ML20216F332
| ML20216F332 | |
| Person / Time | |
|---|---|
| Site: | Waltz Mill |
| Issue date: | 05/27/1960 |
| From: | Mann M, Sears J US ATOMIC ENERGY COMMISSION (AEC) |
| To: | |
| Shared Package | |
| ML20216F225 | List:
|
| References | |
| FOIA-98-100 CF-169, NUDOCS 9804170035 | |
| Download: ML20216F332 (13) | |
Text
.
1
, 1, '*
g; sn
- n ..
U. S. ATCIIIC EEEtGT COIRESSIDE .w DITISIDE W ZEEPECTION '
REPGtT CF.169 ,
By: John R. Sears, Inspection Division Eow York Operations Office Date MAY 27 W mad Marvin M. Maan Division of Inspection, Eq.
Title:
INESTIGATION OF RADIATICE IECIDENT AT THE WESTIMEDUSE TENIDO REACTGt, LICEBSE E0. 5 2, WAIg7. ICIL, FEENTIXAIIA IMBODUCTICE The incident at the Westi-f a Testing Beacter, which eseurred se April 3,1960, was initially reported to the Ier York Operations Office Insporties by Divisies by Westinghouse en April 4, 1960, by t depheme and a teletype.
A writtaa report of the insident was troAasitted by
- Westinghouse April 4, 1960. to the Dirwetor, Divisies of Licenstag ad Regulation, on '
Qa April 13 and 14, Jeek R. Reeder rA Jeha 1. Sears of the the E Inspecties Divisica, eseemsted a p lia Sery investigation of insident. On the basis of this pre 11miu ;ir investigatioe, the inci-dont was elassified as a Class "A" insteest, beeense of the estimated reacter nova time, and the estimated eest of eleanup operations. An ,
investienting eeunittee was then established, oomsisting of the foller.
ing:
. 4 Dr. Marvia L Maan .
Assistaat Director for Complianoe Divisies of Inspection, Eg.
Chairmaa. - .
John R. Sears . Inspection Division Bev York Operations Office l Moreon Booth - !
Division of Licensing and Regulation, Eq. '
Vincent A. Walker .
Division of Inspection, Eq.
The committee April 22, 1960 met at the reeetor site at Walts Mill, Pennsylvania, on Mr. Richara cunningham of the Division of Licensing and Regulation, was present Auring the cessaittee's deliberations. The follor- i ing Westi=f a staff members were interviewed during the investigation:
Mr. E. T. Morris, Manager, Westinghouse Testing Roseter Dr. M. A. Behalts, anginaaring Maanger, WTR Dr. A. J. Pressesky, Manager, Scientific Sgport Sectica Mr. R. J. Cat 11a, Manager, Esalth Phrsies Section Distribution: General Manager E. L. Fries, DIR -
I Dr. C. K. Desk, DIR L. E. Johnsen, DIR General Counsel (Mr. Emaon) 3 l
l han .
Mr.GeorgeGeisler,PlantSuperinkendebt1 Mr. E. E. Beamerle, Supervisor Technical Assistance Mr. L. Pollono, Shift Supervisor Mr. Bianchi, Reactor Operator Mr. D. Catlin, Health Physics Supervisor, Mr. D. Villholt, Health Physicist on duty at the time of the incident, and Mr. D. Gala, Benior Scientist re-sponsible for water cleanup, were interviewed by J. R. Roeder of the ET Inspection Division, in the preliminary investigation.
OPERATIONS PRIG 1 TO THE INCIDENT
- 1. Prior to this incident, an experiment had been performed in an attempt to determine the void coefficient of the reactor. A report en-titled ' Preliminary Report On An Experinct To Illustrate The Simulation And Detection Of Boiling In The Westingbouse Testing Reactor", WIR-SS-TA.
258, Appendix I, was transmitted to the Director, Division of Licensing and Regulation on November 11, 1959 This report describes an experiment in which bubbles of helium gas were introduced into the reactor through two 1/8" 0.D. aluminum tubes. Bubble size could be regulated by appro-priate valving, and could be observed through a length of transparent tygon tubing into which the aluminum tubes discharged. The sensing in-6trumentation consisted of a compensated ionisation chamber connected 3 to a bucking circuit, and then into a high gain D.C. amplifier and a Brush recorder. The D.C. output of the chamber was biased out by the {
bucking circuit. Fluctuating signale due to voids passing through the l 1 resetor appear as deflections on the Brush recorder.
- 2. The sensing instrumentation for the bubbler experiment was left installed in the shield outside of the pressure vessel during the escala-tion in power from 20 megawatts to 60 megawatts. A series of reduced flow
{
experiments had been made in February 1960, at various power levels and at various primary coolant flows, in order to determine the point at which there was the first indication of perturbations due to bubble formation (boiling). The traces of a number of these experimental runs are included i as Exhibit A. It should be noted that during these runs, there is the notation, " Access Tubes - In". Dr. Schultz explained that these access tubes were capped thimbles which had been placed both in the core and in the reflector region, to hold some small experimental samples. The water in these tubes had been stagnant, and so it was felt that during these reduced flow experiments the water in the access tubes was indeed boiling.
An experiment was done to prove this point. In one access tube in core position L-$-8, a thermocouple was lowered, and during the run temperature readings were taken as the thermocouple was raised through the secess tube.
It was said that these readings indicated that there was boiling inside .
the tubes during these runs. In view of this, the decision was made to repeat these experiments with the access tubes removed from the reactor in order to determine the relation between coolant flow . d the initiation of boiling. This is the type of experiment which was being run on the night of the incident.
3 The minutes of the Westinghouse Testing Reactor Safeguards Com-mittee meeting indicate that the reduced flow experiments had been re-viewed by the Comunittee and had been approved. A test specification for the escalation in power from 20 to 60 megawatts was prepared and approved by the Chairman of the Safeguards Committee. This test spect-fication No. 5 1, dated February 19, 1960, is included as Exhibit B.
It places definite restrictions upon operation, in compliance with the Westingbouse license. This test specification, No. $-1, also specified that, if rod motion equivalent to 0.114 reactivity is suddenly required, this would indicate that these prescribed limitations had been exceeded, and that the reactor power level should be decreased by 10 MW. However, i
I
e the investigating committee determined that these restrictions spelled out in test specification No. 5-1 were never translated tato operattag procedures for the reactor operator at the console.
- b. For the particular experiment being performed on the evening of the incident, test specification 30 5-2 had been prepared. A copy is included as Exhibit C. Since in this experiment the flow was to be cut back to 5000 gym, and the maxians power level was to be kb mega-watts, the screa and cutback settings had to be readjusted. This specifiestion did not spell out any anticipated aboomal conditions which might prevail in the reactor as a result of the experiment. A i copy of the specification had been studied by the reactor supervisor -l and the operator and was available at the console. Specification 5-1 i also was in force but a copy was not sait to be available at the con-sole. , ,
i
$. RumMaation of Westinghouse's log book indicated that there had been difficulty in procuring acceptable fuel elements. A large per-contage of elements had been rejected because of surface imperfections.
Dr. Schults said that the depth of imperfection which determined rejec.
tion was 0.01$*.
- 6. Several weeks before the incident, a fuel element in location I L-7-6* had been instrumented with aluminum sheathed thermocouples. These thermocouples had drifted up over a period of three weeks. Calculations of flow and best conditions indicated that the thermocouple readings must be in error, and on April 1,1960, the fusielement in position L.$-6 was installed with 6 thersecouples. These were stainless steel sheathed, ,
F' magnesium oxide insulated thermocouples. These thermocouples were coa. ,
, neeted to read out on a data logger. Exhibit D is a reproduction of the
[- thermocouple readings on these two fuel elements in loostions L-7-6 and i L.$-6 just prior to and during the incident. The readings on L-7 6 went as high0as 3110 F., whereas the highest reading on L-$-6, at any point, was 265 F. l Test specification 5-1 had placed an operating restriction on ;,
the first fuel ring bulk water outlet temperature at 2200 F. The thermo-couple readings of L-$-6 indicate that at no time, even during the inci-dent, was this exceeded.
7 'A final ites which possibly might have a bearing on the incident was mentioned by Dr. Schults durir.g the investigation. About a week pre-vious to the incident, velders had been at work inside the head tank. They ,
had worn plastic shoe covers. Dr. Schultz mentioned the possibility that l a shoe cover or bits of welding rod might have gotten into the primary '
water. However, he also said that there is a strainer in the line before the reactor, and that it should keep particles of foreign material larger than about one quarter inch from the reactor.
It was stated that V'IR personnel did not inspect the head tank during or after the welding operation.
TB AFRIL THIRD IEIDElrF On April 3,1960, the reactor had been operated at a steady state for ap- !
Proximately one and a half days at 40 megawatts with the primary cooling ;
flow at 15,000 gym. At approximately 6:30 p.m., a meeting had been held '
- See Exhibit F for fuel element and thimble positions.
3
4
)
in the control room to outline the experiment to be done that evening.
The ocoditions for scram and outbaek settings, as specified in test specification Bo. 5 2, were set forth. At 7:00 p.m., the reactor power was redneed to 30 angmentts, and serem and outback settings were ad-justed to permit experiments at a flow of 5000 gpa.
In the control room at h time were Mr. Bianchi, a licensed reactor '
operator, at the seasole, and Mr. Polloso, the Shift S w ervisor.
Dr. A presseaky, Meanger, Scientific Support Section, was also in the control room for the purpose of ==Hng some calenlations during the run.
Mr. George Geisler was in the Shift Superviser's office behind the con-trol room. Dr. Schaltz was observing the Brush recordings from the void ,
ocefficient instrumentation on the balcony within the contain==nt shell. '
{
Mr. Niek Shorich, a licensed pile operator, was in h containment shall observing tangeratures on the only operating loop. The neutree power level was being reeerded in the control room, and s,..t dios of the chart is attaehod as Exhibit E. At approxiantely 8:00 p.m. the flow was reemoed to $250 syn. Thiswasdoesbytheoperatoratlheconsole.
As the flow was redmoed, the temperatures insreased and the negative too.
perature ocefficient redueed the power slightly, and the automatic een-trollar breeght the power beek at a level of 30 magmaatts. At 8:20 p.m.,
h operator was instrusted to raise the poser level to 35 messwatts. 4 This he 414 by theauto power demand. The thermal power was omleulated 1 l
to be 37 messenets, and the Shift Bayervisor (Po11ono), instructed the operator (Bisashi) to lower the power to 35 megawatts. This he did. At spyroximately 8:33 p.m., h operator was instructed to raise h power level to b0 megewetts. Es set the automatic power demand, sad the reactor started up. As the mostros power level rooorder went to approximately 37 megawatts at about 8 34 p.m., the recorder pen dropped within a few seconds to apprart==tely 17 messwatts. The f9 rod, which was shoot two thirds with- ,
drown, was thea withdrawn to about 85% by the automatic ecotrol, and h operator switched to $8. The fB rod siso was withdreva about 8% by the autematic eostrol, and the operator attempted to regain reactivity by withdrawing the fl 234 =====Mr. Both Po11ono and Bianchi stated that ,
l they saw the drop in power level on the recorder. Bianchi stated that et the time he thought that the drop may have been due to the effect of sither a tagerature or a void coefficient. The power recorder h n began to indicate incrossing ever, as a result of the control rod withdrawmis, and Bianchi said that the period indicator on the console indicated an ap-proximate 60 second period.
Bianchi recalls ht as the power ceae up, a radiation monitor sounded.
He said ht he had been puzzled by having to use both the f8 and f9 rods, and he had been instructed to help in returning the Power level to its former value by pulling fl. Be claimed h t he did not pull any more rods during the return to power. The reactor was evidently responding to the power demand setting on the auto controller. During the last part of the climb back to power, more radiation monitors went off, and Bianchi was in-structed by Po11ono to cut back manually at about 8:40 p.m. The indicated j l
resetor power was than apprart==tely 38 megawatts. The reactor power was cut book to about 15 angavatts, and then as still more radiation monitors i j
clarmed, p011000 instructed Bianchi to screa the reactor at 8:44 p.m. I BEALTE PHYSICS I
Mr. Donald W111hoit was the Bealth Physicist on duty on the night of the incidant. He said ht approximately 12 people were working in the WIR facility. W111hoit was in the health physics office at about 8:36 p.a.
when he heard the first alara, which subsequently turned out to be the ;
-A. l
E
- 4 v s,. .
i' 1 11 .a l P:
g '.
radiation monitor on the reactor top. He very soon after beard another I alarm and immediately called the control room to determine vbat %
had At this point, all of the alarms sounded practically ',
s ly. Willboit then took a Cutie Pie survey meter and pro- ?
coeded to the control room, f $;
g The following radiation monitor readings were obtained from Mr. George Geisler. Be said that these readings had been taken from the recorder !' i in the control room at about 8:45 p.m.
Approximate Road- o ,
Channel No. Description ing in Counts /S *' '
1 Critical Test Station 100 , 4 2 Gamma Radiation Station 40 h T, :
3 6
Canal Floor Area 100 1& >
Waste Disposal Facility 300 1,, I[
8 Reactor Top 1000 ? g 9 Beam Bole 600 x 10 Rabbit Facility 50 \'i 11 Elev. 32 - North 600 t 13 Elev 16 - South 25 M- 1c 14 D.= haralised E# 700 ; ,
17 Stack Particulate 1000 1 9 18 Reactor Particulate >1000 5 'kj N,i + -l, Immediately after the scram, Willholt left the control room and began ! k.. '
g taking readings in various parts of the building. He then proceeded to ( @ .,
( the reactor building, only to be met by the reactor personnel, who were '
> r- -
E cvacuating the building.
If l
~y Mr. Donald Collins, Bealth Physics Supervisor and Wil2hoit's immediate .
supervisor, said that he received a telephone call at his home at about .-
8:45 p.m. He vne informed of the situation, and reportedly recommanded that the building be evacuated. He arrived at the site at about 9:00 p.m.
At about 8:45 p.m., according to Willhoit, the evacuation order was sounded .
over the Fenco intercom system. Willhoit took some readings in the reactor a building, and then vent out the south door of the reactor building to the $ l Process Building. Herecalledthatheobtainedareadingofabout80ar/hr , i in this area which he erroneously believed was due to Argon kl comin6 from i the head tank. In the process building, he found three men working (two in the machine shop, one in the boiler room), vbo had not beard the Fenco alarm i due to the noise of the machines in the building. He directed them to evne-uste to the control room. Immediately thereafter, the general evseustion clarm was sounded in all of the buildings. W111hoit stated that this could be clearly heard over the noise of the machinery in the Process Building.
He then proceeded to the head tank monitor and at two meters away, obtained j 6readingof5r/hr. He vent to the main gate at about.9:00 p.m., vbere the readingwasabout200nr/hr. He next vent back to the control room where c11 personnel had assembled prior to leaving the site. All persons, with the exception of W111boit and the reactor supervisors, evacuated the site, and assembled at the main gate. W111boit took some more readings shortly after 9:00 p.m., the results of which follow: a t
Reading in ar/hr Area I 3
1 h !
l H. P. Office 2 .. 4 !
R. P. Imb , 5 -
Hallway outside H. P. Office 5 l l Outside South Door 50 l
\
-3 \ s q
u
. +
- V. n g -)
N-
. u, , : a.
- g. *t-Area Reading in ar/hr
-,- .a At Decontaminatica Pad 230
~ Two Meters from Esad Tank Monitor $000 Reactor Control Room 40 Main Gate 200 WREC _
500 Eead Tank Exclusion Fence 4000 Film Badge Rack
- 2 Reactor hp - 1000
- At about 9:00 p.m., it was reported that all film badges from the film badge rack were removed and stored in a radiation. free area.
It should be noted that all of the readings noted above were subsequently '
- determined to be from direct radiatica emanating from the elevated head tank. Subsequent calculations determined that this tank housed approri. ?5 >
mately $000 curies of fission products, all of which was contained in the
. Primary coolant water. Eowever, inasmuch as the source of radiation was l
not known at the time of the incident, W111hoit and Collins reported that all persees in the building were immediately required to wear cannister-
! type masks. ' i l
Because the reading at the main gate (the first assembly area) was'200 A h- ar/hr, it was decided that all personnel, with the exception of W111hoit 1' and Collins, should get in their cars and drive to a guest house, owned "
+
^ by Westinghouse, about 1/3 of a mile away, known as the "Seubert Bouse". .*
Readings taken there by some of the evacuated personnel were found to be p nomal background. In the meantime, Collins and W111hoit, both wearing h masks, began (1) taking air samples throughout the buildings, (2) taking E smear easeles throughout the buildings, and (3) assembling instruments I for the purpose of setting up a counting srea in the " Fabian Eouse", a K seIond guest house that is about 1/4 of a mile away, and also owned by Westinghouse. At about 9:30 p.m., they were joined by R. J. Catlin, Manager of the Realth Physicist Group, who had been informed of the inci. !
delt by telephone. ,
In the meantime, presmably between 9:00 and 9:30 p.m., the security guards sat up roadblocks and flares on all roads leading to the reactor site.
Actingon orders from the supervisory personnel, the guards re. routed traffic. In addition, they were instructed to inform all non.est.ential personnel, who would be arriving in a few hours to begin the third shift, to remain at their homes. ;
I At the same time, the first air samples were taken outside the buildings, l Es were more direct reading surveys. Portable instruments and counting (quipment were set up at the Fabian Rouse. In addition, arrangements for the taking of urine semples of all persons were completed. Catlin reported that it was decided that all persons who were still at the Suebert Eouse
- should proceed by ear to the Fabian Eouse. Upon arrival at the Fabian Bouse, the following steps were takens (1) direct reading surveys were
- tak:n of all employees' cars. These included readings taken on the out.
sida, on the seats, on the steericg wheel, the clutch, the brake, and l
. the gas pedal; (2) smear samples were taken on the teps, the hoods and the trunks of all care; and (3) urine samples were taken from all persons who had been in the buildings at the time of the incident.
The above surveys were reportedly conducted by Carl Frita and the results cf the smear and air samples were counted in the Fabian Eouse. The results ,
of these surveys were negative, the direct reading surveys showed no in.
cretse in normal background count, and the smear samples indicated no re.
movable contamination.
l \
i 6
l
1 t
+
i' ,
j u , .
I l
8 !
At some time during the survey, W111hoit stated that he discovered a com-pany car that had been parked near the head tank and the exhaust stack at the time of the incident, but had been overlooked. He immediately checked it for contemination, but bone was found.
Under the direction of the E. P. personnel > various teams, in their own cars, setareas.
outlying out at about 10:00 p.a. to conduct direct reading surveys in It was reported that one team went north to Madison, Pennsylvania, a second went southeast to New Stanton, and a third went west to Yukon. The.results of all of these direct reading surveys were negative.
At the same time, air semples were taken in the downwind direction, Cetivity.
the results of which also revealed no detectable airborne (It should be noted, at this point, that at the time of the incident,thewindwasfromtheNortheastatabout3to5 miles /hr. There was also a slow drizzle of rain. Collins and Willhoit stated that the '.*-
! actual wind velocity was not obtained until the next day inasmuch as the l
indicating device was not working. They added that the 3 to 5 miles /hr l i
figure was obtained the next day from the Pittsburgh Airport.)
l Collins and Willhoit reported that the time from 12 midnight on was spent l taking various environmental surveys, and analyzing results. Surveys were cleo oceducted in the builAings as personnel were beginning to re enter to begin evaluation and cleanup operations.
The results of two direct read-ing surveys conducted during the morning of April 4,1960 follows )
j 3:15 A.M. >
j Lea Reading in ar/hr Gate Outside South Door of RSB 11
- Road #2 15 35 Road f3 Outside WREC Door 55 Outside Rear of Shop 45 On Road Adjacent to Head Tank Fence 170 on Road in Passageway Between Annez 800 and Boiler Room 80-90 5
- 30 A.M.
Gate 17 Road #1 20 South Door (RSB) 20 Road #2 26 Road #3 44 10' NW of WREC Door 38 At Outside of WREC Door On Road 3, Rear of Shop 32 l
On Road 3, SE of Bead Tank 150 On Road 3, IE of Head Tank 700 700 On Road NW of Process Building 90 On Road, Main Passage Between Boiler Room and Annex Area l
85 7
I l
f a d
. .g TREA'!9Girr OF C0f9tAMINATED WATER As was previously noted, all of the radiation levels emanated from the head tank, which was estimated to house about 5000 curies of fission Products in approximately 60,000 gallons of water.
During the preWinary investigation, the inspectors contacted f ,
Mr. David Gala, Socior Scientist, who had been given the responsibility of analysing, storing, and eventually disposing of the contaminated process water, the total of which amounts to about 250,000 gallons.
Gala stated that he received a phone call at his boas at about 11:30 p.m.
on April 3,1960 He stated that he arrived at the Sunbert House at ,
approximately 12:00 p.m. Be stated that it was immediately decided to obtainsomeE/andOE"ionexchangeresins. At approximately 1:00 a.m., .
a local representative of Rohm & Esas was contacted in order to have this resin for a PC deioniser ready in Philadelphia at about 8:00 a.m.
At about 2:00 e..m.,
a truck was dispatched to Philadelphia in order to s . .
Pick up the resin.
Gals stated that the first semple from the head tank was taken at 9:30 a.m.
on April 4, 1960 The sample was counted in a well-type countar which also listed below revealed 681,406 counts /$see.
figure revealed an activity of about /al. 2.4 x 107 aps/al.had an efficiency)of abou conversion of this On the afternocn of the date of the initial investigation (April 14,1960), ,
Gala provided the inspectors with the results of the latest analysis, -
which revealed that the activity had been decreased to apprartmately kx10-3;ae/al. Gals stated that this activity was still too high for release c,f the water and added that twelve 20,000 gallon tanks (which were on order) would be used to store the water.
STACK EFFIDElfT DISCHAROE [
capacity of the blower for the surge and head tanks is about The 2000rateg/ainute, ft. and the concentration of the effluent is enluated by a Kanne Chamber. At the time of the incident, the recorder of the Chamber was on the 10-9 amps sealm. Immediately after the fission product release, the pen vent off scale and it remained off scale until the selector switch was changed by a factor of 1000 to the 10-6 amps scale, some five minutes later. At this point, it immediately began recording and remained on this scale for about 35 minutes at which time the fan va shut off All of the activity released was believed to be due to A41,KrEandXel$3-135 inasmuch as environmental studies have been negative.
Catlin said that he had attempted to calculate the amount of activity re-leased from the head tank vent in the first burst by woriting back from anal /ses of activity in water samples taken after the incident. He put an upper limit of 350 curies on this release, but said that he believed that the actual amount should be closer to 200 curies.
ENVIRONMElffAL STUDIES As was previously noted, preliminary environmental studies taken between I approximately 9:00 p.m. on April 3,1960, and 9:00 a.m. on April k, 1960, revealed no contamination. It was reported that on the morning of April k, ,
I i
I i
I l
i r
j 1
a 1960, I.P. Personnel began taking further environmental samples and analyzing them. W111hoit and Collins stated that approximately 25 vegeta-tion samples, 25 and samples, and 12 standing water easpies (resulting from the drizzling rain) were taken and analysed. In addition, the licensee's fallout detectors (consisting of gun paper) were analyzed. Analyses of l all of these samples were negative. )
On April 4,1960, Nuclear Science and Engineering Company was contacted !
and informed of the incident. It was reported that this company was ,,
chosen to perform au independent environmental survey for two reasons:
(1) ISkEC had performsd a preliminary on-site and off-site study, and (2) Westinghouse wanted an independent etudy taken.
PERSONNEL MONITORIE i
The licensee engloys a voekly beta, gamma and a weekly neutron badge that is supplied by R. S. landauer and Compmay. In addition, pocket chambers are also used. 4 I
W111hoit and Collins presented the inspectors with copies of film badge !
and pocket chamber results for the period involving the incident. The crposures listed below are those involving persons who received an excess cf 150 mrom.
- l Badge Reading Pocket 7-Day Total For Period Chamber For Baas 3/28 . A/6/60 Reading h/2-h/8/60 Cangilla 320 mrea 55 mr 55 nr Chapla 190 219 " 254 "
Collins 185 191 " 191 "
Cox "
155 141
- 136 "
Lynch 185 196 " s cardner 195 195 " 150 "
Semethy "
185 150 " 265 "
W111hoit 240 181 " 206 "
Cheshire "
150 125 " 106 "
e In addition, personnel were given medical examinations by Westinghouse's Medical Department. These examinations included a routine physical, a blood test, a urinalysis, a thyroid sean, and a nasal swab count. All rGaults were reportedly negative. . (See Exhibit " ".)
IFORMATION GATHERED DURIE COPHITTEE MEETIES At the investigation committee meeting with the W'm staff, further informa-tion relative to the indicent developed.
- 1. Dr. Schultz exhibited the Brush recordings taken immediately prior
- to, and during the incident. Exhibit "A" had been taken prior to this 16-cident. The greatest amount of noise seen in previous runs had occurred during a 40 megawatt run at a flow of 7000 gal / min. On the evening of the ircident vben the pile power initially reached 35 megawatts at the flow of ,
5250 gym, the perturbations on the Brush recorder were approximately equal I to those achieved in the 40 megawatt, 7000 gpm operation. Vben the power ;
1svel went initially to 38 megavatts, the Brush re* corder perturbations i
i 5
a
}t .
33 p i. y }
'a i
, . s c 3
i ersased tt 600 kw for approximately 6 eeoceds. For apprart-ately minuter, the t%sh recorder indiented,at first, very little perturba- } }
tion, and then a gradual increase again, and as the reactor power rose g to 36 megawatts for the second time, the Brush recordings indicated 500 kw perturbations. b ,
- 1 a+
2 Mr. Boumierle has done some preliminary calenlations that indicate '
that the initial drop in power was due to a sudden void being formed in , * ~
'"y the reactor equivalent to approximately the total volume of one fuel element. He has checked his calculations of the loss in reactivity due g
to this void formation against the worth of the rod withirawal necessary ,
to stop the descent in power and to bring the reactor back up to power J L'
again. Be said that these two effects are about equal and opposite in magnitude. Therodmovementwasworthapproximately/.k%,whereasvoiding {)
a channel equals approxicately .3%, and the displacement of fuel in the channel may be worth another . 1%. t
{ !l I
[4 The M staff stated that at this time they cannot be sure of the effect on the initial molting of the fuel element or of the return to 38 T megaestts for the second time. They are considering the possibility that this return may have re eelted and then re-solidified the broken half of ( ,
the element. i
- 3. Both Bianci and Pollono said that they had not been given any ,
i special instruction prior to this experiment with respect to sudden re- ,
activity changes. i
(
Pollono had been at the M for only about three months. He had ;
formerly been at the Westinghouse plant in Kansas City and had been in ,' 3 g
charge of a testing operation there. When he originally came to the M ,
he was put on shift for approximately one month with Mr. George Geisler, t who was then acting as a shift supervisor. In answer to the question ,
whether he had any q=mina about Pollono's qualifications, Geislar replied 4 that he had confidence in Pollono since Pollono followed instructions ex- 4 P11citly when he (Pollono) did not know the answer himself. ;
Bianci's background consisted of four years working as both an I clectronic technician, and as an extra operator at Bettis criticals, the T production test facility, and the reactivity facility. He has been at the M site for approximately l} years and has been licensed to operate the M since October, 1959. ,
1 Bianci said that during normal operation, if he observed such a drop in power on the neutron power level recorder, he would g have brought the reactor back to power.
- k. Af+er the incident, circulation of purge water through the primary i system brought the radiation levels at the pile top down to acceptable work-ing limits. The reactor head was removed and cleaned. Protective paper and polyethylene sheets were placed over all the equipment inside the con-tainment shell. Discharge operations were initiated, and when element No. L 5-6 was discharged, only the top half of the element came out. It was placed in the canal and then into a hot cell. The committee examined the piece of this element. The bottom part of it indicated melting had occurred over a length of several inches. The cladding of the outer cylinder appeared as though it had peeled back from the meat, indicating the possibility of a poor metallurgical bond between the cladding and the meat.
In the center of this element, there had been a 1-1/8" 0.D. basket containinga1/2"0.D.aluminumrodwith1/8"0.D.quartzcapsules. These p - .
t n
p .,.
j espoules held clippings of nickel wire which were being used to get a flus profile through this element. There was to evidesee of this material in the portion of the fuel element observed in the hot cell.
The specifications for the fuel elements were eraMned by the com-mittee. They call for a surface finish free from scratches more than 003" deep, and in deviation from straightness of no more than .015". They further specify that the bond c1mading shall be free of blisters following a standard 9300 7, 3-hour blister anneal.
The shop inspection procedures state that outside diameter measure.
ments and wall thickness measurements aball b taken every six inches.
Finally, there shall be a measurement of the amount of straightness or bow-ing of the element.
The fuel element records indicated that this element had been re-ceived on November 20, 1959, and it had been inspected and approved by Westinghouse.
Mr. Rice said that initially, in the elements made by Clovite, there had been voids in the braze as long as 10" - 12". However, he said that this problem had been solved and u new process employed. Be stated that he had personally inspected several elements made with the new process
- and that there had been approved. Mr. Rice continued that he had on two 1 occasions made an unannounced inspection in Clevite's plant to check on the fuel fabrication. ,
2 Mr. Rice said that in the first shipment of 8 elements, there had been marked bowing noted, that is, more than .020". These elements had been inspected in Clevite's plant prior to shipment, and had been certified to be straight. In a simple experiment it was then determined that dropping the elements a short distance could bow them since the elements are made of soft altamintaa.
As a result of the bowing of the elements in this first shipment, ~
further shipments were ande by station vagon rather than by public carrier, and an inspection was made just prior and immediately after the shipment.
Mr. Rice was of the opinion that this element was a satisfactory element as far as Westinghouse's inspection was concerned.
The coenaittee considered the possibility that the element might not have been blister-annealed, and a call was made to Mr. Weber of Clavite during the committee meeting. Mr. Weber reportedly stated to Mr. Rice that every plate is blister-annealed as part of the regular manufacturing process, and be felt that there is no way that any one plate could have missed the annealing operation. However, it was said that no records are maintained that vould verify that each plate is annealed. {
VESTIfCHOUSE PLAN 3 To insure that this particular accident does not occur to this reactor in the same fashion again, the WR staff is planning the following steps:
- 1. Dr. Schultz said that he does not plan to reebarge any of the i
fuel elements into the reactor which had been there at the time of the incident. He said that there exists the possibility that in the discharg.
Eng operations, the elements may become scratched or damaged in some fashion, .
and that he vill not risk the possibility of piscing a damaged fuel element ,
in the reactor.
1 l
}
1
,5I.
. y. . ,
h ~
A There are approxiantely 100 new fuel elements manufactured by Clovite in the M storage vanit. Before installation of any of these elecents into the reactor, the locking pins and the end spiders will be removed and all of the fuel cylinders of all elements will be examined.
In the future, Westinghouse will have a resident inspector at the fabricator's plant.
2 Dr. Schultz stated that he is inviting a n eber of heat transfer and fluid flow experts to examine all of the data associated with this incident. A committee of these people will be formed.
Dr. Schultz also said that be intends having metallurgists C=ina the discharged fuel element in an effort to determine whether there had been more than the specified enount of uranium in this element, -4: 1 or whether there had been a good metallurgical bond between the cladding and the meat, or whether there is apr evidence of any foreign unterial .
which might indicate a flow blockage. t >
3 Dr. Schultz said that the procedure manual used by the operating group would be reviewed and brought up to date so that supervisors and operators are aware of the dangers in bringing a reactor back up to power
- after a drop in power from an m astermined cause.
I
- 4. The present plans for cleanup after this incident are being closely aminad by W3 8s insurers. Westinghouse scientists and consultants from the insurance group are working together on the problem of disposal of ap.
- proxiantely one quarter million gallana of contaminated water. At present, ,
j the water is being circulated from the retention basin through ion exchange t- etiumns to holdup tanks. The final cleanup of this veter may have to be I l done by evaporation in order to bring the effluent down to the limits .[
i allowed by the State of Pennsylvania for discharge of radioactive waste.
- t i-SUPPIJNNfARY IWGulATION ,
t Subsequent to the investigating committee's meeting, visits to the M site by John R. Sears and phone calls from the M staff have disclosed the following inforantion:
- 1. As of May 6, 1960, it is reported that the core has been completely unloaded of fuel, control rods and experiments. There had been four capsule experiments in the reactor at the time of the incident and one loop had been operatin5 The capsule experiments had been either in reflector or outer fuel ring locat$ons. These have now all been discharged from the reactor and ev==ia G in the canal. There is no apparent daange, and the M staff reports that they are positive that the experiments were not involved in i
the incident.
i l
! 2. The group of heat transfer and fluid flow consultants has met at l the M site. This group included Messrs. Iatourneau, Johnson and Butler l l- from Bettis, Binford from GtNL, and others. )
I l
! The group first calculated that a good fuel element, in this !
location at the stated conditions, would be a factor of 5 away from burn- !
out. The group then assumed the worst combination of variables; for ex-ample they assumed the flow through this che.nnel was cut by 50%, and they used 6,0% of the best values in the Savannah River correlations, and they still calculated that this element was a factor of 2 away from burn out. 1 It is reported that the group feels that the incident was due to a cladding fcilure. j
-12 l
l i'
,; s. .c wD.
. G;,. ... e
- , 1 ; e d. ,
l 3 The melted botton quarter of the ruptured' element has been drilled out of its shroud tube and is now in the bot cell. Elements from locations L-7-5, L-5-5, and L-7-6, from the center fuel ring, and the element from location L-6-4, adjacent to the ruptured element location, have been placed in the same hot cell. It is reported that all channels in all elements
! appear clear and normal. Element from location L-7-6, which was the original instrumented element and had been in the reactor through a few l cycles, is reported to have a slightly reddish-brown color. m people l bare seen such discoloration on elements before, however, and feel that it is due simply to some impurity in the water and is not related to any l burning of fuel aleasnts.
I
- k. Messrs. Dan Thoems of Bettis and Jack Cunningham of CENL have )
Cxamined the elements in the bot cell. Cunningban will check the outside j
[
dimensions and the straightness of the four whole elements in the bot cell. - 1 It is planned that the top half of the ruptured element vill be sectioned to examine the bond between clad and meat. Further metallurgical studies any then be done on the remnants of the ruptured element.
5 As of May 5, cleanup operations of the primary coolant system outside the containment shell had progresse1 to the point that this part j of the system was considered clean. Rowever, it is reported that some particles of highly active material have evidently become lodged in piping joints in the primary piping inside the containment shell, and further flushing vill be necessary to clean up the system. l I
- 6. The m staff is considering moving the location of the fission ;
product monitor in an effort to improve its response. -
l They are considering methods for decreasing the amount of activity i discharged from the head tank vent in an incident, but they have not yet
- l arrived at a satisfactory solution. ,
i 4
Attachments; Exhibits A thru F l
i l
I e
!