235 /1.1 kg U 235 or 0.415. l
. For our assumed system, this conservative assumption not only takes into consideration an optimum reactive geometry but also neglects parasitic neutron capture in the stainless-steel clad, Sm-Al burnable poison wafers, etc. and assumes that the graphite end reflectors are replaced by water - a more effective neutron reflector.
2-5
4 Using: * ' crit. ,( }
. indicates $[b o=ur assumed system would have a k,ff g 0.746. Therefore, even with the application of the most conservative assumptions, our assumed sys-ten would still not achieve criticality. In fact, if our assumed system had a k,ff = 0.746, then it would be 'subcritical by more than $36.00 (assumes g,ff=0.007).
235 N Based on the minimum critical mass, merit , value of 1.1 kg U obtained 35 fuel loading per ele [nent of 38 gm U235, a minimum of from Figure 2, and a U l 29 AFRRI TRIGA fuel elements arranged in an optimum neutronic configuration would be required for a criticality excursion (*$ .09) to occur.
- 5. Verification of the conservatism of this analysis is provided by data in RSD 5-8(3) . That is, experience has shown that, during actual AFRRI core leading,N 69 stainless-steel TRIGA fuel elements (s2630 grams U-235) are required to achieve criticality. Therefore, since the AFRRI core lattice arrangement is very close to the optimal. neutronic geometry for TRIGA fuel elements, the results of this criticality analysis are conservative by a factor of d 2.4 on a fuel element as well as a U-235 mass basis for criticality.
- 6. In summary, a hypothetical AFRRI fuel element storage rack failure is analyzed from a nuclear criticality safety standpoint. Conservative assumptions are applied wherever possible; yet k,ff and Wm crit. '#
- system are found to be no greater than 0.746 and 0.415, respectively. As a
, result, there is no possibility of a criticality excursion in the unlikely event that a fully-loaded fuel storage rack were to fail in the AFRRI TRIGA reactor facility. W
,/
3 Encis
- /
EP A T ,J.
- 1. Fig. 1 Capt, SAF c 2. Fig. 2 Research Reactor Operations Officer 3 References - Radiation Sources Division Scientific Support Department z-G
r f : [ .- e STAINLESS STEEL TOP END.PIXTURE STAINLE15 9 STIEL a SPACER l-GRAPMITE
. . ir q-l '
.01$ IN.
8URNASLE PQl$0N i
$AMAR!UM wAPER
! ^ @ WRANIUM *
! ,c$;;,;=
CLAODING l TURE f JwT%'eT!"
.+.
j igi. 8 l URANIUM
! TNICKNESS ! V* -'
0.011N - I : 23.31 IN I i l o.1: iH. ZIRCONIUM 1S IE s i l
=*= 1.43 IN.
-c e - 1.4 IM. !
I
'P G
.015 IN.
BURNA8LE POISCN 5AMARIUM WAPER t I l L' STAINLESS STEEL. N
- 3.4A IN.
fx"Tua \GRAPNITE I r 15 h ir . I u ASSEMBLED 015A55EM81.ED Figure 1. Standard AFRRI TRIGA Fuel Element Onel L
. . . ' .$ . .~ .e e . .
60 - - . - - i l lug' l l l l !..a l i LIMITING CRITICAL ENRICMMENT, HETEROGENEOUS
.._{ . 4_ y J j;i> !
A0 lI --p a LIMITING CRITICat, E*JRICHMENTl HOMOGENE0uS !!l
,,' m . . . . - . ,... .j..!
; y \\ '1
" l 20 .
-- - - - - - - - -- 6 '!
\\ L l
\. t
, \\ , l ;
- \\ ,
i ;!i 3 \\ ! I iii w to '
- - -~ . -- -
-*>r o ,i i 4 . . t .4
' ' ; \-i+p g a
'+.. . -.
. _ , ._, j y d~ s I I* \ \h ! l ... - t ... . ..., . .
ln E *\ Y\ . ,l. ' I
- d. . ... . .... l . _ .. j
~
, 4 l l . \. p HOMOGENECUS, WATER ..REFLECTED . . . . ,i ..
9 7._ 71 t- . - A ' \.- = =. 7, . . , i
. . _ ., ; ; 5 r.
l
$ MCMcGENEcus, UNREFt.ECTED ! '
2 -. l%ai . _ I. _ ( l ' i .
, , ; 7 7 '. .
HETEROGENECUS, WATER REFLECTED I isA
!I r!
.-- .._..._7...
l __ 4.
. .' , . . j
, l 0.9
~% . .
- t 4
J...._. 0.6 l ~~--_-r-p. _. [..~.{ _.~ % L
}
1 y.--- d; , 0.5 t 2 5 to 20 50 ai . URANIUM ENRICHMENT, WT*. U 235 Fig. 2 Minirmim ci*itical mass as a function of U2M enrichment in hydrogen-moderated systems I j . l "2, .i
-_ -.__g- _
~,, ., . .. .n.
i REFERENCES
- 1. Paxton, H.C., Thomas, J.T., Callihan, D. , and Johnson, E.B. , Critical Dimensions of Systems Containing U235, p ,239, and U233, TID-7028, Los Alamos Scientific Laboratory and Oak Ridge National Laboratory, Oak Ridge, TN, June 1964.
- 2. O' Dell, R.D. (editor), Nuclaar Criticality Safety, compendium of in-formation presented at the Biannual Nuclear Criticality Safety Short Course in Taos, NM by the University of New Mexico, May 1973, published by Technical Information Center, Office of Information St.rvices, U.S.
Atomic Energy Commission, Washington, D.C., 1973 3 Radiation Sources Division Instruction, RSD 5-8,.AFRRI/SSRS. y Y 0
/
Licensee's response to Intervenor's First-Round Interrogatory #3
- 3. State the source (s) you relied on for your statement in the HSR that it takes approximately S7 closely packed fuel elements to achieve criticality.
- Answer to Question 3.
Answered by: Sholtis, Moore, Smoker A. His reference of experience is contained within AFRRI's internal Radiation Sources Division Instruction, RSD 5-8, " Reactor Core Loading and Unloading Procedures" and states that, "AFRRI-TRIGA Core II(stainless steel clad elements) attained criticality with 69 fuel elements,2630 grams Uranium-235." his statement is based en actusi core De loading experience at AFRRI using the standard 1/M approach to critical procedure. actual number of AFRRI-TRIG A fuel elements required to achieve criticality in the core may vary slightly (i.e.~1 to 2 fuel elements) depending on the loading order actually used. B. RSD 5-8, ' Reactor Core Loading and Unloading Procedures," AFRRI/SSRS, 27 March 1981. A copy of this document is on file with the USNRC, Region I Field Office. C. See general statement. D. See general statement. E. See general statement. l l l O t
/-/3
{l l L
4 RADIAT!W SOLICES DIVISIN 27 March 1981 INSTRLUTIN NilVEER 5-8 c-REAClut ERE IGDIE AND INIIMDIE PIOQDURES
- 1. Purpose. To set forth the procedures to be followed by the Reactor Branch staff in the complete loading and unloading of the AFRRI-1RICA reactor core.
- 2. Applicability. The provisions of this instruction are applicable to the Reactor Branch staff, and the Health Physics Division staff.
- 3. Cancellation. RSD Instruction 5-8 dated 14 December 1976 is hereby cancelled.
- 4. General.
- a. The reactor core load'ing and unloading procedures contained herein apply
, to the preparation phase as well as the actual loading and unloading phases. - These procedures are based on two (2) separate core loadings and one core unloading of the AFRRI-1RICA reactor. Buphasis is placed on following the procedures specified herein to insure continuity of operation and retention of experience within the Reactor Branch.
I
- b. All activities associated with either the loading or unloading of the reactor core will be recorded in the Reactor Operations Logbook.
l l c. The miniman nmber of personnel that will' be required is (1) Physicist-in-Oarge (PIC) or his designee, (2) Olef Supervisory Operator (C90), l
.C (3) One NIC licensed reactor operator, and (4) Health Physics Division representa-tive.
- d. A daily Startup Oecklist will be empleted prior to the novanent of any fuel elenents.
l-
- e. An approved Special Work Pennit wiQ be initiated orior to the movenent of any fuel elenents, if a.d a.s e.pm.J../p -
- f. If any new fuel elenents are t be used, each elenent nust be inspected when received at AFRRI. Each elenent will be renoved fran its shipping container, cleaned, and inspected for visual defects. Length and bow measurements nust also be made and recorded. Snears of the elenent cladding for alpha contanination nust l
be perfonned by the Health Physics Division representative prior to being handled by Reactor Branch personnel.
- g. If any new thennocouple elenents are to be used, a thennocouple calibra-tion will be perfonned. The fuel elenent will be placed in a water bath, and Bnf readings will be recorded over the range 20-100 degrees Centrigade.
l
- h. At no time will more tl'an six (6) new fuel elenents be out of their l shipping containers and on the reactor roon floor level.
~
- i. The Physicist-In-O arge or his designee nust directly supervise all sequences of loading and unloading the reactor core.
v
e- RADIATIN SOUIEES DIVISICN 27 March 1981 INSTRUCTICN NLhBER 5-8
- j. An NEC licensed reactor operator will continuously observe the nuclear instrtrnentation at the control console during all movenents of control. rods and fuel elements. 4
- k. No fuel elenent which has experienced burnup in the core shall be renoved fran the reactor pool unless at least two (2) weeks have transpired since its use in'the core.
- 5. Nuclear Instrinnentation.
. a. The following nuclear instrtunentation is the mininun required for a reactor core loading:
(1) Two ionization chanbers will be located outside the core shroud, along the core centerline, and adjacent to core positions F-4 and F-12,-respec-tively. The readouts for these chenbers will be picomuneters, or equivalent. (2) One W3 or fission chart >er will be located outside the core shroud, along the core centerline, and adjacent to core position F-8. The readout for this chanber will be a scaler unit.
- b. The mininun nuclear instrumentation required for the unloading of the reactor core is:
(1) One W3 or fission chanber will be located outside the core shroud, along the core centerline, and adjacent to core position F-8. The readout for i this chsober will be a scaler unit.
- c. An operational check of the channels will be made as follows prior to the movanent of any fuel 'elenent.
(1) A neutron source (3-5 curies) will be placed in the neutron source i holder, and an increase in the readings will be observed on all channels. (2) The neutron source will be renoved fran the neutron source holder and the readings will be taken and recorded.
. (3) Replace the neutron source in the neutron source holder, and then generate a bias curve for the startup channel ' identified in 5.a.(2) or 5.b. above as appropriate. Record all channel readings with the sourco. These measurenents will be performed several times in order to obtain reasonable reproducibility.
These readings will be the basis for future calculations of source nultiplication only in the loading of a reactor core. The neutron source reading will be the difference between the readings with and without thc neutron source in place in the reactor core. l , V Page 2 of 7 i
RADIATION SOURCES DIVISION 27 March 1981 y INSTRUCTION NUMBER 5-8
- d. The nuclear instrumentation will be turned on and allowed to stabilize prior to
. the movement of any fuel elements, or making measurements of source effect.
- 6. Cc.re Loading.
~
- a. A 1/M curve is obtained by plotting the inverse multiplication vs the amount of fuel added (total amount in the core). The inverse multiplication is the ratio of the source reading to the reading with the fuel added. The loading curve will seldom be a straight line but may be either concave or convex dependent upon the geometry (source-detector distance). Hence, a number of different channels will yield different predictions of criticality. Since not all channels will agree, a conservative approach will be taken and the smallest number of estimated fuel elements required for criticality will be used to dictate future steos.
- b. The fuel elements will be loaded in accordance with Table 1.
TABLE 1 FUEL LO ADING SCHEDULE - STEP # # ELEMENTS REMARKS ADDED TOTAL
#' 1 4 4 Load four thermocouple elements,2 in the B ring and-2 in the C ring.
2 14 18 Complete loading of B and C rings. 3 15 33 Load D ring. 4 15 48 Load E ring positions 1,2,4,6,8.9,10,12, 14,16,17,18,20, 22 and 24. This loading is designed to complete a compact array around the control rods as well as to fill water gaps. 5 9 57 Complete loading of E ring. 6 9 66 Load F ring in positions 1, 5, 9, 13, 17, 21, 22, 23, and 27. l
- c. After each step of the fuelloading, perform the following:
. (1) Record readings.
(2) Withdraw control rods 50%
~
(3) Record readings. U Page 3 of 7
~
RADIATION SOURCES DIVISION 27 March 1981 n: INSTRUCTION NUMBER 5-8 (4) Withdraw control rods 100%. (5) Record readings.
. (6) Calculate M,1/M for the step.
(7) Plot 1/M vs # fuel elements. (8) Plot 1/M vs weight of uranium-235. (9) Plot 1/M vs control rod position (50% and 100%). (10) Predict criticalloadings. (11) Estimate worth of the control rods. (12) INSERT CONTROL RODS TO FULL "IN" POSITION.
- d. AFRRI-TRIGA Core I(aluminum clad elements) attained criticality with 72 fuel elements, 2811.33 grams uranium-235. AFRRI-TRIGA Core II (stainless steel clad elements) attained criticality with 69 fuel elements,2630 grams uranium-235.
- e. Continue the loading sequence as detailed below until criticality is obtained, and -
until the excess reactivity is 40-50 cents: TABLE I (Continued) FUEL LOADING SCHEDULE STEP # # ELEMENTS REMARKS ADDED TOTAL 7 2 68 Load F ring positions 19 and 25. - 8 2 70 Load F ring positions 3 and 11.
- f. Prior to loading the core to an operational configuration, the following measure-ments willbe made:
(1) Control rod calibrations using the rod drop techniques. (2) The worth of fuel elements in the remaining iracancies (E and F ring) vs water, taken one at a time. (3) Estimate the core configuration for an excess reactivity of approximately
$3.20.
- g. The loading sequence will continue in order to attain a critical configuration with the transient rod in the DOWN position. This is the basis for the excess reactivity estimate of approximately $3.20.
V Page 4 of 7
RADIATICN SOUICES DIVISICN - 27 March 1981 '- INSTRUCTICN NLbEER 5-8 TABLE I (Continued) FUEL IIMDIE SWEDULE , STEP # # ELBENTS IBIARKS NTm TUPAL 9 2 72 Load F ring positions 7.and 15. l 10 4 76 Load F ring positions 2, 14, 18, and 29, i Record critical rod Bank position; l Calibrate the lower portion of the !- transient rod (0-25%) via the positive-period technique. 11 4 80 Load F ring positions 8, 10, 24, and 30. Calibrate the middle portion of the transient rod (25-75%) via the positive-period method. 12 2 82 Load F ring positions,16 and 20, and this should carplete the operational configuration as stated above.
- h. Calibrate the four control rods via the positive-period method, and then h corpute the excess reactivity in the reactor core (K-excess nust not exceed
$5.00).
l 1. Carplete the core loading, insuring that the K-excess does not exceed
$5.00.
' TABLE I (Continued)
, i l EUEL IDADIE SWEDULE STEP # # ELBENTS REMARKS Nrm 'IUrAL 13 5 87 Load F ring positions 4, 6, 12, 26, and 28.
- j. Recalibrate the four control rods via the positive-period method, and then carpute the K-excess reactivity in the reactor core.
s
- 7. Core Unloading
- a. The reactor core will be unloaded starting with the F ring and ending with the B ring.
- b. The fuel elements will be individually renoved fran the reactor core, identified by serial ntrnber, and placed either in the fuel storage racks or a shipping cask.
V
- c. If the fuel elenents are to be loaded into a shipping cask, the follow-ing actions will be taken in preparing the shipping casks for loading:
_ _ _ _ _ Page 5 of 7__ _ _ __
# 27 March 1981 RADIATION SOURCES DIVISION j INSTRUCTION NUMBER 5-8 (1) A radiological survey will be made of the shipping cask voon arrival and-
. before it will be removed from the truck.
(2) The cask will be moved from the truck to the Prep Area. (3) The hatches, which provide access from the Prep Area to the Reactor Room, will be opened and the lifting hook to the power hoist lowered to the Prep Area. (4) The power hoist will be operated in accordance with RSD Instruction 5-5. (5) The lifting yoke will be attached to the cask and the cask lifted to the Reactor Room. (6) The lid to the cask will be removed. The cask w'ill be monitored by the Health Physics Division representative while the lid is being removed, to insure that no radioactive materialis inside the cask. (7) The inside of the cask will be smeared for gross aloha and beta contamination. (8) The inside of the cask will be vacuumed. The inside and outside of the cask willbe washed down The water drain line on the cask will be checked to insure that it is not blocked. Also verify the operability of the pressure relief valve and the temperature sensing thermocouple. (9) If more than seven elements are to be loaded into the cask, it will be necessary to verify that a thermal neutron poison is present in the cask to prevent the i loading of a critical mass. (10) Move the cask by crane from the reactor deck and position the cask in the reactor pook
- d. Load the cask with up to as many fuel elements as allowed by the license for the cask. If grid index markings are present in the cask, record which fuel element is placed l In which grid position.
- e. Lower the lid to the cask into the pool, place the lid on the cask, and secure the lid.
- f. Raise the cask from the pool, drain the water from the cask into the pool, and then dry the cask off. The cask will be monitored while being removed from the pool to insure that no radiation hazard exists as a result of a weakness in the shielding in the cask. The cask will be smeared for gross alpha and beta contamination.
[ g. An air sample will be taken from the cask to measure the activity of the air. The data from all radiological surveys will be recorded.
- h. After the air sample has been taken, observe the temperature and pressure inside the cask until the temperature and pressure reach an equilibrium.
v Page 6 of 7 l
._- . _ _ _ _ ~ . _ _ _ - . _ _ _ _ _ _ _ _ _ . . _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
RADIATIW SCLBCES DIVISIW 27 March 1981
#- INSTRUCTIN NlhBER 5-8
. 1. Label the cask accordingly and caTplete the appropriate paperwork either
- for tmporary storage or for transporting.
- j. Move the cask to either a tmporary storage area or to the truck for transporting. If the cask is to be placed in taiporary storage, a criticality nonitor nust be available in accordance with 10 70.
s a / 6:cb
/JOS / ' ,
[/ Reactor Branch 01ief anc Physicist-In-01arge (effective 13 Oct 1981) td - IENALD R. SW3ER er MAJ, EN, USA Q11ef, Radiation Sources Division O e o e (7)
O h ATTACHMENT 4 i I i i l l m l l
AFFIDAVIT OF JOSEPH A. SHOLTIS,3R. Joseph A. Sholtis, Jr., being duly sworn according to law, deposes and says: First, it should be pointed out that experiment failures with an assumed concurrent malfunction of confinement safeguards are explicitly addressed in the AFRRI SAR on pages 6-8 and 6-9, section 6.2.5. (A copy of this SAR analysis section is provided as an attachment to this affidavit.) Therefore, such an accident is not, of a "different kind" than those that r a_re treated in the AFRR1 SAR. This analysis considers the worst-case experiment failure at AFRRI which involves the irradiation of 20 liters of Argon gas at a steady state power level of 1.0 Mw(t) for 1 hour, and results in the production of 5.6 Curies of Ar-41 gas -- all of which is presumed to be released to the unrestricted area as a result of the experiment failure concurrent with a presumed total failure of confinement isolation system safeguards. At this point no probability estimates for such an event were even considered nor were any mechanistic ways of achieving such a series of events considered. That is, the worst experiment failure with total release to the environment was simply assumed to occur without regard to how or with what probability. The associated consequences to the public for this hypothetical. worst-case experiment failure with complete release to the unrestricted environment involves a total whole-body dose of 2.7 mrad to an individual at or beyond 25 meters from the AFRR1 Facility, i.e. at or beyond the AFRRI site boundary. This is hardly an accident having " greater severity" than the designated DBAs for the AFRRI reactor facility which are treated in the AFRRI SAR. Licensee agrees wholeheartedly with the Intervenors that experiment failures can indeed occur and that malfunctions also can occur, but not necessarily concurrently with any
credible probability. However, we would be remiss by not trying to. reduce their likelihood of occurrence, particurlarly for concurrent failures. We would similarly be remiss by not trying to limit the consequences of such a series of events, should they nevertheless occur
- and result in a release to the environment. AFRRI has, in place, an extensive surveillance,
- testing,' and preventive maintenance program to detect equipment malfunctions and failures; in most cases, prior to reactor power operations taking place. . As an example, numerous component, subsystem, and system preoperational functional surveillance checks,
= specifically designed to verify proper system functionality to ensure safe reliable operation,
- are required and must be performed and checked daily before any reactor power operations are permitted. AFRRI also has, in place, additional surveillance checks and preventive maintenance tasks that must be performed weekly, monthly, quarterly, semi-annually, and annually. In general, these are all identified and required within the AFRRI Reactor Technical Specifications. Such a system of surveillance, testing, and preventive maintenance does not preclude even single equipment failures or malfunctions but it does reduce their likelihood and provides reasonable assurance that such failures will be detected in a timely fashion.
Almost all of the malfunctions cited by the Intervenor under this contention were detected i by an operator during the normal reactor preoperational start-up checkout procedure and, , thus, before any power operation actually took place. In addition, almost all of the malfunctions cited by the Intervenors in this contention have nothing whatsoever to do with ! the generation of a source term, the operability of confinement isolation safeguards, or a release to the environment. These specifics will be addressed in more detail below where each malfunction cited by the Intervenor under this contention is discussed individually. 1 Also from an accident probability minimization viewpoint, redundant and independent "back-up" systems are often provided such that aj must fall before functionality is actually lost. 4 ______,__,.._.m.,_ , _ _ _ _ _ _ . . _ . . . _ _ _ _ . ,
This is the case for several of the system malfunctions which Intervenor cites in this contention. This point will also be appropriately discussed in more detail when each of the Intervenor's cited malfunctions are addressed individually below. From the standpoint of limiting the consequences of a release to the environment, should one occur, for example, as a result of an experiment failure with an assumed concurrent total failure of the confinement isolation safeguards, AFRRI .has, in place, a system involving a body of experts for reviewing, approving, and limiting irradiations of various materials on a case-by-case basis for each material or experirnent. This reviewing and approving body is the AFRRI Reactor and Radiation Facility Safety Committee (RRFSC), which is required under the AFRRI Technical Specifications, is composed of technical experts from within and outside of AFRRI. Prior to any experiment utilizing the AFRRI Reactor, the experiment and experimental protocol must be presented before the RRFSC for review and approval. The AFRRI RRFSC subsequently must issue approved "special" or
" routine" authorizations for the intended reactor use. These authorizations establish limits on the quantities of materials which can be irradiated, based on the safety and radiological implications of each, should an experiment failure occur together with a failure of cor.finement isolation such that the entire inventory of activated materials is released to the environment. In addition, these authorizations also establish limits on the use of the various experimental apparatuses and exposure facilities that exist or are proposed for use.
4 Intervenor has admitted that multiple failures or malfunctions are generally required to , , achieve conditions for a release to the unrestricted environment. (See transcript of NRC Staff's deposition of Dr. Irving Stillman on 18 Dec 1982 in New York on page 83 line 4 through page 84 line 15, inclusive, and also on page 85 lines 7 through 24, inclusive.) Further, the Intervenor has admitted that such multiple failures are successive and independent and
l
- must occur concurrently in time. (See page 83 lines 11 through 13, inclusive, of the
~
transcript of Dr. Stillman's Depositon on 18 Dec 82 in New York.) Moreover, the Intervenor has indicated that it has not gone through the thought process to postulate a scenario of events involving malfunctions that would ultimately lead to a release to the unrestricted environment. - (See page 79 line 7 through page 85 line 24, inclusive, of the transcript of Dr. Stillman's Deposition on 18 Dec 82 in New York.) Intervenor has repeatedly failed to supply information to the Licensee upon request showing the relationship between each of the Intervenor's cited malfunctions under this contention and the generation of a source term, their relationship to confinement safeguards, and their relationship to a release to the environment - given that each cited malfunction is presumed to occur. (See, for example, Licensee's Motion to Compel served 15 Jan 1982 and as supplemented 24 Feb 1982.) In fact, to this day, Intervenor has still not provided this information even though Dr. Stillman promised to provide it when asked during his deposition in New York on 18 Dec 1982. (See page 78 line 7 through page 85 line 24, inclusive, of the transcript of the deposition of Dr. Irving Stillman on 18 Dec 82 in New York.) Licensee submits that single independent failures can occur but their probability is reduced by virtue of performing routine scheduled preventive maintenance, and their impact is considerably reduced by virtue of having redundant and independent "back-up" systems and by performing routine recurrent surveillance checks and tests. The probability of single independent failures is admitedly low but nevertheless they are expected, but the probability of multiple independent failures is certainly well below that for single failures which makes them extremely unlikely events, particularly if they are to occur concurrently in time. Still, if they should occur and an environmental release results, limitations set-down by the
- RRFSC would limit the available release source term and, therefore, also limit the
- - -- . . . - - - . . . ~. . . - . _-- .
E 4 consequences of such extremely remote sequences of failures leading to an environmental. release. More importantly, releases' due to experiment failure together with confinement isolation failure have been addressed in the AFRRI SAR. Next, we will focus on each of the malfunctions which Intervenor cites 'as supporting
- f. -
- evidence in this contention. The first of these is a breach of containment caused by missing i
r rubber gasket sealing material on the double doors to the corridor behind the reactor control
- room in 1978. This oversight did constitute a violation of Licensee's' Technical L i Specifications and a notice of violation was issued. However, Licensee was in the process of i installing the gasket material when the condition was noted by NRC, and the reactor room .,
l was still capable of establishing and maintaining a negative pressure even without the gasket I l material in place so long as the doors were closed and the reactor room was isolated by- , closing of the ventilation dampers. Therefore, although this condition did represent a literal i violation of the Technical . Specifications, it did not in actuality significantly negate the f capability of the confinement isolation safeguards. Also, if one assumes that confinement isolation was in fact compromised at this internal doorway point, a second set of doors to corridor 3106, also behind the reactor control room, would have confined any airborne source term to the reactor facility confines since this second set of doors did and does still i have gasket sealing material installed in place. Nevertheless, any Technical Specification
- l. violation, particularly a condition involving missing rubber gaskets on confinement area doors or any'other potential compromise to the confinement isolation boundary, is viewed by
! the Licensee as a valid concern. In this regard, steps have been taken to remedy such ;
, situations and ensure they will not occur again in the future. (See NRC Inspection reports on the AFRRI reactor facility under Docket 50-170 since 1978.)
l . i ! The second malfunction Intervenors cite under this contention is a failure, on 26 August I t i i
, m..-%-,_, __. .--4_ __ _ , _ . -
, , , , . . ,mm,..,-m ,,,,m,-,_,-- ..,,,,mc.-y, ,%-.-..~,--,ym . . - , , -,- -- -
1975, of the reactor room ventilation dampers to close when the Continuous Air Monitor was
- alarmed.' This malfunction has no relationship with the generation of a source term. The reactor room ventilation dampers'are designed to close (by a fail-safe air-actuated solenoid)
- automatically upon receipt of a high-level alarm of the reactor room- Continuous - Air Monitor (CAM) or upon receipt of a manually-initiated signal by an operator in the reactor control room. Part of the normal daily preoperational reactor startup checkout procedure involves manually initiating a high-level CAM alarm (artificially) to check the operability of the CAM as well as the closure of the reactor room ventilation dampers. On this particular day, 26 August 1975, the operator performed this preoperational checkout item, but the dampers failed to close. As a result of this failure, reactor power operations did not occur and were not permitted to take place until repair was effected. This is a prime example of i
exactly why we have such preoperational checks. As a result, this failure could not have contributed to a release to the environment, since it was detected and no operations to produce a potential source term were performed. Even if it had not been detected and power operations did occur and a source term somehow were generated,-~ numerous other radiation monitoring devices are available, with audio and visual alarms, to detect the event and alert the operator so that manual damper closure could be effected. This is true since the operability of the dampers was unaffected by this malfunction; the only effect of this malfunction was the loss of an automatic signal to the ventilation dampers for closure. The third malfunction which Intervenor cites is a failure of the in-pool lead shielding doors to stop opening at the fully opened position in August 1976. This malfunction has no relationship to confinement isolation safeguards, the generation of a source term, or a a release to the environment. The purpose of the in-pool lead shielding doors, when closed, is
, to provide adequate gamma photon attenuation or shielding such that one exposure room can be safely occupied for experiment set-up while the reactor is operating at power at the
other exposure room at the opposite end of the tank. Since there is a potential for physical contact between the core and the lead shielding doors in certain regions of the pool, microswitches, interlocks, slip clutches, a core shroud, a TV monitor, and administrative controls have been installed / established to either preclude such contact or, in the extremely unlikely event that contact nevertheless does occur, to minimize fuel element damage. The malfunction cited by the Intervenor in this case precludes operation of the reactor at power due to an additional and separate interlock that prevents the supplying of current to the standard control rod drives and prevents supplying air to the transient control rod drive unless the lead shielding doors are either fully open or fully closed. In addition, this malfunction precludes moving the core dolly into the mid-pool region where physical contact might occur since an additional interlock system only permits core dolly travel into the mid-pool region when the lead shielding doors are fully open. As a result, there was no potential for a source term or release to the environment by virtue of this malfunction. It should also be pointed out that this malfunction was immediately detected by the operator and repairs were effected before operations were permitted to resume. The fourth malfunction cited by the Intervenors in this contention involved a reactor core position safety interlock malfunction on Feb 1,1973. This malfunction, like the third i discussed above, has no relationship with confinement isolation safeguards, the generation of a source term, or a release to the unrestricted environment. This malfunction also was detected immediately by the operator and repairs were effected before operations were permitted to resume. This malfunction could have resulted in the core shroud physically contacting the lead shield doors near the center of the pool if operator error had i S additionally been involved. However, the core dolly drive slip clutch, which was operational,
, and the core shroud would have minimized any impact damage and not permitted the fuel elements to have been contacted at all. Even if a clad failure were assumed to result from
i such a malfunction and operator error, a source term would have been generated but no S pathway to the environment would have been provided since the confinement isolation . system was unaffected, and even if it too were presumed to fail, a clad failure event has been addressed in the AFRRI SAR and its consequences are minimal. While one can never
- - rule out such a release, its likelihood is made extremely remote by virtue of.the protective systems and design features provided. Nevertheless' the ultimate consequences of a release are analyzed within the AFRRI SAR.
The fifth malfunction that the Intervenor cites in this contention involved a malfunction of High Flux Safety Channel #1 to initiate a scram signal on March 15, 1980. Regardless of what the Intervenor actually says in this contention, this malfunction was detected by an r AFRRI reactor operator, not an NRC inspector, and the detection occurred during a normal
- daily preoperational startup checkout. (See Intervenor's initial and supplemental response to Licensee's first-round interrogatory #16e.) As a result of preoperational detection of this malfunction by an operator, power operation did not take place and, therefore, no potential for an experiment failure or the generation of any other source term existed. Just as in the third and fourth malfunctions discussed above, this malfunction also has no relationship to confinement isolation safeguards, the generation of a source term, or a release to the environment. The purpose of high flux safety channels one and two, which are redundant 1
i and independent, is to provide redundant readouts of reactor power level and to intiate a , reactor scram if a 1.1 MW(t) steady state power level is attained. Each day that reactor t power operations are planned, the reactor operator is required to first perform a i preoperational startup checkout procedure which involves, in part, placing test signals on j' safety channels one and two separately to simulate a power increase and to check that a scram signal is generated by each channel at or below a 1.1 MW(t) indicated power level. It 1 was during just such a preoperational check that this malfunction was detected. Even if it s 1 J w r,- , v-ee .m. ,-.v- - ,w, y n ,s.y.,-..,-- , ,,. - , -~%-- - . - --, .,w--.-w---- - - , --,w, .. - - - - , -- ,.~ - , , --.-,. . - - - , , , , - . - . .-,e-- --,--,---.-..-,-,.,-,-#---e. . , , -
l i. had not been detected and power ' operations had been conducted and a power level of 1.lMW(t) was attained by virtue of operator error, high flux safety channel #2 would still have generated a scram at 1.1 MW(t) or, if it too somehow.were presumed to fail, two independent and redundant fuel temperature channels would.still have been available to i scram the reactor upon attainment of 5000C fuel temperatures and, even if they too were somehow assumed to be malfunctioning, the maximum steady-state power attainable with the AFRRI reactor is only approximately 1.4 MW(t) which has an associated fuel temperature of less than 6000C and, therefore, no damage, source term, or release would ' ! be expected. And even if a release to the reactor room did somehow occur, confinement isolation was still available, and even if it too were assumed to fail, the consequences have l i been determined in the AFRRI SAR and they are minimal. This represents an incredible series of events. l' The sixth malfunction that Intervenor cites in this contention involved a reactor exhaust system malfunction on August 9,1979 caused by an electrical fire in the EF-1 cubicle of the motor control center, in turn caused by a power surge due to a faulty transformer. ~.his malfunction, like numbers 3,4, and 5 above has nothing whatsoever to do with confinement isolation integrity, the generation of a source term, or a release to the environment. This event simply involved a minor electrical fire (not associated with the reactor or its safety systems) and a resultant loss in ability to draw air from the reactor room through high-efficiency particulate air filters and out the AFRRI stack. It had no impact on the ability of the confinement isolation dampers to close nor can it contribute to an experiment failure or the generation of any other source term. This particular malfunction occurred while the reactor was g operating, but if it had been operating, stack flow audio-visual alarms would have alerted the operator that no stack flow was being provided and the operator would have been forced to manually scram the reactor under existing procedures. Even if the operator failed to scram the reactor in the face of both audio and visual alarms, no potential for a l [.-..-.----. _-_ . _ - . . , _ _ . - - - _
-~ . .
source term exists, and even if one was presumed to somehow occur, the confinement isolation system was still operational, and if it too somehow. managed to fail, the - consequences of the release have been determined and they are minimal -- another incredible series of events. The seventh malfunction Intervenors cite in this contention involved a malfunction of the fuel element temperature sensing circuit #2 caused by a floating signal ground on August 1, 1979. In order to simpHfy the discussion from here on, only major pertinent points will be identified henceforth. This malfunction:
- 1) was detected by an operator during a normal preoperational checkout so that power operation did not take place.
- 2) is protected by a redundant and independent fuel temperature sensing channel
#1 and other available safety channels.
- 3) does not have anything to do with confinement isolation, generation of a source term, or a release to the environment.
The eighth malfunction that the Intervenors cite under this contention, involved a malfunction, in July 1979, of the pool water level sensing float switch caused by wear on the Jacketing around the wires leading to the switch. This malfunction:
- 1) was detected by an operator during a normal surveillance check; power operations were curtailed until repairs were effected i
- 2) is protected by numerous radiation monitoring devices that would also
, indicate, via shine, any substantial loss of pool water should that occur
- 3) has nothing whatsoever to do with confinement isolation, generation of a source term, or a release to the environment.
The ninth malfunction that Intervenor cites in this contention involved a malfunction of the Reactor Room CAM to initiate a signal for closing the reactor room ventilation dampers on 26 August 1975. This malfunction is the same malfunction that Intervenor cites earlier and which is discussed above as the second malfunction. This malfunction:
- 1) was detected by an operator during a normal preoperational checkout so that reactor power operations did not take place
- 2) is protected by numerous other available radiation monitoring and detection devices, with audio-visual alarms, and by the capability for an operator to manually initiate damper closure
- 3) has nothing whatsoever to do with generating a potential source term or negating the physical operability and integrity of confinement safeguards.
The tenth malfunction that the Intervenor cites under this contention involved a malfunction of the fuel temperature automatic scram system oa January 29,1974 caused by a build-up of high resistance material on the mechanical contacts of the T2 output meter. This malfunction:
- 1) was detected by an operator during a normal preoperational checkout; no operations were permitted until repair was effected
- 2) is protected by the redundant and independent fuel temperature sensing channel #1 and other available safety channels
- 3) can no longer occur with the existing new console design
- 4) has nothing whatsoever to do with confinement isolation, the generation of a source term, or a release to the environment.
The eleventh and last malfunction that the Intervenor cites under this contention involved a malfunction of the reactor core position safety interlock system on February 1,1973 caused
W by a faulty de-energizing relay. This malfunction is the same malfunction that Intervenor cites earlier and which is discussed above as the fourth malfunction. This malfunction:
- 1) was detected immediately by the operator; no operations were permitted until repair was effected -
- 2) is protected by the core dolly drive slip clutch, the core shroud, TV j monitoring, operator action, interlocks, and administrative controls
- 3) has nothing whatsoever to do with confinement isolation, the generation of a source term, or a release to the environment.
This discussion points out the designed protective features and redundancies provided for the reactor facility. It also demonstrates the extreme improbability, and often impossibility, of actually getting a release to the _ unrestricted environment. Moreover, it illustrates the limitations Licensee. imposes on its own irradiations of materials to restrict the consequences of a release should one occur. l 6Y ~/ H A. S@R. Sworn to and subscribed before me on this Ao? day of Ed ,1983.
% d.o-w **
+ _ __ - _ . ',
Of 6', /7 M g . - i ag _ i t
bection 6.2.5, pp. 6-8 and 6-9 from AFRRI Reactor Safety Analysis Report (SAR), dated June 1981.] 6.2.5 Experiments All experiments performed as part of . the TRIGA reactor operations are
~
reviewed by the Reactor and Radiation Facility Safety Committee and must be authorized prior to their performance. The technical specifications contain requirements that must be met before auch experiments can be performed using __. .the_ AFRRI-TRIGA reactor. Experiments are always supervised by trained, licensed, supervisory personnel. However, failure of an experiment is possible and worst-case conditions can be calculated to determine the postulated consequences. The two worst-case conditions for failure of an experiment could resuit in instantaneous insertion of reactivity or the release of radioactive material from an experiment undergoing activation in the reactor. For an experiment failure in which reactivity could be idded, the worst possible case would be the prompt addition of less than 0.36 4 k/k in either Exposure Room 1 or 2. As discussed for the case of improper fuel leading (Section 6.2.3), the addition of 0.36% 6 k/k would l be within the range of an improper fuel loading condition. Such an addition would ! not result in any damage to the reactor or the fuel. For an experiment failure in which radioactive material could be released i o from the experiment, i.e., activation products, the worst case would be the prompt release of the radioactive material to the atmosphere. An authorized experiment involves the irradiation of 20 liters of argon gas for i hour at a power level of 1 MW. The resulting activation would result in a total Ar-41 activity of 5.6 Ci in the sealed container. If the container should fail and release all of the Ar 41 activity, the resulting total whole body dose would be less than 2.7 mrad to an 1 _--- - ,m, - - . _ - . - -
e individual more than 25 meters from the AFRRI facility (Equation 3, Section 6.3.4.1). The failure of this authorized experiment represents the worst case for radiological consequences from an experiment failure in the. AFRRI-TRIGA reactor. Such a whole body exposure would not represent an undue risk to the health and safety of the general public. N
. m.
S O 5-9 w-f
- - -- m ,__ _.
L ATTACHMENT 5 e
s AFFIDAVIT OF JOSEPH A. SHOLTIS,3R. Joseph A. Sholtis, Jr., being duly sworn according to law, deposes and says: A negative temperature coefficient of reactivity simply means that as the temperature increases, negative reactivity is inserted which acts te reduce / shutdown the fission process. Every reactor designed and operated in this country must have a negative temperature coefficient of reactivity; TRIGA reactors are no exception. For stainless-steel clad TRIGA fuel elements containing a fuel-moderator mixture of U-ZrH 1 ,7, which is the fuel used in the AFRRI reactor, the prompt negative temperature coefficient of reactivity has a value of -1.26 X 10-4 Ak/k/oC or -1.8c/oC. ' This value is contributed to, in a cumulative fashion, by three major separate and independent factors. They are, in order of their importance: 1.) The ZrHx or cell factor,2) the Doppler broadening factor, and 3.) the density decrease or leakage factor - each of which is in and of itself negative with increasing temperature. (See the attached General Atomics Information sheet.) If we define or designate the overall TRIGA reactor's temperature coefficient of reactivity aso(T, then we can write an equation fork which shows its constituent parts or terms as: 4= qrHx+ % oppler+ -cDensity . o v.here: d ZrH isx the ZrHx (or cell) effect term koppler is the Doppler broadening effect term, and Nensity is the dendty decrease (or leakage) effect term.
J Both the Doppler broadening factor and the density decrease (or leakage) factor will individually always be negative with increasing temperatures, regardless of whether the fuel is damaged and hydrogen is presurned lost or not. The Doppler broadening factor is negative and inherent to any thermal reactor system fueled with uranium, which includes TRIGA reactors. Uranium, particularly the uranium-238 isotope, has numerous high value resonance capture peaks in its absorption cross-section at epithermal energies. These resonance capture peaks broaden with increasing temperature, such that fast fission neutrons
~
undergoing moderation toward thermal energies have a lesser chance (as temperature increases) of actually reaching thermal energies (where fission ~ predominantly occurs), without being captured parasitically (via radiative capture) by one of these broadened resonance capture peaks in the uranium. In essence, Doppler broadening acts to reduce the resonance escape probability, p, which is one of the six factors in the six factor formula for k-effective, and thus, introduces negative reactivity in a thermal reactor fueled with uranium as the fuel temperature increases. This factor, which is discussed in detail in numerous nuclear engineering texts, will always be negative, inherent, and unalterable in TRIGA fuel so long as uranium is present. Therefore, this contributing factor to the TRIGA reactor's negative temperature coefficient of reactivity, would be unaffected by the Intervenor's postulated loss of hydrogen from damaged TRIGA fuel and, thus, would always remain negative. This factor alone would ensure automatic reactor shutdown during pulse operations since it by itself is negative and the two remaining factors can be made no larger than zero at best - and even this last postulate would require a negation of the laws of nature. The density decrease (cr leakage) factor similarly will always be negative, inherent, and
. unalterable with increasing temperatures in TRIGA-fueled cores (regardless of whether the fuel is damaged and hydrogen is presumed lost or not), so long as nature continues to ensure
that material _ densities' decrease as temperatures increase. This factor not only involves density decreases in the fuel and the interstitial water between the fuel elements during heatup but also includes expansion effects of the fuel and the fuel elements (and thus the overall core) with increasing temperatures which results in an overall volumetric core increase and the displacement of some of the interstitial water moderator from between the fuel elements, by virtue of individual fuel element expansion, with heatup. These effects, taken collectively, not only'act to reduce the effective moderation of neutrons toward thermal energy (i.e. Es for the moderator decreases with heatup) but also decrease the macroscopic fission cros;-section, Ef,in the fuel (and thUs decrease material buckling, B'm) and also increase the likelihood that neutrons will leak from the core without contributing to fission by virtue of the core volumetric increase with heatup which, in turn causes an !~ increase in the geometric buckling, Bg. Each of these individual effects is negative so that, overall, the density decrease (or leakage) factor must always be negative with increasing temperature, regardless of whether hydrogen is presumed lost from damaged TRIGA fuel or not. Just as in the case of the Doppler broadening factor, the density decrease (or leakage) factor is also sufficient, by itself, to ensures automatic reactor shutdown during pulsing, regardless of whether the fuel is damaged and hydrogen is presumed lost or not, since it (C(Density) is always negative, inherent, and unalterable, and the other two factors cannot
- ever become positive. Morever, as we have demonstrated above, both the Doppler broadening and the Density decrease' factors will always each be negative, regardless of whether the TRIGA fuel is damaged and hydrogen is presumed lost or not, so that collectively they are cumulative in a negative sense, making the TRIGA reactor's temperature coefficient of reactivity even stronger in a negative sense even if the ZrHx (or cell) effect factor is presumed to somehow be forced to its maximum theoretical value of
. zero.
l l i
, _ . _ _m-__
f The zirconium-hydride, ZrHx, (or cell) effect term theoretically could .be forced -to a maximum value of zero (but no larger) by somehow removing or driving out all of the hydrogen from all of the TRIGA fuel loaded in core, but even this action would in no way alter the negativeness of the Doppler broadening factor or the density decrease (or leakage) . 4 factor, and so the TRIGA reactor's overall temperature coefficient of reactivity would still' remain negative regardless of whether hydrogen is presumed lost from damaged TRIGA fuel or not. However, the removal of hydrogen from a TRIGA fuel element will understandably . reduce effective moderation of neutrons to the thermal energy region (where fission predominantly occurs) and this would.'in and of itself, constitute a negative reactivity effect, albeit not associated with a temperature change. In fact, if all of the hydrogen was somehow assumed to be lost or driven from the TRIGA fuel, elements in the core, it would be impossible to even attain criticality. This indicates that the hydrogen contained in the , TRIGA fuel elements contributes to effective neutron moderation and, therefore, is an , integrally important (actually required) moderator in a TRIGA reactor core. Therefore, a presumed total loss of hydrogen from all of the TRIGA fuel elements in core, although it would zero out the ZrHx (or cell) effect term or contribution to the TRIGA reactor's overall negative temperature coefficient of reactivity, would in and of itself constitute a constant and large negative reactivity effect, and even this can not make the overall temperature coefficient of reactivity in a TRIGA reactor ever become zero or positive. Moreover, a i . presumed loss of hydrogen from TRIGA fuel would not permit fission to effectively occur in those elements in which hydrogen was presumed lost. , Up to this point, wholesale hydrogen loss from TRIGA fuel has simply been presumed and l nothing has been said about the feasibility of driving hydrogen completely out of all the fuel
. elements in core (i.e., not just damaged TRIGA fuel elements), even though this is exactly -
what is necessary to force the ZrHx (or cell) effect term to zero in the TRIGA reactor's 1 T
- , , - - . , , _ , - - ,- -.~,, rA- **- , - - - - -- - . ---- =--e - - . *- -- . _ -
overall temperature coefficient of reactivity equation. Let's look at the feasibility of driving hydrogen out of TRIGA fuel on a per-element basis, if, from a worst-case point of view, it was assumed that the stainless-steel clad on a TRIGA fuel element loaded in core were magically removed in total and the reactor was operated such that fuel temperatures in this unclad element were maintained at 6000C (the fuel temperature scram point for the AFRRI reactor), only a small fraction of the hydrogen within this unciad element would actually be lost. This statement is based on the hydrogen equilibrium pressure within TRIGA fuel as a function of temperature. (See Figure 2-9, attached, which is from GA Report # E-117-833, "The U-ZrHx Alloy: Its Properties and Use in TRIGA Fuel," by M.T. Simnad, page 2-13, February 1980.) This attached Figure indicates that at 6000C, the equilibrium hydrogen pressure is only approximately 1 psi. (Note also from this Figure that you would have to go to a temperature of approximately 7750C before a 1 atm equilibrium hydrogen pressure is reached.) Since the core is submerged under approximately 14-18 feet of water, where the pressure exerted by the water on the surface of the fuel at this depth is greater than 1 atmosphere (approximately 1.4 atm), there would be no driving force (except for normal thermal diffusion) to remove significant amounts of hydrogen from a failed TRIGA element. Within a TRIGA element hydrogen tends to migrate from hot to cooler regions and then recombine with free zirconium during power operation. Therefore, since radial and axial temperature gradients are established within individual TRIGA fuel elements during power operation and since the water pressure exerted on an element is greater than the equilibrium hydrogen pressure within an element up to approximately 775-8000C (See attached Figure 2-9), it is expected that hydrogen would simply redistribute within an element (to some extent) from hot to cooler regions, but loss of hydrogen from a failed TRIGA element would be very small since it would be due to thermal diffusion only. This , certainly does not constitute a way of driving or removing hydrogen from a failed TRIGA fuel element in a wholesale fashion. It is, therefore, extremely unlikely that hydrogen could
be removed from failed TRIGA fuel in a gross way. This in turn, indicates that it would be extremely unlikely (and difficult) to even force the ZrHx (or cell) effect term, in the TRIGA reactor's overall temperature coefficient of reactivity equation, to zero - which, even if it were presumed to occur, would still not force the TRIGA reactor's overall temperature coefficient of reactivity to a zero or positive value, i.e., overall, MT would still remain negative, even if the ZrHx term were presumed to go to its theoretical maximum value of zero. Intervenor has admitted that if all the hydrogen contained in the TRIGA cere's fuel elements were presumed lost, criticality could not be achieved. (See page 90 lines 2-5, inclusive, and page 91 lines 4 through 16, inclusive, of the transcript of Dr. Irving Stillman's deposition on 18 Dec 1982 in New York.) This statement alone by Intervenor indicates the importance of the hydrogen contained in the TRIGA fuel as a moderator and, in fact, the requirement of this hydrogen contained within TRIGA fuel for the reactor to operate (via fission). Intervenor has also admitted that this gross effect (i.e. total loss of hydrogen from the TRIGA fuel in core and the associated inability, by virtue of its large negative reactivity effect, to achieve criticality via fission) is contributed to by each TRIGA fuel element assumed to be damaged in which hydrogen is presumed to be lost. (See page 91 line 17 through page 92 line 10, inclusive, of the transcript of Dr. Irving Stillman's deposition on 18 Dec 1982 in New York.) Moreover, Intervenor further has admitted that each damaged TRIGA fuel element, which is presumed to have lost all or part of its hydrogen, contributes to the overall large corewide negative reactivity effect that total hydrogen loss from the entire core's TRIGA fuel elements would introduce. (See page 92 line 11 through page 93 l line 8, inclusive, of the transcript of Dr. Stillman's deposition taken in New York on 18 Dec i 1982.) These admissions by the Intervenor indicate that any damaged TRIGA fuel element loaded in core which is presumed to have lost part (or all) of its hydrogen, will be less and i
.less effective (with increasing hydrogen loss) in contributing to the core's neutron population, power level, and fission density. This indicates that any damaged TRIGA fuel element loaded in core, which is presumed to have lost some (or all) of its hydrogen, will.
have a decreased neutron population, power contribution, fission density, and thus fuel temperature, which becomes further and further reduced as hydrogen loss increases, in comparison with normal undamaged TRIGA fuel elements in core. This, in turn, indicates that locally within the core wherever a damaged TRIGA, fuel element (with presumed hydrogen loss) exists, a constant negative reactivity effect also exists, i.e., this local negative reactivity effect is not temperature dependent. Therefore, overall you would not only have a negative reactivity temperature feedback effect (via the ZrHx or cell effect) as long as some (i.e. any) hydrogen exists within any of the core's TRIGA fuel, but you would also have a negative reactivity temperature feedback effect via the Doppler broadening and density decrease (or leakage) effects, and further you would also have a constant negative reactivity effect in each TRIGA element that is presumed to have lost a significant portion of its hydrogen. Licensee is, thus, at a loss to see how the Intervenor can possibly contend that the TRIGA reactor's negative temperature coefficient of reactivity could ever become zero or positive and, thus, ever fail to shutdown the reactor automatically during a pulse. One last point should be made. If a TRIGA element were to fail (i.e. a clad failure occurs) and hydrogen were presumed to be driven out or lost from the failed element, fission gases would also be released to the pool water. These would be detected via the activation of radiation alarms. Under such a scenario, it is hard to believe that any subsequent reactor power operations would be permitted, i.e. with obviously damaged fuel. And finally, sech accidents involving clad failure ag treated in the AFRRI SAR and, therefore, can not be
considered to be of a different kind or greater severity than those that are treated in the AFRRI SAR.
'A. 5 hog, R. -
Sworn to and subscribed before me on this
#59 day of M ,1983. ,
3 w _ . - L - - _ . ',- _ _ Mpo.' 2', /989 32- 3. f( - i l l l -
9 18 No.404 APPENDIX C 4 A BRIEF DISCUSSION OF THE TRIGA PROMPT NEG TlVE TEMPERATURE' COEFFICIENT OF REACTIVITY Reactors fueled with TRIGA U-ZrH fuel-moderator elements exhibit a strong prompt negative temperature coefficient of reactivity. For the~4stainless steel clad. U-ZrH1 ,7 uel, f the ternperature coefficient is -1. 26 x 10 Sk/k per C. There are several factors contributing to the prompt coefficient as noted below: A RELATIVE MAGNITUDE OF CONTRIBUTING COMPONENTS OF THE PROMPT NEGATIVE TEMPERATURE COEFFICIENT OF TRIGA REACTORS TJ-2'rH1 . 0, Al Clad U-ZrH i ,7, SS Clad (%) (%)
- 1. Cell increased disadvantage factor l
with increased fuel temperature lea. ding to a- decrease in neutron economy 40 60 l 2 Irregularities in the fuellattice due I to control rod positions-essentially 10 10 same effect as 1 above i 3. Doppler broadening of U 238 re sonanc e s - increased resonance capture with .
- increased ftiel temperature 20 15
- 4. Leakage-increased. loss of thermal neutrons from the core when the fuel 30 15 is heated "- ,
l i- *- The low-hydride core is assumed to be reflected by graphite radially, whereas the high hydride core is water reflected radially. The graphite reflector gives ~30% more negative contribution to the leakage component for either core. e
>} -
(Takenfrom: General Atomics Report . E-117-033,8 Tha U ~'rH Alloy: Its Properties and Usa in TRIGA Fuel, ' by M.T. Sinnad, Gener:1 Atonics Corporation, San- Die;o, CA, GA Project !.*o. 4314, Feb
- 1030,page2-13.]
10,000 1000. - m 61
. ~
i w 5 m m - w K a. z w e 100. - o e - o
>=
z - r 3 - ac
. m
.a 3 -
d w I i
- 10. -
I r _ l
~
i l 10 700 800 900 1000 1100 1200 1300 1400
. ZrH l.65
**~
EL-1174 Fig. 2-9. Equilibrium hydrogen pressure over ZrH1.65 versus te Perature 2-13
m h 9 ATTACHMENT 6 e
f AFFIDAVIT OF JOSEPH A. SHOLTIS,3R. Joseph A. Sholtis, Jr., being duly sworn according to law, deposes and says: The Licensee does analyze reactivity transients and a LOCA within sections 6.2.2 and 6.3.3, respectively, of the AFRRI SAR and fails to obtain conditions necessary to achieve even a single fuel element clad failure. In fact, Intervenor has stated in its response to Licensee's interrogatory #24f that clad failures are not expected during a LOCA unless pulsing operation also takes place concurrently. Moreover, the AFRRI SAR, explicitly addresses and-( conservatively analyses the consequences of a fuel element clad. failure accident and . specifically designates such an accident as a DBA. (See sections 6.3.2 'and 6.3.4.2 of the AFRRI SAR.) Although concurrent multiple clad failures are not explicitly addressed in the l AFRRI SAR, due to their extremely low (non-credible) probability, they_ are nevertheless addressed within section 4.3 of the AFRRI Reactor Facility Emergency Plan, and the consequences of such multiple clad failure events can readily be obtained simply by multiplying the consequences of a single element failure, from the AFRRI SAR, by the number of elements presumed involved in the multiple failure event. l l Each of the Intervenor's postulated causal mechanisms for a concurrent multiple clad failure accident under this contention is listed below with a brief discussion of why each is
- l. considered not credible.
l
- a. Material Defects: Not credible due to extremely low probability for concurrent
[ failure. Individual clad failures due to material defect are generally considered to I
- occur as random stochastic failures, therefore, two or more such occurrences
- 1. .. . -- _ . .- - - - --
concurrently in time are viewed as being non-credible. No concurrent multiple clad failure accident due to material defect has ever occurred in the entire history of TRIGA reactors, and even if such an event were to occur, its consequences would be limited since individual clad failures due to material defect have always occurred early in an element's life so that burnup and the accumulation of fission products within such an element would be limited but certainly well below the equilibrium saturation activity utilized for a single clad failure accident in the AFRRI SAR.
- b. Uricontrolled Power Excursion: Not credible since fuel temperatures and pressures necessary to breach the cladding cannot be attained either via steady-state or pulse operations. Note: The AFRRI reactor is incapable of firing pulses in rapid succession as Intervenor claims (see Motion for Summary Dispostion relating to Contention 9 Accidents IV.) The TRIGA reactor's inherent negative temperature coefficient of reactivity will ensure automatic reactor shutdown before fuel damage results. Even if all of Licensee's authorized K-excess of $5.00 were available and inserted in a, step fashion, fuel temperatures and pressures would still remain well below the point where clad failure might be expected. (See section 6.2.2 of the AFRRI SAR.)
- c. LOCA: Not credible since fuel temperatures and pressures necessary to breach the clad cannot be attained. (See Licensee's Application for Amendment to License R-84 in 1964-65, Docket 50-170, where conservative calculations for a LOCA are provided.)
Even if it is postulated that pulsing must occur concurrent with a LOCA, clad failure is not credible since pulsing (single or repetitive) cannot be performed during a LOCA. (See Motion for Summary Dispostion and my affidavit concerning contention 9. Accidents IV.) (See also section 6.3.3 of the AFRRI SAR.)
1
- d. - S'abotage: .Not credible due to er.tremely: low probability.1 The AFRRI Reactor Physical Security Plan, which is protected from public disclosure, provides information and data to illustrate the protection afforded to detect and respond to sabotage' and which, thus, makes such an event non-credible from the standpoint of probability. '(See
- AFRRI Reactor Facility Physical Security Plan.)
- e. Aircraft Collision: Not credible due to extreme remoteness of such an event.
AFRRI is not beneath any scheduled air traffic route. The. vast majority of aircraft crashes occur upon take-off and landing (within a few miles of the airport). AFRRI is 1 not near enough to either National Airport or Dulles Airport to be considered in a high risk area for aircraft crashes. Even if the incredible should occur, any consequent release would be insignificant with respect to the normal consequences of an aircraft crash, such as petroleum fuel fires, structural damage, etc. (See NRC Staff's Safety Evaluation Report on the AFRRI Reactor.)
- f. Natural Act-of-God Accidents: Not credible due to extreme remoteness of such an event. Just as for the case of aircraft collision, an Act-of-God event which is adequate i to cause multiple clad failures would also produce extensive concurrent colateral damage which would far exceed the consequences of a release. (See NRC Staff's Safety Evaluation Report on the AFRRI Reactor.)
In summary, Licensee submit's that multiple clad failures due to the Intervenors postulated causal mechanisms are either not possible of yielding clad failures or are not likely enough to be considered credible. Licensee has provided, in it's SAR and its application for
. relicense, adequate assurance that multiple clad failures are not credible events and the I
. - . ~ . . - . - - - - --
NRC Staff has agreed with the Licensee in this area. (See SER as well as page 53, lines 2 through 9, inclusive, of the transcript of Dr. Stillman's deposition in New York on 18 Dec 1982.) A _ H'A. S - Sworn to and subscribed before me on this Ar e day of FA ,1983. s . Q ^-, -. -^ ^ ^ ^ ^ ^ " - U l', l? YY
. "f .
4 , 6 , * -- -s r e -
R e 4 4 ATTACHMENT 7 e
9 --
-~.,,..e.
l l . . , r-I k l GA-A15384 1 l i If l
!)
1. 1 le . TRIGA LOW-ENRICHED URANIUM ;i ' h FUEL QUENCH TESTS j
.l i
s , i
.: c ; ?-
i i Work Done By: Report Written By: ;, i J. R. BIDDLECOME J. R. LINDGREN !: .~
! H. BOTHMAN M. T. SIMNAD ll P. W. FLYNN i L C. FOSTER i[-
t . il e ' I I li - t o,
! l l:
t i j' I c GENERAL ATOMIC PROJECT 4314 l JULY 1980 l l GENEnnt ATome connrww. . . _i , e
t e o e e
..4 V
1 - a - o
\ , .
a 4
/ CONTENTS li;, }
Lgid j/
. o-
. 1. SUtttARY . . . . . . . . ..... .. ............ 1- j,(
. 2. INTRODUCT!0N ....... .. ............. 2 I.=
r-
~ ',
- 3. DESCRIPTION OF SAMPLES, EXPERIMENT AND EQUIPMENT / APPARATUS . .3
- 4. CHARACTERIZATION OF TRICA FUEL SAMPI.ES PRIOR TO AND AFTER 7 TESTINC . ... ....................... 5 r
- 5. TEST RESULTS ........................ 6 gq 5.1. 800*C Fuel Sample Quench Test Results . ........ 6 g%
A Q 5.2. 1000*C Fuel Sample Quench Test Results . ....... 7 y la i
-O
- 5. 3. 1100*C Fuel Sample Quench Test Results ........ 7 d.
5.4. 1200*C Fuel Sample Quench Test Results ........ 8 IE
'8 k ?,
5.5. Microprobe Analyses . . . . ............. s,
- 5.6. Hydrogen Analyses . . . . . . . . . . . . . . . . . . . 10 .
l U3.> ; 5.7. Meta 11ographic Examination .............. 11. 'iste's,
. , 5.7.1. As-Fabricated Control Specimens . . . . . . . . 11 P. 1 p
) - ..e 5.7.2. Specimens Quenched From 800*C . . . . . . . . . 11 y {.]
5.7.3. Specimens Quenched From 1000*C ........ 11 j? 4{g o, 5.7.4. Specimens Quenched From 1050*C ........ 12 s.,0, 5.7.5. speelmens Quenched From 1100*C ........ 12 '}; 5.7.6. Specimens Quenched From 1150*C . ....... 12 p{ _ ; H' 5.7.7. Specimens Quenched From 1200*C ........ 12
- 6. CONCLUSIONS . ................ ........ 14 5 ACKNOWLEDCHENTS . . . . . . . . . . . . . . . . . . . . . . . . . 16 {
APPENDIX At NICKEL-URANIUM FHASE DIACRAM . ... . . . . . . A-1 ,, s I
= 5 g
APPEPDIK B: CHROMIMURANIIM PHASE DI ACRAh . . . . ....... B-1 ,5 APPENDIX C IRON-URANIUM PHASE DIACRAM . ............ C-1 Y. APPENDIE D: NICKEL-ZIRCONIIM FHASE DI ACRAM . . . . . . . . . . . D.1 N APPENDIX Es CHROMI h ZIRCOMIIM PHASE DIACRAM . . . . . . . < . . E-1 d* APPENDIX F 1RON-21RCONIIM PHASE DIACRAM . . . . . . . . . . . . F-1 ' j APPENDIX C MICROPROBE EXAMINATIONS OF TRICA FUEL SAMPLE 4 . . . C-1 i 'hM Ns + APPENDIE H MICR0 PROBE X-RAY SCANS OF TEN TRICA
- URANIUM-ZIR(XHEIIM /~it
FUEL ELEPENTS ..... .............. N-1 111
~
e e e e
. -- . _ , - - . ~
. I. .
. -.g.
TABLES 75 FIGURES
- 1. Calibration trial resulte . samples 455-532r-1Er-lH . . . . .before
. . . an
. .d .efter
. . . . . . 76 '
18
- 2. Characteristica of . . . ......
Schematic of experimental apparatus for..quench ....... tents on ........... 1 TRICA f uct . . . . .
.......... les .
19 quench tests d 77 Microprobe analyses results for TRICA fuel quench test an 2. SS' nample heated to 1100*C used to cattbrate 00*C thermocoup
. . . 20 3. as-fabricated control samples . . . . . . . . . . ."d. . . . . .
78 Photographs of fuel sample 1 before quench from R 22
't . 800*C . . . . .
Photographs of fuel sample 1 af ter quench tram1000*C . . . . 25
- 4. 11ydrogen analyses results on TgICA fuel quench tes 4
5. Photographs of fuel sample 2 before quench1000*C f rom . . .. 27 6 Photographs of fuel sample 2 af ter quench from 1050*C . . . . 29 7. Photographs of fuel sample 3 before heating to 1050*C showing ' 32 R. Photographs of fuel sample 3 af ter heating to.............. localtred melting on all surfaces 36 1150*C . . .. Photoe.raphs of fuel sample 4 before heating1150*C to 36 0 . . . . ' Photographs of fuct sample 4 af ter heating to. 1050*C_. . . . 38 10 Photegraphs of fuct sample 5 before quench from 11 40 12. photographs of fuel sample ......5 after..quench
........ from 1050*C, 42 thermo-couple in contact with fuel 1100*C . ..
Photographa of fuel sample 6 before heatir.g to 44 1100*C . ... 11. Photographs of fuel asaple 6 af ter heating to 1100*C . . . . 46 14.
- Photographs of fuct sample 7 before quench f rom .. . 48 ,
- 15. 1100*C '
Photographs of f uel sample 7 af ter quench from 50
- 16. 1200*C . ...
Photographs of fuel sample 8 before heating1200*C.. to 52
- 17. .... 1 Photographs of fuel sample 8 af ter heating to 200*C . . . . 54 18.
19 Photographs of f uel sample 9 before quench from .. . 1 56 20.* rhotographs of fuel sample 9 af ter lquench E451R1H f rom
. . . . 1200*C 58 Microstructere of as-f abricated control samp e . .. 60
- 21. l E45111L .
Microstructure of as-fabricated control samp e 61
- 22. 800*C . . . .
Microstructure of sample 1 af ter quenching1000*C from . . . 63 21. Microstructure of sample 2 af ter quenching*C f rom 65
- 24. . . .. .
Hierostructure of sample 5 af ter heating to1100*C 1050 . . . . 67 25. Hierostructure of sample 7 af ter quenching *C......from 69 26 27. Mlerostructure of sample 4 after heating jo 1150 .. 71 of sample 3 af ter heating to 1200*C . . . 73
- 28. Microstructure 1200*C . . . .
29. Hierostructure of sample 9 af ter quenching from v
~
iv ,
. % A l
e s e e 4
. t .
! r
?
9 ..(-
- l
<.'s
',l@ !'
v /J i ff1 g ' P .1 .. a,
- 1. StTHNARY }+ fbifr/ .
- ;r t .c3,;
s Quench tests were performed on TRICA low-enriched uranium (LEU) fuel @Gy 6 }?b,: ,
- , e.s ,
samples from temperatures ranging from 800* to 1200*C. Fuel samples ],7 l
, quenched from 800', 1000', 1100* and 1200*C showed remarkably benign D!y { '. +
.. . g 1 -
response to the test conditions. Minor cracking occurred in some samples; 'H ,s '?; volume shrinkage 1oss of hydrogen, and apparent surface oxidation' occurred $Nh,kib f in all samples. Test results on samples quenched from approximately 1100*C h M were variable; these variations were at first believed to have been caused '). by differences in the fuel homogeneity. The results on some samples were .m. S!us J?ph s$ benign (minor cracking, volume shrinkage, loss of hydrogen, and surface c& it'% ,
' oxidation), while localized melting occurred on other samples when heated k i Mq 4,.-t.L to a measured temperature of 1050*C. The localized setting was caused by .
eutectica formed from reaction of the Inconet-600 thermocouple aheath 'I%rl (!)
-lff ]
with the fuel sample. Samples quenched from 1200*C show variable behavior ei<, ..) i, only because one sample contseted and reacted with the tantalum susceptor q , .y g,, originally uned. The second sample showed very satisfactory benign : i;t ;j y , ( behavior. .a M ,y b.n3 p' ,W,- 1
. m:
.a%c
> It
.cy$y-J' i J / f. :f ' -
; 'g .
7 4 f. q F,~$,3 ;;;.] s
, , .p 4 - . ' . J' s d'l Q$ ' -
.; t.s.L a6 g in m
' r Yd::, .
s v::c A. L
'S
,]'
Q Wf:l a.
),; 77
'~ t
.% s 4 A
- ^ .
l l -
* * .e e
* ~ - - . - . ,, ,
.,y
> y :.
IU 6/ g
'.i .
. 'M; t .a
. I ,d.'
, i.
- 2. INTRODUCTION I 3 DESCRIPTION OF SAMPLES, EXPERIMENT, AND EQUIPMENTfAPPARATUS qjf W
Quench tente from temperatures ranging from 800* to 1200*C were per- The samples used in the quench tests were cut from TRICA LEU fuel rode O' I ti o formed on 202-enriched 45U-53Zr-1Er-1H4 TRICA 1.EU fuel samples to otmulate la pieces 12.9 mm in diameter and ranging from 12 to 18 sun in length. A 2 i ole a conclition, when water ingrene occure upon rupture of the cladding in a hele was drilled into each fuel sample to allow insertio's of the'1.6-eme- .A .> o.d., inconel-600-sheathed. Hg0-insulated Chromel-Alumel thermocouple. '!,k fact rod. The tests were atm11ar to those performed for the same purpose < For the fuel teste at 800* and 1000*C, the sheathed therecouple was in- f' t ., o ' in 1958 on 201-enriched 8U-91Zr-1H TRICA fuel samples and rode. Only } <- minor cracking and surface oxidation occurred in these earlier teste, , certed directly in contact with the hole surfaces. When it was found f that fuel melting and fuel-thermocouple reactions occurred in teste per- ( [, am,1 the current tests were performed to determine if the change to t,he ~ new fuel composition would significantly+ influence fuel behavior under formed at approximately 1100*C and higher temperatures, a molybdenum cup y@.j , . the quench test conditions. was inserted in the fuel hole to prevent reaction between the fuel sample 6 and the Inconel 600 thermocouple sheath. $-
, r%.
>',s p
q
; The experimental equipment used is shown schematically in Fig. I and ,!
in the Fig. 2 photograph. The equipment consists of a vacuuMinert , 'j f' , atmosphere quarts tube furnace la which a molybdenum (initially tantatum) @ susceptor le induction-heated by a Tocco 30-kW motor-generator induction unit. (The susceptor material was changed because the tantalum hydrided J.R , ,
+ when the samples released hydrogens molybdenum does not absorb hydrogen.) rg(;
t,, The susceptor radiant heate the fuel sample. The system to twice , evacuated to 10-5torr and back-filled to atmospheric pressure with com- f, si i mercial argon prior to heating the sample. When the sample le heated to (Q the desired temperature, valve 2 (Fig. 1) to opened momentarily while the h3- i semple to dropped by pushing the sample holder rei through the sliding seal. Typical time at temperature for the sample te 1 min. l
=
The thermocouple was inserted directly into test samples 1 through 5 Y . and an attempt w' as made to drop the sample. The thermocouples inserted
>4
* * }#
45 wt-2 uranium. 53 wt-Z strcontum, I we-2 erbium, I wt-2 hydrogen. A11 figures and tables are at the b*ck of this repetrt.
- 3 i 2
iI.
~
l
*
- sI .
, e . e e s
~~~3 , ,
- mm m.vn=n=1,ce c ' . m x - ~::
f h$ jn sampics 6 and 8. which were not quenched, were heated simultaneously
- with samples 7 and 9. respectively, which were quenched. This was done .
tf en allnw temperature to be m a asured in sibling samples, yet avoid having *
'[4 2
t,3 , the thermcouple interfere with the quenching operation.
- 4. CHARACTER 1ZAT10rt OF TRICA FUEL SAMPLES FRIOR TO AND AFTER TESTINC Immediately af ter the sample is dropped into the quench chamber The experimenter characterized the fuel samplea prior to sad af ter .
bottnm. viive 2 is closed to minimise air ingress into the fu'rnace chamher. testing by photographing their externet appearance (Figs. 3 through 20), The atmospherc in the quench water chamber is exhausted through an y performing dimensional measurements and weightdg them. The dimensional ' 7 elephant trunk to an absolute filter hood. Since the fuel sample cools and weight data before and af ter testing are given in Table 2. The net ; l' quickly the experimenter immediately unfastens the quick-connect flange, changes are also listed. . Test result details are discussed in Section 5. reaches into the water chamber, and retrieves it. The furnace and thermocouples were checked using a 12.9-eas-diameter b, by 12.9-mm-long 316SS cylinder as the test specimen (Fig. 2). The thermo- !l i.-
,w couple junctions were inserted in holes 3.175 and 6.35 mm (szlat center -}q
,1 g l, of 12.9-mm-diameter sample) from the surface. The type 3165S cylinder 'i was used to minimize effects ef*possible thermocouple / sample reactions lIN
,tv during the checkout. ")y Yb In t'he first calibration trial, thermocouple I was inserted in the Mn 1.175-mm-deep hole, and thermocouple 2 was inserted in the 6.35-mm-deep . q ,
hole. For the seennd cattbration trial, the positions of the thermo- ,
'f,
*i couples were reversed. The results upon heating in the test apparatus ; I
,' l are given in Table 1. j
[.q'j () s
' I o' $n The two thermoccuples behaved similarly. Upon heating to approxi- ,
s , mately 1000*C. the temperature difference between the sample 3.175 mm ' , , from the surface and that in the center of the cylinder (6.35 mm from '
' 'n the nurface) was approximately 20*C. This outcome indicates that the I ; *,
i =,;. actual nurf ace temperature is 1020*C when the thermocouple 3.175 mm from , i j the surface shown 1000*C. ,, T . Ti
;a I
Y l 4 5 W -w I
- e e e p 'rT
.N i
Some Positi's pressure change, i.e., 13.788 kra, typteJ t, accurred . in the furnsee atmosphere during the treatment of.the samples, indicating ..
- outgassing and release of hydrogen. The pressure change was slightly
- 5. TEST RESULTS ,
larger for the samples subsequently heated to higher temperatures.~ Some ({l - v of the pressure change was undoubtedly also due to temperature increase s Quench testing at 800*C (sample 1) and 1000*C (sample 2) went as of the argon atmosphere in the furnace. expected with no unusual occurences. However, initial tests attempted 3n at Insn*c and higher temperatures (samples 3, 4, and 5), where the 5.2. 1000*C FUEL SAMPLE QUENCH TEST RESULTS
- Inconel-sheathed thermocouple was in direct contact with the fuel, resulted in loca1Jzed fuel melting. Samples 3 and 4 were from fuel rod }
The 1000*C fuel quench test (sample 2) resulted in a weight loss of F.451 Rill, whereas sample 5 and later samples were from fuel rod E451RIL. . 0.0714 g from a total weight of 15.954 g. This constitutes a 0.45 wt 2 i remanining tests from sample 6 onward (all from E451RIL) did not result' loss. Some surface oxidation appears to have occurred, with a weight , in localized melt.ng. These observatto,s n would indicate that the localized , gain from oxidation offseting some of the hydrogen weight loss. No melting was the result of fuel reactions with the inconel 600 thermocouple , y, g. test visual surf ace cracks were found on the sample. The volume lf , shcath..rather thsn inhomogenettles in the fuel. The major constituents of shrinkage, approximately 4.6%, appears to indicate more hydrogen loss and inconel 600 are nickel, chromium, and iron (72N1-14-?7Cr-6-10Fe). le sintering than found in the sample quenched from 800*C, even though the melting cutecties between uranium and the inconet 600 constituents occur overall fractional weight loss was less. as follows: nickel 740*C, chromium 859'C, and iron 725'C. (See Appendices ' A. R. and C.) The melting points of the eutectics between sitconium and 5.3. 1100*C FUEL SIMPLE QUENCH TEST RESULTS j h/ ' the Inconel 600 constituents are as follows: nickel 961*C, clromium a*, The 1100*C quench fuel teet (sample 7) resulted in a weight lose of *i i wo*c, and iron 934*C. (See Appendices D, E, and F.) 0.7764 g from a total initial weight of 18.3849 s. This constitutes a
- 4.23 wt 2 loss. Since the original sample contafned only 0.88 wt I 5,t. ROO*C FUEL SAMPLE QUENCli TEST RESULTS '
hydrogen, even if all of the hydrogen in the samples were removed during , The ROO*C fuel quench test (sample 1) resulted in a weight loss of heating to 1100*C, an additional 3.35 wt % et loss could possibly be attributed to other constituents having been removed, possibly an oxides. i - 0.115 g from a total weight of 19.9152 3 This constitutes a 0.57 vt I There appeared to be very little test-related surface oxidation llovever, since the sample quenched from 1200*C showed a lower weight loss, Y$ togs. on the sample. en it may be assumed that the weight loss was due to this'te not highly probable (see section 5.4). Althougl's the sample'had The fuel initially contained 0.88 wt I hydrogen, developed a radial crack'on one end, there was very minimal evidence of , , retesse of hydrogen. chipping or other removal of solid material resulting from heating to itence. 0.57/0.88 we % or 64.8% of the initial hydrogen pr,esent in the sample vag lost during the test. The sample shoves only a minor radial 1100*C or from dropping into the quench chamber. f p { crack and volume shrinkage (approzimately 1% differential colume/ volume) that could be attributable to the loss in hydrogen. , [ l l
~
6 . L-
* * , e.
- __m l
m -
! Ii
[. 4. 1200*C FUEL. SAMPI.E QUENCH TEST RESULTS
- 3. It appears that the low-melting eutectics initiated the The 1200*C fuel quench test (sample 9) resulted in a weight loss of localized meltir.g by general surf ace melting of the simples. Q
# 0.0600 m from e total initial weight of 13.9545 g. This constitutes e In surf ace regions distant from the source of chromium, iron. . ,
[ n.375 wt % loss. Since the original sample contained 0.88% hydrogen, and and nickel (the inconel 600 thermocouple sheath) melting also it is expected that virtually ajt of the hydroken would have been released occurred, but the high vapor pressure of these constituents would upon heating to 1200'C. the hydrogen weight lens was compensated in part have caused them to disappear by vaporisation during heating. . l hy a weight gain due to sample surface oxidation. Other than minor leaving only the high uranitsu content to be detected by the ; nurface oxidation, the sample showed only minor markings where the sample microprobe analysis. The grain size,in the fuel is signift. l was supported with a molybdenum wire and no other adverse effects in cently larger throughout the fuel rod where low melting tem- ' > appearance. The volume shrinkage of 5.2% is comparable to that observed perature eutectics have formed versus where no uranium melting on sample.2 which was quenched from 1000*C. has occurred. The higher uranium content regions are due to,the microinhomogenettles which are smaller than the grain size in 5.5. MICR0rPOBE ANALYSES the fuel as f abricated (grain size = au40 to 120 pm). Studies by
- Baldwin have shown that microinhomogenettles are characterized l'
, Mteroprobe analynin indicated that the localized melting on sample by uranium-rich grain boundaries surrounding grains with rela- ,
4 nurfaces, which occurred'wherever the Inconel 600 thermocouple was in tively uniform uranium and sitconium content. This type of f' direct contact with the fuel'at 1150*C had the following characteristics: uranium d(stribution provides a beneficial effect on fission gas I release rates (release rates are lower by one to two orders of [j
- 1. The once-motten heads or blisters were of a high uranium content magnitude) as compared to fuel in which the uranium 16 distri- M and in some cases were nearly pure uranium. *the pits or pockets buted uniformly as a finer dispersion. In the regions where near the once-molten beads or blisters had a hIgh uranium con- melting occurred, the microinhomogeneittee were found within the tent, while the matris surrounding them had a high airconium grains as well as in the grain boundaries. The grain size ranged content. These results appear to indicate general high uranium up to %150 pm as compared to en as-fabricated grain size of 40 content microinhomogenattles with a 10 to 150 pm* size range in to 120 pm. -
iF the fuel, probably increased by the formation of eutectics with the thermocouple r. heath, as de' scribed in 10o. 3 below. j Sample 3. which had been heated to 1200*C in contact with a thermo- N j couple, was sectioned longitudinally so that material away from the sur-
- 2. In regions near the thermocouple location, in what appeared to be i. face could be onlysed. Microprobe analyses showed that, in regions away tia reaction zone between the inconel 600 sheath and the fuel, from the fuel surface and the zone of reaction between the fuel and the i uranium-chromium plate crystals were found, and iron and mag, thermocouple sheath, the primary constituents were uranium and strconium, nesium were also present. These results indicate that low- as expected. The concentration of wranissa varied locally (6% to 80%) but melting cuterties formed between the uranitus'in the fuel and the cwnstituents of the inconel 600 sheath of the thermocouple ,
Cent ral Atomic private commmunication. and the Mgo insulation between the sheath and the thermocouple. * . l
- pm = micrometer = 10-6meters = 1 micron.
j - 9 i 8 . i l 1 l
- e. *
.g.-==~,. _
7~, I p. -
. y ,
, ,i;
. i '
, J + it was generally. from about 40 wt I to se high as 80 wt I uranium (the
' control samples E458RIL and E451RtH. The analyses were performed by the,' *
] nominal average compostefon la 45 wt I uranium). The microprobe electron vacuten fusion method. The analyses accuracy to estimated to be,115% by beam site employed was 5 by 50 tai. while single graine in this sample
- M. Histt. who performed the tests. The results are given in Table 4 I.
z3 were as large.as 150 pm in diameter. Therefore. the variation in compo-i 1 sition observed was very local and to clamelffed as microinhomogeneittee i 5.7. METAI,IACRAPHIC EIANINATION S i 1 rather than general inhomogeneittee. Away from the reaction some in the I
] ;.
sample, the uranium to strconium ratio was approximately unity. Silicon t 5.7.1. As-Fabricated Control Specimena , b .
/, .
from an unknown source was found on the surfaces of the hole in which the ! i,- thermocouple was located and this etlicon reacted with the fuel. l The fuel rode were fabricated ty casting.' sone melting. and hydriding. [<
' , the 45 we Z U-Zr alloy. The microstructures of the specimeno are typical TRICA Fabrication reports that the fuel specimen was sectioned with 3 of hydrided dendritic structures. There are variatione in composition and I an alumina (Aly30 ) abrasive wheel. It is difficult to understand why I grain else on's microscopic scale. both axially and radially. , The arrays n.r '
stitcon rather than aluminum was found on the surfaces of the hole if the , cutting wheel was the source of the silicon. f I of fine pores at grain boundaries and within grains and the general feitures . 1 W. 3 j of the microstructures may be related to the recrystallization, phase 6 i g\ ;l , i i , .t; changes. and grais boundary segregation during hydriding. .The phases that 'f 1 Ml'roprobe e analyses were performed on the surfaces of the above , are present include uranium zirconium hydride. and streontian and uranium t kt-( ji umpics in regions away from where the thermocouple reactions occurred. I carbides. Variations in rates of solidification and of hydriding govern / ;, Pflercprobe analyses were also performed on the surfaces of several sample' the types of microstructural features that are observed. (See Figo 21: ,, in which little or no melting occurred during heating prior to quenchinge 6 and 22.) . >,* as well as on two archive samples which represent the fuel in the as- 'i' fabricated condition. The microprobe results for the surfaces of all l 5.7.2. Specimene Quenched From 800*C ' c f the samples analysed are summarized in Table 3. j
; i p
There le little change in the microstructure of the specimena quenched yj' , The amo..nts of uranium and strconium shown do not add up to 1001 " from 800*C compared to that of the as-fabricated specimene. A slight ... because (1) the microprobe analyses average the uranium and airconium , difference in structure'in the surface area may be related to oxidation content over a 50-um-diameter spot, and (2) the correction factor for l
- and some lose of hydrogen from the surf ace region. (See Fig. 23.)
the X-ray absorption by the matrix has not been applieJ (the absorption , , {% j l factor was calibrated for 50% uranium. 50% streonium but varies con-F stderably with varying amounts of uranium and sitcontion). (The detalle 5.7.3. Specimene Quenched From 1000*C ky j O 4 ,. of microprobe analyses are given in Appendices C and N). , The effects are similar to those observed in the specimens quenched e 4
-l ;
' from 800*C. (See Fig. 24.) . -T' 5.6 HYDROGEN L*lALYSES M>.
;'( ' 3 e
i Hydrogen analynes were performed on samples 1 through 5. 7. and 9 . I af ter quench or heating to test temperature and on two as-fabricated 1 10
- gg
.l p i s
o
- e. . . e e n - -._., ,, . . _ _ . . . . . . - . _ , . __ _ _ ,,,,,_ _ , _
i . . i . 7
' ., j
,'j ' A
'f.} ( 9 5.7.4. Specimens Quenched Frois 1050*C' were in contact with the Inconel-600 thermocouple sheaths. The spectanens '. t which were protected from the thermocouple sheath have a significantly i There is some enhancement of the changes 14 microstructure observed more homogeneone microstructure, although the grain houndaries with arrays ' I; at 800* and 1000*C, particularly in the homogeSization and in the number of fine voids are still in evidence. There has been a large loss of ry li of fine volds which may be associated with increased loss of hydrogen. hydrogen, particularly from the o*strace regions where small voids are (see Fjg. 25.) The homogentsation referred to here is related to the ,vielble in the grains. Surf ace oxidation is also evident, with an oxide composition or local urantina and streonium content. lipon heating at film approximately 6 microne thick. >
3 g' temperatures 5.800*C, the diffusion of the constituents will tend to re- *
< :. Al move concentration gradiente and compc,sition dif ferences in the fuel. ?9 y[il n,
y i Also, long-term heating at high temperatures will tend to pramote grain y,'
;. j growth. and,n grain size larger than that in the as-fabricated condition will reault.
E ll 11 (b i... ! a 5.7.5. Spec 1 mens Quenched From 1100*C g;; Q r
. Q 'l The ef fects are minitar to those observed in the 1050*C quench with yl h.1(h r
further enhancement of the homogenization procese. (See Fig. 26.) .
* $1j t:
5.7.6 Specimens Quenched Frtwa 1150*C ( , r, g 2'i i There is markedly increaped homogenization, although the grain /- ( ~ ,. ; { >j houndaries cong1 sting of fine arrays of voide are still in evidence. } u1 Reaction of the fuel with the inconet-600 thermocouple sheath resulted in Q 'Lh 1 the intmation of entectics. (See Fig. 27.) j'j [ [ c:' 5.7.7. Specimens Quenched From 1200*C [:;h-V w Li Figure 28 nhows the microstructures of specimens which reacted with y-the Inconet-600 thermocouple sheath with the. formation of outectice. . hf." lt z Fip,ure 29 shnwn the microntructures of specimens which were toolated from . diki M.9 the thermocouple sheath by means of molybdenum foil cure. The eutectic ! FG.
)
P evidently penetrated rapidly along grain boundaries in the specimens which 3 N. a
-G Jd >
L;
. ;; e JA 12 13 ha e
- e e a
,m.,,,
-=~ 7 . .. p e +: . 8.
, ds4 i r yj 4 iH fraction of a second. The time is so short that there will not. .
be sufficient diffusion to cause eutectics to form between the fuel and the inconel sheath. Under design operating conditions. 6, CI)NCLUSIONS the presence of the Inconel-sheathed thermocouples in the TRICA , , .1, fuel has been and is expected to continue to be satisf actory. k1 y The following conclusions can be drawn from quench tests on TRIG 4 LEU .Q fuel samples from temperatures ranging from 800* to 1200*C. .N , ,
' S is l)L .,A
- 1. All samples in which there was no contact with the Inconel 600 ;
thermocouple sheath survived the tests in excellent condition. ,Mir-: o y 3 All samples quenched from 800* and fl00*C had minor ersching. rpC.
" ;.D : ..
,<m ;!
- 2. For those samples in which the. fuel was in contact with the ,Qy{(
~
Inconel-600 thermocouple sheath, no obvious reactions occurred i[
.to 1000*C. Above approximately 1050*C. eutectics of uranius .[ '
with nicket, chromium with iron, and possibly airconius with T. k *, ,l nielrel and iron formed, which resulted in locallred melting of ,[d E d the surface of the fuel sample. These results indicate satis- , yy 44 ' kr o factory behavior of TRIGA fuel for temperatures to at least i(.t ki 2,, [: nae ' 1200*C. Under conditions where the clad temperature can approach ;ga 4 ,o the fuct temperature for severst minutes (which may allow forma- ;> .;q tion of cutectics with the clad). the results indicate satis- }ki factory behavior to about 1050*C. This is still about 50* to, ,i ? f s
*I* '
' 100*C higher than the temperature at which internal hydrogee .* r pressure is expected to rupture the clad. should the clad tem- 0) >
er. perature approach that of the fuel. Q f>
- 1. Inconet sheathed thermocouples are used to monitor temperatures l At .
[.,*, in both standard and TRICA 1. Elf fuel. No problems have been Li l', , . " observed from their presence. The lowest temperature eutectic
. [
w' j r in the syst'em is f ron with uranium, which nceurs et 725'C. At ij temperatures <J25'C. the diffusion rates are very low and no h cutectice form For pulsing reactors wheN in the fuel is pulsed , l, . to 11050*C, the time at temperature is very short. a small .
;l,.) ,
r< 14 . g5 W
i . j .
+
\
- i FUELF _EV ENTS FOR Pl11 SFD glggg;gligg a TRIGA* RFSFARCH RFACTORR .,
.s
't P- >LASSOUD T. SIMNAD, FABIAN C. FOUSHEE, and KE4woRDS: TRIGA type reac-e3 '
, onDON B. WEST General Atornic Company, P.O. Box 81608
,,,,; fy,f ,/emena. pulsed reac.
m Diego. Califonnia 92138 tors, tirconium hydrides. design, ); moderstely enriched uranium. .i IS tesdng. chase diaprems. ACPR J reactor, physical properties. irra-diation
;c j,ecetved January 29.1975 ,
g A:cepted for Publication May 9,19't5 3' e
'.j
.l .
UT 3p _ .. - m m._. - - . , ~ ~ _. h. y dc.hecacu.we2hs.uh.n. a, operation. The basic characteristics of the TRIGA -
-g research reactors are:
TRIGA fuel reas developed around the enn* oM ~ nt- I inherent safetv. ci core composition *-e en.mht 1. The use of homogeneous U-Zr hydride solid -r he that had a large prompt negattve temperature fuel-moderator e1ements with a . large ,i y prompt negative temperature coefficient of
-.cerncient or reactivity such that if all the avail- ,u.
ale excess reactivsty were suadenly inserted reactivity, b.
' dn the core, the reenitine fuel temperature would )
- 2. Light-water open-pool design, with natural a:aomatically cau e the power excurswn to termi- convection cooling up to 2 MW and, forced y
'L}
c ute before any core damare resulted. Ex^*vi cooling above this lcvel. er n.ents itave aemonstrated' thnt rivenni'~ h~ w - Y II .'. t .- Nssesses a basic neutron-stiectrum-hardenin, 2. Power pulsing capability, licensed to pulse
.*echanism to produce the desired characteristic. r ndy up to a reacMy inserdon of l ,
g additional advantaroc snetmfe the factc thct ?rH 3.2% 6h/h ($4.60) and peak power level of h- ' U.s a good heat cchacity. that it results in rela- . 6500 MW, providing an integrated neutron
?.* 1 ::rciv smail enre ctree and hieh flux values due to flux of up to ~10 n/cm per pulse. A w W Hrit hydrogen content. that it hae excellent special annular core pulsing (ACPR) TRIGA
' tssion-product retentivity and high chemical sn '
reactor at Sandia Corporation achieves a , t.- T .'nm in water at temtAratures up to 100*C, and- peak pulse of 12 000 MW (Rcis. 4,5, and 6). of :nat it can be used effcetivelv in a rugged fuel The ACPR has a large central irradiation i
." c!cmint eire. space for pulse-testing reactor fuels, ma- =G Tens of thousands o_f routine bulces to !he rance terials, and equipment.
of 500 to BOO *C peah fact temperatures have been A typical ACPR fuel-moderator element is " g performea scztn TitGA ruel, and a core was eul*E- shown in Fig.1. The active section is.15 in. long ; erra to pea # wel remorrdvree in excess of and 1.4 in. in diameter and contains ~12 wt% '! lg WnY ror hundreds of httises before a fe'" *le- uranium enriched to 20% in "U. To h~m&n+o _mente ereceded the conservative tolerances on hydridine. a 0.25-in.-diam hole is drilled threv~h j 'timensional citange. , the center of *ha n r+ive sectiom a zirenaiu- *ad
-ry - I8 i" carted in this hole tfter hydridinc is com-
- 1. . m. e mC: a 1
gig!g Graphite cylinders ~.S.4 in. long and 1.419 in diameter act as top and lemom reflectors. The active fuel section and the top and bottom l .'* NTRODUCTION graphite cylinders are contained in a 0.02-in.- l l thick stainless-steel clad. (The clad is provided g % u ..n u , - + e r ., ,., . ; , . _ ,4 ,,,. n.,4,, ,,, w ..a -i d a J.-!= r- e
.w o --rar 3 -. . % , . , . . , , . . - . ' - w, with internal dimples that act as spacers to en-sure a thermal gap of ~0.010 in. 'actween the fuel j -m - . . - ., .s t r t .. n i n ,,- p einen 1 ore IP.cfs.1 meat and the clad.) The stainicss-steel claddin;:
! 7 p 5). C-ect 6000 fuel elements of 7 distinct is welded to the top and bottom end fittin;,,
?
' M h' 'te Occ . fabricated for the 55 TRIGA re- positiens the top of the fuel element in the top grid
~3 reactor; under construction or already in plate, and is fluted to provide passage fer natural
- 3 ;D GCh??:W,Y VO L 'S JANUARY 1C6 .
1 31 1
,.,w %,:m.m.- .n, t',; ~;*
.rpw:~ zz%. r~"W=.
- - -= .~*~ W '..y.". '*""" ~"'~~~~~2'~'m'"~"****~~~'
. _ ~ . * * *,. ' * ~ ~
,. . a .
y -
l w- , . .. ,~~._.. .,,,Q _, }Q, 3 C--.a. :..a.n .:, .g. 7A 'l. l Simnad et at TR!GA RESEARCH REACTORS I l convection cooling-water flow upward through the
- grid. The bottom end fitting, which is.also fluted
. to provide water flow passages, rests 'and is
- centered on the chamfered hole in the bottom grid
; i plate. The bottom grid plate supports the weigh:
Type 304 of the element, which is T.5 lb, and the 2"U l Stainless Steel content is ~54 g. The e+'nd'rd TRIGA fuel has a* cold cao of on!v , h
,J 1 or 2 mils between the fuel and clad. is slio-fit i .
3d tecether. nnd contains 8.5 to 12 wt% uranium as a l Type 304 -
,h Tine metallic disnersion in a zirconium hydride (
Stainless Steel l i m '" "n' The H/Zr ratio is 1.60 fin the face. j l l l centered cubic (fee) delta chasell The equilibrium ,
- y. ' nyarogen cassociation pressure is governed by the j b z composition and temperature. For ZrH, . +ha d
, j' ecuilibrium hydrowen crecenra is i ,tm e+ -anar* d Graphite r' The sincle-nhnca hich-hydride composition elimi-,l
-nates the eroblems of dencity changes associated ll with chase chnenat nnd thermal diffusion of the e
J l C,, . [ '
- I hydrocen The recently developed Fuel Lifetime '
a Nmn-nvnm. ent n-ne*,m (FLIP fuel contains up to ~ 3.0% erbium as a burnable poison to increase the
*' " core lifetime in the higher power (1- to 14-MW) TRIGA reactors. ff' ores with steady-stnte
- ["
Dower levels ahnv 2 uw sea nne nuim- an-ee i -
-l Zr Rod The calculated core lifetime with FLIP 'fuelin the
?
- Clear. ,2-MW TRIGA is ~9 MW-yr. Over 25 onn nntea= s
.l -
bave been nerfnem ed rith *ka TR f ru fa al a lm . U-Zr H Fuel 15 men's nt General Atnmfe with fuel +a-na-9+nrae _
,-l reaching ceave of ~1150*r' --
2 p* TRIGA fuel was devainead evnned +ha ener an! of inherent safetym A core comonciHen 'm snuch! II that had a larce cromet neentive temoerature
,j Type 304 ,
coefficient of renetivity such that if all the avail-Stainless Steel able excess reactivity were suddenly inserted into
,i f _
the core. the resultiac 'n al temce-ntura wnuld t <
' i automatienlly c,nce tha nnwp* excureian +n *p -
' minate before any core damace reenited. Ex-l
. Graphite periments then in progress demonstrated that l ,
zirconium hydride possesses a basic mechanism l l [l to produce the desired characteristic. Additiona! l t1 advantages included the facts that ZrH has a good .
, g[l heat capacity, that it results in relatively small' core sizes and high flux values due to the high l i Type 304 hydrogen content, and that it can be used ef-l .
Stainless Steet fectively in a rugged fuel element size. [l ' 1
. l Current routine power levels for TRIGA re- F i actors (1 to 2 MW) require operational excess l
\l d reactivitics that carinot be instantaneously in-I I serted into the core with complete' safety. How I ever, the safety of these systems has not bee::l compromised because no single control red cr j experiment is worth thd reactivity necessary to !,
! ,n reach an unsafe level. The ACPR, which if ~
', optimized for maximum pulsing performance, re-i ,
quires the precise timing of the removal of 3 to i I pulse rods to produce the operational pulse per-Fig.1 for mance. Typic 9RIGA fuel element. 32, Q hff .O NUCLEAR TECHNOLOGY VOLIS JANUARY 19
- _g
. . a i - i h
I' Simnad et al. TRICA RESEARCH REACTORS A a I The characteristics of the fuel and the design band of the metal structure, and as being precent .!
, experience 'with pulsed TRIGA in the lattice as H+. This theory describes the
$ndcactors operational are described in this paper. transition metal hydrides as metallic or as, alloys.
h p The alternative theory considers that the hydrogen atom acquires an electron from the conduction 9
' *- ASE SYSTEMS AND DISSOCIATION PRESSURES ,.
band and is present as N~. The depleted conduc- 4 he ZrH and U-ZrH systems are essentially tion band remains to give residual metallic bond- 7;
- 4 a b le eutectoids,' containing at least four sep- ing in the hydride and to account for the metallic i I rte hydride phases in addition to the zirconium properties. This theory describes the hydrides as .
$nd uranium allotropes (Fig. 2). The hydride i nic. It is possible that covalent bonding could be g j
hases consist of the following: introduced into either theory, althcugh- few at- g. y r" tempts have been made to do so. In any case, the
- 1. Alpha phase-a low-temperature terminal small hydrogen atom would be expected to enter n solid solution of hydrogen in the hexagonal the tetrahedral sites in the usually close-packed I close-packed alpha-zirconium lattice, metal structure. Nevertheless, most hydrides do . .
- 2. Beta phase-a solid solution of hydrogen not have their metal atoms in the same positions '
dissolved in the high-temperature body- as in the parent metal. The solubility of hydrogen - , - centered cubic zirconium phase. in zirconium above the eutectoid temperature was found to be increased by the presence of beta-
- 3. Delta phase-an fee hydride pha,se (a delta- stabilizing elements and decreased by alpha- 3 prime phase has also been reported, formed stabilizers. 7 below 240*C from the delta). The rates of hydriding and dehydriding of
- 4. Epsilon phase-a face-centered tetragonal zirconium are markedly influenced -(reduced) by r-(fet) hydride phase with the ratio c/a < 1, the presence of surface oxide or nitride films. l extending beyond the delta phase to ZrH:. The surface films will, therefore, affect the F-The epsilon phase is not a true equilibrium measured hydrogen dissociation pressures unless >
phase and forms from the delta by a mar- precautions are taken to eliminate these films, tensitic reaction. It appears as a banded The hydrogen dissociation pressures of zir-
- twin structure. conium hydrides and of U-ZrH have been mea-sured.'8 The eoncentration of hydrogen is When uranium is present, it appears to be generally reported in terms of either weight partially rejected from solution during the hy- g percent or atoms of hydrogen per em' of fuel ,,
driding process. The uranium rejected is present (N,). The equilibrium dissociation pressures in av a fine uniform dispersion. The effect of the the ZrH system are given in Fig. 3. In the delta uranium addition on the ZrH system is to shift all region, the dissociation pressure equilibria of-the phase boundaries of the ZrH diagram to the zirconium-hydrogen binary can be expressed slightly lower temperatures. For example, the in terms of composition and temperature by the eutectoid temperature is lowered from 547 to relation Gl*C. No new phases and no uranium hydride (i 3 have been detected.** At rather histh uranium. con- log P = Ki + '* . . tents _(25_to 50 wtfo), the behavior with hydrogen i. Tas found to be a breakdown of the intermetallic where *
.f alloy. The zirconium reacted with the hydrogen' tring polyphase regions of uranium, zirconium, Kt = -3.8415 + 38.6433X - 34.2639X* + 9.2821X*
and zirconium hydride phases, mainly the cubic K: = -31.2982 + 23.5741X - 6.0280X* fi celta' hydride. The phase boundaries of the ZrH ' 3 . hagram'were relatively unaffected in the region = pressure, atm j ;f high hydrogen content, but the alpha and beta T = temperature, K s T'tses were markedly shifted. The main effect of h Se additicn of uranium in the low hydrogen con- X = hydrogen-to-zirconium atom ratio. g g {;nt region was to considerably increase the range The heat of solution of hydrogen in the delta l g me alpha phase. Uranium hydride phases were hydrided phase decreases with increasing solute j , tjbse. ced. i concentration, from -46.3K cal / mole, in delta of l Q .here is no generally accepted theoretical composition ZrH .., to -37.7K cal / mole, in epsilen i N ' Wiption of the structure of metal hydrides." of composition ZrH ., (Ref.12). It is significant i i [. " TrEsent, 6vo quite different theories are used that no discontinuity in the function is in evidence ! Meuss metal hydrides. In one, the hydrogen is throughout the entire delta-to-epsilon composition i g Mddle losing its electren to the conduction range, involving an H/Zr composition range cf
.i
- l
('7. W TLOINGLOGY jOL 28 JA'iUARY 19M 33 l n 4 ,C, -s L fu.m; -="O -*4 4Ib ~ T } *'" ' h *"'#^ ' ' '
y~ K u t.;>.f.5d d # A h 'w di m s;b
; :s. 44 uf. w ;.. .w '.' , . m . .. ..
4
.v -J .,e p
-i;';,j;,
T;!
~
.-- -- - m , _,..,..u- ._, _
.w .
Simnad et al. TRIGA RESEARCH REACTORS 2000 . (a) ZlRCONIUM.HYOROGEN PHASE DIAGRAM 1800 - 1' 4 l. / 0 / 1600 - . SNAP REACTOR / OPERATING / RANCE [ 1400 - ! 0+6 ', _1200 -
, '! E * =00 hI 00 6 l; ' { 0.940 -
3 jjl 800 - o +6 +1TI tI-e . l .. y ) 3 0.930 - [! 8 600 - h [ !l l A I
'i i
' 11 3 *3 0.920 -
,' ' !! 400 - O X-RAY OlFFRACTION O e
.'.1 O ELECTRICAL RESISTIVITY 5
- ll A DILATOMETRY " '
, : 7 METALLOCRAPHY
..i. U 0.910 -
p 200 - . Ih 6+e l:i h l
. ; i,.
O
, , ( , , , 0.900 - 6+e Ie 1 0.6 0.8 1.0 1.2 , l.4 1.6 1.8 2. 0 l- ]a
, ;' , NYORCCEN CONTENT (H/2r Atom patio) PHASEstPHASE oa j 1
N I ' ' 0.890 -l i p; . I.70 1.80 1 90 2.00 j g H/1r l 1: .' j l 1 3 s l
! (c) VARIATION CP DEGREE OF ,
!. b P.
s TETROGONALITY OF ( HYORIDE WITH HYDROGEN CONTENT j J i l' -. 4j .!u, 300 l - 1600
' A i.
A 2r-w 25 wtY. u ; 9 0 2r-M i = t */. U f 1500
} 800 -
O )
! U _
O 1400- ' g', . g (bec) y 3700 - - 1300 $
, g 0+6 2 w w I . O -
1:00 I
- O O 600 - A ,
\ I- 1100 (b) 6 ANO (d+ Si ZlRCONIUM HYDRIDE PH ASE BOUNDARY
-4 tC00
? i i i i SCO 0.6 0.8 t.0 1.2 1.4 M/2r Fig. 2. Zirconitrm-hydrogen phase relatier. ships (from Ref. 3).
i 34'
- NL' CLEAR TECHNOLOGY VO L :8
- - - -~
LWL'ARY Fe
Simnad et al. TRICA RESEARCl! REACTORS s l 10 0 - ,1 i
,i ,1 ,
,1
/ / / / / - / /
OXYGEN. 51400 ppm ' ' ' ' ' ' NITROGEN 5 200 ppm j j j j j j j jj CARBON 5 600 ppm
/ / / / / / / //!
/ / / / / / / //s
/ / / / / / / /////$
~
to s'
/ '
/ / s' / s'/ l/ s'/ //// ll/ /
$j O
,' /
' / / / / / / / / / / / / / /O; i / / / / / // / / / /
1 / / / /
~
/ / / / / / / / / // / // / $
/ / / / / / / / / //////
/ / / / ///////// *$
E 1 $ / / / DE(TA' BETA' O.1 s s s s s s s ss t # s s # e v4 o O l e t # o ' - ' '
,y*
0.01 750 800 850 900 950 450 500 550 600 650 700 , TEMPERATURE (*Cl Fig. 3. Dissociation pressure isochores of, zirconium hydride (expressed as H/Zr atom ratios) (from Ref. 9). C -1.4 to 1.9. This is compatible with the transition material of stainless-steel or nicket alloys will Too from fec-delta to fel-epsilon, involving a con- provide a. satisfactory diffusion barrier to hydro-tinuous anisotropic expansion of the cubic phase. . gen at long-term (several years) sustained clad-l The isochores of the delta-epsilon regions of the ding temperatures below ~300*C in a water or ZrH system exhibit a progressively increasing steam environment.
~
j change in spacing with~ increasing hydrogen con- The equilibrium dissociation pressures in the H/Zr composition range of 1.4 to 1.7 at tempera-
] centration.
gression is Any deviation attributed from this type to significant of pro- tures up to 1300*C have been measured.' The contamination
,. of the binary with oxygen, nitrogen, carbon, etc., results for an H/Zr range of 1.55 to 1.7 agree to form a ternary or higher-order alloy system, closely with the values obtained from' extripola-The hicka. _h..Mia en-n em n- e tw/*- % 1 O tion of the reported data that extend to 950*C.
~ ~
. are sincle ch,se (de1ta ofeesilon) add are not_ However, the data for an H/Zr range of 1.4 to 1.5
- subieet to thermal rh,ee separation or the-,1 indicate that the hydrogen dissociation pressures eveline. For a composition of about 7ru. 'k a- for these compositions are considerably' lower
'w a e c o r a is than the values extrapolated from the tempera-
! inuilibrium nyorocen di s e nm ' N n, tures below 950*C, probably as a result of phase h 1 atm at ~i60-C._ This alicw= considerable varia-
- $ unn in fuel central temneratures without t,uilding_ changes at the elevated temperatures. For ex-y ~ un hich internal cas craceurac in tha fuet element, ample, at the H/Zr ratio of 1.5, the measured disscciation pressure at 1100*C is 7.7 atm versus 3
ihe ahea-ca nr e a r~ d ekgea in tha hichar hv-l the extrapolated value of 25.2 atm, and at 1200"C i e yrides eliminates the problem of_ larce volume is 11.5 atm measured versus 70 atm extrapolated. h
- j
{j _associ.aea 3es H0 C trwun naase trancmrmatwns at
"'a tower nvd rida enmnneiticas. _ Simi- The influence of carbon on the dissociation j _ _ r" tha -*-aara nf sicnificant therm?! hiffucina pressures of hydregen in carbcn-modified U-ZrH
- 4 n /drwm in the hicher hydridas nrecludaa co' - fuels has been m ea sured.b" The dissociation l y 3, , o , . -i..-,, ., , ,, - r e
,ad crr ekiac._ yhe glad pressures were found to be predictably higher i
P 2 TEAR TLCH::r:.OGY VOL 3- J A.NUARY 1976 - 35 l
S
/ /
l
--v'""*2
- E ,-- n i h.---.
- ,. . c6.
u . . ,sy . . , . p- , 6, ,,, 5.7 . . p ',
,,_ . . , tk y .- . m,.., .-... m,.. 7.m,
; g% -
. . Y. ~ ' ~~~~~Q*3
- ;. Q.
.. @ "c.i % d e d M D3. g hj M M q.';$ 4 O @ %.h d d W.%$78
-y??.WMi:.yM.%Ykin%; ;~p1=25!%=W.Wh& =**W.W5%W.ihsf^.-yb T[
3N N N.
- 35. .T.,. .. \ s. . . - . - . .
- m. _.-ym .9 ;.u.p.m p: u q.g -
, .g . yQ
'l %
i ! I Simnad et al. TRIGA RESEARCH REACTORS
, than the dissociation pressures of the carbon-free been described." No aianificant hvdrogen redis. (
! hydrides. The hydrogen dissociation pressures tribution was observed in the delta- or opsilon- '
are expressed as a function of temperature and phase hydrides, in the lower hvdrMe* - het"a-ar, e composition: _ extensive mic*'tian nf hvd*~nn Nk place.
.l P = K exp(-aH'/RT) mwg vioaor, a w a - - ., m e --m n, 4, a its
~^
t rim onium hydride havet been m en cured in th e'
- where the value of K is governed by composition. temperature range of G50 to 800*C (Ref.15). Thg '
The carbon associates with zirconium on a 1-to-1 nvarocen aosorotion follows a carnbolic time hw ratio. 'and the rate constant is proportional to the con-i All available evidence indicates that the addi- centration difference and to the square root of the 1 tion of erbium to the U-ZrH introduces no de- diffusion consta'nt. The temperature dependence leterious effects to the fuel. Erbium has a high of diffusion is given by boiling point and a relatively low vapor pressure so ' that it can be melted into the uranium. D = 0.25 exp((-17 800/RT )] . zirconium uniformly. The erbium is incorporated The di'rne< n, a k-a-ma, :, -4. 4n- sva.<aa into the fuel during the. melting process. All the analyses that have been made on the alloy show found to be indeoendent of concentrntion. that the erbium is dispersed uniformly, as is the uranium. Erbium is a metal and forms a metallic solution with the uranium-zirconium; thus there is PHYSICAL. MECHANICAL, AND no reason to believe that therc will be any segre- CORROSION PROPERTIES gation of the erbium. Erbium forms a stable
f hydride (as stable as zirconium hydride), which _The density of U-ZrH decreases with An,A- n also indicates that the erbium will remain uni-crease in the hydrogen content, as shown in Fig. 4 formly dispersed through the alloy. Also, since (Itef.15). The density change is quite high (15%)
neutron capture in erbium is an n-y reaction, _un to the delta chase (H/Zr = 1.M. a nd then there are no recoil products. changes little with further increases in hydrogen. The erbium cannot migrate or segregate in the The Enermai conductivity measurements hav6 fuel at the temperatures and times involved since been made over a. range of temperatures. A the diffusion rates are much too low. Inter-I problem in carrying out these measurements by metallic diffusion rates follow an exponential conventional methods is the . disturbing eff.ect of , relationship with temperature and are extremely hydrogen migration under the thermal gradients ~ ! low at the operating temperatures for this type of insposed on the specimens during the experiments. i alloy. Thus, with a conservative diffusion coeffi- This has been minimized by using a short pulse-cient of 10-" cm*/see at 800*C, the diffusion heating technique to dete mine the thermal dif-distance would be ~0.1 mm/yr. Hence, there fusivity, and -hence permitting calculation of the could not be any significant migration during the thermal conductivity. A value of 0.042 + 1.79 x lifetime of the fuel. 10 T cal /sec-cm *C is usea zor the tha~il e cw.uucu ay for TRIGA design calculatione, g i ~ The meenamra ! -rann-Hrm of U-7rH nra diffi- g l Y' MinR ATION OF $4YDROGFN nonem cult to measure because of its h*ittia -m-o, i
-ypen..e. conDIENTS l
Howeve r . at elevnted +amneratures it exhibits e I ! sienificant ductility and creen deformation. rna Fual ale ~ ant anaration in a reactor is not creep strength is markedly influenced by the isothermal, and hvdreven micrates to coida* +am- structure, as shown in Fig. 5 (Refs.18 and 10). 1
,oerature racions from hich-temneratura rawinns. T_ha bate rhnse has a much lower creep etra--th Tha acuilibrium cissociation oressure obtained th'n the datts ohnea. Thic 4e nn imonrtant factor y+ = n the racistricu ton is comolete is lower than E +ka -a'-H "alv arentar irrndistion ' stability of f the dissocin*{on cressure before redistribution._ tha dalta ohnsa at hich temoeratures. r The cimencionat chan7es of fua) rods due to. The hydrida fuel has excellent correcien re- '
hvcrczen ~fe"-***"_'*a of mi- nr imoortance in sis'tance l'n w' tam Bara fuel soecimens have been
.sn -ni+- ~a ., - c ' a "
"eaa. In the SNAP-lon subiaeted tn 9 ~-ac erri-ad t mta- mm ranment 9t l reacter, the small amount of hydrcgen_-adictrihu- '570*F 'ed 1m ei dv-i-w ' /.00-h neriod i- in
( '
.tien in the high hydride was found to be de- 'autoc19va? Tha , -o - - n ----aei~ c'ta wne vo'
~
' tar ~iaad by t e m r a rr tv-a wr'diarts 2,thi" 'ha- mc/cm 2-mnn'th weicht cain ._ accomoanied k- '
,ala7% _
-n r~.a e, nn nf t8a ey-e n a ie..a. nr twe hydriaa ec The results of studies at General Atomic on an adharent oxide filmm The maximum ext ent of l thermr4 migration of hydrogen in U-ZrH fuel have ~ corrosion eenetration after .:00 h w = (2 mil.
1 .
- o_ o
Simnad a 21. TRIG 4 PJ. SEARCH REACTORS l , isa. O BECK (LAR 10) On- 7,4 e CALCULATE 0 DENSITIES , cr' O KCRST (NAA-5R-6880) L P E TOY, VETRANO (NAA-$R-4244(93-7)] lta 7 2' 4 AVERAGE OF 68 SER FUEL ROOS (93-7) I th7 A NAA 116-1, -2 EXPERIMENTS , Q TAYLOR, AMBROSE (NAA-$R-9782) 3e a's, 7,o _ j on- d the. nee 6.8 d
$ 80 Zr - 20 0
.%_ 6. 6
~ '
ide {m g 5 6,4 - 85 Zr - 15 u a _ 6.2 - 90 Zr - 10 U A'" "l'A'====
'S in- 6*0 -
aBOUNDARY 93 Zr - 7 U AT EUTECT010 **d i k*b D TEMPERATURE g PHASE
. 15__Zr - 5 U_
a __. hen 58 -
"^ SOLUBILITY OMRE
. ;en. EUTECT010 BOUNDARY 4+e I ave TEMPERATURE 100 Zr r j
'~
l I by "+3+7 *0e - ! I' f f f t I I I I It e e 9
- t of 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 ants -
H/Zr ats. . se (, Fig. 4. bensity diagram (from Ref.16). 79 x ' gyna.4 ,e-+e -,~4na ,.4 ,+ c'aaa-a' "'- sive irrndfation tests in the SNAP rew+ne nrn-mal c u n i, tt,n + +' e -irranti.m hydrida e~et a-e
-a ' ' gram have led to empirical correlations between l l 'reH+1 vale Inw 1 eactivityin. mater. steam. and air. _swening unoer steady-state operation and th_e
! iffi- 2
+ac+e w,"a important variables of temperature, fuel composi-
- re. ml ' jn'om* amna-atures i..r a m. ,-- u ~.un %.-f =Ta,.
o r ganar',j er it . Theen 7 -u e -a - L tion, burnup, neutron flux, and fluence." The bits 5 Cae n f6 a. k e +4 aw *n n e bich n e A *0'f'. offset crowth durine ea rly life (uo to ~O.1 matal . The I~ at.% burnuo) is ascribed to the vacanev-enne'an-l the sation-tvoe crowth phenomenon. Future deveinn-19). IRRADIATION EFFECTS ments in the cuisine U-Zr hydride fuels' will hew"ma e' agth include etudies of the in%aaaa n'
'Eva, a _ e,+ a nmra. -+sa kew aa- < +we <n a t
- tor The U.S. Atomic Enerev ('nw i"i nn W r')
y cf E set a limit of 9M mm (0.1 in I fne ' wi'"d N1 "ada- mieinw amad'H aae-Instrumented pulsing fuel elements hav'e been h growth of fuel elements for all outsine reactors. fabricated to determine the temperature dist:ribu-re- y l evn-+ s a ! a e c m al alam a-+e w -- a n-a-etad inta ct
- een fiir lone carinds in TRIGA reactors with cladding tions in the fuels and claddings and to record the 2 at y gas pressures in the fuel elements.
g i etonentions of to to M -m Wugoslavia, non- In the ACPR fuel elements a small gas gap
. an ;.j ! ;ulsingiTIuGA). _A maah'nical rntaha""" ~aab-(3"5 gm,0.015 in.), prc ided by menns of dimples 350 g ' 2.-ism a,i cad th ce a imre" I a!nnantions . wh!"b -"n in the cladding, introduces a thermal resistance to a y g au-4 -+ a hv enitable fuel element desien.
9 10 :1 Burnues of un +n ~0.52 total metal at.% (75c-c control heat flow rates from the fuel immediately
' Of 3I dnun of the 11) have been attained successfully after The pulsed operation and to prevent film boiling.*
fuel elements can be pulsed to temperatures
.:h TRIO A ("al ala-a~e. The results of c:cen-Al liti AR TECl!NCLOGY VOL;S JANUARY 1976 37 3
2
.,1 j* .
'(,
.w
? ,,
m... -
'"'s-y>.
. , ~ ,. llC , ' *},:g ,*.}
. . ., 2 . - ; ; ec .y y . . f, . Q{w .,..n . g,..-;- .~ -. :[a".~[
_ ,. , ,: . :;y .
.. -&,):&. ,a
.~.Y,C.-~__,_.:m
';T..l. y w'p; '~Q*C. , . T-Y.?-w& *,~~.
-~- ~ _p . gq * *'"V
- 7 R*':'h, .;.:L;. ;w-
. , . . . ,' *l; - L *. .~~
. f!lL.; :.V
F .
, . . . _ ./ ._ w mE
-~ - '
_.__m .u.-.. u,.a;;.r ,. a._- , . ,,. a: ;f2_ g._w 17 h0[E 'I - i
" Sirnnad et al. TRIGA RESEARCH REACTORS
-3 1o neutron fluxos under pulsing and steady-si power conditions.
ZrH, at 4000 psi (p) Afuch information on* irradiation effects hydride fuels has been generated in the SN.;
, reactor program." The swelling of the L' -
o ZrH, at 3000 psi (D) (Ref. 18) hydride fuels at high burn,g is governed by th basic mechanisms: 10
=
~
- 1. The accommodation of solid fission prodt resulting from fission of *"U. This lec
= -
an early estimated growth of ~1.2% to L
) -
av/V per metal at.% burnup. This et ti: - anism is relatively temperature insensit:
.s
[10 5 2. The aggiorneration of fission gases at c Q
= vated temperatures (above 1300*F). T I takes place by diffusion of the xenon ,
$ krypton to form gas bubbles.
e - 0 g, g A saturable cavity nucleation phenome:
+
zrH,', at 4000 psi (6) 3., that results from the nucleation an
! i of irradiation-formed vacancies into vc
'O
-6
=
/ -
; over a certain range of temperatures wht
- ; the voids are stable. The saturation ZrH,,,, at 4000 ps i (6 + e )
;l j growth by this mechanism wasIe'rmed offj
_ ! swelling. It was deduced from the~i% _ [ decrease in fuel-to-cladding AT experiend during the early part of the irradiation. T
,o -7 t , , ,' saturation was reached in.~1500 h.
- TCMPC ATURC ( ) - [* The highest swelling occurs in the beta phr
, 1 ' at elevated temperatures by means of the fist
' 'A Fig. 5. Creep properties of zirconium hydride: com-I gas agglomeration, because the low creep stren:
parison of p with 6 and 6 + c phase material of the beta phase cannot accommodate the fissi (from Ref.17). gas pressures in the gas bubbles. Sweeping
"!. fission gases can occur by phase boundary mot]
if the beta phase forms in the irradiated fuel i i
>1150*C without exceedine the safe level of the For example, beta-phase fuel specimens (H/2:
l internal hydrocen oreseure. Test elements with 1.2 and 1.4) werd postirradiation-annealed ah
' hot spots of ~117W knva avbibited Icent swelling low-temperature (700*F) irradiation. Anneali after ~200 to 400 enlees. The swelline resultari for 211 h at 1300*F produced small amounts
; from inta ~ ' narosity fnem ed hv tha -nduni shrinkage, whereas annealing for 75 h at 160C nue!entian, crowth, and micration of hydrocen
~hiikkinc +ne va ma curfnce in the hot-sect recion, !puase produced fuel15 to 25%
(H/Zr = 1.6 swelling. to 1.9)Anneall.ng under the del: sa: J I "In ~ standard noncapped TRIGA fuel, the etendv- ,! conditions produced small amounts of shrinl: state power temperature levels increase after f or swelling-<1.5% in all cases." The shrink: pulsme . ...s ...e '~m - -**-'*"*"d * ' " l Iof the fuel on postirradiation annealing is ascrit
, w.% s , - . - -n kr+mpo, tho f,,a l and to recovery of the matrix from damage causec
) c !n ddNw. The cao fermation is caused bv the temperatures lower than those employed dur' l- raoid. fuel exoansion d"H-w the a"ise heatinw of annealing. Anomalous shrinkage can also be t h a 4" a t ._ _ _ . tributed to hydrogen loss.
An in-pile high-temperature King furnace ** ; The samples exhibiting large decreases provides a means to investigate the behavior of : density _showed cracks and. voids .that. .s_ugr l reactor fucis in high-temperature transients [e.g., ' fission gas agglomeration." The void cluste-high-temperature gas-cooled r e a c t o r (HTGR) high-hydride samples were correlated with c: ccated fuel particles) under transient heating len-phase banding, which led to the conclusien
; conditions by neutrca pulting to over 3000*C. some damage mechanism (other than fission 1-A rhedium self-powered neutron flux detector agglomeration) takes place based on an app 2:
has been installed in the Sandia Corporation delta-epsilen-phase boundary damage phenc TRIGA -AC PR fuel elements to determine the non.25 The epsilon-phase irradiation data indi
.38 NUCLEAR TECHNOLOGY VOL. :S JANUARY Q 9 - _.- _ . - _ _ - _ _ _ _ _ _ _ -
e f
.s
' Simnad e2 al. TRIGA RESEARCH REACTORS TI
.Y. , t i he presence of stress-related fuel growth phe-tests, except for the NAA-121 capsule test in wrnwth (= h wir *naini nri- which the temperature profiles were similar, and A, I:n:menaAnsince the consideration in integrating the data correlate well with the S8DR data.
important j!?i
- g. .pd. A model of swelling based on burnup and
.. *~fI" radiation results is the presence of thermal .Q(,; i
.-ac:ents in the fuel samp~les in the tests. It is temperature led to the relationship for volumetric
~
Nfe'EJ"fr~oYn' the~dat'a Yhat the fuel in the reactor growth of** i
.csts and in capsule tests, which clos ly simulated ,
jj [p
.ne.5 NAP reactor operating conditions and ther- % = a3 + S exp(-A/T) ,
.nal gradients, showed a different temperature-Q where swelling ,, relationship c ompa r e d to the other ! lb ,
e*Esule fuel experiments where the fuel tempera- -
-.: i-n
'I tures were more uniform.se
%SV V
=2AD
#+E AL 4 N ,1 j A number of attempts have been made to p carrelate the measured swelling of SNAP-reactor B = burnup, metal at.% F" l fuels with burnup, temperature, and hydrogen con-t ent. Large uncertainties in each of these param-T = bulk average fuel temperature f)i;,
eters have made it necessary to use a statistical a' S' and A = constants. h pp sample *' based on clusters of points rather than en individual datum points. The fuel swelling data are usually plotted as f The observed dimensional changes indicate log (corrected volume growth) versus 1/T. The h variatior.s in the ratio of volumetric change to corrected volume growth for offset swelling has P diameter change large enough to establish the im- been given variously by portance of diameter as an engineering variable.*' [dr ay hL 7pn t.,.i y e. + se sea +*nnie only when the r4 tin cf fri r to AD/D is 3._The fual w-aw+h (= n*a'a--a T - 2* 03 hh VI T- me radial diractinn when the ratio is between l and 3 and in the axial direction when the ratin h MV - 2.8B . f rever thnn 4 Ayin t shrinbee civam v4=a tm e 4[
-e af M The mean value of the tr_ tin h +ka and more recently b p:
te!m ehnen i= 2.8 n; we fnr +5e . =tinn ew,,o
- i~,n.nn a $y pi
[p .
' :n the swelling correlations used for the SNAP- T - 3B reactor fuels, the bulk average temperature is [1
- f. "l insidered to be identical to the arithmetic av- (the Bonzer-Swenson correlation).
~
Enge 5 ti1[siens c'e'nterliric temperature aniti[e. The S8DR data tend to show a greater tempera-ture sensitivity than that shown by the Bonzer- {j ;
.< urface temperatur,e..,2' The time-variation of this y
'emperature was calculated from the beginning- Swenson correlation, but the bulk of the data does j h
'i-life (BOL) temperature, based on the cladding fall within the Bonzer-Swenson scatter band and b i
*hermocouple readings, the measured end-of-life indicates a higher temperature sensitivity only if
.EOL) fuel AD/D, and th'e following fuel swelling considered as a separate data set.** The S8DR i!
I
*n ed el- data show a correlation of [(aV/F) - 3B)] of 0.8% at 1250*F, 0.4% at 1200*F, and 0.2% at 1150*F. The S.
av/V = 3B + exp(-K/T) , total correlation is as follows:
~ [Af f
#' av *
~21.5 1860 -1
= 5.5 exp - .3b exp
--3 T
-2 l .; V ,
l r B = burnup, metal at.% 1, I ! f 6 , E = constant (~30 000) \ C
+ 21.5(/1860 - 1) i ,
T = bulk average fuel temperature (in degrees 7 [' R). l t where P I
.s stated earlier, the greatest success has
,g .
b = burnup, metal at.S (- ichieved by using the offset, or equilib- '
~
m, bulk average fuel temperature that the fuel
- pr , .g/10 000-h operat:.on j - ees af:er offset swelling has. been com-
The feel swelling cbserved m the SNAP T = absolute cperating temperature cf fuel at I
'a., ;nd SCDit reactor experiments was gen- the time offset fuel grnwth has heen com-1, l # ~ " lower than predicted from the capsule pleted , "R.
3 39
' " w v u w. vet :s :,mem m6
?
o _ _ A'- - = . mm._mmm., w _
~ -
y,, .,g s._ n _'_ _. .__ __< _ _. m._ _ _ .f . , z. m, 1 ma,r _ J-
.m -.,s
~
- ~ ~ ' ~~
Simnad et al. TRIGA RESEARCH REACTORS V
' PULSE HEATING lf g-nulses can permit an increase in bubble size ovt
- a ceriod of time! this wouM tend to iacrence o
. The U-ZrH fuel elements used in th'e TRIG g disructive force while wo*anine the reet-nind reactor are capable of operation under conditions f: matrir thus crndun11v nrnducine the coreus e' of transient experiments for delivery of high- panded fuel that constitutes the grav catch _. Oth<
!! irlensity bursts of neutrons. For these exper- possible mechamsms wer.e considered but has iments, the reactor is equipped with a special been tentatively rejected on the basis of evidenc I
control rod mechanism that provides a method obtained in the postexperiment analys es. of obtaining a step reactivity change of pre- There was no indication that exet.ssive hen-I determined magnitude in the reactor. Theine the transfer rates contributed to cladding distortic l nuclear pulse, nearly all the enerev is stored as Cladding material cut from the fuel elemer thermal energy in the fuel material. This results appeared to be straight and true. When the initi. ! l in an almost_ instantaneous rise in the temnevature longitudinal cut was made, the cladding spra: I ! cf the fuel body. These fuel elements have oper- open slightly to a. uniform gap of ~Tirin., as wo1
! ated repeatecly in the Advanced TRIGA Prototype be expected from the residual stress remaini:
. Reactor (ATPR) to peak power levels 'of over due to the action of the die in the final drawir l'
8000 MW, providing a neutron fluence per pulse of operation during fabrication. The thermal-stres ! ~10" n/cm'. distortion due to excess heat flux would has
" tended to produce residual stress in the opposi' t The ATPR fuel elements have been subf ae+ad +6 thousands of mileac af 9000 MW 's-4 ~ n-a Jhg direction. Additionally, the external surface di:
nuclear safety stams fenm the larce cromnt neen- coloration was far less than for a normal elemen d' tive temnernt"-a r e a '" ri a -+ mf resetivity nf tha after either normal maximum steady-state c Cl uranium-7irconium hvdrMe fn a l - -d a -e
- a- ~e- pulsed operation.
H
..i terin t . The inherent prompt shutdown mechnnism Internal gas nracen-a "a c i-d'-e +ad +^ '
of TRIGA reactors has been demonstrated ex- negligible compared with the via'd anint 'ne H e tensively durine the tem of thnnennHe nf n"I c a c claddinc- The pressure-transducer calibratic
]j Icondneted na " '" -aea+^*e These tests in-volved step insertions of reactivity of up to was rechecked and found satisfactory after t1 test. In additio.n, a pressure-instrumented elt 3fo 6k/h. An in-pile test has been performed on ment (identified as 2E) was pre'ssurized to 50 g
! fuel elements of a modified desien (gapped) for after the test and maintained pressure overnigD.
high performance in the TRIGA ACPR. As ex- which verified that there was no measurable le:
;]l'. pected, there was satisfactory fuel body per- in the element. Note that gas chromatography w:
1 formance after 400 pulses at temperatures up to performed for gas extracted from element 2E.
"' ! the design point of 1000*C (Ref. 2). There wns no this test no hydrogen gas was detected, althou!
I evidence of interaction between the c1nd end the the instrument has a high sensitivity for hydroge' The mechanisms of nucleation, growth, at I ] fuel The +-n n sient cas cressure in the sence 1 between the fuel and the clad was measured durinc migration of gas bubbles in solids have bet j. 3 the rul=e. and cank aracences va-a <mma +n wa
<40 osia-well below the emna r bcnna imnlina studie,d extensively in recent years, mainly connection with fission gas formation and swellir I I the ec"ilibrium 'ressure data. 'As testing o in nuclear reactor fuels and helium formation t
!3 l by at h i gh e r temperatures continued. there wn= nuclear transmutation in alloys. This informatic hj l some evicence that at hot-soot regions, where is most useful in elucidating the damage meet PU ' : the temoerature n an - +ba 'nal surface renchan anism. Barnes and Nelson, Nichols, and Lav b[4 .I ~1200'C , the fuel tradually ewalled =15chtiv ova- s tarce number of outea= n daw +% < n-a,ra nr+ha ton et 21.#* have presented reviews of this subjec The conclusions reached from the results of rc "i ' hyriranan nressura in entl hubbles th9t nucleated cent studies can be summarized as follows: T' I
y _in the hot ena*= I-19w m -u _ > , ~ ' behavior of the gas bubbles determines how mut
. L; -formed a crnv n9tch_ gas is released and how much swelling is prt The bssic c9use nf fuel bed
- di c t *a c e a r*m- duced t,y the gas retained. The h"hbi== c
! mnea than 200 oulses neeears to ba Inan t nver- micrnta bodily under the influenca of variot h "' tine of tha fuel body as a result of tha-~ L drivine forces and by various mechnnisms.
gggy fley neef- * , "-' t a r a a a h n e chnnnal most enses, tne buchlae ~f e n '- - di -a-vi,-as +n +% r"p o an ichad esaciel teet ala- tien determined by temoerature cradiente. =tra: ments The mechanism by which this apparently 1 c_radient e . snd m n + die'~ e "-ae - - - cecurs is as follows: If the internal temperature ~ beunda ries. 3nsivsis indic*as that ems!! MM arri hydrocen concentrations are sulticiew w are dominated by the behavior of dialaaat!~ Iiu n rncr e a ht:5bles m nvcrocen cas. otastic nr craan _ncwever, as their size increases tha 'a=*a-,t-yicidin~ M tha '"a! -a-"' -a- ~--c - w t-n ci ent coenmae -a-a '--" Models bar 40 '
- NUCLEA R TECHNOLOGY VOL 28 JANUARY D
-9 0 _ _ _ _ _ . _ _
,c . ,
- j. ,
Sunn.nl a a!. TRIGA RESEARGi REACTORS I fold. First, small . power reactors using U-ZrH wer- j on the behavior of dislocation lines and grain the boundaries are appropriate when the temperature fuel can safely sustain accidental power excur-
- dng gradients are small; little gas escapes and swell- sions to high fuel temperature. Second, and per-ex- ing is then the main consequence. . haps more significant, high-level pulsing reactors
- her On the other hand,'it has been found that whara (fluences of 10" n/cm 2) can be operated with
- ave steep temeeratura ~ -a d * * = nea"* tha haha-inr U-ZrH fuel with a reasonable fuel lifetime. Fur-mce is more comnlen diffa-a + Wa1= beine recuired thermore, with regard to standard tvoes of TRIGA researen reactors, it is evident that n-amad aa*
for each temcerat"re maa- A__t hich
- a m a a -+"-e =
!at- ne cradient can drne b"bklae 'an ~ d N ' b E- = fuel temneratures are conservnH"a ion., ,rd the bubbles micete no tha temna-m+"-a ' - - ' - The results of rapid dehydriding tests indicate ents dients, becoming trapped on crain bounda ries that the endothermic nature of hydrogen loss
.tial -.there cas is reieaseo oerindica llv. Barnes and slows down the rate of temperature rise. When
-ang Selcon" nave postulated that bubbles migrate the hydride specimens are rapidly heated to ele-suld predominantly by a surfr:e diffusion mechanism, vated temperatures in a dynamic vatuum system,
.ing g while the main Darammers determininc tha be- large-scale . internal cracking takes place, where-ving '
havior of the bubbles are temperature cradient. as when a backpressure of hydrogen is maintained ess bubble radius, surface ainuston. vapor pressur_e_ (as in a clad fuel element), the hydride fuel body tave -nf the solid, and surlace tensinn. 11 the material contains relatively small bubbles that are associ-site is stressed, the moving dislocations will drag the ated with the grain structure and substructure of dis- , bubbles. The temperature cradient la rcalv da- the material. In the SNAPTRAN (Ref.' 32), TREAT (Ref. 33),
'~
ent, *arminac the critient si'es of the bubbles. or These observations are in ,line with the con- and KIWI-TNT (Ref. 34) tests, high-hydride modi-clusions of the present authors regarding the fied U-Zr hydrides have been pulse-heated to be , mechanism of formation of the distressed area in destruction. In these specimens the hydrogen content (1.82 wt%) was very high, and the temper-j the l the hot spots in the pulsed special test fuel. This atures were high enough to rupture or granulate i tion j area evidently was subjected to a temperature the fuel. In the KIWI-TNT transient tests, the l the . i range and to cycles 'of thermal gradients and cle - i stress gradients such as to favor the nucleation, specimens were exposed to a large nuclear tran-psi migration, and growth of hydrogen bubbles toward' sient under the following conditions **: the surface. ght, ! I As tbIa h"hbles crew in size. the internni 1. In containers designed to withstand internal - leak was N hydrocen cressure en"M nnt be neenmmodnts k - pressures of 120 000 psi - l In !4 -'i h a mat m e trir which consecuentiv crachnlly 2. In chambers sealed with prestressed rup-tugh 25 iia!d ad ,rd e raHed until n mado contact wHh the
- ture disks calibrated to burst within 5% of ;cn. g cAg. With enhsequent pulses. _t_he cladditur specific pressure values and i* itsalf ie emeHed to make provision for fuci reen n4 swa!Hnc bv ite own exnnn=ina Tha' W 'b" 3. In a container in which the hydrogen gas r in j eladding will be deformed by swelliac nf the enci released from the fuel acted on a free piston
' ling g body at temperatures below those where crono or that impacted a copper anvil and produced 1 by M +a-an -= oreceure can enuse the claddinc t > an indentation calibrated to give a measure tion 'M This phenomenon follows from the fact of the gas pressure as a function of time. g
- ch- q that in a pulse, the rapidly heated fuel expands l aw- g thermally more than the cladding and thereby in type 3 tests, extensive cracking of the fuel ect. "
forces the cladding to expand once the initial gap took place since large void space' was availaole. re- ' has been bridged by swelling of the fuel body. Much less fractilre occurred in type 1 and type 2 The TMe , c c o"d e rather weH fn* the fact that the tests since the internal gas pressure was bal-tuch ch?ncae nre coverned by a proeressive nrnenes anced. In some of the TREAT tests, the fuel h irradiation temperatures were high enough to melt
)ro- '
~3~
M de na tnt e nbee curine a few pulses, l 'c an. 3' Frnm the enent*e nf " nen *c m " ~ be- the fuel (~1800*C). l .cus N cenalnded th,t it-7rH feal af amente can be safelv Measurements and calculations have bden re-
" '" a 1 + a -- t ""c e W ported of hydrogen loss from hydrided 10 wt%
l In ! Tulsed even tn va-
+n f"al temnarn_- U-Zr fuel elements (1.25 in. diam x 1.0 in. Iong) ec ' ' r'il a t ta- ~~ "~ ann -oi e e' a "' that were rapidly heated by induction to tempera-es%f tures in excess of 1100*C did measuramede rain g n:n two of thn riva test elements exceeded the tures near the melting point." Results indicated Jes, .,
conservntive dimensional tolomnnes. In the first~ that within ~75 sec, the surface temperature in a ' 'es: 'l
. 200 pulses, there was .no external evidence of nonoxidizing atmosphere reached 1700 to 1780*F l ure -han;;e in any of the five special test elements. with only minor hydrogen evolution. Abruptly
% p'ractical consequences of this are several- thereafter, the surface was observed to crach Id
^ VOL.3 J A51'ARY 19h 4l
' G L\ R 11 GlNOI OGY i
- i d
.=vmer.- =rnwumr->veam:s% . .a . ., . e r=*mrm-rw. -ww _
w;'7" ~ '
- . ..rwY ,'
-= _ - .ft. w w I'.f.P.
c.~ J ^"..w F,
- f .f:2 '-9.:. . ,te.MM xm .6. .u.-Sa.7,'y.,
j : n,,.e. $.
- ...f,0.ic>
,. g 3,. .,P:-
r_ n_, y v. . . . . ~ :~ ' . <.
- = - = ....a. -
;;7 - - .--,m . -
, ;.a 3 ?m .
~
,yy'.**'*--" g
* ^"
- cu
~ -
.,.n m___~v- n _.mc. A ; % /r_.nt i"),; q q d ri a . & fw 4 - ~ a h * 'NY L i, _
i - t i I Simnad et al. TRIGA RESEARCH REACTORS pa'r allel with the cylindrical axis, with strong out-recoil into the can between the fuel nno c8 d_me gassing rates, and the temperature dropped.. After effect credomia,+ne in '" n ' f +amneratures un to a few seconds the temperature again began to rise ~400*C; the rarail -ala,en mte is docendent ca and outgassing continued. After ~3 min at surface -the f" a t entf,ra +a-vnb m a -ntio but is inaa. temperatures of 2010 to 20207, the specimen was cooled. Subsequent analysis she red large amounts controlline cendent nf feel temocrnture. Above ~.iOOP the ma-h,nie~ <a*
" m -" -a ' a, e e of residual hydrogen. In another series, tempera- Trnm TRIGA fuel is diffucion 'nd tha nmena+ -a.
tures up to 3400T were reached before power was leasad i= denandent na +5a r~av *n-na-,+n-a th. shut off. In these, almost all the hydrogen was _ fuel surface-to-volume entio. the time of irradin. driven off. The volume of the sa=ple was found to tion, and the isotone half-lifa_ I have decreased, and the surface cracks' visibly The results of the TRIGA experiments and , healed as the temperature rose above 20007. i measurements by others of fission-product re-lease from SN AP fuel have bee'n compared and FISSION PRODUCT RETENTION fo nd 6 be in gd agmme1 Tha fractional relasse. 6 of fission-eroduct
- i. A number of experiments ha e been performed cases into the eno betwaan Nat vd a M fann s
+n d a+ a -+ a a the awant to wmen nssion oroduc+= full-size standard TRIGA fuel element is riven br ar. -a+,5-aa w. it 7.u Nel These experiments e = 1.5 x 10"
- Tvere conducted over a period cf 11 yr and under a variety of conditions. Results erove that only a + 3.6 x 10' exci-1.34 x 10*/(7' A 273)l ' ,
/
. small fractinn of +ha fis sima a-~incts tre en- _where T is the fuel temoer,ture P Thi c '"na .
la,=ed- even in com oletelv u--',4 TT 'zrH fuel, tion is n!c+ted i, F!w. A. The first term af +W d The ralea se 'rnc+ inn vn -ine f-~-- i M v in" fni in t irrndintion tam-a-a+nra of 150P in ~10-8 nt A00*r h (Ref. 36). The experiments on fission-product to, ,
- l b release include:
'( ===
THEORETICAL l I 1.1960-the measurement -:.f the quantity of a MAXIMUM single fission-product iso:cpe released from y a full-size TRIGA element during irradia- to -
; tion. e.
- o L
2.1966-the measurement of the fractional re- , M lease of severalisotopes from small speci.
's mens of TRIGA fuel caterial during and 10 2
E L after irradiation at temperatures ranging M s
.i from ~25 to 1100*C. 5 m
3 ::: si 3.1971-the measurement cf the quantities of " ~ l l several fission-product isotopes released f 10 -
'i, ,;
from a full-size TRIGA fuel element during S : i l irradiation in a duplicarica of the 1960 ex- $ - l j periment. 5 0 Jr Postirradiation-annealing measurements -4 jg _ o of the release from small fuel samples _ l l l heated to 400*C. 5 Postirradiation-annealis; release mea- I
;f surements from a small previously irradi-
' '~
, ated fuel sample that had experienced fuel 10 ) o 1966, EXPERIMENT burnup to ~5.5% of the 2 "U. e 1971 EXPERIMENT, FOSTlRRADI ATICN ANNEAL
- 4. SNAP-measurements =ade as part of the o SNAP, POSTIRRADIATICN ANNEAL Space Nuclear Auxiliary Po rer reactor pro- m 1966 EXPERlPENT,, POSTIRRADI ATICNI
!{
l - g -am. -6 fANNEAL , , ! .l 10
- c. 400 Sco 12c0
' 1600 2000 !
7s. .,.-a 2~.a-+- e k e... .k-. .k. .. tv n l, TEMPERATURE (*0) mechanisms involved in the re:eise of fiscion Fig. 6. trecu a :rnm i nn e a' -- -f vhich ore- Fractional release of gaseous fissicn product ccminates over a di f f a ra r.? - - c -a ' n - a r'""a. from TRIGA fuel showirg theoretical ::axt-p,o n e. c,. u -,- 4,. .u-, ,' mum, and exprimental vs. lues above .iO O 'C
' e einn f ra nma"t corrected to ir. finite irradiation.
42 ' l
- ~
Nt' CLEAR TEC}fNoLOGY POI 28 JANt!ARY M h 9 __ _ ___E.__ _ _ - _ _ - _ _ _ ___
Simnsd er 21. TRIGA RESEARCH REACTORS i
-a-~ in steady-state coeratiort as the effect of neci-function is a constant for low-t em aa m *"~a
~ dental ran c+4"i*r a ba ~a= occurrmw from exceri-s o lease: the second term is tha M ah *- n a -' t"-a '~~ mental devices in the core is creatly red _u_c.ed._
- n. cor:fon. - - * ~ ~ " " ' ~ The basic physical processes that occur when
.T h a
- a la n e a " a+i "- e b'"a ' " M a a e ,o a standard full ei 'a '"a1 al a- a + -e -- e*- "'- the fuel-mcderator elements are heated can be e thouch Individ"ni -a'e"-+-
-e-a ~+da -" + h described as follows: The rice in temnernture nf
' the hydride i-~a' eee "Ia erobability that a ther-lifferent cea-a+=v ,
'"al alement mal neutron in the fuel element will cain enercy e The curve ia "~ c -"Hae +n n
- : hat has baan irediated for a- time sufficientiv ' Tram an excited state of an oscillatine hydrocen 4e 't annitih- ~ atom in the 1sttici. As the neutrons enia ana-~
. Ione that all fleeinn-crodna+ 'r+4"i+
d * "C n and the release fraction is at its theoratient from the ZrH. their mean-free-cath le in raneed
- Taximur 1 Ficure 6 shows that the menenred iporeciablv. This is shown cualitativelv in Fic. 7 d 'lalues of frFetional releases fnll well below the for a standard TRIGA can Since the avernce
+% a '" a ' a'a~e-* 4e- an- n' - b'a
'374. Therefore. for safety enneide-ntinns this chord lencth in
- t
' -"-"a ciWe "e-v conservative value= for +ke hieh- with a mean-free-cath tha a -ak- ki " +~. i^~-~.
a 'e-merature t rela'ea fram TRIGA fuel from the fuel element before cacture is increased. 27 ~ Also worthy of note are the following con- In the water (where the temperature remains clusions from the TRIGA fission-product release relatively constant), the neutrons are rapidiv re-experiments: thermalized so that the cacture nnd esc'ne nrah- , abilities are relatively insensitive to the eneru
- 1. Parn"se the samoles were unciad. the hich- with which the a-+a*e +h a 'rn t a r. h ne"_+-na .g ,g g g ,,,g
,T,hg, i
y ~ -' * "
- a -a*=urements were mace on essen- ,,,,n, % , , .
ls-i tially dahvdrided U-2r_. Postirraciatinn 'nnealine e neer= he crectr"m +n hntden more in the fuel , d ak d -id m.e n*n- - _ measurements m. dicate thn* *ba cess did not sicmficantiv affect the rela'ee rate. -than ntu r e -c inanthe eadwater. As a result, ent disadvantn ca '9 c+ne there fneis*haatemaer_-
" nit
_._ 2. Part of the 1971 experiments was the mea- cell in th a enea that decra,ces the ratio of. surement of the release from.a postirradiation absoretions in the fuel to total-cell absoretions as is ine-a'e ad Tp;fis., anneal of a sample of fuel that had been irradiated the fuel element temoeratura a+ne h*ince ohnut n shift to a burnup of ~5.5% of the "U (or 1.1% of the ' chance in disadvnn+nca ' total uranium atoms). 'Tha -asults of this part of in the core n e"+ rnn kninnna civine a inns at
~
reactivity. and is termed the cell affact. - the avna i-~+ ia&*+e4 that the effects of long- ~ The prompt negative temperature coefficien* term i rrn din ti nn of the faci en fission-oroduct Telae c a are small. at least for total burnuo__ for the TRIGA-FLIP core is based on the same core spectrum hardening characteristic that oc- _@_ia'laat of the maximum that has been achievec.- curs in a standard TRIGA core. However, for a t 3. "'he ralanse frnction -for nealdaat conditions n -- ,1 - a- ,+i - +a,n-TRIGA-FLIP (70% enriched) fuel element, the
'? i ie che ,,+ 4 e+1r er +wn uranium loading is ~3.5 times that of a standard
' ' f
$a -, *" - a a nt +h a ' n-norntu re durine accident TRIGA element, and this causes the neutron i' i a ~ d W -e Thi c is because the fission oroducts mean-free-cath in the FLIP element to be much released as ? -aen!t af ' "o' a d i i"
- e 're shorte r. For this reason, the escace probability Gneo +w,+ x , .. n - nco+nd in +wa syoi .ol,d can l duri~e an=~n t ena-ntion.
- 4. Since the fra? *a-na*"t"*a distt ibution i= 100 net is' n+ h a -~ n i it is ancessarv to intecrate the_ y '
400*C Temroe+"-a d7,A+ -M ea en 'm a+i na nvar the { < 80 - g Ia m ee rat"-a die +*ibutie in a fuel element. -j $ l L 23'C "h 60 - Y lE PROMPT NEGATIVE TEMPERATURE COEFFICIENT e$ om 40 -
.I "g h 9
,m The basic nar matar tb,t ' M a '." e *h a TMOA E e 5 20 - l -
- reactor system to crerate safelv during either g l_
'c0 j credy e+-en ne t -, n c i n., e n -a m na e 'e t ne promet_ . ,
i 0.1 1.0
- r. c -, , i" n +nmenr,+nrn enoniaiant of reactivity as- 0.01
[ NEUTRCM ENERGY (eV) 3 j 'Reinted veith tha TRIGA fuel ar:d core oesten.
* ' " ~ ~ ' " ' " " ' " " " ' ~ ' ~
- Fig. * . Tra.nsport cross section for hydrogen in :tr-2- [ nf +ha w hnvior of a TRIGA core conium hydride and average spectra in TRIGA
'C j 3,2a ~" i"" M w ., en -% --a , Hows arcat frecdom ZrHr. 'uel elemcnt for 03 and 400*C fuel.
43 ' ' *3 7 E AP. TFCEOLOGY VGL .4 J AXI.' AFY 19~6 2 A , , _ ,
*T.wy a > 'Y===w n.e y.. g . ~a , . . _ .m.,,,. ,- _,
**RI'YW'*F 3 ' . . ..-~-- ~-~
);> ..'y n - -
- 3. - -, . :. ,
-.7.-J w . K 4:.'.Y :-
m **:G'<. . .& -::. .- -;- 3. , .' ~~ .Z-. 8:*Q QQ ,
%, .bj ,
j[1.f N ,_ , ' ** . . ^ ~ ~ * ^ ." f - 3
. .L - -
--- n . . - .
f - -
.s ..l
, . , _ . . : -" _ .~ w-::. " -
mmurm;;sn :,:.,Llia j pj.M A *, , j - t
~
Y Simnad et al. TRIGA RESEARCH REACTORS r 1 r too 1o* the *"Er resonance. The tempera _ture-depe_nde character of the temperature Coefficient ofl
" FUEL TEMPERATURE = 23*C #a *ER
} WATER TEMPERATUR = 23*C [ TRIGAS_ LIP .codis advantageo.us in that a mig imum_, r._eactivity. loss is incurred in reachir
/ s q normal operating _. temperatures, but any sizab E - E increases in the average core temperature ress
\ 7 '0' E y 10 ,e in'a sizably increased prompt negative temperd i 2
,/ s
? ture coefficient to act as a shutdown.mechanisq E
Calculations show the temperature coefficient d FUEL TEMPERATURE = 7ao*C \ be insensitive to the change in configuration fro WATER TEMPERATURE = 23*C ' a compact core to the operational cor'e with fog 3.o , , , ,I , . .I 4 - ion flux traps containing either water or typical e5 0 001 0 01 01 10 periments. Burnup calculations indicate that aft d ENERGY (eV) 3000 mwd of operation, the *"U concentratik Fig. 8. Thermal-neutren spectra versus fuel tempera- averaged over the core is ~67% and the *"Er cop 4 ture relative to e, versus energy for "Er. centration is ~33% of the BOL values. Tempera 9 ,. ' ture coefficient calculaticns for the burned d [' core, including fission products, gave results al [, , _ for neutrons in the fuel is not greatly enhanced as shown in Fig. 9. The EOL coefficient is leg l the fuel-moderator material is heated. In the temperature dependent than the BOL coefficier j TRIGA-FLIP fuel,the temperature hardened spec- because of the sizable' loss of "Er and the resu12 i :- trum is used to decrease reactivity thrnurth its ing increased transparency of the ~0.5-eV ress I interaction with a low enercy resonance materini, nance region to thermal neutrons.
. Thus erbium, with its double resonance at The temperature coefficient, therefore, da
f ~
~0.5 eV. is,used in the TRIGA-FLE_ fuel _both as,a burnabl.e .. poison and as a material to enhance pends on spatial variations of the thermal-neutre spectrum over distances of the order of a mean l
the prompt _ negative temperature coefficient. The free-path with large changes of meap-free-paa neutron spectrum shift, pushing more of the occurring because of the energy change in thermal neutrons into the '"Er resonance as the fuel temperature increases, is illustrated in is Fig. 8, where the cold and hot core spectra are plotted along with the energy-dependent absorption cross section for *"Er.'As with a standard 18 - emw TRIGA core,the temperature coefficient is prompt CORE LIFE
. because the fuelis intimately mixed with a large 34 -
4 portion of the moderator, and thus fuel and y solid moderator temperatures rise simultaneous- -- h 12
- i ly, producing the temperature-dependent spectrum i shift. E fa f For the re,enan inst discussed. mnre than 50% 4 to -
,; of the tem eerq +u re enentriaa+ for n sta ndn ed gf
; 'TRIGA core comes fra- *ka ta pa m+ % dg . 52
*2 8 -
'. pendent disadvantace factor, or cell effect. and cG FNo-oF.
2r)% an r5- '- ~ h m'a- 5-aada + e nf the zaaty E{ core LIFE resen~5reat '8 " d + ^ ~ " ^ * * * " " aU ^ " a " d a "
- Ia' Imp hy 6 -
from the core._.Thaea a"ar+e -- ^ a a + n- a -, - g ture coefficient of ~9.5 x 10"/'C, which is rather *-
. ccnstant witn temperatura. On the other hand, for I a TRIGA-FL1P core, the 'effect of cell structure WATER TEMPERATURE = 23*C on the temperature coefficient is small. Almost 2 -
the entire coefficient comes from temperature-dependent changes in nf within the core, and ~80% , I I I l i I l of this effect is independent of the' cell structure. o 100 200 000 400 500 600 7 The calculated BOL temperature coefficient is FUEL TEMPERATURE (*C) shown in Fig. 9 for '70% enriched TRIGA-FLIP i fuel. It increases rapidly as a function of fuel Fig. 9. Calculated prompt negative temperature cc I temperature because of the steadily increasing efficient versus fuel temperature at BOL ar. number of thermal neutrons being pushed into EOL-70% ent:ched TRIGA-FLIP bel. [4 Nt' CLEAR TEClfNOLOGY VOL. S JANUARY 19-9 2- -. . _ ___ _
f .
, Simnad et al. TRIGA RESEARCH REACTORS p'
single collision. A quantitative description of group (~100 groups) cross sections, stored on
- ofdent'fa r these pr'ocesses requires a knowledge of the tape for all commonly used isotopes, are averaged
- nin- p differential slow-neutron energy transfer cross , over a spatially independent flux derived by solu-
. ;hi4 L section in _ water and zirconium hydride, the en- tion of the B-1 equations for each discrete reactor.
ergy dependence of the* transport cross section of region composition. This code and its related
- abla h cross-section library predict the age of each of asult [- hydrogen as bound in water and zirconium hy-era- dride, the energy dependence of the capture and the common moderating materials to within a few ism, h fission cross sections of all relevant materials, percent of the experimentally deterinined values, it to - and a multfgroup transport theory reactor de- The resonance integral method of Adler et al."is
- r.ptm__ scription that allows for the coupling of grout 9 by used to generate cross s'ections for resonance four . speeding up as well as by slowing down. materials.
The core thermal cross sections were gen-ex- Qualitatively, the scattering _of slow neutrons iftar by zirconium hydride can be described,by.a model erated using the multigroup cross-section GTF in wiiich the' hydr' ogen atom ' motion is treated code." GTF computes the spatially dependent
~
tion as in ~ isotropic _lia[monic ' oscillator' with energy thermal spectra at each mesh point in the cell,
- con-
' era- transfer quantized in multiples of ~0.14 eV. More using the discrete ordinate's method and the fine-d-up p precisely, the SUMMIT (Ref. 37) mcdel uses a group (58-point) cross-section data contained in _ s as ; frequency spectrum with two branches, one for the the thermal portion of the GGC-5 code. . optical modes for energy transfer with the bound Cell-averaged broad-group cross sections are less [
- isnt L proton, and the other for the acoustical modes for those obtained by averaging the 58-point cross sult- energy transfer with the lattice as a whole. The sections over the space-dependent spectrum. In eso- {
y optical modes _.are ..r_eprespied_a_s t broad fre- the past, cell-averaged thermal-group cross sec-y quency, band . centered at 0.14 eV., and with the tions have been generated by first obtaining da- g, width adjusted to fit the cross-section data of broad-group cross sections averaged over a 58-Woods et al." The 1 o w-f r e q.u.e_n c y acoustical point spectrum for the homogenized (space-inde-
. tron
- emn-
[ modes are assumed to have a Debye spectrum pendent) cell. Using these cross sections, a E
-path j with a cutoff of 0.02 eV 'and_g_weightdetermined separate cell calculation was then done to obtain broad-group disadvantage factors for each of the in a [ b_y an effective _ mass.of 360.'
This structure allows a neutron to thermalize regions in the cell. The broad-group disadvantige ['
. by losing energy in units of ~0.14 eV as long as factors were then used' in a space-independent ' j h its energy is above 0.14 eV. Below 0.14 eV the - spectrum calculation to generate the cell-averaged
- / 4 neutron can still lose energy by the inefficient cross sections. The use of 58 thermal-group
! process of exciting acoustic Debye-type modes in cross sections in the GTF code versus the-broad- [y which the hydrogen atoms move in phase with the group cell method just described, results in a 6 zirconium atoms, which in turn move. In phase more accurate ratio of a.( "Er)/a.(**U) for the with one another. These modes therefore cor- cell-averaged broad groups in the erbium reso- . f respond to the motion of a group of atoms having a nance range. This ratio can affect the calculated 7 y, mass much greater than that of hydrogen, and reactivity of the core, but more important is its
/. ij -
indeed even greater than the mass of zirconium, effect on the calculated life of the core. Cross sections calculated with GTF have a smaller ratio f Because of the large effective mass, these moder of a.('"Er)/a,(""U) than those from a standard fs are very inefficient for thermalizing neutrons, but 7 for neutron energies below 0.14 eV, they provide broad-group cell calculation. This smaller ratio y the only mechanism for neutron slowing down. (In gives a shorter calculated lifetime for the core. ( a TRIGA core, the water provides for ample Scattering kernels were used to describe the i . neutron thermalization below 0.14 eV.) In addi- interactions of the neutrons with the chemically 2 [ tion,it is possible for a neutron in thelr_H_to gain bound moderator atoms. The bound hydrogen l [ one or more energy _ units _of ~0.14 eV in one or kernels for hydrogen in water were generated by
. g@ 5everal scatterings, from excited Einstein oscil- the THERMIDOR code," while those for hydrogen lators. Since the number of excited oscillators in zirconium hydride were generated by SUMMIT.
- ,Q present in a ZrHjattice increases with tempera- These scattering models have been used to ade-ture, this process of neutr6n speeding up' is quately predict the water and hydride (tempera-
, .1 J strongly tempe'rature depndent and plays an im- ture-dependent) s p e c t r a as measured at the !J7:a j portant role in the behavior of ZrH-moderated General Atomic linear accelerator." Figure 10
?
; reactors, illustrates the agreement between calculations and l
- j For calculations of the prompt negative tem- experiments for thermal-neutron spactra in ZrH.
- o- i prature coefficient, all neutron cross sections TRIGA temperature coefficients have been de-l termined numerically" by calculating the change
' 2nd Nr energies above thermal (>1.125 eV) were g} 2eersted using the GGC-5 code," where fine- in reactivity associated with a uniform heating of 1 45
% } (LW ILCHNOLOGY VOI 25 JANU/sRY 1976
' ~
U ,
~
J . + - - .
._. .y _ _ .. __......~ _ , .- .- , - ... -
.r- . u. .._o . m rec.,a ;ggs..uu.6
..s , .
H
;\
'l
, i. Simnad et al. TRICA RESEARCH REACTORS io s , , , . . . . , . . , . . . . . . . . . . . .
The Serr values were derived from reactj
" calculations where the rea.ctivity was first coq
, puted with the. prompt fission spectrum alone a i -
QESTURE - then recalculated with the fission spectrum a, io s , counting for both prompt and delayed neutroa
'E- $ The two k,rr values thus obtained, kp and k,, wel
- ? used in the relationship 50'c I 4 h,(1+So) '
10 r : Seff = -1 , m k-p t:
= .
. ' .s.
[ 316'C h where So is the actual delayed neutron fractis
, 5 3 (0.0065).
, @ 10 r .
-3 The prompt neutron lifetime was calculated .
i.
\-
the 1/u absorber method, where a very sm-i amount of boron is homogeneously distribut: 1
., i g 468'c , ,
. throughout the system, and the resulting cihnge p e.
,o z . \. .
% ., reactivity is related to the neutron lifetime h :! *\ . ~
follows: lii . %
.j : 'r
! = 5k,rr ,
dq ( Zr H i,75 BORON POISONED
' W ROGEN N
,$(,f %
. where w is related to the boron atomic density b SUuuiT
.lj; g CALCULATION yboro,, = 7o Uo = 6.024 x 10"(x 10**) x w ,
p3 . 7p
- j w a. i me -
., A O.0 01 0.01 0.1 1.0 10.0 and f8 b NEUTRCN ENERGY (eV) iW N3= boron atomic density Fig.10. Experimental and theoretical neutron spectra
.. j:
from ZrHt.rs showing the offect of temperature w = integer va lation calculations done with the SUMMIT va = 220 000 cm/sec
.' ' ] ca= 755 b (2200 m/sec boron cross section).
n e i the fuel-moderator elements, with the core water N and reflector materials assumed to remain at and 5k,rf is the difference in reactivity between 23*C. The effects of variation from a uniform - P*I# of calculatfor;s-one in which the system cog -
;; temperature for both the cell and the entire core tains no boron and one in which it does. Tg 0 have been investigated with the results that the calculation was found to be insens,itive to changs I t cell effect is reduced by ~5%, but this decrease is in w between 1 and 100.
, offset by an increase in the core le'akage contribu- The enthalpy of TRIGA fuel material as l ;b tion to the prompt negative temperature coeffi- function of temperature has been determined frot i 5 cient of ~10?c when zones of different temperature data and fundamental considerations given in 4 are incorporated in the cell and reactor calcula- Paper by Douglas, which leiids itself fairl U tions. readily to different conditions of composition.
'T Douglas measured the heat content of differes
, J samples of zirconium hydride ranging, in hydre W OTHER PULSING PARAMETERS gen atomic percent, from 0 to 55.5 (Refs. 46,4' and 48). From the results of these measure To perform kinetics calculations and interpret ments, and knowing the compositions at the phas experiments, calculations have been made t,o de- boundaries, one can extrapolate the Douglas dat termine both the effective delayed neutron frac- to values of x (hydrogen atoms per :frconiu:
. tion, Serr, and the prompt neutron lifetim e, 1. atom) >1.25 (the highest value of the sample I Calculated talues are shown in Table I, where it studiedi. This was done to derive an expressit l
is seen that the effective delayed neutren fraction for the heat content (above 25'C) of the 5 phas was found to be insensitive to the reflector 2rH,, which is approximated well by the followin material.
- relationship:
4 NUCLEAR TECHNOLOGY VOL21 JANUARY.
. v -
3- ,
.3 d - Simnad et al. TRIGA RESEARCH REACTORS LO
.1 actor TABLE I h
com- h1 Calculated Values of I and Segg e and 1 ac- Reflector rons, Fuel Element Water Graphite wa*r3 c g Stainless-steel clad Berr = 0.0070_ #ett = 0.0070
, ~
$ b 8 wt% U-ZrHa.t ! = 39 x 10"' sec_ l = 43 x 10 sec e
E (20% enriched U) s Aluminum clad perf = 0.0073 # cts = 0.0073 8 wt% U-ZrHz.o I = 45 x 10 sec ! = 60 x 10~' see
- ed by f.
J j (20% enriched U) small ibutad h* t Stainless-steel clad Beff = 0.0073
>[ 12 wt% U-ZrHz.s I = 32 x 10 sec g e in h i (20% enriched U) '18 23
[ i ACPR
*t t 3 l Stainless-steel clad Berg = 0.0071 h I s
! = 16 x 10sec (BOL) j 1.6 wt% Er.
I 8.5 wt% U-ZrHt.s 20 x 10sec (EOL) ity by : , (70% enriched U) \ TRIGA-FLIP f a (H - H:3)2,n,= 0.03488 T* For the 385-cm* (15-in.-high) fueled portion of a TRIGA fuel elemer.t, one obtains l + [34.446 + 14.8071(x - 1.65)}T 3
- 882.95 - 370.18(x - 1.65)J/ mole . Cp = 825 + 1.61(T - 25'C)W-see/*C elemen't h
The temperature is in *C. (from 25'C) . [ The enthalpy of uranium metal was derived [ fr m the specific heat data given by Etherington"
.on) g and is given by . LIMITING DESIGN B AMIM P A D A*9CTrn AMr) VM t tFe_ _
L -
/een a (H - H is)U = (0.6525 x 10** T' n Fuel-moderator temnevnture in the knete I'mu acon- a
. Tha ,I + 0.1094 T - 2.776)J/g . . of TRIGA reactnr nne ca H n- 'This limit =tems
'---"7-u
- tanges k from the out-c'e=ine nf 5
- r~
q Using the expression and the subsequent streds produced in the fuel o element c!nd "ar4'I- The strene+h 67 +wa M,a as a T pz,g, = 6.49 - 0.55 H/Zr g/cm , y ,, f,,, g g ,,,,,,,,,,,,,,,,,,,,,,,,,,% , _ g 3 imit on the fuel temoernture. A fuel temperature 7 1 la a E a safety limit of 1150*C for mieine stainless-steel fairly
- where H/Zr density < 1.6 and of 5.610 g/cmpZrH, s
was <computed 1.6 = pZrHfor .e'he 1 t y-n 0 'O-2rH , r. or Er-U-ZrHia. fuel is used as a t l fferent ( ZrHi ... Using 18.9 g/cm' for the u.ranium density, the density of the 8.5 wt.o c U-ZrH .. is metal i deci 4,'m value, , , , ,, , to
, g creclude
, , _ , . , , , , ,the
, , m loss
%,_,,,of clad e n n., intecritv_
7 yd#4 ,,, h 5.9768 cm'. Then the volumetric heat content of ,,,,,,,,_.__,__,..,_,%,.,,,,,._,,,,,,,,,,,
]
3.5 wt?c U-2rHi.. alloy is calculated to be b r~ ' ' - -' - ' ^*^
~~~~'"' #" " ' '"~ ~'"""'"'"'~~
eM M
' ***'#"~"*~
l phisc 1 11 - Hi s(8.5U - ZrHi..) = 2.08 x 10-3T ' + 2.0 4 T "" "*r h--"' -- - ' - -
- d a -' H -- -' '---d'-
s data % .
, ~4 ri e e'-- ~ed"e t -! ~1o c oa ena t e-m.-e 5 , ~4
- 52.2 W-sec/cm' e nn
- 0nium d .
~
4 a n' a -- - " - This is a time- and temperature-amales 3 the solumetric specific heat is ession -dependent fuel growth as discussed earlier. .A.,_
""' ~ ~ ~ ' " " ^ " " ' ' ' "
l , phase 3 Cp = 2.04 + 4.17 x 10-' T W-sec/cm' 'C 8tna r~~-"--' da '~~ "-*c'n--a--+"-" " -"-r awing .{ '~' 'r-~-"-n- - - '-
- e e " 2 " ~- --c---r (from O'C) .
.i.
' EW 11CE OLOG.' VOL :8 JANUARY 1976 47 y i93 j
} -
s I
-&%me; + W .. ; m ,_ . .-
w-
.u .:. c.:,5. n .xM- &u n: :.=-w.?.a? .
,. m w T m ' : ??.; .-=.. 3 -m. ,e
, g.., 3 _.?n . c. ,;.cug. u;;;.~-z~. ":a. :.W:
. , . . ::";:g~5'M.G ~ =*V ~
... . . . . - . ~ _ - . . . - -
- - . . _ ~ , - ~ .. . . . -
..a._,w,._.u.-.~....,,. -
3._. .A.. 1~i.44 } ', ,,
. ~ . w ,, ,. .
Siinnad et al. TRICA RESEARCl! REACTORS k _ inslenificant c'Icu12ted' fuel crowth from temoera- 105 ture-dependent irradiation effects. (For ACPR -
, iuet. uner -'arnuo is enremelv low. the steniiv-
! itate coeration91 fn al t a-ra-a+ ,-a ' daciy e-t-t n .i n o r en e 1 The dissociation orpssura of the +i va nniu m . ULTIMATE TEf4SILE
! hydrocen system is the princiral e nnt wik"t o r +n the fuel element intaran t a-ace"-a a+ 'u n i 'n--^--
n t"rac a bnve ~RPO*C. Below ~A0n*C tracoed' air 0.2% YlELO and ficcion-croduct esses can be the mnior enn. ~
, ,.s w,,+ n r e +n n,n 4-+n.-,, n nc e..
- a. At ecullibrium
_conditinn ' hie n*ece"re is a stronc func*in' af "c* g only tamrarSt""o h"* nico the ratio of hvdenwam tn .S i zirconium atoms and the carbon content of the $ 10' - mate-i,' _The current uroer limit for +ba bed-n- E cen-to-zirconium ratio is L65 the decim vntna a le 1* The carbon content is currently ~0.2"o (2000 ppm). The equilibrium hydrogen pressure q+ as a function of temperature for the fuelis shown in Fig. 11. Figure 12 shows the temperature-dependent strength curves currently used for
; stainless steel in TRIGA design work.
- i For the ACPR fuel, optimized for pulsing with a built-in thermal barrier (0.01-in gap) letween the fuel and clad, the clad temperature does not 108 ! ! !
f*
, 10 000 400 500 600 700 800 900 1000 1i 6
TEMPERATURE (*C) Fig.12. Strength of Type 304 stainless steel as a funi
,, g a tion of temperature.
51000 - - e'xceed 280*F (138'C). At 250*F (138*C) the yie
-y,. ;
strength of Type 304L stainless steelis 38 000 p: g and the ultimate strength is 68 000 psi (Ref. 3C s j The stress imposed on the clad S by the intern: I ". E Q. pressure is z l [d d
. g:. .
0 e 100 S=[rPa* (:
,9 @ l where
.f =
s -
;,j'i ,
r, = clad radius l 3 2 < f g . te = clad thickness (in the same units as v.) s
'M l 5 P, = hydrocen pressure._
O 10 -
, f For the dimensions of the clad, the maximu:
, ', T l 1 allowable hvarocen pressura te 1' f I 1 l 38 000
. P,, = n ., g n , _ , = 10 2 5 n =1
'c en prnA"ea vield end 10ico sco sco 1000 1100 1200 cco ITo~o
. ZrH. ,, TEWE RATURE <*C) P= 68 000 s = 1840 psi Fig.11. . Equilib r:u m hyd rogen pressy re cver 7 rHi.e _to rut *ure tee!Sd. From Fig.11 it can be si j versus temperature.
that for U-7rH i.n the fuet tenw.aures tnat 48 / -
, NLTLEAP TEdiNOLOGY VOL. 23 J AN L'A RY
, ~
*1. -
y I Simnad et al. TRIGA RESEARCH REACTORS 4 I
!; nroduce these cressures. under ecullibrium enn- -where R is the cas cons +~+ and T is the 71 r-I mtions , are 1080 --d l unar' conium hydride temperature in Y' -
[ The equilibrium condition defined above never r quation (6) describes the escape of gas from a E occurs. nowever. ha-'"=a *ha '"a' ie " n*
- n cylinder through diffusion until some final concen-q egnstant tem na-'ture ova- +ha whole volum e. tration is achieved. Actually, in the closed sys-h tonsecuentiv. the hvd*nca" n-a=="ra= will ha tem considered here, not only does the hydrogen y u,,,, u,...e +w- ,o.e - ,m s,.4n., ,,,i,,.. c,leulated_ diffuse into the fuel / fuel-clad gap, but it also
,. . w , ,,u-,,, *e-- ~.,+.,-a As hydrogen is diffuses back into the fuelin the regions of lower h
$ ieleased from the hot fuel regions,it is taken up fuel temperature. When the diffusion rates are r' in the cooler regions, and the equilibrium that is equal, an equilibrium condition will exist. To obtained is characteristic of some temperature account for this, Eq. (6) was modified by substi-
' [* lower than the maximum. To evaluate this re- tuting for the concentration ratios the ratio of the
[ .duced pressure, diffusion theory is used to calcu- hydrogen pressure in the gap P. to the equilibrium i late the rate at which hydrogen is evolved and hydrogen pressure P,. Thus, Ea. (6) is rewritten E reabsorbed at the fuel surface. as ( Ordinary diffusion theory'* provides an expres- [3 sion for describing the time-dependent loss of gas from a cylinder: g ) , d( c/c,) dt _ , p dx E c-cf 4 'dD (4) _where the hydrogen pressure Pg(t) is now a func-I 7 exp - ' tion of time and P. Is the eauilibrium hydrocen { C~C/ = i E
""' $s o cre c:enra nver the ?irconium h"d-ida which 4e n where functinn nf +b a '"a t +a-a-e+"va f" The rate of chance of the internal hvdrecen K c, eg, cf = average, initial, and final gas con. pressure in esi. inside the fuel element cladding g centration in the cylinder, respec- ,,i,s,,,,,,,
i tively
'T"oo p J (, = roots of the equation Jo(x) = 0 dPa 14.7f(t)ns
-=
- 2. + 273
, (10) r di 6.02 x 10** V, 273 -
g D = diffusion coefficient for the gas in . anc- r* the cylinder , where d of .
, t = time n. = number of molecules M H, in the fuel 71 eld
% = radius of the cylinder. T = gas temperature. *C j
pI 3 Setting the term on the right side of Eq. (4) equal f(t) = fractional loe= rate from Ec. (9) 0)* to e, one can rewrite Eq. (4) as (2rnal j' V, = free volume inside the fuel clad in liters. l c/cl = cf/c, + (1 - cf/c,)x , (5) l h As the atom density of h;'drogen in ZrH .a is i l (1) .d and the derivative in time is given by ~5.60 x8 10** atoms /cm' and the fuel volume is [ 366 cm , na is ~1.02 x 1085 molecules (H2 ). The L d( c/cj) gap volume is assumed to consist of a 10-mil l dt
" - #1 !#'} dx-dt
' I annulus 15 in. long plus a cylindrical volume, at h the top of the element, t in, high with a diameter Equation (6) represents the fractional release rate
}- of 1.438 in., for a total of 14.36 cm'. Also, the l [g of hydrogen from the cylinder,f(t). The dertva- temperature of the hydrogen in the gap was
,; tive of the series in the right side of Eq. (6) was assumed to be the temperature of the clad. The mum approximated by effect of changing these two assumptions was S gy tested by calculations in which the gap volume was 4 r = -[7.339 cxp(-8.34c) + 29.88 exp(-249e)]de/dt , decreased by 905, and the temperature of the
^
(O ~ ' g hydrogen in the gap was set equal to the maximum fuel temperature. Neither of these changes re-
+
4 there e = Dt/r$. sulted in maximum pressures different from those y The em-un ,.enfficir.,t for hydrncon in ?i = based on the original assumptions although the (T .) r p.m s..w a f' in rhich the H/7r -"a C initial rate of pressure increase was greater. I* W " ". 1.M and 1.R6 is niven bv For those cenilM~e
~
t
~ C "i ,'
i D - 0.25 excf-17 800 /n f 7 - ?7'n1 .., (8) P, = 1.406 x 10'( T + 273) [f(t)dt (11) l
! . TT( m.ni n';Y VOL S J etJAP.Y l976 49 1 . .
.a '
j -. h- -- m
+ > w.. . u ..w. :. : n' ...a , m n . .<.& .~ m_ .m -9. m . s . w m =8.. n .
M .=_m, ? ..>.f% - m. :&-nm-a.r.u.c.w-nsws&L=g:w. ng;M...g,y,ns; [f%.', .
. ip c
b o Simnad et al TRIGA RESEARCH REACTORS ~ o il is 11c m for which the equilibrium hydronan The fuel temperature used in Eq. (8) to eval-
- uate the diffusion coefficient is expressed as pressure in ZrH, .. is 20R0 nsi The calculation indicates, however, that the internal pressure
[ z) = To; t < 0 increases for ~0.3 see at which time the pressure T(z) = To + (T., - To)cos[2.4504(z - 0.5)}; la 0 , is ~420 psi, or ~20% of the e'quilibrium value. After this time, the pressure slowly decreases as (12) the hydrogen continues to be redistributed along [ where the length of the element from the hot regions to
- j. the cooler regions. Calculations were also.made T,,, = peak fuel temperature, *C for , step increases in power to the peak fuel To = clad temperature, *C temperatures of 1250 and 1350*C. Over this range i
.the time to the peak pressure and the fraction of z = axial distance expressed as a fraction of the equilibrium pressure value achieved were the fuellength approximately the same as for the 1150*C case.
t = time after step increase in power. Thus, if the clad remains below ~138'C, the maxi-I mum internal pressure that would produce the It r,c eeen-ad +hnt the fuel temreratura vne yield stress in the clad is 1025 psi [see Eq. (2)}' i nvn ri, n* mth r, ain s. The hydrogen pressure and the corresponding equilibrium hydrogen pres-over the r.irconium-hydride surface when equilih- sure could be 5 times greater, or ~5000 psi.
~
rium prevails is strongly temperature dependent, From Eq. (14) (or Fle.11) this pressure cor-as shown in Fig.11, and for ZrH i .a can be ex- resnnnds to a maximum fuel to-aa-""*a pressed by ~1240*C in Z r H , ,, simi M -1v na aa"'"'2"- hydro 2en crossure could be 5 Y 1 A4n ** 0900,-e!- P, = 2.59 x 10'exp[-1.997 x 10*/(T + 273)}' . (13)
.. . d( before the ultimate clad strewth was reached
?[ The coefficients have been derived from the data develo~ ped by Johnson. The rate at which corresnondine to a fuel temperature of ~1 tone 73, yi,,,,,, ,, ,,, w y ,8, r e ,4 s,,..,er,
~ '[' hydrogen is released (or reabsorbed) takes the -
slowly en ~57 000 est it 500P. Nee the nros-ji form sure is a stron~ ome+!nn of fuel temne'+nva Ho
~
~ fuel tamna-e n e to ernduce ructure decrences t
, g(t,z) =
f(t,z) , (14) very slowly over this ranae. ra-nini-~ " 1900P i P,(z) . for , ri,a ta-na iture un to 900*r nra dar-a,ei g t" ~ ' '"d 17"2 Y f"" "hd '*""*""""" "IO where na enn c -r e -o r + s .. a '" For nongapped TRIGA
': g-
. f(t,z) = derivative given in Eq. (9) with respect pulsing fuel elements, the clad temperature during r to time evaluated at the axial position z heac flow from a pulse is greater than the 138'C ACPR value but normally <500*C.
l
'p 5 P.(t) = hydrogen pressure in the gap at time t P,(z) =~ equilibrium hydrogen pressure at the Measurements of hydrogen pressure in TRIGA fuel elements during steady-state operation have ZrH temperature at position z. not been made. However, measurements have
.,6 .
;- ~ The internal hydrogen pressure is then
,3 P.(t) = 1.406 x 10 8 (To + 273) (( [,'g(t,z)d:dt .
}
% (15) $4 a- '
7, f 4 This equation was approximated by 3 o.1- N
, .i 4
8 N
- P,(t,) = 1.406 x 10 ( To + 273) e 2 ~
Y = *
~
P,,(t s_i )~ S x E..aE,.i 1- f(t, , 2,)a z at , (16) "' o o 1.- P,( tj ) . pi {4 2 I where the internal summation is over the fuel j element length increments and the external sum- Ec coi aco oi mation is over time. I ,,g _ TE R %CRE ASES IN TE*.*FE RATU E (sec)
' N *% --+4m of the fuel element internni l ,
Ja ! o rec eu ro in 'k a e a"'P -i"m hvd rva- araeevaa '*- Fig.13. Fual element internal pressure versus ur-g e8 --'c' _ '"rrtion of time after a cten increncem after 2 step increase in maxarnum 'uel temp: in !? .ce'ratu re. The maximum :uei temoarnt"ra_ atu re.
-t/
WCLE A R TECHNOLOGY VOL..3, JANUARY ' 10 s
4
, V'. Simnad et al. TRICA RESEARCH REACTORS U
been made during trarisient operations and com- 18) lower than would be necessary for elad en k tallure. A lactor ol a is more than adecunta to on
.re
{E pared with the results of an analysis similar that described here. These meaeurmede to i ~'i - ' account for uncertainties in clad strencth and manulacturing tolerances. The integrity of the tre cated that in a pulse in which the maximum ' claa nas oeen demonstrated by TRIGA reactor h -temperature in the fuel was >1000'C. the mn*um 2e. *h as '5ressure was only ~@ of the ecuilibrium value_ pulse experiments to fuel temoeratures =:11wC? Under any condition in which the clad temcera-mg 7 valuated at the peak temperature. Calculations ture increases above cuu C, such as durine a to , f li the r"-a**9 re resultinc from such a cu!=e ioss-ot -c ooiant accicent or uncer ' film boiline ide P. "Eine the methods describad nbove c tve calculnted _ conditions. the temoarature ufarv limit m et h a_
~ ~8 ""** "a,tav *b in os a - aa -
nel k Tra*9Ure v'l"*a - decreased as the clad material loses much of ite_ age h '_.e d "M"ae u_ strencth at elevated temoeratures. To establish _ i of C An instantaneous increase in fuel temperature N will produce the most severe pressure cori- this limit. it is nesumed thnt the fuel nnd the ebd era are at the same temoerature. An analysis for this
.s e. 'ditions. When a peak fuel temperature of 1150*C is reached by increasing the power over a finite condition indicates that at a fuel and chd temeer-fxi- ature of ~950*C, the eouilibrium hydrogen pres-the period of time, the resulting pressure will be no sure produces a stress on 'the clad ecual to its_
2)], $ creater than that for the step change in power ' ultimate strength. There are no conceivable cie-es- i analyzed above. As the temperature rise times cumstances that could give rise to a situation in psi. E become long compared with the diffusion time of wnica tne ciaa temperature was hicher than th or- hydrogen, the pressure will become increasingly of less than for the case of a step change in power. . fuel. The reason for this is that the pressure in the The same argument about the redistribution of lum psi, y clad element results from the hot fuel dehydriding the hydrogen within the fuel presented earlier is
- htd faster than the cooler fuel rehydrides (takes up valid for this case also. In additine at elevated the excess hydrogen to reach an equilibrium with temoeratures the clad becomes permeable to
- 2. O tsas ) tne hydrogen over-pressure in the can). The Evdrocen. Thus, not only will hydrogen redis- neah ras- ? slower the rise to peak temperature, the lower the tribute itself within tne luet to recuce tna shre. but some hydrogen will also esmne from the
,tha j pressure because of the additional time available
' for rehydriding. svstam entirelv. _
ises The use of the ultimate strength of the clad 00*C ; An assessment of the effect of some of the material in the establishinent of the safety limit sing assumptions used in this analysis is given below: under these conditions is justified because of the
'400 d 1. It was assumed that the peak fuel tempera-ture was constant with radius for evaluating the transient nature of such accidents.
' tlGA L ' tring te hydrogen diffusion coefficient. This overestimates
,he ' average fuel temperature by ~15%. As the CHARACTERISTIC PERFORMANCE VALUES 38*C f FOR A TmGA ACPR t diffusion rate is very temparature sensitive,this UGA $5 .issumption provides a degree of conservatism The core characteristics projected for the have considerably in excess of 15%. performance of the TRIGA ACPR containing fuel l have N 2. The diffusion model used does not rigor-l y optimiaed for pulsing operation are given in p ously . account for the changing boundary condition Table II.
3, imposed by the hydrogen confined to the fuel-clad The standard experiment used for the analysis
,l pl cap. The modification to the model to account for of the system, with other than air in the irradia-
! [ diffusion back into the fuel is an approximation tion region, consisted of a mixture of 37.5 vol%
that is reasonable as long as the total fraction of CH2, 12.5 vol% stainless steel, and 50 vo1% void' { the hydrogen lost from the fuel is small. At the which was homogentaed to fill the volume of the N g 'ime of the maximum hydrogen pressure, this test cavity. This standard experiment was con-
;; fraction was calculated to be ~6 x 10**. From venient in that it had a calculated reactivity worth y his it is concluded that the model should be valid (~$5.50) in the range of interest as the design j , g in calculations of maximum pressure. upper limit for routine experiments. The final Tb forecoing analysis cives n = tron" i d'm recommended reactivity worth upper limit for I j '"aC routine experiments is set equal to the reactivity j _ Mn that the c iaa will not be ruch'-ad " mee
-,.n,, n r t h e n u *% insertion necessary to produce a 1000*C maximum
. y m ne ra ,,,-n, ,-n -n..-
'O 3 n.in e n i nnne y,.n,.w ., % r.i,4 temnoentura fuel temperature.
H b "A"C Mowever, a conservative safety limit _ Note fro;n Table II that the reactivity insertion
~ !, ll50P has been enosen ter this condition. g necessary to produce a 1000*C maximum temper-ature is about the same ;chether an experime is j a e at this e,fotv hmn tomnomera tha m-e ,.. - . ._.-% s c u ., , . . . . -- f
- n the e:.perimental cavit; or not. These values
! i 51 l yn W I L( h" GLOGY VOL. 25 J ANI.' A RY 19% 1 a l 1
;l .
v--# m mm% -
% __.s ,
^
~ -
; Simnad et al. TRIGA RESEARCII REACTORS o.
TABLE II [ Pulse Performance and Parameters for ACPR it i~ f} Parameter , Value -
.I .. ' ! Fuel material ' 12 wt% U-Zrlit.e'(U is 20% enriched) -
.. Critical mass 117 elements includes 6 fueled followers.
+
,ij 6.20 kg 8 '*U
!! Operationalloading (~$10 excess 154 fuel elements. 6 fueled followers.
11 reactivity) ' up to 5 aluminun-followed pulse rods .,
!f- Worth of pulse rods $4.80. (min)
*l Worth of bank rods ' $8.15 (min) '
ti~ Effective delayed neutron fraction i;
;j (/f fr) 0.0073 U Prompt neutron lifetime G)
{l 32 usec
! Prompt temp coefficient (a)
(average between 23 and 700*C) Air in irradiation hole. 6 bank f' rods half in, pulse rods out -9.G x 10"* ok/AT 3
=
Standard experiment in irradia-tion hole, all rods out
,.7
-9.3 x 10 6k/AT
~
i Irradiation Region Contents I
.1 '
J Air Standard Experiment
. il' '
.' f. . .
P/P ' I Axial - 1.25 1.25 . Radial (power in T,nar cell /P core) 1.07 0.98 ! Cell 1.76 1.76 -i
,.l Total 2.35 2.16 Performance Peak adiabatic fuel temp (*C) 1 000 1 000
.: Average adiabatic core temp (*C) 555 590 i* ok ($) 4.80 Core energy release (1 sec) (MW-sec) 4.90 ,j 106 115 :i N
j n/cm* >10 kev in hole (175 MeV/
' fission) 1.20 x 10** ---
.i Peak power (MW) 20 000 j 22 000 ,
Min period (msec) 1.2 1.2 11 k' - l .
. S '
]; are calculated to be about the same because, even reflecting the acceleration of the rod. The scram
', though the temperature coefficient is somewhat occurs 1 see after the pulse begins.
, reduced with the standard experiment in the Figure 14 shows the reactor power and energy 4 I trradiation region (actually.a result of the bank release as a function of time after the initiation of rods being withdrawn to compensate for the stan- the pulse. The maximum power is 20 000 MW.
dard experiment worth), there is a compensating the prompt burst energy rel~ ease (~0.1 sec) is
. .f i'
change in the peak-to-average power generation.in ~100 MW-sec, and, within i see,106.MW-see l' the core. energy has been released. Figures 14 through 17 present the values of an Figure 15 illustrates the maximum fuel tem-analysis for a $4.20 reactivity insertion followed perature in the fuel element in which the power by a scram. The total reactivity is added within . density ~ is greatest, the maximum clad tempera-
, j 0.085 sec,. with thd rate at each point in time ture the coolant velocity, and the maximum hes:
52 *
, ,, Ntl CLEAR TEC}{NOLOGY VOL. 25 JANt'ARY W.
,_ ___ - ._a
)* .
[ . 'Simnad et al. TRIGA RESEARCH REACTORS 7 i ,10' . flux. The peak fuel temperature of ~1000*C oc-j io .
. . curs 0.1 see after the beginning of the pulse and I quickly falls off to 880*C within 1.0 sec. The clad
! temperature does 'not begin to increase signifi-I '
cantly until 0.5 see after pulse initiation, at which
'O' time it begins to increase to its maximum value of f 10',
180*C at ~10 sec. This also corresponds to the
- time of the maximum heat flux,12.5 W/cm', and
. .i .
maximum coolant velocity,220 kg/h'. Figure 16 shows the temperature at.the ther- { e j iG T ,E mocouple location as a function of time after the 5 e :* E l' E c . hI e ; a e o
!! !,,,l,_ ,__ ._ ENE R,GY,pW{sgq,,,,,,u ,o, ' 1MO, f
. l i
~
rl l' ,6 800L - b F , r us j zj i- s 600 4 I ,8' s POWER (MW) h ' - to' 10' t- l g400
-{- *
~
; e s l l # 2M-
.j 10*8 --6 ' 108 10 10" 10* th 103 0.07 0 .0' 8 0.09- 0.'10 0.11 0.'12 0.13 0.14 TIME FROM BEGINNING OF ROo MOTION (sec)
. TIME FROM BEGINNING OF ROD MOTION (sec)
Fig.16. Standard pulse thermocouple temperature for Fig.14. Standard pulse power and energy released , maximum power cell versus time from ~first
. versus time from first rod motion. rod motion. - . p s
li
.N 0
1000 , , , , . , , , . . ., 25 { . ., ,
,g g e *
. c MAXtMUM FUEL TEMPERATURE (.C) 20 f (~ 800 -
o E 600 - - 15 8 MAXIMUM HEAT FLUX (W/cm ) [
~
5 't! l $ 400 - h - 10 h ft u o ram > o a 1-E E COOLANT VELOCITY (kG/h) N 3
'rgy ' i , ~ -
5
$ -l' P. o( [ 'W' T: E [ MAXIMUM CLAD TEMPERATURE (*C) ft j is ~ l 4 '
J t 3Cc ,y o .. i ,e , ! , , ! , I0 10 ' 10" 10' 10' 108 . [j TIME FROM EEGif4NING OF RCD MOTrON (sec) ver ;1 ca- 'I 4 11 Standard pulse hot-element at gore edge-maximum temperature, hest flux. and coolant velocity versus c r.: j{ 'the f rom first rod motion.
- ~
-,g ' f I \E il CllNot OGY VOL. 28 JANUARY 1976 53 m
j /. - - 81B ows..;~ a'.ap s. . -* ,--.. . ,.. c; . _,- wm
-4
*.eg f.
-_ -c 1 _
? - - - - - -
=t g , ;
.A_ u.. - -
- - m_%.
wu
- F Q e .
i ;i Simnad et al. .TRIGA RESEARCH REACTORS l' 9 1000 & . . . . , y value. They can all be at varying, relatively small differences from the real values, but the
? - l[ -
combination of errors for the quantities as they li are related to influence the pulsing performance i _ is within a reasonable accuracy for the measure. ; 900 - i- ments of the pulsing performance. i
. ',- For a comparison of flux in the experimental I
.i cavity, Sandia quotes 1.55 x 10 25 n/cm2 (>10 kev) -
!! / for a $4.40 pulse. Using a 100-MW-sec' energy i e' s release, to include energy ~ beyond 10 sec that !
6 800 - TIME FROM FIRsT / s - could be measured by the detectors, the flux-watt .
] RCO MOTION ,',- #
(>10 kev) is 1.55 x 10'. This compares to the '
- I $ ,' / computed value of ~1.1 x 10'/W. The calculated ,
E *- ' flux value assumes an energy release during fis- {
,- y 3[o ,e - -
,( j sion of 190 MeV/ fission. If the delayed energy .
- j. j 3 7ao .
/ ,e . , , ' _ from fission is not included, as would be the case )
,, , . during pulsing, the calculated neutron fluence [
. . , ,' 1.1 sec value could increase by as much as 10?o. '
'l .; ,
, 1.9 see ti ~ / 0.69 sec ji 0.29 sec I'
O.097 see .I 600 - e,' ., i 3*!} REFERENCES kf '1 g e
- 1. M. T. SIMNAD, G. HOPICNS. and J. SHOPTAUGH.
E f*.[ , , , , , , "Fuci Elements for the TRIGA Mark III Pulsing Reac . . _ E o.1 0.2 0.3 0.4 0.5 0.6 0.7 tor," Trans. Am. Nucl. Soc., 7, 110 (1964). .
? ' *
' RADIUS (cm) .
- 2. M. T. SIMNAD and J. B. DEE, " Equilibrium Disso- t Fig.17. Standard pulse radial temperature distribution clation Pressures and Performance of Pulsed U-Zril !
! in a fuel rod at core edge at various times from Fuels at Elevated Temperatures," Proc. Symp. Ther- ~
first rod motion. modynamics of Nuclear .ifaterials, Vienna. Austria.
.. International Atomic Energy Agency (1967).
I h beginning of the ' pulse. It rises rapidly as the 3. M. T. SIMNAD and R. CHESWORTH, "TRIGA Re- .
! prompt energy is deposited in the fuel and in- search Reactor Experimental Instrumentation," Proc.
i creases to its peak value (845'C) at ~15 sec. Symp. Research Reactor Instrumentation, Tehran, Iran. e l The redistribution of the energy within the fuel International Atomic Energy Agency (1972). ! is shown in. Fig.17. Here the temperature is 4. W. f NHITTEMORE et al., " Stability of U-ZrHi. plotted as a function of radial position (at the axial TRIGA Fuel Subjected to Large Reactivity Insertions." centerline) for several times after the pulse GA-6874. General Atomic (1965). - i l initiation. At 0.097 sec, the temperature distribu-
;j tion reflects, essentially, the adiabatic energy S. W. L. WHITTEMORE et al., " Characteristics of l 'a deposition in the pulse. By 0.29 sec, a significant Large Reactivity insertions in a High Performance ,
amount of hea't has flowed toward the clad and the TRIGA Core." GA-6216. General Atomic .(1965). 1 fuel center (particularly the unheated. r.irconium
! rod). Within 40 sec, the initial distribution has 6. F. A. HASENKAMP " Measured Performance of the 4 " ~^ ' ' "
been erased and conduction to the coolant domi- A""fs g, 96S(
, l nates.
4 Comparisons of calculated performance, using 7. M. T. SIMNAD. " Study of FLIP Fuels for TRIGA d parameters equivalent to those in this analysis, Reactors." Gulf-GA-A9910. General Atomic (1970). l with experimental values from the Sandia Labora-
~! tories ACPR have generally shown agreement 9. G. B. WEST and J. SHOPTAUGH, '.' Experimental within ~5?c, with ~ the calculated values always Results from Tests of 19 TRIGA-FLIP Fuel Elements m
. the Torrey Pines Mark F Reactor." GA-9330, Genera:
being larger, and with the calcubited peak power Atomic (19G9).
~20% larger than the experimental value. The
., manner in which the kinetics parameters are 9. K. E. MOORE and W. A. YOUNG, " Phase Relatien-
.; interrelated would indicate that no single calcu- ships at High Hydrogen Centents in the SNAP Fuel i
lated v:!ue is greatly different from the real System." NAA-SR-125 d. Atomics International (196H. 54 NUCLEA R TECHNO' LOGY VOL25 JANtlARYI94
; us <v i
[, - - 3 Simnad et al. TRIGA RESEARCH REACTORS f# , . [! t g . .... J. W. RAYMOND and D. T. SHOOP, "The Metallog-
- 26. J. C. LeBLANC, " Prototype 58DR Fuel Element IV
[.apy of Zirconium-Base Alloy Hydrides," NAA-SR- Performance Test (NAA-121-1 Experiment) " AI-AEC-ucmo-10927. North American Aviation (1965). 13002. Atomics International (1971). h. , l :8
- 27. K. R. BIRNEY, " An Empirical Study of SN.*.P Reac- L 5
- 1. K. M. M A C K A Y, Hydrogen Compounds of the tor Fuel Irradiation Behavior," NAA-SR-12284, Atomics I wtallic Elements. F. N. Spoon. Ltd., Iendon (1966).
International (1967). , 31 J. W. RAYMOND, " Equilibrium Dissociation Pres- m ' wres of the Delta and Epsilon Phases in the Zirconium- 28. N. F. DAVIES and R. E. FORRESTER, " Effects of
- Hydregen System " NAA-SR-9374, North American Irradiation on Hydrided Zirconium-Uranium Alloy,' NAA- h.{.
uistion (1964). 120-4 Experiment," AI-AEC-12963. Atomics Interna- 1 I ! I tional (1970). y gfi
- 3. H. E. JOHNSON, " Hydrogen Dissociation Pressures
..f Stodified SNAP Fuel," NAA-SR-9295, Atomics Inter- 29. R. S. BARNES and R. S. NELSON, " Theories of
[ Swelling and Gas Retention in Reactor Materials," g national (1964). . j); a AERE-R-4952, U.K. Atomic Energy Authority (1965). , L ;;. U. MERTEN and J. BOKROS, " Thermal Migration . , p, l ,
! ..( Hydrogen in Zirconium-Uranium-Hydrogen Alloys," 30. F. A. NICHOLS, " Behavior of Gaseous ' Fission i
; .,, j h I. Nucl. Mater., 10, 3, 201 (1963). Products in Oxide Fuel Elements," 'WAPD-TM-570, p f ,
Westinghouse Atomic Power Division (1966). p ; J- 15. P. PAETZ and K. LUCKE "On the Kinetics of H3 dregen Engassing of Delta-Zirconium Hydride," 2. 31. H. LAWTON et al., "The Irradiation Behavior of
.wtallkde., 62. 9, 657 (1971). Plutonium-Bearing Ceramic Fuel," Symp. Fast Breeder ,
Reactors. British Nuc1 ear Energy Socicty, London l
- 16. L. BERNATH, Ed., " SNAP-4 Summary Report," (1966). H NAA-SR-8590, Atomics International (1963). g~ .
- 32. W. E. KESSLER G. F. BROCKETT, and G. E. - -
I 17. W. A. YOUNG et al., "Thermophysical Properties BINGHAM, " Zirconium-Hydride Fuel Behavior in the ! !'
?'
) ut t;nitradiated SNAP Fuels," NAA-SR-12607 Atomics SNAPTRAN Transient Tests," Trans. Am. Nuct Soc.,
' 'l f I International (1968). 9, 155 (1966). ;
I f l f3 1*. J. C. BOKROS," Creep Properties of a Zirconium-Hydrogen Uranium Alloy," J. Nucl. Mater., 3, 216
- 33. K. J. MILLER, " Post-Irradiation Examination of Treat Capsulo No. 5 Through No. 9," NAA-SR-Memo- ,
' ; i l
j 8 t XII. 11374 North American Aviation (1965,). 3 g,
. S.
l' 7 13. J. T. BERLING and G. D. JOHNSON, " Elevated 34. BERNARD M. LEADON, WILLIAM H. GALLAHER, g". . j Nmperature Tensile Creep Properties of Zr-10 wt% U- GEORGE T. R AYFIELD, and RONALD W. OBERMEYER. , Mloy Hydrides," NAA-SR-Memo-11649, Atomics Inter- " Measurements and Calculations of Hydrogen Loss from - ;* g g national (1965). Ilydrided Zirconium-Uranium Fuel Elements During j Transient Heating to Temperatures Near the Melting .!
*A lo. J.' D. GYLFE et al., " Evaluation of Zirconium Point," Trans. Am. Nucl. Soc., 8, 547 (1965). j' li :!ydride as Moderator in Integral Boiling Water-Super-
- t. cat Reactors," NAA-SR-5943, North American Avia- 35. R. E. TAYLOR, " Pulse Heating of Modified Zr-H j-i w>n (1962). Hydride," USAEC Report NAA-SR-7736. North Ameri- , ,
g can Aviation (1962); see also NAA-SRs7398 (1962) and y f 21. A. F. LILLIE et al., " Zirconium Hydride Fuel " Compounds of Interest in Nuclear Technology," AIME j ;l. ~ E!ctnent Perfor'mance Characteristics," Al-AEC-13084. Monograph (1964). ! Atomics International (1973). d
$ 36. F. C. FOUSHEE and R. H. PETERS, " Summary of . ;
i j 22. E. E. ANDERSON. S. LANGER, N. L. BALDWIN, TRIGA Fuel Fission Product Release Experiments," ;. . g anil F. E. VANSLAGER, " An In-Core Furnace for the Gulf EES-A10801, Vol. II, General Atomic (1971); see ,i Hwi-Temperature Irradiation Testing of Reactor also S. LANGER and N. L. BALDWIN," Fission Product 'j e Fuels." Nucl. Technol., 11, 250 (1071). Release Experiments on Uranium-Zirconium Hydride ,
'd Fuels " Gulf-GA-A10781, Vol. !, General Atomic (1971). "
y a F. M. MORRIS, " Applications and Experience with jl
.- :. hw lastrumented Fuel Element," Sandia Laboratories 37. J. BELL, " SUMMIT, An IBM-7090 Program for the g , 'D70) Computation of Crystalline Scattering Kernels," GA- ,
+ 2492, General Atomic (1962).
- $g y E E. KRUPP," Post-Irradiation Annealing of SNAP r al Irradiated at Lew Temperatures-NAA-116 Exper- 38. A. D. B. WOODS et al., " Energy Distribution of Ntt." NA A-SR-12039, Atomics International (1967). Neutrons Scattered from Graphite, Light and lleavy
$' Water Ice, Zirconium Hydride, Lithium Hydride, Sodi-R P. E. FORRESTER and W. J. ROBERTS, "In-Pile um Hydride, .nd Ammonium Chloride, by the Beryllium '
. nauer of SN AP-1 Experimental Reactor Type Sub- . Detector Method "in Proc. Symp. Irelastic Scattering of W Tucl Elements (N AA-115-2 Experiment)," NAA- Nc:.!rens in Solids and Liquids, Vicnna. Austria. Oct.
j d2 25 Atomics internattor.al (1068). 11-14. nternational Atcmic Energy Agency (1900). , ,
,u,
. ' Li A P. 7FCHNOLOGY VOL. 23 J ANUARY 1976 55
[ e ~.- b.; q!
,y
v.
. ,;., n,_ _ -. ..~ _. :-. m.- - m . - , o.m. . m - , . . w - . ~ .e m _.g h- - *[' ym, yu 1 l
l
. !:i .
.: Sinmad et al. TRIGA RESEARCH REACTORS U
- 39. D.' R. MATREWS et al., "GGC-5 A Computer Pro- 45. G. B. WEST et al., " Kinetic Behavior of TRIG 4 gram for Calculating Neutron Spectra'and Group Con- Reactors," GA-7882. General A,tomic (1067).
*i stants," GA-8871. General Atomic (1971).
I!'
- 46. T. D. DOUGLAS, "The Zirconium-Hydrogen Sysc
- 40. F. T. DLER, G. W. HINMAN, and L. W. NORDHEIM, tem: Some Thermodynamic Properties from a Hea:
"De Quantitative Evaluation of Resonance Integrals... Content Study " J. Am. Chem. Soc., 80, 5043 (1958).
in Proc. 2nd. Inten:. Conf. Peaceful Uses At. Energy
~
(A/ Conf.15/P/1958), Geneva, International Atomic En- 47. T. B. DOUGLAS and A. C. VICTOR, " Heat Contere ergy Agency (1958). of Zirconium and of Five Compositions of Zirconiuq
;. Hydride from O' to 900 C " J. Res. NBS 611, Research
!l
' Paper 2S78, pp.13-23 (1958)..
- 41. R. ARCHIBALD and K. D. LATHROP, "GTF, A Ii Space-Dependent Thermal Spectrum Code " GA-8775, 48 T. B. DOUGLAS, "High Temperature Thermodye General Atomic (1968). namic Functions for Zirconium and Unsaturated Ziry
-l l conium Hydrides," J. Res. NBS 5,67A,403 (1963).
42..H. D. BROWN, Jr., General /.tomic Company, "nermidor-A FORTRAN II Code for Calculating the 49. H. E T H E RIN GT O N. Ed., Nuclear Engineerit 7, ; Nelkin Scattering Kernel for Bound Hydrogen (A Modif1- Handbook, McGraw-ilill B o o k Company, New York cation of Robespierre)." unpublished data. (1958). 4 aj
- 43. J. R. BEYSTER et al., " Neutron Thermalization in Reactor Handbook, Vol.1, p. 569.
Zirconium Hydride," GA-4581. General Atomic (1963).
- 51. W. JOST, Diffission Academic Press, New York
-] (1952).
- 44. J. R. BEYSTER, J. L. WOOD, W. M. LOPEZ, and f'
-fu R. B. WALTON, " Measurements of Neutron Spectra in Water Polyethylene, and Zirconium Hydride," Nucl.
- 52. W. M. ALBRECIIT and W. D. GOODE, Jr., "The Diffusion of !!ydrogein in Zirconium Hydride," BMI-
! 3; Sci. Ent., 9, 168 (1961). .
1476 Battelle Memorial Institute.(1960). r ti.
~
bi[; -
. .2:
o ,,
"lf e
'd .
.'} h
-,- .1
'h
; -s q.
t i'.
~;;
b Ng .
.C
. ,n l -
! I
'k, in
!N6 4
. 1 1- .
i b, I 56 NUCLEAR TECHNOLOGY VOL. .S JANUARY We
. . ~ r
9
)
n ATTACHMENT 3 t e t I l \
AFFIDAVIT OF
' JOSEPH A. SHOLTIS,3R.
Joseph A. Sholtis,3r., being duly sworn according to law, deposes and says:
~
The Intervenor has on several' occasions stated that' repetitive pulsing during a LOCA is necessary to achieve Intervenor's postulated sudden temperature elevation such that -
. multiple clad failures would result. (See page 128 lines 2 through 9, inclusive and page 115 lines 2 through 11, inclusive, of the transcript of Dr. Stillman's deposition in New York on 18 Dec 1982; see also Intervenor's responses to Licensees first-round interrogatories 33b,24d, and 24e.; also see Intervenor's response to Licensee's first-round interrogatory #24f where Intervenor states, "If the 1 MW TRIGA reactor was not capable of the pulsing operation, it is unlikely that cladding failures would result from a LOCA involving that reactor.") It should be noted that.this last statement by Intervenor has important implications to Contention 2, Accidents II.4 as well as this subject contention.
The important question is whether repetitive pulses can indeed be fired frequently enough during a LOCA and result in adequate temperature elevation for clad failures to occur. Licensee submits that pulsing during a LOCA, particularly repetitive pulsing with a frequency between pulses of a fraction of a second (the frequency which Intervenor has stipulated during Dr. Stillman's deposition in New York on 18 Dec 1982 -- see page 74 lines 11 through 14, inclusive, of the transcript of Dr. Stillman's deposition in New York on 18 Dec 1982), simply is not possible in the AFRRI reactor. First, the only way in which the core could possibly become uncovered with water is if a rupture of the tank occurs at an elevation below the top of the core. (Note: Because of the higher elevation of all plumbing associated with the primary coolant system, the core physically cannot become uncovered
-. ____-___A
via drainage or pumping through breaches 'in ; primary coolant lines.) Upon loss of-approximately 4-6 inches of water from the normal pool water level, a scram signal will be , automatically initiated by actuation of the pool water level float switch. This action would 4 . terminate any power operation that happened to be in progress already and would also preclude any subsequent reactor power operations from taking place. Therefore, operation of the reactor during a pool water loss situation would require a total malfunction of the pool water level float switch such that a scram signal is not generated. Even if this malfunction were presumed to occur during a pool water loss situation and reactor power operation was also presumed to occur during the water loss, radiation alarms would alert the operator to an off-normal situation well before the core actually becomes uncovered since less and less water shielding would be available as the water loss progresses and direct gamma shine from the core would become evident. At this' point, to recapitulate the , - scenario, water is being lost from the pool at a maximum rate of 250 gals / min (the rate ! - which Intervenor stipulates in its response to Licensee's first-round interrogatory #33a and which is reiterated on page 45 lines 3 through 10, inclusive, of the transcript of Dr. Stillman's deposition in New York on 18 Dec 1982) which equates to a water level drop rate of approximately 4 inches / minute, the pool water level float switch has been presumed to fall such that no scram signal is generated, reactor power operations are presumed to take
- place during the water loss, and numerous audio-visual radiation alarms sound due to direct gamma shine before the core actually becomes uncovered. It is hard to believe that an i
operator would continue reactor power operations in the face of numerous radiation alarms sounding, or conversely, that the radiation detection system would fail totally and not provide alarms during a pool water loss situation. Nevertheless, we will still assume that reactor power operations continue to be performed as the water loss progresses toward
, ultimate core uncover' . At this point, a discussion of how pulses are fired at AFRR1 is necessary before continuing.
4
,- - . ~ - - - - . - _ . . _ . - _ . . - - - - . . . . . - . . - - ,- , . - -
First, pulses are fired from a low power steady state condition, usually at approximately 15 watts but certainly never .above 1 kilowatt. In fact, the AFRRI reactor has a built-in interlock system which prevents firing the transient control rod out of core if the power level is greater than or equal to I kilowatt. Therefore, in addition to the already mentioned safeguards which must be presumed to fail in order to operate, this 1 kilowatt interlock must also be assumed to be non-operational in order to permit .the firing of successive repetitive pulses. This is so because the power level of the reactor immediately after a pulse is fired and continuing for about ten minutes thereafter will always be greater than 1 kilowatt because of delayed neutrons produced from the pulse which ultimately die away on a negative 80 second period. Next, let's assume that the operator has attained a steady state power level of 15 watts by virtue of having manually withdrawn the three standard control rods. Typically to attain a 15 watt steady state power level at AFRRI the " shim" and " safety" control rods must be fully withdrawn while the " regulating" control rod is withdrawn approximately 80% At this point, the transient control rod anvil is - raised (without any air supply to the transient rod drive) to the desired withdrawal point, the range select switch is turned to the "3 MW-pulse" setting and the mode select switch is turned to the " pulse-hi" or " pulse-lo" setting (at which point, the " pulse-fire" button will light up if the j 1 KW interlock is satisfied; then, by depressing the " pulse-fire" button a pulse can be fired). Upon depressing the lit " pulse-fire" button, a pulse timer, which is normally set at 0.5 seconds, begins counting as the transient control rod is driven upward (to meet the anvil stop) which initiates the pulse. It takes approximately 100 msec for the transient control rod to reach its upper limit of travel when the lit " pulse-fire" button is depressed and it l l s takes an additional 100 msec (maximum) for the pulse to occur and shut itself off via the action of the negative temperature coefficient of reactivity. When the pulse timer finally reaches 0.5 seconds or 500 msec af ter pushing the lit " pulse-fire" button (i.e. approximately 300 msec, minimum, after the pulse is already over), a signal is generated automatically i l
which scrams all the control rods which consequently fall' back into the core. Once the O sta..dard control rods are back in the core (i.e. after approximately 500 msec after the pulse timer initiates the scram signal), the standard control rod drives.begin driving "down" automatically to meet the already inserted control rods and this automatic lowering action, in and of itself, takes approximately.30 seconds. Once the standard control rod drives are fully "down" and again in contact with the standard control rods, then and only then can the standard control rods be manually withdrawn again in preparation for a second pulse. It takes.approximately 3 minutes to manually drive these control rods back out of core to c reestablish a steady state power condition. What all this means is that if the pulse timer is operational, it is impossible to fire successive pulses at a frequency faster than about one every 4 to .5 minutes and this relatively quick pulse repetition rate can only be achieved if the i kilowatt interlock is non-operational and operator error is also assumed. In order to fire repetitive pulses faster than one every 4 to 5 minutes, the 1 kilowatt interlock must fail, the pulse timer must also fail to scram the control rods and gross operator error must additionally be involved. And even for this incredible series of events,-the pulse n petition rate could be no faster than one every 600 msec. That is, it takes physically about 500 msec for the transient controi rod to drop back into core after the first pulse is initiated and an additional 100 msee to drive it back out for the second pulse. This raises an interesting question. If the pulse timer must fall to inititate a scram signal in order to be able to fire successive pulses every 600 msec, then how does the transient control rod get back into the core for firing it out the second time. This demands not only a malfunction of the pulse timer but a selective and particular malfunction of the pulse timer which somehow leaves , the standard control rods remaining withdrawn but nevertheless scrams the transient control rod so that it can be redriven out of core for the second and subsequent pulses. All of this actually becomes rhetorical anyway since the second and subsequent pulses will not occur even if the transient control rod could be driven selectively in and out of core at will at a
very fast rate. The reason subsequent pulses will, in fact, not occur is due to the extremely large amount of negative reactivity that is introduced as a' normal matter of course as a result of the first pulse and its associated fuel temperature increase; this temperature increase occurs and persists for several to tens of seconds. Therefore, even though it might be possible to selectively drive the transient control rod in and out of the core at will at a very high repetitive rate, the core will still be well suberitical (many dollars subcritical) as a result of the first pulse and the temperature heat-up which occurs and persists. Therefore, the transient control rod worth will be insufficient to overcome the core's large subcriticality to even attain criticality (let alone fire a second pulse) even if the transient control rod could indeed be selectively driven out again immediately af ter the first pulse. Let's assume AFRRI fires a $3.28 pulse. Such a pulse will result in a peak power of about 2500 MW(t) and a fuel temperature rise of approximately 5500C. The pulse will have a width at its half-maximum power level of approximately 10 msec and the temperature increase will decrease with time but persist at significant levels for approximately 10 j seconds or longer. Such a pulse would introduce $9.90 of negative reactivity, since the prompt negative temperature coefficient of reactivity has a value of -1.8c for every 10C of 1 temperature rise. Therefore, since the transient control rod must be presumed to selectively scram immediately after the pulse (in order to permit subsequent rapid withdrawal of the transient control rod), the core would be $9.90 below the delayed critical state at the time the 5500C fuel temperature rise actually was attained and would decrease slowly to a zero value over the following approximately 10 second or longer time interval. t Since AFRRI's transient control rod total integral reactivity worth is only $3.35, even if it could be driven out of core totally and immediately after the first pulse, the core would still be $6.55 below the delayed critical state. This means that even though safeguards and the operator might fall and permit selective and frequent repetitive transient rod firing in and out of core, successive pulses could not actually be fired more rapidly than about one every
ten seconds (but certainly not on the timeframe of fractions of a second as intevenor has stated) and at these limited rates, the fuel temperatures would have (and, in fact, must have) recovered to near ambient conditions. J
. Up to this point, nothing has been said about the feasibility of firing a single or multiple series of pulses with the core partially or completely uncovered as a result of the presumed pool water loss. Intervenor has admitted that criticality cannot even be attained if the core is completely devoid of water. (See page 115 line 15 through page 116 line 3, inclusive of the transcript of Dr. Stillman's deposition teken in New York' on 18 December 1982.)
Further, when asked by Licensee,"Can you give us some feeling about how much water must be in the core to still be able to attain criticality and fire a pulse?", Dr. Stillman indicated that Intervenor had performed such a calculation, viewed it as an essential point, and that this information would be forthcoming. (See page 116 lines 4 though 14, inclusive, of the transcript of Dr. Stillman's deposition taken in New York on 18 Dec 1982.) To date, this vital information has not yet been provided to the Licensee. Without this information, Licensee is at a loss as to how criticality could be attained and a pulse could be fired when the core is uncovered to any extent. Licensee must therefore imagine on its own the arguments which substantiate Intervenor's claim that criticality and pulsing can occur during a LOCA without benefit of any " insight" from the Intervenor. ( i First, it should be recalled that a low power steady state condition is first established by the manual withdrawal of the standard control rods in preparing to fire a pulse. It was also s pointed out that to attain an approximate 15 watt steady state power level in anticipation of firing a pulse the " safety" and " shim" control rods had to be fully withdrawn while the
" regulating" rod had to be withdrawn approximately 80% This leaves only the upper 20% of the regulating rod (or about 30c of reactivity) which could be used to overcome any negative i
reactivity as a result of a water void in the core and still pe mit the firing of a pulse from the steady-state condition. This is true since the transient control rod must be fully "down" to intiate a pulse. This 30c of reactivity available in the regulating rod would be completely a used up if only approximately 3-5% of the core water was missing. We might, therefore, only fire a pulse (any pulse) when the core is provided with 95% (or more) of its total normal water inventory. However, even this limit is open to question. The interstitial water between the fuel elements in core is necessary as a moderator to ensure that neutrons, in fact, reach thermal energy (where fission predominantly occurs) effectively. When water is removed, neutrons cannot reach thermal energy and this is true wherever water is presumed to be missing. Therefore, if we assume, for example, that some fraction of the upper portion of the active core region were devoid of water, then the uncovered fueled region of each fuel element that is devoid of water has no input of thermal neutrons to initiate fission. This means that all uncovered fuel element regions will be largely incapable of effectively contributing to fission, power, neutron population, fission density, and thus even a fuel temperature increase since inadequate neutron moderation is provided to such uncovered regions. This indicates that locally within the core wherever water is not provided, conditions cannot be aggravated beyond those conditions already in existence because of the water loss or LOCA by itself. It should also be pointed out that each fueled region of the core which is presumed to be l l devoid of water will have an associated higher than normal fuel temperature because of the missing water together with the internal source of heat being generated by virtue of the in-place fission products that are undergoing natural radioactive decay, and this increased fuel
- temperature locally will automatically introduce negative reactivity (also locally) which l
; would also act to suppress effective fission in those uncovered regions.
l { i
This discussion indicates that attainment of criticality, by itself, would be seriously in question even for relatively small or minor water void fractions. And certainly if criticality is not possible then pulsing (even the firing of a single pulse) would be totally out of the J question. I l' In summary, Licensee has demonstrated that numerous safeguards must fail (extremely unlikely) and gross operator error must be assumed to permit the transient control rod to be selectively fired repetitively out of core during a LOCA. Even if all of this were presumed to occur, actual successive pulses could not, in fact, occur at a frequency faster than one approximately every 10 seconds, i.e., until fuel temperatures have basically recovered to ambient conditions. Licensee further has demonstrated that uncovered fuel regions cannot effectively contribute to fission. Thus, fuel temperatures in these uncovered regions cannot become aggravated beyond those conditions which already exist by virtue of the water loss alone. Licensee submits, therefore, that multiple or even single clad failures during a LOCA are not expected since conditions necessary for clad failure cannot be attained. The record amassed to date, particularly the analyses of the LOCA as provided in the AFRRI SAR and as formerly provided in the 1965 license amendment LOCA analysis, are valid and provide a true, realistic, and, in fact, conservative picture of such an occurence (i.e., a presumed LOCA at AFRRI).
, ,/
2 A.SHO J W~ Sworn to and subscribed before me on this JLG day of J4,1983. M ' EY ~ _ . - __ __ -
I
, UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION
- BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of ARMED FORCES RADIOBIOT.DGY Docket No. 50-170 RESEARCH' INSTITUTE (Renewal of Facility (TRIGA-Type Research Reactor) License No. R-84)
CERTIFICATE OF SERVICE OF DUPLICATE SIGNED COPIES OF 25 FEBRUARY 1983 FILING I hereby certify that true and correct copies of the foregoing " LICENSEE'S MOTION FOR PARTIAL
SUMMARY
DISPOSITION" were mailed this 25th day of February, 1983, by United States Mail, First Class, to the following: Judge Helen Hoyt Atomic Safety and Licensing Board Panel U.S. Nuclear Regulatory Commission Washington, D.C. 20555 Mr. Ernest E. Hill Administrative Judge Lawrence Livermore Laboratory University of California P.O. Box 808, L-123 Livermore, CA 94550 Dr. David R. Schink Administrative Judge Department of Oceanography Texas A&M University College Station, TX 77840 Mr. Richard G. Bachmann, Esq. Counsel for NRC Staff U.S. Nuclear Regulatory Commission Washington, D.C. 20555
Elizabeth B. Entwisle, Esq. ( 237. Hunt Road Pittsburgh, PA. 15215 e, Atomic Safety and Licensing Board Panel U.S. Nuclear Regulatory Commission Washington, D.C. 20555 Secretary (3) U.S. Nuclear Regulatory Commission ATTN: Chief, Docketing and Service Section Washington, D.C. 20555 s l $ VID C. RI ARD Counsel for Licensee I t 2 l}}