NUREG/CR-4981, Forwards Fr Notice on Denial of Committee to Bridge Gap Petition for Rulemaking Re Fire Response Plans for Graphite Fires & BNL Rept,NUREG/CR-4981 Re Safety Assessment of Use of Graphite in Nuclear Reactors Licensed by NRC

Forwards Fr Notice on Denial of Committee to Bridge Gap Petition for Rulemaking Re Fire Response Plans for Graphite Fires & BNL Rept,NUREG/CR-4981 Re Safety Assessment of Use of Graphite in Nuclear Reactors Licensed by NRC

ML20236E589
Person / Time
Issue date:	10/23/1987
From:	Michaels T Office of Nuclear Reactor Regulation
To:	Andrea Johnson Oregon State University, CORVALLIS, OR
References
RTR-NUREG-CR-4981 NUDOCS 8710290279
Download: ML20236E589 (50)

Text

c - ,-

• \, . ,. . s.

~

October 23, 1987 Dr. Arthur Johnson Chairman, TRTR' ,

Radiation Center. '

Oregon State University i Corwallis, Oregon 97731 '

l

Dear Dr. Johnson:

Enclosed for your information and for distMoution to members of the TRTR, as appropriate, are the following generic documents, which are relevant to non-power reactors.

1. Federal. Register Notice on the denial of the CBG petition for rulemaking regarding fire response plans for graphite fires (52 FR 3732, October 6, 1987).

2. A BNL Report, NUREG/CR-4981, A Safety Assessment of the Use of Graphite in Nuclear Reactors Licensed by the U.S. NRC.

Sincerely, original signed by Theodore S. Michaels, Project Manager Standardization and Non-Power Reactor Project Directorate

-Division of Reactor Projects - III, IV, V and Special Projects Office of Nuclear Reactor Regulation

Enclosure:

As stated 1 DISTRIBUTION:

icentralfile.,

.NRC PDR PDSNP Reading LRubenstein EHylton a7102 g h a71023 ppg ,

LRubenstein TMichaels PDR CRaggi R fl.fO PDSNP TMichaels PS E ton PlijS

~

LRube(nstein l P ,-

h

?O/14/87 10 87 Ip, 10/R3/87 I'

1 1

]

a JJ Y '

f~% j j_

h

' UNITED STATES NUCLEAR REGULATORY COMMISSION W ASHINGTON, D, C. 20655

% I- October 23, 1987

%, * * . . + /

Dr. Arthur Johnson Chairman, TRTR Radiation Center Oregon State University -

Corwallis, Oregon 97731

Dear Dr. Johnson:

Enclosed for your.information and for distribution to members of the TRTR, as appropriate, are the following generic documents, which are relevant to non-power reactors.

1. Federal Register Notice on the denial of the CBG petition for J rulemaking regarding fire. response plans for graphite fires g

(52FR.3732, October 6,1987).

2. A BNL Report, NUREG/CR-4981, A Safety Assessment of the Use of Graphite in Nuclear Reactors Licensed by the U.S. NRC. ,

Sincerely, f

[ 4 4 f. 7p kl Theodore S. Michaels, Project Manager Standardization and Non-Power Reactor Project Directorate Division of Reactor Projects - III, IV, Y and Special Projects Office of Nuclear Reactor Regulation l,

Enclosure:

As stated r

__________-_-__a i

( _ . _ _ _ _ _ -

q' , .

Fedital Registtr / Vol. 52. No.103 / Tuesday, October 6,1987 / Proposed Rules 37321 k 'The State of Montana."in alphabetical its potential is essentially independent

• The NRC failed to required basic j order. of stored energy in graphite. Empirical safety measures that could help tn <

Done in Washington DC, on this 1st day of measurements of stored energy in reduce the threat of such a fire.

October.1987 graphite are not needed to perform an Licensees whose reactors use graphite.

B.C. Johnson, evaluation of the releasable stored including dozens of non-power reactors Act/ng Deputy Administrofor, Veterinary energy. Furthermore, the requirement for and one commercial power reactor, have Services, Anima /ondPlant //eo/th Inspection such measurements could result in no fire response plans for combating Service. personnel exposures that would be graphite fires in their reactors. Non.

l [m Doc. 87 23108 Filed 104-87; 8:45 am) inconsistent with NRC's as low as is power reactor lice,nsees do not have I senseo caos uSu.a. reasonably achievable (ALARA) adequate emergency plans to evacuate principle. members of the publicin the event of a ADoRESSES: Copies of the petition, graphite fire or other severe accident.

NUCLEAR REGULATORY public comments and abstracts of the For these reasons, the petitioner COMMISSION comments received on the petition, and would require aillicensees whose the Brookhaven National Laboratory reactors employ graphite as a neutron Pad 50 Report NUREG/CR-4981 are available moderator or reflector and whose (Docket No. PRW-50-44) for inspection and copying under Docket licensed power is greater than 100 W to:

No. PRM-50-44 in the NRC Public (a) Formulate and submit for NRC Committee To Bridge the GAP; Denial Document Room.171711 Street NW., approval fire response plans for i of Petition for Rulemaking Washington, DC. Copies of NUREG/CR- combating a reactor fire involving 4981 may be purchased through the U.S. graphite and other constituent reactor I AOENCY: Nuclear Regulatory parts (e.g., fuel) which might be involved Government Printing Office by calling i* ' Commission ^ In such a fire, taking into consideration (202) 275-2000 or by writing to the U.S.

ACTION: Denial of petition for Government Printing Office, P.O. Box the potential for explosive reactions.

rulemaking. 37082. Washington, DC 20013-7082. Response plans shallidentify precisely }

Copies may also be purchased from the which materials will be used to suppress suuuARy:%e Nuc! car Regulatory National Technical information Service, a fire without tncreasing the risk of Commission (NRC) is denying a petition for rulemaking submitted by the U.S. Department of Commerce,5285 Port explosion, and shallindicate where and Committee To Bridge the Cap.The Royal Road, Springfield, VA 22161. In what quantities these materials will petitioner requested that the be stored.

FOR FURTHER INFORMATION CONTACTt

. Commission amend its regulations to Theodore S.Michaels Standardization (b) Formulate and submit for NRC { '

require all licensees whose reactors and Non Power Reactor Project approval evacuation plans for a reactor employ graphite as a neutron moderator Directorate Office of Nuclear Reactor fire. Plans should include evacuation out  !

t or reflector and whose licensed power is Regulation, U.S. Nuclear Regulatory t a sufficient distance from the reactor such that no, member of the public il

. greater than 100 W to: (1) Formulate and Commission, Washington, DC 20555, I submit for NRC approval fire response receives a dose to the thyroid greater '

Telephone (301) 492-8251.

plans for combating a reactor fire than 5 rem, assuming a release to the sUWMENTARY IMPORTATION

{w involving graphite and other constituent environment of 25% of the equilibrium radioactive iodine inventory.

I reactor parts (e.g., fuel); (2) formulate %e Petition

>' and submit for NRC approval (c) Perform measurements of the evacuation plans in case of a reactor A petition for rulemaking was filed by "Wigner energy" stored in the graphite i fire: and (3) perform measurements of the Committee To Bridge the CAP (CBC) of their reactor, and submit these l the Wigner energy atored in the graphite on July 7,1980. The petition was measurements to NRC for review i of their reactors and submit these docketed by the Commission on July 7 together with a revised safety analysis, 1' measurements to the NRC for review, 1986 and was assigned Docket No. which shall address the risks and together with a revised safety analysis PRM-50-44. A notice requesting consequences of a reactor fire. A that shall address the riska and comments on the petition was printed in sufficient number of graphite samples consequences of a reactor fire. the Federal Register on September 3, shall be measured to identify the The petitioner believes these 1986 (51 FR 31341). The petition requests location of maximum stored energy, and j requirements are necessary because the that the Commission amend its to determine the maximum quantity of previous NRC safety evaluations of regulations- stored energy within 110%.

3 these reactors allegedly were based on a Basis for the Request Public Comments on the Petition i belief that graphite fires were not

credible and on an inability of the NRC The petitioner offered the following On September 3,1986, the l

., and its contractors to properly calculate justification for the proposeod revision Commission published a notice in the j

^

Wigner energy in the graphite. The of the regulations: Federal Register (51 FR 31341) j Commission is denying the petition

• The occurrence of a graphite fire at requesting comments on the petition.

I because Fort St. Vrain Nuclear the Chernobyl plant in the Soviet Union The NRC received nine requests for an

- Generating Station and all NRC licensed demonstrates that such fires are credible extension of the comment period. An

- research and test (non. power) reactors events. The NRC and its licensees have extension of the comment period was have approved plans for dealing with mistakenly dismissed graphite fires as granted, changing the closing date for emergencies in accordance with existing noncredible events. the comments from November 3,1986, to regulations. The protective actions are + New experimental data show that February 2.1987. A total of 27 comments based on conservative dose calculations NRC's generic analysis of stored energy were received, six of which supported consistent with those proposed by the in research reactor graphite significantly the petition and 21 of which opposed the petitioner. underestimates the actual amount of petition.Of the six commenters  ;

Graphite burning is a very low- stored energy, and thus underestimates supporting the petitiv, two were probability (i.e., noncredible) event und the associated risk of grnphite fire, individual citizens ed four were from i l J '

1

, y q .s 37322 Iqdiral Register / Vol.'52, No.193 / Tuesday, October 6,1987 / Proposed Rules citizen's groups. Of the 21 commenters ur:necessary exposure of reactor scientists at BNL consider the graphite opposed to the petition.15 were personne). burning a secondary or corollary event universities or university related

• CBG fails to provide a technical re sulting from the explosions that >I organizations, four were companies basis for any of the petition's proposed occurred as a result of a very rapid involved with the nuclear industry, one requirements. reactivity insertion that overheated the was a state government agency, and one The comments opposing the petititon fuel and cladding.The explosion created was an individual citizen. are too numerous to address the conditioru necessary to initiate and Of the comments in support of the individually. Ilowever, each comment sustain graphite burning (e.g.,

petition, none offered any specific has been considered by the staff and its - fragmentation of fuel and graphite.

technical insights but rather sirnply ^ / contractors in analyzing the petition and rupture of the moderator inert gas endorsed the information and basis of ' in developing the NRC position. boundary, admission of air, a favorable the petition.These comments covered Abstracts of all comments received and ratio of graphite volume to surface area, general concerns that include: the full text are available at the NRC sustained heat input from asphalt fires,

• The potential for graphite fires, Public Document Room in the Docket and decay heat). Although the petition

• Training of firefighters to manage file PRM-50-44, as noted in the address considers the Chernobyl accident a graphite fires, section above, demonstration of graphite fire

• Evacuation of persons on-site and credibility, the accident confirms that in nearby areas in the event of an Analysis of the Petition initiation and sustained burning of accident- (1)The petitioner asserts that "the graphite require the existence of a liighlights from the comments occurrence of a graphite fire at the compley combination of ideal opposmg the petition are as follows:

Chernobyl plant demonstrates that such conditior.s whIch are extremely difficult fires are indeed credible events." to achieve in any real situation and are reac ors t he ob ( K)

CBG filed its petition on July 7,1986. virtually incredible m the reactors being resctor ignores the extreme differences considered under this petition.Th,e in power level, core size, tinion product Consequently, only fragmentary words

• credible and ' incredible have inventory, operating temperature, information, mostly conjecture, was availabic before the petition was D!ed, been used in many AEC/NRC safety resctor control systems, and inherent , More detailed and definitive information , analyses. As used by the staff, these

, design characteristics. was first made available, outside the words have always been a qualitative

• CBG's inference that graphite fires . statement of the likelihood or were the initiating events in both the Soviet Union, during a meeting held by

' the laternational Atomic Energy Agency probability of an event or condition Chernobyl and Windscale accidents.

(IAEA)in Vienna, Austri%cn August 25 occurrias. Accordingly, the staff's ccnnot be substantiated. conclusion that sustained or self-

~

• The operating temperature of the to 29,1986. Without the beuefit of the detailed Soviet report, the basis of the sustamed graphite burning is not a Chernobyl graphite (700*C) dismisses credible event in NRC licensed reactors

.l CBC's contention that stored energy in petititon is seriously flawed.

In response to the CDG assertion is still valid (i.e. the random the irradiated graphite played any role simultaneous occurrence of the several in the Chernobyl accident. regarding the Chernobyl event, the NRC l conditions necessary for sustained

• CDG ignores the necessity for an selected Brookhaven National Laboratory (BNL).operatorof the graphite burning or self. sustained initiating event to raise the graphite graphite burning is an event with a very I

temperature 50C*-100C' above its Brookhaven Graphite Research Reactor, normal operating temperature before whose staff is recognized internationally small probability of occurring). The staff for its research on reactor-grade thus concurs in the conclusion reached '

any Wigner (stored) energy in graphite can be released. graphite end its properties, to review the in the BNL report:"There is no new evider.ce associated with the analyses

• CBG ignores the fact that only the published information and de:ccmine its relevancy to the ne of graphite in NRC. of either the Windscale accident or the releasable stored energy, not the total Chernobyl accident that indicates a stored energy, in graphite,in accordance licensed reactms in addition, BNL credible potential for a graphite burning with the annealing temperature, can personnel reviewed the Chernobyl and Windscale eccidats and the role. it accident in any of the reactort, contribute to a graphite temperature considered in this review, Nor is there increase, any, of the graphite moderator in these events.The results of this review are any new evidence that detailed case by-

• The conditions necessary for coatnined in NUREG/CR 4901,"A case safety analysis of the role of graphite burning do not exist nor can graphite in NRC-licensed resclors are they be created by random events in Safety Assessment of the Ur.e of Graphite in Nuclear Reactors Licensed warranted." Accordingly, there has been non-power reactors. no change in the staffs assessment of

• The conditions necessary for by the U.S. NRC." july 1987.This report i is available as noted in the address graphite burning, the Chernobyl graphite burning do not exist in the Fort accident notwithstanding. in NRC-St. Vrain reactor. section above.

The staff has used the BNL report, licensed reactors, and no changes are

• Operating temperatures of the comments received from the public, and required in the staffs previous findings graphite in the Fort St. Vrain reactor in the safety evaluation reports prepared preclude the accumulation of any its own understanding of and expertise significant quantity of stored energy relevant to the use of grephite in non, for these reactors.

I (i.e., the graphite is self annealin ). power reactors and Fort St. Vrain to (2) The petitioner states that "the NRC

• NRC-epproved emergency p ans evaluate and respond to the assertions has failed to reqm' basic safdy (required by to CFR part 50, Appendix and prop) sed requirements of the CCB l,t E) are in place at all NRC licensed petition (pRM-50-44). "',[,"p[.

gr p , educe the threat of a i reactors and are adequate and in their evaluations of the Chernobyl acceptable. accident, both Soviet and international The petitioner did not identify the

• Measurement of stored energy is scientists argee that graphite burning " basic measures" the NRC has failed to not consistent with the ALARA d!d occur during this accident. However, require and provided no basis for this philosophy, since it requires the most of the experts, including the statement.The staff considers that the

3 ,

i federal Register / Vol. 52, No.193 / Tuesday, October 6,1987 / Proposed Rules 37323 elements of the NRC regulatory und Protection Program for Nuclear Power (4) The petitioner asserts that "non.

licensing process represent the basic Facilities Operstmg Prior to January 1, power reactors do not have adequate aafety measures required of licensees to 1979." sets forth fire protection features emergency plans to evacuate members 4

ensure the safe design and operation of required to satisfy Criterion 3 of of the public in the event of a graphite i their reactors as well as to provide Appendix A to 10 CFR Part 50. These fire."  ! ,

, specific plans and procedures for NRC requirements include the " basic Neither the petitioner nor any of the 4 managing and responding to off-normal safety measures to reduce the threat of a citizens' groups or individuals conditions and accidents. Some . . fire, supporting the petition provided a basis examples that are relevant to fire It is the staff's judgment that the NRC in support of this assertion.The staff detection, protection, and mitigation are has required adequate basic safety has reconsidered the need to provide a listed below: measures to reduce the threat of fire as plan to evacuate members of the public

• Safety reviews of non-power well as to mitigate the consequences of located off site in the very unlikely reactorn include an assessment of the any fires that do occur.These measures event of a graphite fire and,in the nre protection systems at each facility. have been reviewed. approved, and course of evalua'ing this petition, has F/e detection. fire extmguishers, fire implemented for alllicensed reactors. not identified any such need.

aliirms, fire prevention, fire fighting They generally apply to all fires and truming of facility personnel, and onsite As stated in Regu!atory Guide 2.6' have been found to provide acceptable Revision 1:

cnd offsite response to fire alarms are protection for the health and safety of typical areas included in the safety In the judgment of the NRC staff, the the public.

review. Inadequacies identified during potential radiological hazards to the pubhc the review must be corrected before a (3) The petihoner alleges that " licensees associated with the operation of research and have no fire response plans for graphite test reactors are considerably less than thn=e license is grunted.

inv tved with nuclear power plants In

• Each non-power reactor licensee is fire s." addition. because there are many different sqquired by con $li tions of the license As discussed in item 2, above, all ionds of non. power reactors. the' potential for

('l echnical Specifications) to provide a licensees have NRC-approved emegency swions ansing and the scfety review for experiments to be e neequnas thmof van Imm faduy to emergency plans in accordance with 10 mserted in their reactors and for fa hty.These differences and variations are changes in reactor operation. Among CFR 50.54[q) and 10 CFR Part 50- expected to be reflected reahstically in the l,t Appendix E. These plans provide for emergency plans and procedures developed L many other safety considerations, an msponse to fires, for training of fire for each research and test reactor facihty.

assessment of fire potential (e g., l fighting personnel, and for periodic lls flammable materials) is included. demonstr e prop peration o e Accordingly, each non power reactor I

• Each non-power reactor licensee licensee has developed an emergency $

has responded to the requirements of 10 plant based on the identified p a a

,. CFR 50.54(q) and to CFR Part 50, characteristics of its reactor facility. To 1

, , th p t on Appendix E, in submitting an emergency assist licensees in meeting the plan for NRC review and approval. All reported that the offsite fire fighters and p their supervisors were regularly trained requirements of 10 CFR Part 50, licensed non power reactors now have Appendix E. Regulatory Guide 2.6 4.

approved emergency plans and the in fire fighting procedures for their facilities and that the fire fighters were ( ANSI /ANS-15.16.-1982. Table 2) necessary implementing procedures. provides an " Alternate Method for These plans were reviewed against confident that they were prepared t 4 Determining the Size of an Emergency ANS!/ANS-15.16-1982 and Regulatory deal with the type of fires they could Planning Zone (EPZ)." Table 2 is based I Guide 2.6, proposed Revision 1, as encounter, including a fire involving on highly conservative does calculations

( outlined in NUREG-0649, " Standard graphite.This is consistent with BNL that are generically applicable to non-research,8 which recommends a basic

) Review Plan for the Review and power reactors. These calculations i Evaluation of Emergency Plans for fire fighting technique for graphite fires. include the very conservative

• Research and Test Reactors." that is, exclude air or oxygen and cool. assumption for non power reactors that  !

Examples of the evaluation items that the graphite. Success in using this basic 25% of the equihbrium radioactive I are relevant to " basic safety measures " cool and-smother technique was iodine is gaseous and will escape from l to reduce the threat of. . . fire" are demonstrated during the Chernobyl the reactor building into the hated below: accident. Gold nitrogen gas was pumped environment. it is the current and i (a) The (emergency) plan should also into the bottom of the reactor t standard practice of the NRC staff to use [

describe non. radiological monitors or successfully cool the graphite and fuel the 25% lodine source term with regard indicators * * * (2) Fire detectors * * . debris while excluding oxygen t to 10 CFR Part 20 recommended dose

, (b) The emergency plan should smother any burning. Also at Chernobly, considerations in its safety evaluations 6 scribe an initial training and periodic graphite blocks were successfully of non power reactors. Table 2, which is ,

retraining program designed to maintain quenched usmg water (NUREG-1250, pp. based on power level, recommends that ,

4-12,4-21, and 7-23). Since this basic

! the ability of emergency response cool-and smother technique is effective reactors with power levels less than or i i* personnel to perform assigned functions equal to 2 MW use their " operations 1 for the follo>ving: for most fires, the staff has concluded '

boundry" for their EpZs, which

) * *

• f. Police security, ambulance, that the licensee existing emergency essentially recognizes that a reactor of i and fire fighting personnel * *

• plans provide an adquate response for this power level will only need to (NUREG-0849, Sections 8.0 and 10.0) graphile fires as well as any other type initiate protective actions for members The licensee for Fort St. Vrain has of fire. of the general p"blic on site and will not satisfactorily met the requirements of 10 pose an unacceptable radiological CFR Part 50 48 and 10 CFR Part 50, ' R W Powell. R.A Meyer, and R.C. Dourdeau. hazard to members of the pubhc off site.

Appendix R. Appendix R. " Fire "ceimi R, d. tion meets in a Graphite Ree< tor There are only five licensed non power sin,csure. rwed-n of abe second unned Aore, laternonono/ contence on 4A, reacefuf reactors containing graphite that have j i i covers eil iype, of nre. ir.ct.ains graphne fire. 0.c. of Atomic Enem. vol. r tow. p m power levels greater than 2 MW. Three l

• l1  !

l

.p a- I l

[ .

b 37320 Fedorzl Ragister / Vol. 52, No.193 / Tuesday, October 6,1987 / Propcsed Rules l 4

of the reactors have power levels less information that demonstrates that, paper by Ashbaugh, Ostrander, and than 10 My, one has a power level of !a reven in the remote c2se of graphite Pearlman : at the American Nuclear MW, and Jne has a power level of 20 - burning there is a r<eed to modify any Society annual meeting in june 1980.

MW TurM2 recommends an Epf.cf100 edsting emergency plans.

• Stored energy decreases with meters of non power reactors with increusingastance from the fuel region h(5)The petitioner states that (e.g.,5.01 r al/gm at 18 inches,1.34 cal /

power levels greater than 2 MW and neric analysis of stored energy in "NRC's equal to or less than 10 MW, and 400 lesearch reactor graphite significantly ym at 22 h ches, and an unmeasurable meters for those with power levels amwnt at 26 inches).

underestimates the actual amount of (Within the graphite island, stored gmster than 10 MW and equal to or less stored energy and thus underestimates than 20 MW.The licensee for each of energy decreases from 33.3 cal /gm at the the associated risk of graphite fire." 1 these restors has an NRC-approved fuel box graphite interfact to 19.2 cal /

emergency plan that takes into The conditions necessary for e,tored gm about 3 inches from the fuel box consideration the specific energency releases in graphite are toward the center of the graphite island.

characteristics of each reactor (e.g., described in section 3 of the BNL report. These resul:s illustrate the principles l' fision product inventory and engineered The staff agrees with the methodology associated with the proposed safety featuras)in the development of derived for estimating the stored energy requirement to measure the Wigner the ection levels, procedures, and that can be released from paphite and energy stored in the research and test protective actions necessary to protect in the analysis applied to the estimation reactor praphite.The significant changes all rnembers of the public within its EPZ. of stored energy releasec. in Section 6 of in stored energy with relatively small  ;

Regulatory Guides 1.3 and 1.4 the BNL report. differer.cea In location demonstrate the i f recommend the usr. of the 25%

in section 2 of the BNL report, the difficulty tri selecting the locations and I radioactive irdine source term in necessary conditions for graphite to the number of samples needed to determining the compliance of power burn are discussed in detu1 A s characterize the " maximum stored  !

reactors with the siting, coMninment, reassessment of the literature onthe energy end to determine the maximum '

end dose gul:letines of 10 CTR Port 100, experiments previously perf srmed at quantity of stored energy to within '

The staff bel ever the current regulatory BNL and the nported details of the gagy pr:ctices cre suitable to ensure. that the Windscale and Chernobyl ace; dents are The buses fore storage and release of l

f basic statutory requirement, fot included in the 8% & study. The Wigner energy in graphite are  !

adequate protection of phbMe Frrelth and ev.clusions reached as a result ofinsse delineated in the BNL report, which I safety, is met, smlyses are: shows that thre is no unique i These e nergency plarmina (T)he potential to initiste or maintain a connection between total stored energy p

[ considerations are appropriate fg graphite burnin;;ia%nt is essentially and the releasable energy.Thus, reactors utilizing graphite componMts. independent of Ge stored enerRy in the establishing the magnitude of the stored Because the graphite contains no fission graphite, and depsnds on other factors that energy in non-power reactor graphits by '

are unique for em.h research reactor and for empirical measurements would not products and very few activation l products, eun the remote possibility of fo St. Vr n 1 o o havv self ustained provide the information needed to j the graphite burning would not reactors. certain necessary conditions of evaluate this potential. Because the contribute to the radiological source sometry, temperature, oxygen supply. releasable stored energy saturates, an ,

term.Therefore, a graphite fire m and of react on product removal and a favorable upper bound on the stored energy that l

itself presents essentially no heat balance must be maintained. There is no can be released to 700*C can be I radiological hazard to the publ5. nm evidence associated with either the determined from existing data.

Because of the major differences in Wmdscale A:rident of the Chernobyl Therefore,no measurement of stored Accident that Indicates a credible potential energy is required.

design, power level, co'e s;ze, fission

• for a graphite burning accident in any of the Also,because of the several product inventory, reactor control

"' actors considered m this review.

systems, and inherent reactor neutronica,e.omparison of the Chernobyl On the buis of its review of the DNL burning in addition to a graphite accident and its consequences with report, the literature on DNL temperature of 650*C, the potential to i uccidents and the resultmg experiment 3 und the information on the nitiate or maintain a graphite burning consequences for non power reactors is Windscale and Chernobyl events, the incident is esentially independent of not appropriate. nor is it meaningful- staff finds that tec conclusions reached stored energy in the graphite. This

Many of the comments received in by DNL are correct and adopts them as further supports the conclusion that no opposition to the petition speak of the its own. measurement of stored energy is impropriety of comparing NRC. licensed (0) The petitioner asserts that " actual need'ed non. power reactors with the Chernobyl empirical measurements of Wigner Many of the commenters who RBMK-1000 reactor, oppond the petition cited a violation of p energy will be required to assess the j The petitioner he

.ot provided any proof of inadequacy in the emergency magnitude of the energy stored in rese r:h reactor graphite."

ALARA considerations because stored energy measurements would not provide l plans for non-power reactors. On the needed information. but would incur basis of a review of the guidance for Measurements of stored energy in its radiological exposures. The emergency planning contained in research reactor graphite were made by impracticality of taking the comples and Regulatory Guide 2.6 and ANS!/ANS the University of California, Los making the measurements was also 15.1th1982 and the requirements of to Angeles. in the course of pointed out. For exarnple, sampling the CFR Part 50 Appendix E, the staff has decommissioning its Argonaut research graphite reflector pieces in the ends of a l

concluded that the emergency plans reactor, Several thing. tearned from its

j. previously approved by NRC are still program of sampling and measuring '

appropriate and adequate. Neither the streed energy were reported by a "Craph% Stored Eneepy in the UCt.A Research l

petitioner nor the commenters com.nenter who opposed the petition. e.,,,cim, 7,,,,,n, ,f ,3, aus, voi, st isas, p.

supporting the petition hve supplied This information was also reported in a vz l

I

S .

f* Federal Register / Vol. 52, No.193 / Tuesday, October 6,1987 / Proposed Rules 37325 1 (4  ;

TRICA fuel pin would require breaching in response to un NRC request, Public organization plans, and procedures to

{ the fuel pin cladding as well as Service Company of Colorado privide the necessary protection of the jg providing shielding against the fuel pin's addressed the implications of the health and safety of the public even in lq radioactivity. Similar challenges would Chernobyl accident for the Fort St. the very unhkely event of a graphite fire. .)

'g be associated in taking a sample from graphite reflector components clad with Vrain.The licensee submitted a final Be s.is for Denial f l .1 report entitled " Design Differences. Air

metal. In addition,it was pointed out ingress and Graphite Oxidation, and The NRC denies the petitioner's

,} that numerous samples would be Steam Ingress and Water Gas request to amend 10 CFR Part 50 to required to establish the true magnitude Generation" (P-86041, December 4, require licensees whose reactors employ of stored energy in the various graphite 1986). The staff has reviewed the report graphite as a neutron moderator or components. and concludes that the only significant reflector and whose licensed power is The staff has considered the relevant similarity between Chernobyl and Fort greater than 100 W to:

DNL findings and the comments St. Vrain reactors is that they both (1) Formulate and submit for NRC received and has concluded that contain a large amount of graphite approval fire response plans for empirical measurement of stored energy moderator.There are design differences combating a reactor fire involving in non-power reactor graphite , between these reactors that preclude an graphite and other constituent reactor components is not practical nor is it accident similar to the Chernobyl parts (e.g., fuel) necessary to ensure the health and accident at Fort St. Vrain. (2) Formulate and submit for NRC safety of the public. Furthermore, on the bas!s of its approval evacuation plans in case c' a reviews, the staff concluded that the reactor fire; and n r I ower ea 1 ."In icatmg muchal WNy M & M h hn Mnn measummenu d em pmstmsd conce mador vesM g@ner eng stod M Me sW kW that it has no fire response plans for w old be maintained dunng and after their reactors, and submit these combating graphite fires. The petitioner rneasurements to the NRC for review the assumed accident scenarina.

also states that " graphite is used as a together mth a revised safety ana:ym Although the mitiating events me moderator in the Fort St. Vrain nuclear beyond the plant s original deugn basis, that shall address the risk and power plant in Colorado." j the plant design appears to have an consequences of a reactor fire. ,

Other than the lack of graphite fire adequate margin of safety to withstand This denialis a based on the response plans, the petitioner does not these events. following: ,

identify specific concerns related to Fort The staffs comments and conclusions (1) Each licensee of a non. power 1 St. Vrain. Ilowever, it is implied that all can be found in the NRC Public reactor has submitted an emergency l reactors using graphite components are Document Room under Docket No. 50- plan that has been approved as meeting subject to CDG's concerns and 267,in a letter dated Aprill,1987, the requirements of to CFR Part 50, i assertions. In reality, the petition and Accession No. 8704090248. Appendix E. The petitioner has not

- requirements are really directed at NRC- The petitioner's assertion that demonstrated that these plans do not

} licensed non power reactors. graphite burning and oxidation were not provided an appropriate level of Fort St. Vrain is a high-temperature included in the staffs evaluation for Fort protection of the health and safety of the 1 fe gas cooled reactor (llTGR) owned and St. Vrain is in error.This subject was public. {

U operated by Public Stryice Company of thoroughly reviewed in both the (2) The licensee for Fort St. Vrain has Colorado. Its design capacity is 330 construction permit and operating an approved emergency plan that meets J MWe. it uses a ceramic fuel particle license safety evaluations.These staff the requirements of 10 CFR Part 50 3

+

(uranium and thorium carbide) clad with evaluations may be found in the Public Appendix E, as well as an approved fire y silicon carbide and multiple layers of Document Room in the 50-267 docket protection program that meets the i pyrolytic carbon.The fuel particles are file.The licensee's updated Fort St. requirements of 10 CFR Part 50,

.! compacted into small rods and installed Vrain Final Safety Analysis Report, Appendix R. In addition, at the request 3 in fuel holes in the hexagonal graphite section 14, contains much of the of the NRG the licensee has submitted a fuel blocks. Including the reflectors there information and analyses submitted for report adaessing the implications of the are 500 tons of reactor graphite in the NRC review.The staff concluded that Chernobyl accident for Fort St. Vrain.

! core. The reactor coolant is helium with significant graphite oxidation at Fort St. The report hps been reviewed and i en average inlet temperature of 762* F Vrain was not credible. (Note:In approved by the staff.The petitioner has (405'C) and an outlet temperature of addition to the previously discussed not provided a technical basis that

.l '

1445*F (785'C). The average graphite conditions necessary for graphite would show that an additional fire i moderator temperature is 1380*F (749'C). burning. Fort St. Vrain must suffer response plan would enhance the These characteristics are far different simultaneous independent structural protection provided for the health and

, than those of the non power reactors. failures resulting in the release of the safety of the public by the existing

BNL has reviewed Fort St. Vrain inert helium and the subsequent supply emergency plan and fire protection j parameters in relation to graphite stored of an adequate air / oxygen flow).The program.

energy and concludes in section 7 of its staff finds no ba

Is for changing its (3) Measurement of maximum stored report. ' Tort St. Vrain operates at previous conclusions.The licensee for energy in non power reactors are not temperatures that preclude Fort St. Vrain has met the requirements necessary to ascertain the releasable accumulation of stored energy.There of 10 CFR Part 50. Appendix R (which atored energy in graphite components are no know problema associated with sets forth fire protection features below B50*C. Existing knowledge stored energy in graphite for operating required to satisfy Crturion 3 of10 CFR provides this information which is temperatures associated with liTCRs." Part 50, Appendix A) and has an NRC- adequate for a safety evaluation of the

. The staff agrees with BNL's conclusion approved emergency plan that meets to effect of stored energy on the potential and can find no reason to empirically 10 CFR Part 50. Appendix E. The Fort St. for graphite burning and the associated measure the stored energy in Fort St. Vrain fire protection program and danger to the health and safety of the

' public. Additionally, such measurements Vrain's graphite components. emergency plan specify the necessary f

i

i i

i .

I '3732G Federal Register / Vol. 52. NO.193 / Tuesday, October 6.1987 / Proposed Rules are contrary to the NRC's ALARA encyclopet las and related products and Regulatory Commission,825 North principle, since unneeded knowledge services di ret to the consumer by Capitol Stree NE., Washington, DC would be sought at the expense of means ofin cine, over.the counter, 20426,(202) 3 -8293.

unn cessary personnel exposure, direct mail spd telephone sales . SUPPLEMENTS y INFORMATION:

Accordingly, the Commission denies solicitation, yhe order modification the petition. . request is bayd on claimed char ges of I. introduction Dateds at Bethesda, Maryland, this 23 fact and inw.Jhe supplemental petition y j day of September 1987, was placed ongthe public record on ~ the federal n Regulat y For the Nuclear Regulatory Cornmission. September 22,1987. Commission (Copmission) hereby victue stelle,1c' institutes its founth annual proceeding to Executive Directorfor Opemtions. List of Subjectsfn to CFR Part 13 determine:(1) Al{ estimate of the

[FR Doc. 87-23073 Filed 10-5-87; 8.45 aml Encyclopediasales Trade practices. average cost of common e ulty for the anuno coes rmos u Emily 11. Rock. f jurisdictional operations o public S*Cf8'G rY' W utilities for the ye)r ending June 30,

{FR Doc. 87-23014 F,iled 10-15-87; 8.45 am) 1987; and (2) a quarterly indexing FEDERAL TRADE COMMISSION swwo coos ers*f' procedure to establish benchmark rates

~ t' f return on commgn equity for use in 16 CFR Part 13 individual rate cas s.

DEPARTMENTS ENERGY IDocket D-s9kl The benchmark r tes of return resulting from the fiist three annual Federal Energy Rjgulatory Prthibited Tr de Practices. proceedings were advisory. The l

EncyclopaediaBritannica, Inc., et al. Commission  ;

Commission proposis to make the l

y i 18 CFR Part 37 ( benchmark rates of turn established A2 ENCY: Federal Trade Commission.

8 '

cCTi:N: Notice 5f period for public [ Docket No. RM87-3bl f comment on petilion to reopen the

• II. Discussion -!

proceedmg and modify the order. Generic Determination of Rate of e Return on CommorgEquity for Public A. Base Year A veragg Cost of Common

SUMMARY

Encyclgpaedia Britannica, a Utilities P dRate ofReturn corporate respondent in the order in /

Issued. september 30 1987.

Equity:

The Commission Market pro Requig' poses to ad Docket No. D-8908,ijs prohibited from "

prese tions hi e AoEwev: Federal Enef$y Regulatory O de os 20 H2 A 140 . T} e I

{is, ,, ,

promoting merchandise or services, or mmissi n. DOE. 1 Commission believes th(t the inethod ACTION: Notice of proposed rulemaking. adopted in those prior orders has attempting to collect liebts, and filed a received a full alnng of tl)e issues and hI petition on April 2.1907 requesting that suuuARY: The Federaf Energy represents the most reasonable way to i the Commission reopen the proceeding Regulatory Commission hereby determine the benchmark kate of return.

and either set aside the, order, now or at institutes a proceeding /under Part 37 of Therefore, the Co nmisstor{ proposes to a fixed future date, or niodify the order, its regulations.The purpose of this reh n the followmg const&nt growth A nuppicmental request fo reopen the proceeding is to deterngne an estimate dacounted cash flow [DCF)!rnodel to proceeding has been filed on September of the average cost of common equity for determine the average mark'st required 22,1907.This document apnounces the the juricdictionaloperat}ons of public l rate f retum f r electric utils les for the public comment period orIJhe utilities for the year endng lune 30,1987 year ending June 30,1987: p supplemental petition. (

DATE:The deadline for filitig comments and a quarterly indexin . procedure to establish benchmark ra s of return on k = 0 +.58) y + s 4 on this matter is October 3q 1987. common equity for use in individual rate where: U ADDRESS: Comments should be sent to cases. It is proposed thattthese k= murkrt required rate of return I the Office of the Secretary,%deral benchrnark rates of retur@ remain y = current dhidend yield (current)nnual ,

advisory only,These benEhmark rates of dividend rate divided by current market  !

Trade Commission.6th Streef and pnce)

Penn9lvania Avenue NW., Washington. return on equity establish &d as the {

t ould be used g = dividend growth rate f DC 20580. result of this proceeding, a)d interveners

Requests for copies of the petition as a guide to companies ag g in individual rate cases and as a i in the third annual benchmark rate proceedtns should be sent Dranch, Room 130.to Public Refetche reference point for the Cor(mission in its the NOPR proposed to presumpuvely set the t

F02 FURTHER INFORMATION CON ACT:

deliberations. The Commis(ion may g ed' da $ f'I$ u a. C u*krAtd)' '"

m

' lock K.Chung Enforcement Diviilon. take official notice of them p individual 6n effect at the time a company filed. See Notice of Dureau of Consumer Protection, Federal rate proceedmgs. g propo.ed Rotem. king ceneric peierminadon of Trade Commission, Washington. DC DATE: Comments addressin in this proceeding areNovember glRgong eE due o(the issues ,

(

s1 oso Uuly 20580,(202) 326-2984. f 21. tone 1.The fin.1 rule..fier con ider tigo of SUPPLEMENT ARY INFORM AT10N:Thd* 5.1987. comments nied. ellowed the benchmark rates of return to reimain advisory only. See Orded No. 461.

ord:r in Docket No. D-4908 was ADDRESS:All filings should ference R. R.

published at 41 FR 17884 on April 2 f Docket No. RM87-3M100 an should be Qg* "",%u,on t .,, ana 1970. A correction to the order was q addressed to: Office of the Secretary. Unmary 2. ssee).

published ai 41 FR19301 on Msy 12, 3 Federal Energy Regulatory Cosimission, e Order No. 420. Generic Determinatforiof Rate of Return on common Equiry for rubisc uuttues, so PR 1970. The original request to reopen the 825 North Capitol Street NE., e Washington, DC 20426. P y,so2g )

,;gN proceeding was published at 52 FR a; 12430 on April 16,1987. The petitioner.$ FoR FURTHER INFORMATION CONTACT:

Encyclopedia Britannica. sells ( Ronald L. Rattey. Federal Energy for Pubhc order Utlhties, No sei. eee suem in si1. FR 125o5 (lune 211f, ).

i,

1 NUR$G/CR-4981:

. BNL-NUREG-52092 1 I

1 A Safety Assessment of the Use of Graphite in Nuclear Reactors Licensed by the1 U.S. NRC Prepared by D. G. Schweitzer,~ D. H, Gurinsky, E. Kaplan, C. Sastre .

' Brookhaven National Laboratory 1

l

- Prspared for I U.S. Nuclear Regulatory -)

Commission )

l I

I

lh

-l NOTICE This report was prepared as an account of work sponsored by an agency of the United States Government.. Neither the United States Government nor any agency thereofl or any of their L employees, makes any warranty,. expressed or , implied, or assumes any legal liability of re- i

. sponsibility for any third party's use, or the results of such use, of any information, apparatus,. {

product or process disclosed in this report, or represents that'its use by such third party would ~

.I not infringe privately owned rights. ]

l 1

1j 1

i 1

NOTICE i l

Availability of Reference Materials Cited in NRC Publications j Most documents cited in NRC publications will be available from one of the following sources: f 1, The NRC Public Document Room,1717 H Street, N.W. f Washington, DC 20555

2. The Superintendent of Documents, U.S. Government Printing Of fice, Post Of fice Box 37082 Washington, DC 20013 7082

3. The National Technical Inforrnation Service, Springfield, VA 22181 othough the listing that follows represents the majority of documents cited in NRC pubhcations, it is not intended to be exhaustive.

Referenced documents available for .nspection and copying for a fee from the NRC Public Docu ment Room include NRC correspondence and internal NRC memoranda; NRC Of fice of Inspection and Enforcement bulletins, circulars, information notices, inspection and investigation notices; Licensee Event Reports; vendor reports and correspondence; Commission papers; and applicant and licensee documents and correspondence.

The following documents in the NUREG series are available for purchase from the GPO Sales '

Program: formal NRC staff and contractor reports, NRC-sponsored conference proceedings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regulation * ' ' the Code of Federal Regulations, and Nuclear Regulatory Commission issuat es.

Documents available from the National Technical Informata Service include NUREG series reports and technical reports prepared by other federal agencies and reports prepared by the Atomic -

Energy Commission, forerunner agency to the Nuclear Regulatory Commission. l Documents available from public and special technical libraries include all open literature items, i such as books, journal and periodical articles, and transactions. Federal Register notices, federal and state legislation, and congressional reports can usually be obtained from these libraries. f Documents such as theses, dissertations, foreign reports and translations, and non NRC conference proceedings are available for purchase from the organization sponsoring the publication cited Single copies of NP.C draft reports are available free, to the extent of supply, upon written request to the Division of information Support Services, Distribution Section, U.S. Nuclear Regulatory Commission, Washington, DC 20555.

Copies of industry codes and standards used in a substantive manner in the N RC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, and are available there for reference use by the public. Codes and standards are usually copyrighted and may be purchased from the originating organization or, if they are American National Standards, from the j American National Standards Institute,1430 Broadway, New York, NY 10018.

NUREG/CR-4981 BNL-NUREG-52092 A Safety Assessment of the Use of Graphite in Nuclear Reactors Licensed by the U.S. NRC a0eYu shed Sep e er l

,YSch eitzer, D. H. Gurinsky, E. Kaplan, C. Sastre S'g?/n" J '" "v"2""e?""'"' l Prepared for i Office of Nuclear Reactor Regulation l U.S. Nuclear Regulatory Commission Washington, DC 20555 NRC FIN A3855 1

p, ,

L . .

i

,s o a I'

CONTENTS H

d ABSTRACT .'. ... . . . .-. .'. . . . . . . . . . ......... 1 1.., INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . .

2

2. GRAPHITE B URNING . . . . . . . . . . . . . . . - . . . . . . . . . .-. 2

'3.: ' STO RE D E NERGY . . . . . . . . . . . . . . . . . . . . . . . . ' . . . . 7-3.1 Summary . .. .. . . . ...................

7

3.2 Wigner Energy - Its ' Generation and Buildup' ........ 7 j c

3.3 -Stored Energy Releases-. . . . .......... ..... 12 '

3.4 Calculational Approaches . ................. 16

4. THE CHERNOBYL ACCIDENT . . . . . . . . . . . . . . .:. . . .-. . .- 17

5. ~" ACCIDENT AT WINDSCALE-NO. 1 PILE ON 10th 0F OCTOBER, 1957" ... 18 l

-i t

6. U. S. RESEARCH REACTORS . . . . . . . . . . . . . . . . . . . . . 19 i 6.1 Criteria.for Stored Energy in Graphite . . ......... 19

~6.2 Stored Energy in Graphite. . . . . . . . . . . . . . . . . . 20.

~6.3' . Gra phi t e Bu rning . . . . . . . . . . . . . . . . . . . . . . 22

7. FORT ST. VRAIN - GRAPHITE STORED ENERGY . ............ 24

8.

SUMMARY

.:. . . . . . . . . . ....... . . . . . . . . . . ' . 24 8.1 Graphite Burning . . . . . . . . . . . . . . . . . . . . . . 24 8.2 Stored Energy in Graphite ................. 25 8.3 Safety Assessment - . .................... 27

9. CONCLUSIONS .. . . . . . . ........... .. ...... .27

10. ' GLOSSARY . . . . . . . . . . ........ ........... 28

11. REFERENCES . . . . . . . . . . . . . . . . . . . . . . ...... 29

12. BIBLIOGRAPHY . . . . . . . ............... ..... 33 I

I 1

L l

I V

l iii

.J

.i

-' R

.3 FIGURES 3..

Figure 1. Graphite burn configuration' . .. . . .. . . . . . . . . .. 5 Figure'2a. Total:ve released storedjenergy . . ... . . ..'... . . . .. 10

. Irradiation =.30*C, Tanneal = 800*C Figure 2b."Totalavs released' stored energy . . . . . ' . . . . . . . . . 11.

Irradiation = 30*C,'Tanneal = 400*C I Figure 3. -Stored energy released . . . . . . . . . . . . . . . . . . . 13 4

. Figure 4 Cumulative energy release;.. . . .. . . . . . . . . . . . . 14 exposure of 500 mwd /AT or less i

Figure 5.- Cumulative energy release; ................. 15 l' irradiations at 70*C and above l 1

. 1 i

[.

i l

V i

i A Safety Assessment of the Use of Graphite in Nuclear Reactors Licensed by the U.S. NRC D. G. Schweitzer, D. H. Curinsky, E. Kaplan and C. Sastre ABSTRACT i

This report reviews existing literature and knowledge on graphite burning and on' stored energy accumulation and releases in order to assess what role, if'any, a stored energy release can have in initiating or con hypotheticalgraphiteburningscenariosinresearchreactors.{ributingto-It also addresses the question of graphite ignition and self-sustained combustion in the event of a loss-of-coolant accident (LOCA).

The conditions necessary to initiate and maintain graphite burning are j

summarized and discussed. From analyses of existing information it is con-cluded that only stored energy accumulations and releases below the burning l temperature (650*C) are pertinent. After reviewing the existing knowledge on I stored energy it is possible to show that stored energy releases do not occur spontaneously, and that the maximum stored energy that can be released from any reactor containing graphite is a very small fraction of the energy l produced during the first few minutes of a burning incident. l The Windscale and Chernobyl accidents are summarized and reviewed. It is shown that there is no evidence from the Chernobyl' event that stored energy releases played a role either initiating or contributing to this accident. An i improperly controlled process of annealing the graphite at Windscale with'nu-clear heat resulted in damage to the fuel elements that initiated fuel burning which resulted in~a graphite fire. Stored energy releases did not initiate or contribute to this accident either.

The conclusions from these analyses are that the potential to initiate or maintain a graphite burning incident is essentially independent of the stored energy in the graphite,' and depends on other factors that are unique for each research reactor and for Fort St. Vrain. In order to have self-sustained rapid graphite oxidation in any of these reactors, certain necessary condi-tions of geometry, temperature, oxygen supply, reaction product removal, and a favorable heat balance must be maintained. There is no new evidence associ-ated with either the Windscale Accident or the Chernobyl Accident that indi-cates a credible potential for a graphite burning accident in any of the 3 reactors considered in this review. I

1. Research reactors as used herein means research, test, and training l reactors.

l l

1

1. ' INTRODUCTION On September 3, 1986 the NRC published in the Federal Register [51FR3134, 1986] a notice of receipt of a petition for rule making filed by The Committee to Bridge The Gap to consider the subject of graphite fires in U.S. research nuclear reactors. Under contract with the NRC staff, Brookhaven National Laboratory staff with past experience in safety evaluation of graphite burning and stored energy releases initiated a reevaluation of graphite burning and stored energy information. The objective of this evaluation was to develop an analysis of tho potential role of stored energy releases in initiating or con-tributing to graphite burnir.g scenarios, as well as an analyses of graphite ignition and self-sustained combustion in the event of a LOCA accident.

The 1986 accident at Chernobyl motivated studies describing the causes i for the accident. As a result of this new information, BNL has undertaken a reevaluation of the Windscale Accident, graphite burning scudies, and stored energy information that might be relevant to hypothetical graphite burning .

scenarios in nuclear reactors.

Prior to a detailed analysis of the Windscale Accident, the British mis-takenly assumed that the accident might have been initiated by a stored energy release that took place during the anneal of the reactor. Subsequent work by both the team at Brookhaven National Laboratory and the British showed that this was not true, and that the accident was triggered by an uranium. fire. In the Prime Minister's report to Parliament, [ Penney, 1957], the following statement was made,

"...the most likely cause of the accident was the combined effect of the rapid (nuclear) heating and the high temperature reached by the fuel elements in the lower front part of the pile. In all probability, one or more end caps of the cans of fuel elements were pushed off, and uranium exposed."

As a result of the extensive full scale work carried out at BKL, a great deal of detailed information was developed on the factors affecting both the burning of graphite and the stored energy releases that occurred during anneals [Schweitzer, 1962c; Kosiba, 1953].

2. GRAPHITE BURNING For reasons that are well understood, graphite is considerably more dif-ficult to burn than is coal, coke, or charcoal. Graphite has a much higher thermal conductivity than have coals, cokes or charcoals, making it easier to dissipate the heat produced by the burning and consequently making it more difficult to keep the graphite hot. Concomitantly, coals, cokes and charcoals develop a porous white ash on the burning surfaces which greatly reduces radi-ation heat losses while simultaneously allowing air to reach the carbon sur-faces and maintain the burning. In addition, coals, cokes and charcoals are heavily loaded with impurities which catalyze the oxidation processes.

Nuclear graphite is one of the purest substances produced in massive quantities.

2

7m7 . .. . .

l The literature on the oxidation of graphite under a very wide range of conditions 10 extensive. Effects of temperature, radiation, impurities, por-osity, etc., have been studied in great detail. for many dif ferent types of graphites and carbons [ Nightingale, 1962]. This information served as a foun- -]

dation for.the full scale detailed studies on graphite burning accidents in j air-cooled reactors initiated and completed at Brookhaven National Laboratory j

[Schweitzer, 1962a-f]. After British experimenters at Harwell. confirmed the l results obtained at BKL [ Lewis, 1963] there appeared to be no neu conclusions 1 from additional work in this field. The aspects of the work pertinent to f evaluating the potential for graphite burning accidents are described here in i some detail. fi Burning, as used here, is defined as self-sustained combustion of graph-ite. Combustion is defined as rapid oxidation of graphite at high tempera-tures. Self-sustained combustion produces enough heat to maintain the react-ing species at a fixed temperature or is sufficient to increase the tempera-ture under actual conditions where heat can be lost by conduction, convection, and radiation. In the case where the temperature of the reaction increases,

.the temperature will continue to rise until the rate of heat loss is just equal to the rate of heat production. Sustained combustion is distinguished from self-sustained combustion when, in t}e first case, the combustion is sus-tained by a heat source other than the graphite oxygen reactions (e.g., decay heat'from reactor fuel).

Early attempts to model the events at Windscale [ Robinson, 1961; Nairn, 1961] were followed by the BNL work described here.

Some 50 experiments on graphite burning and oxidation were carried out in 10-foot long graphite channels at temperatures from 600*C to above 800*C. To obtain a lower bound on the minimum temperature at which burning could occur, the experiments were specifically designed to minimize heat losses from radia- '

tion, conduction, and convection.

The objectives of the full scale channel experiments were to determine under what conditions burning might initiate in the Brookhaven Graphite Research Reactor (BGRR) and how it could be controlled if it did start. Chan-nels 10-feet long were machined from the standard 4 in. x 4 in. blocks of AGOT 2 graphite used in the original construction. The internal diameter of the BGRR channel was 2.63 inches. Experiments were also carried out on chan-nel diameters of one to three inches on 10-foot long test channels in order to obtain generic information. The full length of the channels was heated by a temperature controlled furnace and was insulated from conductive heat losses.

At intervals along the length there were penetrations in the furnace through which thermocouple used to read the temperature of the graphite and air were introduced, and from which air and air combustion products were sampled. A l

preheater at the inlet of the graphite channel was used to adjust the air to j the desired temperature. The volume of air was controlled and monitored by flow meters to allow flow measurements in both laminar and turbulent flow conditions.

I

'2. Trade name for nuclear graphite used in the BCRR. l 3

L_____-__ ____ __ _

t.

. l' a

-In. a typical? experimental. run the graphite was first heated to a prese-  ;

- 1ected temperature. .The external heaters were kept on to minimize heat losses -j

-by conduction and radiation. The temperature changes along.the graphite'chan-nel were then measured-for each flow rate as a function of. time:with the

' heaters kept on. It was . observed that below 675'C it was not possible lto -

obtain. temperature rises along the channel if the heat transfer coefficient (h) was! greater than 10- cal /cm-sec *C. Below 650*C it was not/possible,to.-

get large temperature rises along the' channel with 30*C inlet air temperatures-at any' flow rate. For h values lower than'10 " cal /cm-sec *C maximum tempera--

ture rises were 0-50*C andLremained essentially constant for long periods of time (five hours). For h values greater than 10 4 cal /cm-sec *C the full length of the-channel was cooled rapidly.

There were two chemical reactions occurring along channels. At. low tem-

.-peratures.the reaction C + O2 to form C0: predominated. As-the temperature

- increased 1along the channel C0 formed ei',ner directly at the surf ace of the ,.

channel'or by the reaction CO2 + C. At temperatures'above 700*C,:C0' reacts in

)

the gaseous phase to form CO2 with accompaniment of a visible flame. It was observed that the unstable conditions which were accompanied by large and l rapid increases in temperature involved the gas phase reaction CO + O2 and. J occurred only for h values- below 10 " cal /cm-sec *C below 750*C. Temperature

-rises associated with the formation of CO2 from C + O2 were smaller than those ]

due to CO + O2 and decreased with time. They too occurred at h values below i 10 4 cal /cm-sec *C.

In a channel which was' held above 650*C there was an entrance region-run-ning some distance down.the channel which was always cooled. A position was  !

reached where the heat lost to the flowing gas and the heat lost by radial condu'ction through the graphite was exactly equal to the heat generated by the oxidation of'the graphite and of the CO. This position remained essentially constant with time. Beyond this point rapid oxidation of graphite occurred with the accompaniment of a flame (due to the CO-0 gas phase reaction). Under

~

conditions of burning, the phenomena were essentially. independent of the bulk graphite chemical' reactivity. Rate controlling reactions' during burning were '

determined by surface mass transport of reactants and products. i The experiments were used to develop an equation which expressed the length of channel that can be cooled as a function of temperature, flow rate (heat transfer coefficient), diameter and reactivity of the graphite. It was found that the maximum temperature at which thermal equilibrium (between heat generated by graphite oxidation and heat removed by the air stream) will occur in a channel can be predicted from the heat transfer coefficient, the energy of. activation and a single value of the graphite reactivity at any tempera-ture. Above this maximum temperature the total length of channel is unstable and graphite will burn. The studies show that the bounding conditions needed to' initiate burning are:

1. Graphite must be heated to at least 650*C.

2. This temperature must be maintained either by the heat of combustion or some outside energy source.

4

1

.)

1

'l

• ]

.n

. 3. - There must.be an adequate supply of oxidant (air.or oxygen)..

.i s

4. The. gaseous source of oxidant must flow at a rate' capable of removing

. gaseous reaction products- without excessive cooling of the graphite surface.

5. .In the case of-a channel cooled by air these conditions can be met.

.However,'where such a configuration is'not' built linto the structure

.'it is necessary,for a geometry to develop to maintain an adequate flow of oxidant and removal of the combustion products from the reacting surface. Otherwise, the reaction ceases.

.+ To illustrate how difficult it.is to." burn" graphite th excerpted f rom a report by Woodruff and Bogert . [ Reich,1986}g . These testsfollowing was cwere carried out in.a search for methods for extending the useful life of the N-Reactor. (The following is quoted directly 'from text of the report.):

" Dry Burning Test: 'Tliree pieces of graphite were weighed and stacked together as indicated in Figure 1. Grafoil and carbon felt were placed under and around the blocks. This wrapping material was used as thermal insulation to hold heat in the blocks, and as a buf- .

fer to prevent catalysis by. contact with the stainless steel tank  !

used to contain the test. Thermocouple were placed at 5' locations in the blocks to monitor' temperatures through the test. . Two oxy-acetylengtorchesdeliveringacombinedheat output of approximately 2.7 x 10 BTU /Hr. through rosebud nozzles were positioned about 2 inches above the graphite'. Oxygen flow rates to the torches were .

' , orc h tot ationu

$ % IU' ic #1 i

."/

-/ Af e.s ,, 9

/

.. _ . / '

/

%l Il* Jl Figure 1. Graphite burn configuration.

3. The receipt of this report from Mr. W. Quapp of United Nuclear Corporation, Inc. is gratefully acknowledged.

l 5

adjusted to produce nearly neutral flames. Still photographs and a video tape were made to visually record the test.

"Five minutes after ignition, the surface of the top block in regions directly below the torches was glowing yellow-white at an estimated temperature of 1832*F (1000*C).

" Twenty-five minutes after ignition, the lower blocks were also red over their entire surface. Block temperatures continued to rise at rates of a few degrees centigrade per minute until fuel to the torch over the thermocouple was shut off 57 minutes into the test.

The peak recorded temperature for thermocouple #1 was 2300*F (1260*C). Other temperatures appear in Table 1. Using an optical pyrometer, the blocks maximum surface temperature was estimated to

{ be approximately 3000*F (1650*C) directly under the torches.

TABLE 1: PEAK TEMPERATURE DATA The rmocouple Dry Test TC #1 2300*F (1260*C)

TC #2 2140*F (ll70*C)

TC #3 1890*F (1030*C)

TC #4 1615'F ( 880*C)

TC #5 1515'F ( 825'C)

FUEL AND BLOCK WEIGHT DATA Acetylene Consumed: 13.0 lb (2.69 x 10 5BTU)

Oxygen Consumed: 20.0 lb Total Block Weight Loss: 1.314 lb BTU /lb Weight Loss: 2.05 x 10 5 "With the acetylene to one torch shut off, oxygen was being blown onto the hot block at a rate of approximately 0.16 pounds per minute (1.9 cfm). The oxygen alone could not sustain a reaction with the graphite and the region below the nozzle cooled quickly. Sixty-five minutes after starting the test, both torches were r . moved, and the blocks were allowed to cool. When cool, the blockr were reweighed to determine weight loss.

"In the dry burn test, small craters were formed directly beneath each of the two torches. They are approximately 2 inches in diam-eter and their bottoms average 3/8 inch below the original graphite level. These craters account for only a small portion of the total weight loss. The remainder of the weight loso is the result of oxidation on the blocks surfaces that were exposed to air.

"In the interface areas where one block rested on top of or beside another, there are no visible signs of oxidation.

6

~

" DISCUSSION: i i

There is a common perception taken from our experiences with coal and charcoal that when a mass of these' fuels' achieves a glowing red condition a self-sustaining combustion is underway. Transferring this perception to graphite has led to repeated references-to " burning" graphite when in fact a self-sustaining reaction was not-in progress. The test sequences described in these tests demonstrate how difficult it'can be to achieve conditions for self-sustained combustion of graphite." -i I

3. STORED ENERGY 3.1 Summary j lA review was made of existing literature and knowledge'on stored energy j accumulation and releases in order to assess what role, if any, a stored- j energy release can have in initiating or contributing to hypothetical grsphite i burning scenarios in research reactors. j

]

From analyses of existing information it is concluded that only' stored )

energy accumulations.and releases below the burning temperature (650*C) are l pertinent. A review of existing information on stored energy has shown that -l stored energy releases do not occur spontaneously but are initiated by mecha- )

nisms that raise'the graphite temperature above the irradiation temperature.  !

Moreover,.the maximum releasable graphite stored energy that could be produced .

by combustion from any reactor containing graphite is a very small fraction of d the energy produced if graphite burning were to occur. I Conclusions from these analyses are that the potential to initiate or maintain a graphite burning incident is essentially independent of the stored energy in the graphite. .

3.2 Wigner Energy -- Its Generation and Buildup From the earliest days of the Manhattan Project, E. P.'Wigner [Wigner, 1946] recognized th.t if graphite was used as a moderator in nuclear reactors used to produce plutonium, "the collision of neutrons with the atoms of any j substanco placed into the pile (reactor) will cause displacement of these atoms. ... The matter has great scientific interest because pile irradia-tions should permit the artificial formation of displacements in definito num-bers and a study of the effect of these on thermal and electrical conductivity, tensile strength, ductility, etc. as demanded by theory."

The theoretical prediction has been amplified by the work of F. Seitz

[Seitz, 1958), the experimental work of Burton (Burton, 1956] and many others. One of the many observed effects of neutron bombardment of graphite in slowing down the fast neutrons produced in fission to thernal energies is the production of large numbers of displaced carbon atoms and vacancies. Many of these displaced atoms of carbon come to rest in between the planes which l constitute the structure of the graphite. The rest of the displaced atoms may j 7

either wander back to their equivalent positions in the lattice, or to crystal boundaries. This introduction of new atoms between the planes increases the spacing between the original planes. This can be measured by the increase in the dimensions of the C-axis. This change in C-axis dimensions is reflected by a change in the gross dimensions of the graphite specimen. Distortion of the lattice results in an increased energy of the overall system.' This l

increase in lattice energy is called the Wigner energy or stored energy. I It was recognized that these two effects, dimensional changes and Wigner energy, might prove to be troublesome in the operation of graphite moderated reactors. The total stored energy of the graphite increases with neutron exposure and is a function of the temperature of the exposure, and the energy distribution of the neutrons. The stored energy that can be released is spread over a range of temperatures. It has been shown that when graphite irradiated at moderate temperatures (less than 100*C) is heated above the irradiation temperature some of the stored energy is released as heat when the temperature of the test specimen is raised some 50-100*C above the irradiation temperature. Increases in exposure to fast neutrons increases the total energy stored. Eventually the stored energy which is releasable up to a tem-perature of 700*C saturates even though the total stored energy can continue to accumulate with increasing exposure. Total stored energy can be determined by combustion of the sample. . Stored energy releases also can be measured by differential thermal analysis where the difference in behavior of an unitra-disted specimen and an irradiated specimen are compared in a calorimeter by increasing the temperature in a pre-determined manner.

Broad experimental programs were undertaken during the Manhattan Proj ect. This work was followed by basic and applied programs in the late forties and fifties. Much of this early work was presented at the first Geneva Conference on The Peaceful Uses of Atomic Energy held in Geneva in 1954

[ Woods, 1956]. By the early fifties it was known that large dimensional expansions take place in reactor graphite structures and that stored energy accumulated. The British decided to control the stored energy of the Wind-scale reactor by heating up the graphite moderator (annealing). This process was carried out at' regular intervals. The Brookhaven graphite gas cooled research reactor (BGRR) was annealed to reduce the dimensional changes (growth) caused by irradiation and to release the stored energy. Prior to carrying out this work considerable experimental work was carried out to determine the rate of growth and the rate of buildup of stored energy as a function of irradiation exposure and temperature of exposure.

A large body of complex literature exists on the accumulation of stored energy at different irradiation temperatures and fast neutron exposures. Much of this work is not pertinent to the problem of how much stored energy can be released below a given temperature. In this report we have analyzed existing information in order to identify the factors needed to determine the quantity of stored energy that can be released below the bounding temperature (650*C) needed to initiate graphite burning.

8

1 i

j i

l i

The energy required to raise graphite from some initial temperature >To to some higher temperature, T, is the enthalpy, which is calculated from the integral of the specific heat at constant pressure over the temperature inter-val of interest [ Schick, 1966}. Consider a starting temperature of 30*C, and  ;

a final temperature of 650*C, the minimum temperature required for graphite to l burn. The energy required to go from 30'C to 650*C-is 202 calories per gram.

Energies required to reach 650*C from various starting temperatures are shown below:

Starting Final Temperature Temperature Enthalpy (C) (C) (cal /g) 30 650 202 650 195 l 50 150 650 175  !

200 650 160 Observed stored energy accumulation is non-linear, and depends upon irra-diation temperatures, levels of exposures to fast neutron fluxes, neutron energy spectra, spatial distribution of the flux, properties of specific ~'

graphites, geometries of individual reactors, etc.

At low temperatures and at low exposures, the displaced carbon atoms move into interstitial positions [Kircher, 1964; Schweitzer, 1962a], and the re-sulting forces between these displaced atoms and planes in the lattice force  ;

the lattice apart, leading to expansions that are initially linear with fast neutron exposure. As neutron irradiation continues, the number of simple defects increases until they begin interacting and result in the formation of larger complexes [Schweitzer, 1964b]. Similarly, initial stored energy in-creases are linear with neutron irradiation, until a dose is eventually reached at which the stored energy tends to saturete.

Figure 2a shows that a sample exposed for 5000 mwd /AT" at 30'C has e total stored energy of 620 cal /g, but only 275 cal /g is released in annealing temperatures up to 800*C [Davidson, 1959, in Nightingale, 1962]. Similar results for other exposures and annealing temperatures up to 400*C are shown in Figure 2b [Kinchin, 1956]. >

Results of calorimetric and heating experiments show that stored energy will not be released until the annealing temperature exceeds the irradiation temperature by some specific amount. This threshold temperature increase has been reported between 50*C to 100*C above irradiation temperatures [Kircher, 1964; Cottrell, 1958; Woods, 1956].

4. Units of neutron dosage are reported in different units by different authors. Forthisreportwegeneralgusetheconversiononemegawatt-day per adjacent ton [ mwd /AT] = 3.9 x 10 thermal neutrons per square centi-l meter [nyt(th)). FordatafromKinchin[Kighin,nyt(th).

1956] and Bridge For these data we I [ Bridge, 1962), we use 1 mwd /AT = 5.56 x 10

{ were unable to obtain conversion factors for fast neutron flux.

i 9 f i

{

(

[. -

.6'0- - - -- -- -- -

4___-~~"'~"

600 -

/

550 - /

/

/

^ /

'h b0 500 -

/

/

/

v /

v

. 450 .- /

/

h &

/

[

v 400 -

.Y r

g i

!j t /

.o . 350 - /

1. /

6') / ,

/ l IP-"-

~~~~ "~ -- f1 300~" </

,, s

• f', -

250 - ..

200 ----

T- - r-- - - r- - - - , - ---- -- T -- T r--~~--

1 2 3 4 5 (Thousondn)

Dposure (t&d/At)

D F!eleased + Total Figure 2a. Total vs released stored energy [ Nightingale, 1958],

Irradiation = 30*C, Tanneal = 800*C.

10

150 --: - -

f

/

I E;0 -

1. j/

170 - /

160 -

I t.0 - ,s

/

140 - ,/

30 130 -

/-

o x

120 -

110 -

l'/ .. -a 100 - / f M

,b u

{

c'O E.0 -

/ ![

j/

E) 7 0 -- 23 4/

0 EO -

5.0 -

" //

/

4--- r 20 - i 7- - r-- i~~----

0 200 400 600 bposure (MNd/At) 0 Released + Total Figure 2b. Total vs released stored energy [Kinchin, 1956].

Irradiation = 30*C, Tanneal = 400'C.

11 1

L. _ ~A. _ _ _ _ _ .~

At irradiation temperatures above about 150*C the rate of accumulation of total stored energy is very low [ Bridge, 1962; Neubert, 1957; Nightingale, 1958, 1962). At about 30*C and at low total exposures, the total stored energy increases with exposure at a near linear rate of about 40 i 10 cal /g per 100 mwd /AT. As the~ exposure continues, the rate of accumulation of total stored energy decreases, and the stored energy that can be released below the minimum bounding temperature to initiate graphite burning (i.e. 650*C) satu-rates and then appears to decrease. An upper bound on the stored energy that can be released to 700*C can be found from existing data. Figure 3 shows this as about 120 cal /g for an irradiation in the temperaturggrange of 35-70*C at an exposure of 930 mwd /AT (equivalent to about 3.6 x 10 nyt (thermal)

[Neubert, 1957]. (This is about 1/60 the heat of combustion of graphite.)

3.3 Stored Energy Releases A great deal of evidence exists demonstrating that stored energy is released through a series of complex and interactive thermally activated pro-ceases. Relesse of stored energy is generically attributed to the recombina-tion of various interstitial defects with vacancies, or the annealing of the interstitial to edge atoms or other voids in the graphite crystal. Removal of interstitial species from between the graphite planes reduces the stored energy, lattice parameter increases, and other forms of radiation damage.

Existing views of irradiation changes in graphite support the claim that

. irradiation produces different defects that thermally anneal with different activation energies (i.e. different energies are required to initiate the releases). The type of defects and their respective quantities depend upon the magnitude of the irradiation, the temperature of the irradiation, and whether or not the graphite was subjected to anneals between irradiations. In the latter cases [Schweitzer, 1964a, 1964b} data show that defects interact with each other and that changes that occur during such anneals are very different from the changes observed after a single irradiation.

At any given temperature the stored energy that can be released with time can result from several different processes whose rates decrease as the de-fects anneal. No evidence exists that stored energy releases are spontane-ous. The observation that a 50-100*C increase above the irradiation temper-ature is required to observe finite release rates is consistent with the exponential changes in release rates with reciprocal temperature associated with thermally activated processes.

From our review of the literature on Wigner energy we have compiled data on releaseable stored energy at various combinations of exposures, and irradi-ation and annealing temperatures and have plotted this information in Figure 4 and Figure 5. In both figures a curve is shown of the amount of energy re-J quired for a sample of carbon to go from 100*C to the particular temperature i of interest (i.e., the enthalpy between 100*C and some temperature T). Also shown are curves entitled " envelope of releases," which simply delineate an upper bound on stored energy releases found in the technical literature. Data above the enthalpy curve indicate a region where a sample in an adiabatic environment would heat up to the upper intersection of the enthalpy curve and the envelope of releases. Figure 4 shows that the maximum releasable stored 12

\

L _____ m_

130 - - -

+ ,

120 -

'y I 110 - '-

u 100 - /

5 8 4 0 --

v -

vu

'f: -

80 -

8 6 70.-

t:

s /

e 60 -

il b 50 -

u E

40 -

.A  %  %

3 0  % ._

30 -

20 -

cf

/[

10 - -- , , 1 i r--- - r -r - r- - -- - r-- - - -

0 0.2 0.4 0.6 0.3 1 (Thmn.onds)

&posure (IMd/At)

Figure 3. Stored energy released [Neubert, 1957],

Irradiation = 30-70*c.

O Tanneal - 250*c + T,one,1 - 500*c oTanne,1 - 700*c 13

4 o

Eo -

cncu .

N d

u ra n atsu of .. t.a w-wise tav.t.p. f e.e..... ,4

\ '

~n -

U) '#

L ' O

.e Cc LLJ G -

,' O T

/

/E O g

O

@ p M Z U O

g /g U O

-e MO O ge O e

$ O B B e m'0 T /0

's 0?Mh g ' c 00 U O i

/

f'". -

U B

O o .A O  ! , "

O S00 1000 Annealing Temperature deg. C. est es-ce-es Figure 4. Cumulative energy release; exposure of 500 mwd /AT or less.

l

[

1 14 1

r

~ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

t i

j l

s l Eo .1 cncu .

N _l

- 1 -

m U

. 6.i ,.t ...... u-.n ..

n en ....i.,..e........

c.  !

-O l e ***

Co LD G -

\,

v

,~,,& O \

a

/$ a 8

m a G - ',O O D

'O e 0 0

& /

O  !

O $ -

D' N C =

0 500 1000 )

Annealing Temperature deg. C. j BNL CS-CB-04 1 l

Figure 5. Cumulative energy release; irradiations at 70*C and above.

.]

l i

15

energy in irradiations below 500 mwd /AT (irrespective of irradiation temper-atures) is sufficient to raise the carbon temperature from 100'C to about 450*C. Figure 5illustratestheamountofreleasablestoredggergyatexpg-sures in the range 16-5700 mwd /AT (equivalent to about 9 x 10 - 3.2 x 10 nyt (thermal)} and irradiation temperatures greater than 70*C. Figure 5 indi-cates that irradiations at 70'C or above (irrespective of exposures) have resulted in temperature rises f rom 100*C to no more than about 300'C.

3.4 Calculational Approaches Buildup of stored energy in graphite is a result of the formation o'f a large number of ill-defined defects each of which can be associated with a stored energy release of unknown specific magnitude, unknown activation energy and unknown temperature range. Since the sum total of these defects deter-mines the accumulated stored energy and since this in turn depends upon the level of the irradiation, the temperature of the irradiation, and the history of irradiations and anneals, BNL does not believe that any of the calcula-tional approaches involved in the past UCLA license renewal hearings can be defended. Other calculational approaches such as the bounding method used by Spinrad [Spinrad, 1986) rely heavily on a number of empirical correlations which involve appreciable uncertainties. These include determining the frac-tion of energy transferred to carbon atoms by neutron moderation that goes into atomic displacement energy. This must be combined with the fraction of stored energy that self-anneals at various irradiation temperatures. Aside l f rom the direct dependence of this method on measurements showing a great deal of uncertainty, these models cannot account for the non-linear buildup of stored energy, the saturation effects, the temperature dependence of releases, l the exposure dependence of releases and the complex consequences of irradiations combined with several anneals.

After review and analyses of existing information on estimating stored energy pertinent to graphite burning scenarios, we believe the approach pro-posed in this report.is consistent with existing data and is acceptable for safety assessments. Total stored energy accumulation has no overall correla-tion with the stored energy that can be released at temperatures below 650*C.

The stored energy that can be released below this temperature saturates at a value that can be bounded from existing knowledge. The dependence of the sat-uration value of the stored energy released on irradiation temperature can also be bounded from existing data. This approach allows for safety analyses irrespective of the uncertainties in total exposure and total accumulated stored energy.

We emphasize again, that the adiaba' tic assumption that all the released stored energy goes into heating the graphite is bounding but unrealistic.

Under adiabatic conditions where the decay heat is transfered from the nuclear fuel to the graphite, steady increases in the graphite temperature could occur I that are much larger than those due to the hypothetical single spike from the release of stored energy.

l 16

Because heating graphite to at least 650*C is necessary but not suffi-cient to initiate burning, the conclusion of these analyses is that the poten-tial to initiate or maintain a graphite burning incident is essentially independent of the stored energy in the graphite.

4 THE CHERNOBYL ACCIDENT BNL has examined recent studies analyzing the Chernobyl accident to determine if any additional information on graphite burning has been devel-oped. The accident summary described here has been taken from Kouts [Kouts, 1986):

On April 25-26, 1986, "The accident took place during an experiment con-ducted at the start of a normal reactor shutdown scheduled for routine main-tenance. The operating staff had prepared to do what they considered to be a test of some electrical control equipment that was meant to serve a safety purpose."

The objective of the experiment was to see whether the coastdown of the turbine of the nuclear reactor system would supply power long enough to allow for start-up of the standby diesels. The test required that the reactor power had to be reduced to a level (700 MW[th]) just above the value which was known to be low enough to become unstable. In approaching this level, a series of unfortunate operations were carried out in which many safety systems were intentionally by passed for unknown reasons. In one of these operations, the power level began to decrease rapidly, and fell to an estimated 30 MW(th) before the. operator could halt the drop by control rod motion. After the operator had stopped the rapid drop, he managed to achieve some measure of control at 200 MW(th). At this point, the number of control rods in the reactor were far ler.s than regulations permitted.

Further manipulation of the cooling and feed-water systems resulted in other problems eventually leading to a rapid power surge estimated at 300,000 MW(th). Six violations of safety requirements, eventually resulted in a steam explosion that blew off the top of the reactor. The explosion disintegrated the fuel elements, fragmented the graphite, and exposed the graphite and fuel to air. The force of the steam explosion blew pieces of the core and fuel through the roof of the reactor building. A second explosion lifted the cover plate shearing the fuel channels releasing primary system steam pressure to the exterior. Falling hot projectiles ignited asphalt roofing materials causing extensive fires.

Graphite burned for many days supported by asphalt fires and decay heat from the buried fuel. Soviet teams tried to put out the fires by dropping massive amounts of materials from helicopters. The attempts were not success-ful presumably because the dropped material insulated the hot debris. Even-tually liquid nitrogen was used to cool and inert the burning debris. No evidence exists that stored energy in graphite played any role in this accident.

17

5. " ACCIDENT AT WINDSCALE NO. 1 PILE ON 10th 0F OCTOBER, 1957"5 Windscale Pile No. 1, was a graphite moderated, air cooled reactor, fueled by natural uranium metal encased in sealed aluminum cans to prevent the uranium from reacting with the components of the air and to contain the gaseous and solid fission products produced in fission. In 1952, the Wigner

'(stored) energy was found to be releasing on a shutdown of this reactor be-cause the graphite temperature rose above its normal operating temperature when the forced cooling was reduced on reactor shutdown.

l To avoid a recurrence of such an incident the Windscale piles were there-fore regularly heated above their normal operating temperature to bring about a controlled release of the Wigner energy. The accident developed during the course of one of these controlled releases on October 7th, the day of the start of the Wigner release. Nuclear heating was used, but with cooling essentially shut down to increase the temperature of the graphite above its normal operating temperature. In this instance the first nuclear heating was thought to have inadequately heated enough of the core graphite. To bring about a more uniform temperature throughout the graphite structure the reactor was " pulsed again" but according to the investigators of the accident the rate of increase of nuclear energy input was too rapid, and caused the uranium cladding to break and expose uranium to air. Uranium is an extremely reactive

. metal. It reacts readily with oxygen, nitrogen, and hydrogen with the release of a large amount of heat. There is also the possibility that the initiating event in this accident may have been the failure of some aluminum clad magnesium lithium cartridges which were in the reactor at the time.

The operator of the reactor was not aware of the cladding failure due to an inadequate number of thermocouple and inadequate radioactive sensing de-vices at the outlet of the cooling channels. Radioactivity sensing was done at a point some distance from the channel. Since the anneal procedure re-quired allowing the heat to be conducted through the graphite structure by maintaining the cooling shutdown for a day or longer the failed slugs heated adjacent ones and they too f ailed. Finally after a couple of days during which the temperatures of portions of the reactor were noted to be rising, efforts were made to cool the reactor by admitting air. These efforts failed to cool the hot sections of the reactor. On October 10th a plug in the charg-ing wall of the reactor was removed. The uranium cartridges in the four chan-nels which could be viewed were at red heat. Water was finally used to cool down the reactor after other efforts failed.

There is no evidence that stored energy releases initiated or played a significant role in the evolution of the Windscale accident.

5. Title of a report presented to Parliament by the Prime Minister by command of Her Majesty, November 1957. Other sources on this accident - " Final Report of the [ Alexander Fleck] Committee Appointed by the Prime Minister to Make a Technical Evaluation of Information Relating to the Design and Operation of the Windscale Piles and to Review the Factors Involved in the Controlled Release of Wigner Energy." Presented to Parliament by the Prime Minister by command of Her Majesty, July 1958.

18

Y

6. U.S. RESEARCH REACTORS 6.1 Criteria for Stored Energy in Graphite Analyses of existing infpynttion indicate that _the conditions associated with the initiation and maintenance of graphite burnteg scenarios are essen-tially independent of the stored energy in the graphite, irrespective of its value.

As shown in Section 3, if the irradiation temperature of the graphite was 70*C or above, the maximum stored energy releasable below 650*C for any level of irradiation cannot raise the graphite temperature to the minimum value which would be required for;1aitiating a self-sustained burning reaction. For graphiteirradiaggontegperaturesbelow70*Ctotalexposuresofabout500 mwd /AT (3.5 x 10 nyt) are required to continue to heat the graphite from about 100*C to 650*C if an e'xternal heat source can raise the graphite from its ambient temperature to 100*C. We have assumed that if the stored energy in the graphite cannot be bounded, any process that heats the graphite to 100*C should be treated as if it heats the graphite to at least 650*C.

The analyses and conclusions on stored energy releases and graphite burn-ing conditions described above provide a meaningful method of categorizing nuclear reactors with respect'to stored energy releases below 650*C (the threshold temperature for graphite burning) as follows:

(1) Any reactor containing graphite in which tha lowest irradiation is 70*C or higher, can be excluded from stored energy safety concerns.

(2) Any reactor in which the graphite is irradiated at temperatures below70*Cbuthasreceivedatotalgastneutronexposurethatis nyt) can be excluded from much less than 500 mwd /AT (3.5 x 10 stored energy safety concerns.

(3) Thosereactorswhichgavegraphitethathasreceivedmorethanabout nyt) of fast neutron irradiation below 70*C 500 mwd /AT /3.5 x 10 without thermal anneals or subsequent reirradiation at higher tem-peratures would require detailed heat transfer analyses to determine if the graphite were capable of reaching 650*C following an event that raised its ambient temperature to about 100*C. It is important to recognize that even under conditions that allow the graphite to reach 650*C or above, this is nut sufficient to initiate burning.

In order to separate reactors into these categories, it is necessary to determine only the total fast neutron exposure reached by graphites irradiated at temperatures below 70*C.

6. Estimated fast neutron fluggce was converted to HWd/AT using the conversion factor: 7 x 10 nyt = 1 mwd /AT.

19

I l

6.2 ' Stored Energy in Graphite The significance of stored energy for U.S. research reactors under NRC's licensing authority was assessed in light of criteria in Section 6.1. The information used in the assessment was obtained from Safety Analysis Reports (SAR's) and other readily available data representing the main types of these reactors. The objective of the assessment was to determine if . stored energy releases can initiate or significantly contribute to the evolution of graphite burning accidents, and if graphite would play a role in previously reviewed potential accident scenarios.

For the purpose of overall screening of the research reactors, rough estimates of' the graphite exposure were made. Only operating research reac-tors containing graphite and licensed to oparate at powers greater than 100 W were included in the survey.

For TRIGA reactors GA Technologies publication GA-4361 [ West, 1963] was used to derive a maximum neutron fast flux (above 0.1 MeV) in the side reflec-tor. In addition, an analysis performed by GA Technologies [GA Technologies, 1987] shows, for three out of the four locations where graphite is found in the reactor (i.e., graphite reflectors in the top and bottom of the fuel ele-ments and in the radial graphite reflector) that stored energy would not be sufficient to raise the graphite temperature to 650*C. The reason for this is that these locations satisfy, in essence, either criterion 1 or 2 in Section 6.1. The dummy elements, which are not in suery TRIGA reactor, were found to have enough stored energy such that the graphite could reach 650*C if the tem-perature of the graphite is elevated to at least 120*C. However, no normal or abnormal operation would produce an initiation temperature of 120*C. Even if this temperature were reached, water cooling of the aluminum clad surrounding the graphite would preserve the integrity uf the clad and prevent exposure of the graphite. Additional discusLion on the significance of stored energy in TRIGA reactors is found in Section 6.3.

The remaining research reactors were reviewed to assess their stored energy accumulation. These reactors are listed in Table 2. Values of fast flux at the graphite were obtained from the licensees. Where licensee data were not availabic, peak fast neutron flux data for the reactor core compiled by the American Nuclear Society [ Burn, 1983] were used, keeping in mind that the neutron flux that could be expected at a graphite reflector located close to the core would be about a factor of 2 to 10 lower. In the case of MTR reactors, the published data on power and fast flux in the ANS compilation were correlated, removing an outlier, to arrive at a flux-to power conversion factor.

The total neutron exposure in some reactors was available from the licen-sees in terms of mwd of operation. In those few cases where these data were not directly available they were estimated based on data of first full power operation and reported equivalent days of full power operation for 1983.

From the survey (see Table 2) it appears that four reactors (General Electric, North Carolina State University, University of Lowell, and University of Virginia) have stored energy greater than 500 mwd. However, the 20

(

a presence of stored energy above the 500 mwd threshold in parte of the reactor '

graphite is not by itself taken as a safety concern, as discussed in greater detail in the preceding sections of this report and in Section 6.3. -

1' '

l l

Table 2. Stored energy calculations in graphite for non-TRIGA

• Research Reactors ,

'9 Fast Irradiated Power Duty. Total, Flux Dose Temperature Reactor identifier Type kW Year h/) r.  : mwd n/casq/s nyt mwd /AT 'C m

General Electric Co. Spec. 1.00E+02 ' 100.0i 5.00E+11 4.3E+19 617 Westinghouse Electric Spec. 1.00E+01 -

3.00E+11 - -

N. Catalina State U. Pulstar 1.00E+03 403.0 1.30E+12 4.5E+19 647 Georgia inst. Tech. PffR D2 0 5.00E+03 708.0 5.00E+10 6.lE+17 9 H.I.T. MTR D2 0 5.00E+03

• 160.00 National Bureau Stds. HTR D2 0 2.00E+04 52013.0 2.00E+09 4.5E+17 6 Cintichem HTR 5.00E+03 1961 7600 42250.0 2.80E+08 2.0E+17 3 Ohio State U. MTR I.00E+01 1961 20D 2.2 2.60E+11 4.9E+18 70

'Purdue U. MTR 1.00E+0! - - -

Rhode Island MTR 2.00E+03

• 148.00 U. Lowell NTR 1.00E+03 140.0 5.00E+12 6.0E+19 864 U. Missouri (Columbia) er.t 1.00E+04

• 100.00 U. Missouri (Rolla) hit. 2.00E+02 1962 62 12.9 4.86E+12 2.7E+19 387 - [0.

U. Virginia MTR 2.00E+01 1702.0 3.50E+12 2.6E+20 3676 Worcester Poly. MTR 1.00E+01 1960 100 - - -

lowa State U. Argonaut 1.00E+0! - - -

U. Florida Argonaut 1.00E+02 1959 213 24.9 1.30E+11 2.8E+18 40 U. Washington Argonaut 1.0Z42 1967 100 8.3 1.30E+11 9.4E+17 13 NOTESt Yaat - Year of init.41 operation at (at least) one half of full power. -

Duty w Number c.C h ere of operation per year, reported for 198J.

Total - Total MW Jafs of operation to date. -

Fast Flux - Peak f ast neutron flux in the core or graphite ".uflector.

Dose , Product af years of operation, duty. and frist flux. Q resents maximum possible dose to any Artphite.

mwd /AT PEquivalent dose in mwd /AT. Factor 7e16 nvt = 1 mwd /AT.

Irra/f ated Temperature - Normal maximum operat.ing ten:perature of exposed graphite.

t - De graphite in the General Electric Co. Teactor wa? annealed in 1976 when the reactor fuel container was repiated for a leth in the weld area. T4tal mwd since that anneal is 44 mwd.

- - Not significant because of low Muer. /

• - Since irradiated temperature is above 70'c stored energy was not estimated.

>/ .

21

)

4

6.3 Graphite Burning Research reactors which use graphite in or near their cores and are licensed to operate at power levels greater than 100 watts (thermal) were categorized with respect to:

1. Quantity and location of graphite in and near the core,

2. Geometry,

3. Accident conditions considered by the NRC staff in the licensing bases of the reactors,

4. Fast neutron flux,

5. Normal operating sequence, and

6. Graphite irradiation temperatures.

Although present information indicates a great deal of variation in fast flux, operating sequences and graphite temperatures for reactors within a given type, our analyses of existing information shows that these factors are not significant to those factors related to graphite burning. In scenarios that postulate graphite burning, the quantity of graphite that can burn is an important factor in determining the consequences of burning. However, the credibility associated with a postulated burning accident depends upon the existence of all of the conditions necessary for graphite burning, including the capability to heat the graphite to temperatures above 650*C and maintain-ing this temperature in the presence of much cooler flowing air. In any given reactor, this not only depends upon the original geometry, but also upon the geometry resulting from the accident that allowed the graphite to heat up in the presence of air.

In assessing the potential for graphite burning in the research reactors licensed by NRC, consideration has been given to conditions during normal operation and conditions that may exist following a LOCA. The LOCA was selected as having conditions most likely to result in high temperatures in the fuel and graphite and, therafore, most likely to release the graphite stored energy and to result in conditions with the potential for graphite burning.

All TRICA reactors operate in water pools. Since graphite does not burn under water, all accidents in which the core and graphite reflector remain submerged will not be subject to graphite burning. GA Technologies [GA Tech-nologies, 1987] has estimated in a response submitted to the NRC on January 28, 1987 that aluminum clad graphite in dummy elements could, under loss of coolant conditions for some of the reactors, reach 770'C and result in melting of the cladding. GA Technologies claims that the hot graphite at 770*C cannot burn because the specific requirements for graphite burning cannot be met since the graphite radiates its energy rapidly and quickly cools to the ambi-ent air temperature. Our assessment of this claim is based on the experiments discussed in Section 2. That is, radiant heat losses to the cooler 22

--s- . - - . ,_ _ _ _ - . _ _ _ _ _ _ . _ _ _ . _ _ _ _ , _ _ _ _ _ _ _ _ _ _ _ _ ___ _ _ _ _ _ _ , _ _ _ _

surrounding structures coupled with convective cooling by the cooler air surrounding the graphite could cool the graphite and preclude its burning.

Analysis of a LOCA in an Argonaut reactor predicts peak fuel temperatures of about 120*C [Chen, 1981]. This, coupled with the insignificant stored energy of the graphite suggests no change in the conclusions already reached during the evaluation related to license renewal. The likelihood of graphite fires was reviewed in NUREG/CR-2079 [Hawley,1981).

Reactors with MTR fuel and the PULSTAR reactor have their fuel located in i a water pool. In accidents in which the water level in the pool remains above l the core top the graphite could not burn. During a LOCA the maximum fuel I

plate surface temperature for any of these reactors is 500*C and for many it is much lower except for two cases where it has been calculated to reach 510*C and 582*C. In these two cases, however, emergency core cooling spray systems are activated during a LOCA and the actual fuel temperature would be much lower than the calculated fuel temperatures [NUREG 0928, Section 14.1.3, p.

14-3; NUREC 1059, Section 14.1, p. 14-2]. The stored energy is unlikely to raise the temperature to 650*C under non-adiabatic conditions that exist.

Also, the graphite will not burn if the conditions to austain burning are not present. If the fuel plate surface temperature is always less than 500*C, the heat losses from the graphite by radiation to the cooler structures of the pool coupled with convective cooling by the cooler air in contact with the graphite should preclude conditions necessary for graphite burning.

The Safety Analysis Report [GE, 1981] for the General Electric Nuclear Test Reactor was reviewed f or potential impacts of graphite stored energy on the safety analysis of the reactor. The loss-of-coolant accident analysis in the report predicts maximum fuel temperatures of 300-320*C depending on assumptions about peaking factors. Such temperatures pose no danger to the aluminum clad fuel. However, there is no indication that the loss in thermal conductivity of irradiated graphite, or the releasable stored energy in the irradiated graphite, have been included in the thermal analysis. The reduced thermal conductivity could in principle lead to higher local graphite temper-atures which in turn could result in some stored energy release. Since in this postulated accident the graphite acts as an effective heat sink, the potentially higher graphite temperatures could have an impact on maximum fuel temperatures. Without a numerical analysis accounting for the space depend-ence of the thermal conductivity, for the time dependence of the rate of energy release, and for the concomitant changes in thermal conductivity of the graphite, it is not possible to estimate the impact of the irradiated graphite on the course of this postulated accident. However, in connection with Amend-ment No. 9 to the General Electric license, the NRC staff evaluated the conse-quences of a postulated maximum hypothetical accident which assumed, nonmech-anistically, that all of the fuel in the core melted (NRC Safety Evaluation, Section 3.4, dated June 30, 1969). This scenario encompasses any potential impact of degraded thermal properties of irradiated graphite on the conse-quences of a loss-of-coolant accident. The resulting radiological doses to an individual at the site boundary under the extremely conservative assumptions of the analysis were well below the allowable 10 CFR Part 100 guidelines.

23

l l The MTR-D 2 0 reactors have the graphite located away from the core, in a cavity.with restricted air interchange. In the analysis of loss-of-coolant ,

scenarios of the SAR for the National Bureau of Standards reactor [NRC, 1983c], NRC staff agreed that a LOCA will not result in melting of the fuel.

Under such conditions it appears implausible that the graphite could be subjected to temperatures compatible with burning.

( . 7. FORT ST. VRAIN - GRAPHITE STORED ENERGY Fort St. Vrain operates at temperatures that preclude accumulation of stored energy. There are no known ' problems associated with stored energy in graphite for operating temperatures associated with HTGR's.

8.

SUMMARY

8.1 Graphite Burning The factors needed to determine whether or not graphite can burn in air are the graphite temperature, the air temperature, the air flow rates, and the ratio of heat lost by all possible mechanisms to the heat produced by the burning reactions [Schweitzer, 1962a-f]. In the absence of adequate air flow,

, graphite will not burn at any temperature. Rapid graphite oxidation in air qr removes oxygen and produces CO2 and CO which, along with the residual nitrogen, suffocate the reaction causing the gra'phite to cool through unavoidable heat loss mechanisms. Self-suatained rapid graphite oxidation cannot occur unless a geometry is maintained that allows the gaseous reaction products to be removed from the surf ace of the graphite and be replaced by fresh reactant. This necessary gas flow of incoming reactant and outgoing products is intrinsically associated with a heat transfer mechanism. When the incoming air is lower in temperature than the reacting graphite, the flow rate is a deciding factor in determining whether the graphite cools or continues to heat. Experimental studies on graphite burning have shown that for all the geometries tested which involved the conditions of small radiation and

conduction heat losses, it was not possible to develop self-sustained rapid

}

oxidation for graphite temperatures below about 650*C when the air temperatures were below the graphite temperature. At both high and low flow rates, the graphite was cooled by heat losses to the gas stream even under conditions where other heat loss mechanisms such as radiation and conduction were negligible.

At temperatures above about 650*C, in realistic geometries where radia-tion is a major heat loss mechanism, graphite will burn only in a limited range of flow rates of air and only when the air temperatures are high. At low flow rates, inadequate ingress of air restricts burning. At high flow rates, the rate of cooling by the flowing gas can exceed the rate of heat produced by oxidation.

O

Studies have shown that burning will not occur when there is no mechanism to raise the graphite temperature to about 650*C [Schweitzer, 1962a-f]. If l

the temperature is raised above 650*C, burning will not occur unless a flow pattern is maintained that provides enough air to sustain combustion but not ,

enough to cause cooling. Since the experiments were designed to minimize all heat losses other than those associated with the air flow, 650*C can be '

considered a-lower bound for burning. I 8.2 Stored Energy in Craphite Fast neutron irradiation of graphite results in the development of stored (Wigner) energy. For a research reactor that has accumulated 30 cal /g of graphite after years of operation, this energy corresponds to about 1/250 of the energy released by combustion. Existing data show that for graphite irradiated at temperatures of 30'C or above, the stored energy that can be released at 650'C saturates at a value that is less than 1/30 of the combustion energy.

Analyses of the Windscale Accident and the Chernobyl Accident have shown s that stored energy releases were not initiating events nor did they play any significant role in the evolution of the accidents. Although precise details of the buildup and release of stored energy vary with reactor geometry and factors relating to reactor operation, this review and analysis did not un-cover any substantiated evidence or credible scenario in which stored energy releases were responsible for an accident leading to graphite burning [ Fleck, -

1958; Kouts, 1986].

In assessing the role of stored energy releases in graphite burning sce-narios only the stored energy released below the burning temperature was con-sidered pertinent. Stored energies released at or above the burning tempera-ture are a small fraction of the energy released by the burning process.

A large volume of literature exists on the accumulation of stored energy at different irradiation temperatures and different fast neutron exposures.

Total accumulation of stored energy is a complex phenomenon that depends upon  %

many factors related to reactor geometries, fast flux distributions, graphite properties, reactor operating schedules and other conditions. At irradiation temperatures above about 150*C, the rate of accumulation of total stored energy is very low with negligible releases occurring if the graphite tempera-ture remains below the graphite threshold burning temperature of 650*C. At about 30'C and at low total exposures, the total stored energy increases at a near linear rate of about 401 10 cal /g per 100 mwd /AT (Nightingale, 1962].

As the exposure continues, the rate of accumulation of total stored energy decreases, and the stored energy that can be released below 650*C saturates and then appears to decrease [ Nightingale, 1962; Neubert, 1957; Woods, 1956].

From existing data, an upper bound on the stored energy that can be released below 800*C is 280 cal /g if the graphite was irradiated at 30*C. If the graphite was irradiated at 70*C, data indicate that the maximum stored energy releasable below 700*C is about 150 cal /g. The saturation value for an irradiation temperature of 135'c is about 50 cal /g released below 700*C.

25

1 l

Although there appears to be significant differences in the estimates of l total accumulated stored energy calculated in the past (Hawley, 1981; NRC, 1983a, 1983b], these values have little relevance to graphite burning condi-tions.' The total stored energy is always greater 8han, and is not directly proportional to, the stored energy that can be released below the threshold temperature associated with graphite burning. It requires about 200 cal /g of stored energy to raise the graphite temperature from 30*C to 650'C if there are no-heat losses., Similarly, it requires about 190 cal /g'to raise the graphite temperature from 70*C to 650*C and 180 cal /g to raise it from 130*C to 650*C. The evidence on maximum stored energy releasable below 650*C shows that if graphite is irradiated at 70*C, or above, the maximum energy released below 650*C is not sufficient to raise the temperature to the burning tempera-ture even under the hypothetical conditions of a spontaneous release under totally adiabatic conditions. In an assumed adiabatic LOCA scenario, the decay heat in any nuclear reactor should be the major source for raising.

graphite temperatures.

The analyses and conclusions on stored energy releases and graphite burn-ing conditions described above provide a meaningful method of categorizing nuclear reactors with respect to stored energy releases.below graphite burning temperatures:

(1) Any reactor containing graphite in which the lowest irradiation tem-perature is 70*C or higher, can be excluded from stored energy safety concerns.

(2) Any reactor in which the graphite is irradiated at temperatures below70*Cbuthasreceiveda{gtalfastneutronexposurethatis less than 500 mwd /AT (3.5 x 10 nyt) can be excluded from stored energy safety concerns.

(3) Thosereactorswhichgavegraphitethathasreceivedmorethanabout nyt) of fast neutron irradiation below 70*C 500 mwd /AT (3.5 x 10 without thermal anneals or subsequent re-irradiation at higher tem-peratures require detailed heat transfer analyses to determine if the graphite is capable of reaching 650*C in an accident that heated it initially to about 100*C. We emphasize again that graphite tem-peratures exceeding 650*C are necessary but not sufficient conditions to initiate and support burning.

In order to separate reactors into these categories, it is necessary to determine only the total fast neutron exposure reached by graphites irradiated at temperatures below 70*C.

One pound of graphite releasing a stored energy of 200 cal /g is equiva-lent to running a 100-watt light bulb for one hour. Recognizing that such releases cannot occur unless another energy source raises the graphite temper-ature above its operating temperature, spontaneous stored energy releases can-not be considered credible initiating events for graphite burning phenomena.

Since the maximum energy that can be stored below 700'C is about 1/30 of the 26

combustion energy, the single release of stored energy that might occur during a graphite burning accident is an insignificant portion of the total energy released in the first few minutes of burning reactions. These conclusions are consistent with analyses of both the Windscale and Chernobyl accidents.

8.3 Safety Assessment Consequences of graphite burning accidents depend upon the amount of graphite that can burn, and the inventory of radionuclides that can be re-leased. Both the amounts of graphite and the inventories of radionuclides in the Chernobyl and Windscale reactors were many orders of magnitude greater than in NRC-licensed research reactors operating in the U.S.

Analyses of the actual reactor accidents in which graphite burning occur-red and analyses of hypothetical accidents show that some mechanism must lead to either fuel or graphite heatup under conditions where air is available.

The review of a number of research reactors representing the various classes or types of research reactors currently licensed to operate in the U.S. (e.g.

the TRIGAs, ARGONAUTS, PULSTAR, GE-NTR, MTR-D 2 0, and HTRs) found that undar normal operating conditions their design features and/or environments should preclude graphite being heated to a temperature at which burning could be ini-tiated. In addition, under LOCA conditions it was judged to be plausible that the potential for cooling the graphite by passive means (e.g. . radiation, conduction, natural convection) also should preclude graphite burning.

9. CONCLUSIONS After review and analyses of existing information on graphite burning, stored energy accumulations and releases, and causes of the Windscale and Chernobyl accidents, we have concluded that the above phenomena are suffi-ciently well understood to allow the following evaluations of U.S. research reactors and Fort St. Vrain.

The conclusions of these analyses are that the potential to initiate or maintain a graphite burning incident is essentially independent of the stored energy in the graphite and depends on other factors that are unique for each research reactor and for Fort St. Vrain. However, in order to have self-sus-tained rapid graphite oxidation in any of these reactors certain necessary conditions of geometry, temperature, oxygen supply, reaction product removal and favorable heat balance must exist.

The reactors considered in this review have all undergone safety evalua-tions and have been granted operating licenses by the NRC. There is no new evidence associated with the analyses of either the Windscale Accident or the Chernobyl Accident that indicates a credible potential for a graphite burning accident in any of the reactors considered in this review. Nor is there any new evidence that suggests that detailed case-by-case safety analyses of the role of graphite in NRC licensed reactors are warranted.

27

7__----__.--____-______

i l

l

10. GLOSSARY l BGRR Brookhaven Graphite Research Reactor BNL Brookhaven National Laboratory BTU /Hr British Thermal Units per Hour cal /g Calories per gram CBG Committee to Bridge the Gap CO Carbon monoxide CO2 Ca' bon dioxide FSAR F1nal Safety Analysis Report LOCA Loss-of-coolant accident mwd /AT Megawatt days per adjacent ton NRC Nuclear Regulatory Commission 2

nyt(th) Exposure in terms of thermal neutrons per em 02 0xygen SAR Safety Analysis Report 1

28

11. REFERENCES

[ Bridge, 1962]

H. Bridge, R. T. Kelly and B. S. Gray, " Stored Energy and' Dimensional Changes in Reactor Graphite," Proceedings of the Fifth Conference on Carbon, Volume 1, 1962.

[ Burn, 1983]

R. R. Burn, Editor, Research. Training, Test and Production Reactor Directory, American Nuclear Society, Second Edition, LaGrange Park, Illinois, 1983.

[ Burton, 1956]

R. Burton and T. J. Neubert, "Effect of Fast Neutron Bombardment on Physical Properties of. Graphite: A Review of Early Work at the Metallurgical Laboratory," J. Appl. Phys, 27, 557-572 (1956).

[Chen,'1981]

W. Chen, " Final Safety Analysis Report Submitted to the U. S. Nuclear Regula-tory Commission in Partial Fulfillment of the Requirements for a Class 104 License," College of Engineering, University of Florida, Gainesville, Florida, January 1981.

[Cottrell, 1958]

A. H. Cottrell, J. C. Bell, G. B. Grenough, W. M. Lomer and J. H. W. Simmons,

" Theory of Annealing Kinetics Applied to the Release of Stored Energy From Irradiated Graphite in Air-cooled Reactors," Proceedings of the Second United Nations International Conference on the Peaceful Uses of Atomic Energy, Volume 7 (1958).

(Davidson, 1959]

J. M. Davidson, " Stored Energy in Irradiated Graphite," US/UK Graphite Con-ference, Held at St. Giles Court, London, December 16-18, 1957. U. S. Atomic Energy Commission Report TID-7565 (Pt. 1), pp 11-20, March 16, 1959.

[Emmons, 1965]

A. H. Emmons, D. G. Fitzgerald and E. L. Cox, " University of Missouri Research Reactor Facility Hazards Summary Report, In Support of an Application to the United States Atomic Energy Commission for a Class 104 Utilization Facility License," University of Missouri, Columbia, Missouri, July 1, 1965.

[ Federal Register, 1986]

51FR31341, 51:170, Wednesday, September 3, 1986.

[ Fleck, 1958]

Sir Alexander Fleck, J. Cockcroft, W. Penney, R. Spence. J. Diamond, J. M. Kay and H. W. B. Skinner, " Final Report of the Committee Appointed by the Prime Minister to Make a Technical Evaluation of Information Relating to the Design and Operation of Windscale Piles and to Review the Factors Involved in the Controlled Release of Wigner Energy," presented to Parliament by the Prime Minister by command of Her Majesty, London, UK, 1958.

29

sv^ .

,1 ,

s> .n T

[GA Technologies,J1987). .

CA Technologies, P.:0.. Box 85608, San Diego, California, 92138, Letter to

-U.-S.1 Nuclear Regulatory Commission, GEN-1010,fJanuary 28,.1987.'

.[GE,c1981]

=

General Electric Company, " General. Electric Nuclear Test Reactor, Safety-Analysis : Report," NEDO-12727, -80NED029, Class I, April 1981.

M [Hawley,fl981]. . . .

LS. C.~ Hawley R. L. Kathren and M. A. Robkin, " Analysis of_. Credible Accidents for Argonaut Reactors,". NUREG/CR-2079, Pacific Northwest Laboratory report '

.(PNL-3691) prepared for U. S. Nuclear Regulatory Commission, April 1981.-

'[Kinchin, 1956]

G.11. Kinchin, "The Effects of Irradiation on Graphite," Proceedings of the -

Second. United Nations International Conference'on the Peaceful Uses of Atomic

. Energy,LVolume 7, pp. 472-478 (1956). .]

-[Kircher, 1964]

J .

F.7 Kircher, and R._E. Bowman, Editors, Effects of Radiation on Materials and Components, Reinhold Publishing Corporation, New York, 1964.

[Kostba, 1953] .

l W. L., Kosiba, G. J.!Dienes, D. H. Gurinsky, "Some Fffects Produced in Graphite -l

.byLNeutron. Irradiation in BNL Reactor," Second Conf ince on Carbon, pp.

-143-148, 1953.

[Kouts, : 1986 ]

H. J. ~ C. Kouts , "The Chernobyl Accident," Brookhaven Lecture No. 227,  ;

BNL-52033, 1986.

[ Lewis, 1963]

LJ. B.-Lewis, "The Thermal Reactivity of Nuclear Grade Graphites to Oxygen,"  !

' Progress in Nuclear Energy Series IV, Volume 5, p. 145, 1963.

[Nairn, 1961] .

J. S. Nairn, and V. J. Wilkinson, "The Prediction of Conditions for Self-Sustaining Graphite Combustion in Air," Proc. of the US/UK Meeting on the Compatibility Problems of Gas-Cooled Reactors, TID-7597, '1961. ,

1 i

[Neubert, 1957]

T. J. Neubert, and R.' B. Lees, " Stored Energy in Neutron-Bombarded Graphite,"

Nuclear Science and Engineering, 2:748-767 (1957).

.[ Nightingale, 1958]

R. E. Nightingale, J. M. Davidson and W. A. Snyder, " Damage to Craphite Irra-diated up to 1000*C," Proceedings of Second United Nations International  !

Conference on the Peaceful Uses of Atomic Energy, Volume 7, 1958.

) [ Nightingale, 1962]

L R. E. Nightingale, Nuclear Graphite, Academic Press, New York, 1962.

l 30 E

1 1

[NRC, 1983a]

Nuclear Regulatory Commission, "In the Matter of UCLA Research Reactor (Proposed Renewal of Facility License)." Docket No. 50-1420L, page 1870, California, July 23, 1983.

[NRC, 1983b]

Nuclear Regulatory Commission, " Testimony of CBG Panel II - Chemical Reac-tions, In the Matter of UCLA Research Reactor (Proposed Renewal of Facility License)," Docket No. 50-142 OL, page 13, California, October 13, 1983.

[NRC, 1983c]

Nuclear Regulatory Commission, " Safety Evaluation Report Related to the Renew-

, al of the Operating License for the National Bureau of Standards Reactor,"

l Office of Nuclear Reactor Regulation, Washington, D. C., NUREG-1007, September 1983.

[NRC, 1984]

Nuclear Regulatory Commission, " Safety Evaluation Report Related to the Renew-al of the Operating License for the General Electric-Nuclear Test Reactor (CE-NTR)," Office of Nuclear Reactor Regulation, Washington, D. C.,

NUREG-1069, September 1984.

[ Penney, 1957]

Sir William Penney, B. F. J. Schonland, J. M. Kay and J. Diamond, " Accident at Windscale No. 1 Pilo on 10th October 1957," Presented to Parliament by the Prime Minister by command of Her Majesty, London, November 1957.

[ Reich, 1986]

F. R. Reich, and R. J. Nicklas, "An Evaluation of Methods for Reducing or Delaying the Ef fects of Graphite Distortion to Extend the Production Life of the N-Reactor Core, Appendix E-2," United Nuclear Corporation, Inc., UNI 3680 September 1986.

[ Robinson, 1961]

P. J. Robinson, and J. C. Taylor, " Thermal Instability Due to oxidation of a Graphite Channel Carrying an Air Flow," Industrial Group Headquarters, Risley, Warrington, Lancashire, UK, IGR-R/W- 302; also in " Proc. of the US/UK Meeting on the Compatibility Problems of Gas-Cooled Reactors," TID-7597, p. 471, 1961.

[ Schick, 1966]

H. L. Schick, Editor, Thermodynamics of Certain Refractory Compounds, Volume II, Academic Press, New York, 1966.

[Schweitzer, 1962a]

D. G. Schweitzer, " Activation Energy for Annealing Single Interstitial in Neutron-Irradiated Graphite and the Absolute Rate of Formation of Displaced Atoms," Phys. Review, 128:2, pp.556-559, October 15, 1962.

[Schweitzer, 1962b]

D. G. Schweitzer, " Oxidation and Heat Transfer Studies in Graphite Channels IV," Nuclear Sci. and Eng., 12:59, 1962.

31

l

[Schweitzer, 1962c]

D. G. Schweitzer, ' Fundamental Studies of Radiation Damage in Graphite," BNL Lecture Series, No. 16, BNL-745, April 17, 1962.

[Schweitzer, 1962d]

D. G. Schweitzer, and D.H. Curinsky, "0xidation and Heat Transfer Studies in Graphite Channels II," Nuclear Sci. and Eng., 12:46, 1962.

[Schweitzer, 1962e]

D. G. Schweitzer, G. C. Hrabak and R. M. Singer, " Oxidation and Heat Transfer Studies in Graphite Channels I," Nuclear Sci. and Eng., 12:39, 1962.

[Schweitzer, 1962f]

D. G. Schweitzer, and R. M. Singer, " Oxidation and Heat Transfer Studies in Graphite Channels IV," Nuclear Sci. and Eng., 12:51, 1962.

[Schweitzer, 1964a]

D. G. Schweitzer, and R. M. Singer, " Removal of Radiation Damage from Graphite by Alternate Reirradiations and Low-Temperature Anneals," Nuclear Sci. and Eng., 19, p. 385, 1964.

[Schweitzer, 1964b]

D. G. Schweitzer, R. M. Singer, S. Aronson, J."Sadofsky and D. H. Gurinsky,

" Decomposition of Defects by Neutrons in Reirradiated Graphite," Nuclear Sci.

and Eng., 18, p. 400, 1964.

[Seitz, 1958]

F. Seitz, and J. S. Koehler, "The Theory of Lattice Displacements Produced During Irradiation," Proceeding of the Second United Nations International Conference on Peaceful Uses of Atomic Energy 1958, Vol. 7, 615-633, USA, 1958.

[Spinrad, 1986]

B. I. Spinrad, " Stored Energy in Reflector Graphite of Research Reactors,"

correspondence to Nuclear Regulatory Commission in reply to notice of rule making, Washington, D. C., 1986.

[UCLA, 1981]

University of California at Los Angeles, " Safety Evaluation Report Related to Renewal of the Operating License for the Research Reactor at the University of California at Les Angeles, Los Angeles, California, July 1981.

[ West, 1963]

G. B. West, and J. E. Larsen, " Calculated Fluxes and Cross Sections for TRIGA Reactors," General Atomic, Division of General Dynamics, San Diego, California, GA-4361, August 14, 1963.

[Wigner, 1946]

E. P. Wigner, " Theoretical Physics in the Metallurgical Laboratory of Chicago," J. Applied Physics, Vol. 17, No. 1, November 1946, p. 857; also an address presented to the Am. Physical Soc. at the Chicago Meeting, June 22, 1946.

32

[ Woods, 1956]

W. K. Woods, L. B. Bupp and J. F. Fletcher, " Irradiation Damage to Artificial Graphite," Proceedings of the International Conference on the Peaceful Uses of Atomic Energy, Volume.7, pp. 455-471, United Nations, 1956.

12. BIBLIOGRAPHY Beattie, J. R. , J. B. Lewis and R. Lind, " Graphite Oxidation and Reactor Safety," Proceedings of the Third. United Nations International Conference on the Peaceful Uses of Atomic Energy, Volume 13, Paper P/185, 1965.

Dalmasso, C and G. F. Nardelli, "The Wigner Release in Graphite-Moderated Reactors," Energia Nucleore, English translation in USAEC Report AEC-tr-4545, May 1961.

Dickson, J. L, G. H. Kinchin, R. F. Jackson, W. M. Lomer and J. H. W. Simmons ,

"BEPO Wigner Energy Release," Proceedings of the Second United Nations -

International Conference on the Peaceful Uses of Atomic Energy, Volume 7, 1958.

Fox, M. and R. W. Powell, "The Annealing of the Graphite Moderator Structure in the BNL Reactor, BNL-275, January 1954.

Hawley, S. C., R. L. Kathren, " Credible Accident Analyses for TRIGA and TRIGA-Fueled Reactors, NUREG/CR-2387, Pacific Northwest Laboratory report (PNL-4028)' prepared for U.S. Nuclear Regulatory Commission, Ap."1 1982.

Kosiba W. L., D. H. Gurinsky and G. J. Dienes, " Evaluation of bdL Pile Graphite," BNL-255. October 5, 1953.

Kosiba, W. L. and G. J. Dienes, "Effect of Displaced Atoms and Ionizing Radiation on the oxidation of Graphite," Advances in Catalysis, Academic Press, New York, 1957.

Lewis , J. B. , P. Hawtin and R. Murdoch, " Thermal Oxidation of Nuclear Graphite," J. British Nuclear Energy Society, pp. 95-98, April 1964.

Meyer, W. A., Jr., " Stored Energy in Irradiated Graphite," University of Missouri, Research Reactor Facility, December 10, 1986.

Nightingale, R. E., " Record of Proceedings of Session E-21," Proceedings of Second United Nations International Conference on the Peaceful Uses of Atomic Energy, Volume 7, 1958.

Powell, R. W., R. A. Meyer, and R. G. Bourdeau, " Control of Radiation Effects in a Graphite Reactor Structure," Proceedings of the Second United Nations International Conference on Peaceful Uses of Atomic Energy, Vol. 7, P/462, USA, pp. 282-294, 1958.

33

Rimer, D. E. and W. M. Lomer, " Calculations on the Release of Stored Energy in Graphite," Atomic Energy Research Establishment, Harwell, U.K., A.E.R.E.

M/R.2063, June 1958.

Rimer, D. E., "The Validity of the Constant Activation Energy Model for the Release of Stored Energy in Graphite," Atomic Energy Research Establishment, Harwell, U.K., A.E.R.E.-R.3061, August 1959.

Schweitzer, D. G. and R. M. Singer, "Effect of Irradiation Temperature and Annealing Temperature on Expansions and Contractions in Alternatively Irradiated and Annealed Graphite," Trans. Amer. Nuclear Society, 6:2, pp.

383-384, November 1963.

Schweitzer, D. G. and R. M. Singer, " Anomalous Stored Energy and c-Axis Changes in Alternatively Irradiated and Annealed Graphite," Carbon, 2:4, pp. 185-191, 1964.

Schweitzer, D. G., " Determination of the Single Interstitial Migration. Energy From Stored Energy and Thermal Resistivity Changes in Irradiated Graphite,"

Carbon, 2:4, pp. 404-412, 1965.

34

a gac,eoa= m u s. Nuc6 An anoutAion, co in.o= . a. Pus e Nuusin .uw. ., noc. m v., a. . ., m.

"' NUREC/CR-4981 IEo'i 8 BIBLIOGRAPHIC DATA SHEET

$40 INSTMucVIONS ON THE mtvinst BNL-NUREG-52092 4 all t ANif MJillit t i J LL AVL hl ANE A Safety Assessment of the Use of Graphite in Nuclear Reactors Licensed by the U. S. NRC.

. oan n. uni couPtin o uuNen vsAn

. uinonm l-May 1987 D. C. Schweitzer, D.11. Gurinsky,

.E. Kaplan and C. Sastre September 1987

p. ea,.sonMimo onsANidanoN Naut ANo MAsuno Aoomass ##aesumie s.ms e enoaactsi Asarnona uNel NuMetR 1 Brookhaven National Laboratory Upton, NY 11973 ,,,,,,,,c,,,,,,,,,,,,,,  ;

A3855 f

la 67QNsoHINQ QHGANRJA f tGN NAME AND MAILING ADDRilS (deesfia## ge Cepet ilsIVPtOFREPorif I z

Of fice of Nuclear Reactor Regulation  !

U.S. Nuclear Regulatory Commission Technical Washington, DC 20555 13, SUPPL $ MENT ARV NQf 31

,,A....A , . ,

I This report reviews existing literature and knowledge on graphite burning and on stored energy accumulation and releases in order to assess what role, if any, a storedenergyreleasecanhaveininitiatingorcontributingtohypotheticalgraph-ite burning scenarios in research reactors. It also addresses the question of graphite ignition and self-sustained combustion in the event of a loss-of-coolant accident (LOCA). The conditions necessary to initiate and maintain graphite burn-ing are summarized and discussed. From analyses of existing information it is con-cluded that only stored energy accumulations and releases below the burning temper-ature (650'C) are pertinent. It is shown that there is no evidence from the Chernobyl event that stored energy releases played a role either initiating or con-tributing to this accident. The conclusions from these analyses are that the potential to initiate or maintain a graphite burning incident is essentially inde-pendent of the stored energy in the graphite, and depends on other factors that are unique for.each research reactor and for Fort St. Vrain. There is no new evidence associated with either the Windscale Accident or the Chernobyl Accident that indi-cates a credible potential for a graphite burning accident in any of the reactors considered in this review.

1. Research reactors as used herein means research, test, and training reactors.

14 QoCuMthf ANALT&44 * . R8 vWoHollDESCRiPfoMS lb AVAIL A8stII g

$fATEMINI

-graphite safety assessment graphite burning research reactors Unlimited Cherynobyl 16 StCunery CL AsseeICAisON iia *G 900e8 m in=rins alioPi= INoeo unus Unclassified u.,,,,,

Unclassified il NuMetM op FAGes le PHatt

• U. S.COVERktiL hi #AlkflhG Of f lCC s1967.181-682 60219

s

't-

~ PhGEf 1 ~U.S.. NOCLE AR :REGOL ATORY, COMMISSIONS

DATE

,10/20 87 p '

PACKING SLIP- -

l a l' . t' l' ' USNRC- . ORDER' NUMB:'A28688-00002-y ..T'G MICHAELSJ ORDER' TYPE: ' 2 '-

[ '~ ,

P234'

'/ '

lc 8 1

DOCUMENT NO.- - QTY S . COST TOTAL COST C0MMENTS!

tNUREGCR4981~ .2

' TOTAL'00CUMbNTS' ORDERED:. -

2' TOTAL.; DOCUMENTS'ENCLO$EO: 2 ,

I n.

l i

\

8' 9e 4

4 9

.4 e

/

3 i

UNITED STATES sescut souamctass nan NUCLEAR REGULATORY COMMISSION. .. Po81*Cjsyc'8

WASHINGTON, D.C. 20555 ,,,,, % o.,

OFFICIAL BUSINESS PENALTY FOR PRIVATE USE,4300

.L i

+ h V)

Fn 3

?

2 m

mm 40

-4 n m$

l. C>

53 i

Z 2

8 m

E a

R 8

a c

S

.5 8

E a

o 5

_ _ _ _ _ - _