ML19343A171

Forwards Info Re Facility Location,Restricted Areas,Emission Monitoring,Sampling Sys & Seismic Test,In Response to NRC 800731 Ltr. Atmospheric Dispersion of Ar-41 from Ucla Nuclear Reactor Encl

ML19343A171
Person / Time
Site:	05000142
Issue date:	09/05/1980
From:	Wegst W CALIFORNIA, UNIV. OF, LOS ANGELES, CA
To:	Tedesco R Office of Nuclear Reactor Regulation
Shared Package
ML19343A172	List: ML19343A171 ML19343A173
References
NUDOCS 8009160321
Download: ML19343A171 (30)

Text

-, ., - . . . - . . .- - , .

M i

' UNIVERSITY OF CALIFORNIA, LOS ANGELES l fl 1, \

BEREELEY

• DAVIS

• IRVINE

• LOS ANCELES

• RIVERSIDE

• SAN DIECO

• SAN FRANCISCO [ '-W ,, f - SANTA BARRARA

• SANTA CRt;Z ll v

. //

OFFICE oF ENVIRONMENTAL HEALTH AND SAFETY T11E CENTER FOR THE liEALTH SCIENCES LOS ANGELES, CALIFORNIA 90024 September 5, 1980 ROS:C1705, 1712G Docket No.

i- 50-142

'g Robert L.-Tedsco, Assistant Director for Licensing i Division of Licensing Office of Nuclear Reactor Regulation

, Washington, D.C. 20555

Dear Mr. Tedesco:

Attached are UCLA's answers to the list of 14 questions received from your office dated July 31, 1980. As you can see we have gone to '

f some great lengths to answer the questions in great detail, and hence have not been able to respond before this time.

j We believe we have clearly answered all of the questions, but if you need clarification of any of our answers, please contact Neill

Ostrander at.(213) 825-2040.

4-

, Very truly yours, Wh f W N4 W Walter F. We se, .h.D

! Director of Research and Occupational Safety  ;

WFW/lc Enclosure-

+

4

-O O SI t (-

i r m

8 0 0 916 03.W I

September 5, 1980 This will certify that the following responses have been reviewed by Professors Catton and Pomraning, Messers Hornor and Zane; and by Dr. Wegst, and therefore by a majority of the Radiation Use Committee.

A. Zane, Secretary 4 r -

y Radiation Use Co=mittee k-D 1

l l

l' Question 1. 8-22-80 In order to consider posting .the roof of the htth/ Science ' building as a "testticted area" it would be necessary to first considet .telocding the meteorology labordory facalties from .that bu.itding. Is it conceivable for the metco.tology facitrties to be moved from the Ma.th/ Science buildhg i to ano.ther buildhg. Would .the a.sttonomy facilities also have to be'.temoved

, or a,te they princ.ipatty alttized when the reactors is not operating?

Ate there advantages to NEL h having the meteorology fac,ilities located on the roof af the Ma.th/ Science budding? Ate student programs ot Univer-

.sity (reactor) programs incorporated in the u.se of this facility?

The meteorology facilities consist of special building features as well as the instrumentation. The original Math / Science building provided a ninth floor structure, the roof (tenth floor) of which bases much of the weather observation equipment. The station is located approximately 172 feet ENE:of the reactor stack and is equiped with telescopic tracking and telem-r etry equipment. The Math / Science Addition includes a nine story (ca.100 ft) precipitation shaft as an integral part of the building. The upper ter-minus of the shaft is located in a tenth floor structure approximately 46 feet north of the reactor stack. The relocation of these facilities is not practical.

The observatory facilities of the Astronomy Department are used only at night when the reactor is not operating. The planetarium, approximately 180 feet east of the stack is used for classes and discussion sessions during the daylight hours. Usage is scheduled by the Astronomy department and is 40 to 56 hours6.481481e-4 days <br />0.0156 hours <br />9.259259e-5 weeks <br />2.1308e-5 months <br /> per quarter during the fall, winter, and spring quarters; 2 to 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> in the summer quarter. Taking three quarters at 60 hours6.944444e-4 days <br />0.0167 hours <br />9.920635e-5 weeks <br />2.283e-5 months <br /> and one quarter at 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />, the usage is 190 hours0.0022 days <br />0.0528 hours <br />3.141534e-4 weeks <br />7.2295e-5 months <br /> per year or 8.1% of the 45 hour5.208333e-4 days <br />0.0125 hours <br />7.440476e-5 weeks <br />1.71225e-5 months <br /> / week, 52 week year.

The NEL uses the meteorological facilities infrequently to spot-check wind direction and speed. The station does not function as a publicly reporting weather station and does not compile tabular records of the strip chart recordings. There are no formal student programs in nuclear engineering that utilize the meteorological facilities.

J i.

o

-r y e ,+ w ..w- =.>w+ err---. ,e--. -, .=,,*e. --

-me-- y *-. e--,- y

Question 2. 8-27-80 When standing on the ,toof of the htth/ Science building and looking at the .tcaeto,t stack, it is not appstent that the veditation syster is opetating. It is fu.tthe.1 not appa, tent that the reactor is operating.

As the Azth/ Science roof e.ea is an open access ::one, it would appeat that inforration attestis g to the status of reaetc.t operation wcaid be hcipfui. Could a blinking red light be instatied togethet with a peAranent sign posted where it can easily be .tsid from the roof of .the Mtth/ Science that (to wit) "when the red light is blinking .the UCLA rescatch reaetor is in ope.tation. Please do not Loitet in this a. tea."

The ventilation system operates 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> per day including weekends and hol idays . Operation is evidenced at the roof top by the waving string tied to the exit grill . On the rare occasions that the system is down for maintanance, the reactor does not operate.

A blinking red liiht and sign are technically feasible. UCLA considers the hel pfulness o / that information to be questionable. It would obviously suggest unspecified hazards associated with the research reactor, a suggestion this is not compatible with the hazard level .

UCLA maintains that the research reactor is operated within the limits of 10 CFR 20.105. During the two-year environmental monitoring program (March 1976 tnrough February 1978) the average value of dosimeters within 300 feet of the stack was about 9 mr/ quarter or 36 mr/yr (y-only).

(See response to question 9.) If attributed entirely to reactor operations and scaled to the 1979 operating level, the projected y dose ra e would be 72 mr/yr. Assuming a semi-infinite cloud in which the y-dose rate is 74.6% of the total dose rate, the estimated total S + y dose rate is about 97 mr/yr. Accordingly, none of the rooftop areas warrant posting under the 500 mr/yr provision of 10 CFR 20.105(a).

The reactor does not operate continuously and the shorter term limits of 10 CFR 20.105 should be considered. During the two year period 1976-1978, the reactor produced an average of 14.5 megawatt hours, or produced at 100 kw for an equivalent operating time of 145 hours0.00168 days <br />0.0403 hours <br />2.397487e-4 weeks <br />5.51725e-5 months <br />.

Assuming that this effective operating time yielded a rooftop dose rate of 35 mr/yr (y), the hourly rate when the reactor is operating is 0.248 mr (y) or 0.33 mr/hr (S + y). This is well within the 2 mr/hr limit of 10 CFR 20.105 (b)(1).

The reactor operates much less than 45 hours5.208333e-4 days <br />0.0125 hours <br />7.440476e-5 weeks <br />1.71225e-5 months <br /> a week, so that the weekly d' ~ a rate is less than 0.33 x 45 = 15 mr/ week. This is well within the limit of 10 CFR 20.105(b)(2).

The UCLA policy in regard to defining restricted areas is expressed in the UCLA Radiation Safety Handbook, p.20, as: ,

"Any areas such as corridors and waiting rooms which are open to non-restricted personnel shall have a radiation level less than .6 millirems per hour. If the level exceeds this, the area must be classed as Restricted."

i a

1. 2 - page 2 This requirement is more stringent than the " Radiation Area" definition of 10 CFR 20.202(b)(2) which stipulates a limit of 5 mrem /hr or 100 mrem in any 5 consecutive days.

UCLA does not wish to provide signs or displays that may stir apprehensions about a particular facility when that facility is operating within legal limits and within the uniform and more restrictive campus policy.

Question 3. 8-22-80 To .teflect tite measures taken by UCLA, tite Teclutical Specifications will specif;{ that titz roof containing tite stack shall be dectated a ".testticted atea".

UCLA will accept the declaration in some form with the understanding that the definition of " restricted area" is unrelated to the " Restricted Area" definition of 10 CFR 20.3(14). The phrasing of the declaration should be such as to avoid future misinterpretation of the declaration. One sug-gested form: "The eighth floor in the immediate vicinity of the venti-lation exhaust stack is posted and controlled to a degree commensurate with the potential exposure risk involved upon entry of that area. It is not a Restricted area in the sense of 10 CFR 20.3(14)."

Presumably this statement will become another paragraph of the proposed Technical Specifications, 3.8.2 " Dose in Unrestricted Areas."

Ouestion 4. 8-22-80 It wa.s noted during the site visit that equ,ipmcast for propcsed Atgon-41 holdup tanks are being obtained and stored. Please provide a schedule for completion af this facd.t.ty.

There is no calendar schedule for the completion of the hold-up system.

The pacing is dictated by the following events:

(a) Approval of License renewal .

(b) Approval of a construction plan by the Campus Architects and Engineers.

(c) Approval of the plan by the Radiation Use Committee.

(d) Preparation and presentation of the proposed Amendment before the Campus Radiatior Safety Committee.

(e) Submission of the proposed Amendment to the Nuclear Regulatory Commission.

(f) Approval of the proposed Amendment by the Nuclear Regulatory Commission.

(g) Completion of the construction.

These steps are not entirely sequential and some overlap is possible in those steps internal to UCLA. The time lapse between steps (e) and (f) is quite uncertain, and not controllable by UCLA.

Question 5. 8-22-80 Wha.t is .the reliability of emissions monitoring equ.ipment and accu):acy of calib.taten standa,tds and methods? Who checks calib.tations? Mcw are records ma.intained?

Reliability of equipment: A 4.3 liter flow-through ion chamber manufac-tured by Carry is used to continuously monitor the ventilation system exhaust when the reactor is operating. The measuring device is a Keithley model 640 vibrating reed electrometer operated in the integrate current mode on the 1 volt range. The electrometer is of solid state design and is highly reliable. Past history of the above equipment shows only preven-tive maintenance performed during the semi-annual-calibrations.

A gamma spectrometer standard model R-34 source kit manuftetured by I.C.N.

• of Irvine Cal was used to calibrate a 3 inch by 3 inch sodium iodide scin-tillator. This standard bears the serial number 534 and was calibrated in 1-7-73.

This kit was used to . determine the efficiency of the scintillator from 60 KEV-to 1.33 MEV. Two curves were determined. One for a disc source on th'e crystal face,_and a second for a 545 ml gas sample bottle using the standards as theoretical point sources simulatino the 545 mi gas sample

- bottle.

A steel cylinder of 1639 cubic centimeter capacity instrumented with ten

. gold foils was filled with pure argon .to a pressure of 225 psig and ir-radiated in the thermal column for 30 minutes at 250 watts.

Grab-samples of the gas were taken and counted on the scintillator. The gas sample was then passed through various ion chambers to be calibrated, and ;their readings and the time recorded. The gold foil data was collected using the scintillator and the average neutron flux determined. The con-4 centration of Ar-41 was then determined and compared against the grab sample previously counted. A concentration vs instrument scale reading for the various ion chambers was thereby determined.

Three calibration runs as described above were conducted. The average

! measured efficiency of the scintillator was 1.20% with a confidence level to 1 sigma of .07%. This compares favorably with a theoretical model which

! predicted 1.2% at 1.29 MEV.

Since the preparation, the experiment, and the data reduction requires i approximately 1 man week, an alternative calibration technique was developed j using a tritium gas standard. This allows the tritium monitor and the

argon monitor to be calibrated simultaneously and requires only 1/2 man 3 day.

The experimental value determined from tritium gas is (2. 350 026) x 108 microcuries per rilliliter per amp for the 4300 ml chamber. This compares

'favoraoly with a theoretical value of 2.354 x 10' microcuries per milli-liter per amp. The large chamber captures nearly all of the energy of the tritium- B's .(18.61 key max) and only a fraction of the energy of the.

2 O

.#5 - page 2 argon-41 S's (1.20 MEV). Nevertheless, that fraction is much larger than the total energy of the tritium S's, and equal curie concentrations of tritium and argon-41 yield larger ion currents with argon-41. Thus, for the 4300 ml ion chamber 1 pCi/ml- (tritium)+ 4.255 x 10. amps 1 pCi/ml ( Ar-41) + 3.425 x 10 8 amps e 10'# ohms and yield a practi-The currents are measured acro $s a resistor of cal scale factor' of 2.92 x 10- pCi/ml per millivolt for argon-41. The tritium calibration checks the performance of ion chamber and electrometer.

The electrometer alone can be (and is) checked against a standard current source.

.The reactor supervisor has performed the tritium calibrations on a semi-annual basis as required by the Technical Specifications. Grab samples both in the exhaust fan. intake plenum (on the eighth floor) and at the output of the ion chamber are also taken periodically and compared with the readings of the argon sampling system. This verifies the chamber cali-bration and allows a correction factor to be applied should the stack con-centration differ from the sample line concentration.

The argon' emissions are determined by the reactor -supervisor who periodi-

- cally reduces the data on the Ar-41 chart using a compensating polar planim-eter and records this data in the reactor supervisor's documentation log.

Question 6. 8-22-80 What was tite .teliabd.ity of the insttwnentation and sampling syste.u used fo,t tite SFgdispe,tsion tests conducted by UCLA?

The SF6 work performed at UCLA was analyzed by Dr. Fred Shair of the California Institute of Technology. Over the last eight years, he has done SF6 dis-persion tests for the EPA, the California Air Resources Board and others.

Although a complete statistical analysis has not been performed, the stated accuracy with eight standards is within +13%.

The sampling and analysis procedures are described in Chapter 2 of fir. Rubin's thesis. A copy of that thesis is included in this response.

It is evident from the scatter in wind direction and speed data (Rubin, Figure 3) that turbulence is pronounced. The sampled concentrations show large fluctuations that might be expected under such turbulent conditions.

In particular, the response to Question 8 indicates that none of the prin-ciple sampling locations (except the stack exit) were on the plume center line and large fluctuations should be expected in the fringe area. The data are highly consistent at the stack, and the general decrease in con-centration with distance (as evidenced by averaged data) are powerful evi-dence of real fluctuations, not random experimental errors.

l l

l l

I G

e

t 4%%

e+ 4 9 io eE Evitu 1,0~

TEST TARGET (MT-3)

$+#s '

l.0 lffEn 814 i 5 m gn mm i-l,l h," bb

.8 l1.25 1.4 1.6 i

e 6" =

MICROCOPY RESOLUTION TEST CHART

1. 4 ' 4$ 4%

• %f>///;/ $4AO

g 5,,

///

L-c o

==. = = =a- .= .

F--

t

+ g> 0

• $ + /g, e4

\\\\! //// s k

++<>p,jj ,M eE Ev <e 1,em r.4([4 TEST TARGET (MT-3) l.0 502 EM

!!$ El I,l

[f IIM .8 1.25 1.4 ' 1.6

==

l

e

MICROCOPY RESOLUTION TEST CHART

~

h' N O /% 4 4g$7k 'S4A4 Azzr f 7,  :

~ ~

- ~ ' '

s, n . x uc.;.:n , a- c-

Question 7. 8-22-80 What is tite reliabdcty arid accuracy of tite fit.m badges at various levels arid types of exposure used by UCLA?

Upon receipt of a new emulsion, calibration curves for y and S exposures are constructed in the following way:

For y exposures from approximately 10 mr to 3 R the calibration curve is established using a 20.16 mg Radium-226 needle. The needle was certified by Radium Corp. of America, Cert. No. 20412, 1962. The needle is placed on a board with ten te::t films in standard metal film badge holders which are placed at calculated distances from the needle. Films are exposed for varying lengths of time to obtain exposures from 17 mr to 3100 mr.

(Actual points are 16.9, 25.4, 50.8,102,136, 254, 508,1016, 3096 mr.)

The films are processed with unexposea (background) films using a standar-dized de:velopment technique. A calibration curve, plotting calculated exposure vs film density, as read on a Photovolt densitometer, is then con-structed. Shifts in emulsion performance (aging, etc.) are determined on a monthly basis by exposing five check film; to 102, 254, 508 and 1016 mr as well as one exposure unknown to the film badge technician. Density readings of these films must agree within 110% of the values on the cali-bration curve or a new- ten point calibration curve must be constructed.

The accuracy for the exposure range of 100 to 1000 mr is therefore 110"..

For y exposures above 3000 mr, a number of films in metal holders, are y

exposed to Cesium-137 y rays on the EH&S vertical calibration range. Exposures are measured with Condenser-R chambers. A calibration curve is estab-lished for high y exposures approximately every three years.

S exposures: A set of 10 films are exposed to beta radiation from a thick slab of depleted uranium (dose rate at the surface of the slab was cer-ti fied by supplier; Eberline Inc. Sept.1972). The calibration curve covers beta exposures from 35 to 22000 mr.

Reliability in terms of badge recovery is nearly 100% for those badges issued to the nuclear energy laboratory. Bad film receipt is rare and is caught in the initial calibration stage. Processing errors are neg-ligible. The threshold sensitivity is approximately 10 mrem (y) and 20 mrem (S), Badges are changed either monthly or quarterly depending upon expected exposure. Thus for low radiation levels, the threshold dose rate is 10 mrem (y) per quarter and 20 mrem (S) per quarter.

t

Question 8. 8-22-80 in the request for Amendment No.10, UCLA wtclized an overati enLssions reduction of 460 .that tats calculated on .the bases of reactor use, available meteorologicai dilution models, wind dLtection btfor.mation, and building occupancy information. In the httervening yeats, meteo.tological .statwas have been established much closer to UCLA providing more applicable informa-tion, more rigo.tous meteorological mixing and ttartsport mcdets a?x avad-able, UCLA has conducted meteorological studies and additional inforana-tion is possibly avaLiable on building occupancy. With .the use of this more curtertt htfo. mation, plea.se prepare calculaticrts to predict the con-centtation of At-41 at the Math /Seience building from stack emissions due to reactor opetation. Utilize the megawatt hours of opetation in 1979 for reactor operating time.

Also please subnit any actual .telease data avaliable such as from smoke or vapor tests, and compate tcith the calculated results.

Use roof occupancy factors of 10% and 100%, and toind direction factors af 100% and 60%. Also consider infhtite immersion, and non-infinite immersion factors . indicated in the footnote associated telth the calculatiotts for At-41 concenttations provided in 10 CFR 20 Appendix B.

Meteorology j The distance from UCLA to reporting meteorological stations has not markedly altered from 1975 to today. The 1975 data was based upon an Air Pollution Control District ( APCD) station located on Westwood Boulevard and Pico Boulevard, approximately 2.3 miles SSE of the reactor site. The availability of alternate and more proximate meteorological data was reviewed in late 1979. No new options could be identified. Availabl e observations are in accord with older qualitative statements by the UCLA Department of Meteorology that a southwesterly wind prevails 60% of the daylight hours at the UCLA campus.

The assistance of the UCLA Department of Meteorology was sought in identifying and characterizing a transport model applicable to the NEL exhaust vent. Professor M. G. Wurtele kindly responded with the description following the text response to Question 8. The implications of that jet model are discussed in the following paragraphs. An alternative Gaussian model is described in subsequent paragraphs.

The Jet Transport and Related Modeh The exhaust velocity from the stack is bar ed upon 14000 CFM over 'a l duct area S im.has by 34 inches or 41.2 ft/sec (12.5 m/s). The average speed of the afternoon SW wind is 6.53 knots (3.36 m/s), at least as observed by Rubin in his 1976 study [1]. The vent exit radius is taken as (A/n)b where A is the cross sectional area. Hence R, = 1.343 feet (0.41 m) .

i l

f

1. 8 - page 2 The further numerical results are based upon a SW wind so that the plume departs to the NE. In Rubin's work [1], the average plume direction was raore easterly'll to 10') and hence further from the MSA int?ke than

~

assumed here.

For the bent-over jet, a location or observer is either within the plume eavelope or is external to the envelope. More specifically, if a point P is located at perpindicular distance r from the jet centerline, then P(r) will be within the plume if r < R and external to the plume if r>R. The plume radius is R computed as yz =.5 z, where z is the plume elevation relative to the stack. The equations provided by Professor Wurtele have been evaluated for three locations; the location letter names of Rubin are adopted.

Location E. At the parapet, NE of the stack at a distance of 12.28 meters. The head of a six foot observer will be 7 feet (2.13m) above the stack.

Location G. The center of the MSA intake ventilator at an elevation of 3.5 feet (1.07m) relative to the stack exit.

The location is 20.88 meters from the stack and 15* off (North) from the prevailing wind direction.

Location H. The head of a 6 foot observer standing on the highest leve area of the meteorological staticn. The observer is .,2 feet (52.42m) from the stack, the elevation is 6.95m relative to the stack, and the angular offset is 23" east of the plume path.

The following table presents the plume rise z, the radius R, the vertical component z-h and horizontal component Y of the distance r from the plume centerline. The distance r is given by r=[(z-h)2+y23h ,

Jet Plume Geometry for Selected Observers

( All quantities in meters)

Location z R z-h Y r r-R E 7.03 3 .51 4.90 0 4.90 1.39 G 8.00 4.00 6.93 5.40 8.79 4.79 H li.10 5.55 4.15 20.48 20.90 15.35 l

In all cases, r > R and.the locations (observers) are outside the plume. According to the theory, the concentration seen by the observers would be zero.

Gaussian Modification of the Jet Plume With .he simulant SF,, Rubin found (and measured) down wind concentrations of SFs at each of the down wind locations. Eighteen samples were taken (6 per day, 3 different days), at each location. Althougn some were very Y

1. 8 - page 3 low, none were zero.

Under the ocean breeze that characterizes the usual south westerly wind, the lower atmosphere is extremely turbulent. Wind speed and direction fluctuate markedly so tnat the jet 91ume is driven randomly in elevation and azimuth. On the basis of observed dir:ctional fluctuations (c = 29.5'), the atmosphere would be classified as stabil;*y class A, extremely unstable.

With the assumption that the turbulence perturbes the uniform jet into a Gaussian radial (Rayleigh) distribution about the mean jet centerline, it is interesting to compare the radius of the jet with the Gaussian o as given in Reg. Guide 1.111 at 100 meters.

R (jet,100m) = 7.07 m o (class A,100m) = 16.20 m Evidently if the plume has significant Gaussian characteristics, its characteristic dimension is given by a = 2.3R.

The (inverse) dilution factor is then given by:

2 -

f ., c(r) , 2V (R e

-r 2/2a 2 +e -r /20 2 2 c a c _

The distribution extends to infinity and the second term represents an image plume at elevation -(z +h) relative to the observer.

The preceding equation for the dilution factor yields results that compare well with average values of the Rubin work.

Comparison of Guassian Model With Experimental Observations (Rubin 1976) location a r r1 f(Eq) f( Exptl )

• E 8.07 4.90 9.16 154 198 G 9.20 8.79 10.56 235 303 H 12.76 20.90 27.30 1430 1670

• Rubin uses the reciprocal dilution factor. The value here is the reciprocal of tFe mean value of 18 dilution factors (f 2) at each location.

O,:-

l

1. 8 - page 4~

j Considering the complexity of the actual geometry, the agreement is really excellent. The Gaussian model will slightly over-predict the downwind concentrations relative to the observational data.

1 The. foregoing results are not directly comparable to the emis-i sions reduction factor of 460 calculated in the application for Amendment 10. The 460 figure represented an over-all average effect that included a reactor utilization factor, wind factors (speed and direction frequencies), and occupancy factors. In calculating the concentration at the ventilator intake of the Math-Science Addition, the predicted direct dilution at that location was estimated to be 1/0.0411 = 24.3 with the wind from the SW, the reactor operating, and

.'. the occupancy 100%. The 24.3 figure may be compared with the 235 to i

303 figure shown in the preceeding table for location G. The latter results indicate a 10-fold reduction in concentration relative to

that used in the application for Amendment 10.

Assuming 5% operating time, and using the more conservative

, dilution factors of the Gaussian model, the effective dilution factors and annually everaged concentrations of argon-41 at the three locations are:

Dilution Factor Concentration f pCi/ml E (at the parapet) 2340 5.1 x 10~'

G (Vent Intake) 4700 2.6 x 10~'

H (Met Station) 28600 4.2 x 10~"

These assume a SW wind frequency of 100% and 100% occupancy. The

concentrations are relative to 1.2 x 10~5 pCi/ml at the stack, and would

, be proportionately lower for. reduced wind frequency and/or occupancy.

Immersion in semi-infinite clouds of these concentrations would yield the following dose rates. '

E (at the parapet)  : 64 mr/yr

!- G (Vent Intake)  : 33 mr/yr H (Met Station)  : 5 mr/yr Again, these are for 5% reactor utilization,100% SW wind, and 100%

- occupancy.

For the interior occupants of the Math Science Addition, the annual average concentration is assumed to be uniform and equal to the concentration at the ventilation intake (2.6 x 10 SpCi/ml.) The S-dose 4.

rate is_ largely due to' the local concentration and is therefore estimated to be:

127(2.6 x 10/4 x 10~ 8) = 8.3 mr/yr (S)

L J

,- - - , - . . -..w -- --.,---4 -.g,c.-.-.,v-- .,w... 4y-r w- ~w +-et

1. 8 - page 5 2

The gamma dose rate depends upon the size of the cloud (room). If

. a sphere of radius R contains uniformly diitributed sources emitting S photons /cm'-sec., the photon flux at.the center of the sphere is F = SR. If the source strength is given in pCi/ml, and converting from photon flux (of 1.29 Mev argon-41 photons) to dose rate, the equivalent numerical equation is

! 0(mr/hr) = 82.6 C R C is in pCi/ml and the radius R is in centimeters.

I f The 1975~ application for Amendment 10 subdivided the Math Science Addition into 13 cells each characterized by the radius of an equivalent volume spnere. A large room will have a larger dose rate than a small room, and is also likely to contain more people. Thus an effective average radius is a person-weighted radius.

li = [ N R /[ Ng R. and_Ni are the radius and population of the 41 bcell. From the 1975 data, R = 683 cm. The average gamma dose rate is 0(y) = 82.6 x 2.6 x 10~S x 683 = 1.5 x 10~"mr/hr

= 1.3 mr/yr (y).

The total dose rate is 9.6 mr/yr (S + y). Reactor utilization is 5%,

S.W wind frequency and occupancy are both 100%. From the definition of R, the total person-rem dose rate is the individual dose rate multiplied by the population.

J Reference

1. Rubin, Mark Philip " Atmospheric Dispersion of Argon-41 from the UCLA Nuclear Reactor," a UCLA Master's Thesis, School of Engineering, 1976.

1 4

I 4

I 4

, p w -- . -- .- - - - , 3 --..-c , w --r=-' --r-e- ?**

• 1 The Non-Buoyant Bent-0ver Plume  ;

I I

The use of Turner Workbook formulas in the present problem is not appropriate, and they cannot be expected to yield reasonable results.

The only applicable simple model is that of the " bent-over" plume, dis-cussed by Briggs in Chapter 4 and by Csanady in Chapter 6, especially sec-tion 6.12 and Anoendix to Chapter 6, A.6.1. Most-discussions of plumes issuing from cnimneys are concerned with a credominance of buoyancy, since most plumes are emitted at a temperature considerably greater than ambient.

In the present case, the ten.nerature may be taken as equal to ambient and the plume becomes a " jet" in conventional terminology, with an initial condition of a momentum excers, characterized by vertical velocity wg and an initial radius (i.e. radius of chimney Rg).

This jet is emitted vertically into an atmosphere with horizontal velocity V. Two assumptions then apply, as follows.

1) The atmosphere may be turbulent, but it is assumed that the entrain-ment of air into the jet is "self-generated" by the jet, just as if the air were smoothly flowing. The fact that the jet begins vertically and the airflow is horizontal is taken into account only by the use of an increased entrainment coefficient of 0.5, in contrast to the value of about 0.08 for a plume moving horizontally with the wind.

2) Mathematically, it is assumed that one can translate in time according to t = x/V, where x is the horizontal distance traversed by a particle in the jet, even though time is required for the jet to be accelerated to the horizontal wind speed. Similarly, it is assumed that the differen-tial . equations for vertical structure can be transformed into differen-tial equations for horizontal structure by the relations l.'

L

=

d_ , L ,dx , X L dz dx dz w dx l where V is constant and w variable.

} These assumptions are not really satisfactory, but I know of no simple '

theory that avoids them. The-similarity theory, which derives essentially

[ from Morton, Taylor, Turner (Proc. Royal Soc. A234,1956) - yields the

-results cited by Briggs on p. 33 and Csanady on p. 219: the height Z of the jet centerline, and the expanded radius R of the jet are given in terms of the initial variables R g, w ,g V, and the horizontal distance x. (The i stability of the atmosphere does not enter for such short distances. Of course, this stability will greatly affect the turbulence of the atmospheric f flow itself; but, as mentioned above, this does not enter.)

- The relevant formulas are s

Z = 2.3(w2 R 2 )t/s y 2/3 x /s t

= 2.3(wg /V)2/3 (p x)i/s Plume radius is propo tional to-height by similarity theory:

i R = yZ y = entrainment coefficient

- = 0.5 to that dilution of contaminant is inversely proportional to s

4 4 r" ? "w* t m w w w +w- *m+--e- r re-y - te wit--re - v

• tv1-m -'N *'+=T-- ** y=e-S-' Wv*' * *T'FWW-9 ' W~ t**FTr'*TW~T'w'

• T

2 fw t*!' (R oX)2/4

_ o_

bR = (2.3)2 l iV ;

j R2 n g

./:. 2/3

= 1.3 The conservation of contaminant requires that 2 2 cwR gg = CVR cgis the concentration at the stack release point, and C is the downwind concentration where the plume nas expanded to radius R and the velocity V is close to the mean wind speed. Consequently the dilution factor is:

2 1/3 2/3

• I" Rwg These formulas must be interpreted as follows. Take the mean wind direction and plot it on a horizontal map with stack as origin. If at a distance of 22 meters, the intake is further that a distance R from the mean wind direction, theory says intake is outside polluted plume Similarly , determine Z at a distance of 22 m. If the difference (Z - R) is still greater than intake height, intake is outside plume.

1 If the plume has not risen to the height of the 9th floor of M.S. , the theory fails, of course, and a complicated flow situation arises.

If intake is within plume, concentration of pollution is reduced to C from cg at chimney exit. l Thus, in this example this simple theory says that either (1) the !

l air intake is within the plume, in which case it is receiving a high con-  !

centration, or (2) it is outside the plume and receives a concentration 1

I I

s of zero. There is' probably some validity to the concept of a very sharp gradient of concentration (rather than the continuous gradient of the Gaussian plume) s'o near to the exit. It would be nice to have. continuous readings  :

," over time at the -intake.

However, the assumption that all entrainment is due to the plume motion is very important here. In the. typical sea-breeze regime, which is the

! one that produces tne pollution problem in question, this assumption is quite invalid. The ~ air is highly turbulent, gusting with eddies that will distort the plume and give highly variable concentrations at the intake over periods of 10 seconds to a few minutes. The meaningful value of con-centration under such conditions would be a mean over a time scale--say, 20 minutes--sufficient to average out these eddies, but smaller than the j time scale of variation of the mean wind V, or of the statistical inten-sity of the atmospheric turbulence. Such data are not now available.

The next step upward in modeling sophistication would take into account 1

the turbulence of the atmosphere and the transition from a vertical to I a horizontal plume. Numerical procedures and a large computer would be required. Teske et al . (1978) present a two-dimensional model useful for many purposes, but not applicable, as they note, to the simulation of plume-bending in a moving atmosphere. The appropriate model for the problem under-discussion is that of Bennet and Golay (1979), a full three-dimensional, higher-order closure code. Several simulation runs using this code would produce interesting and useful results.

M.G. Wurtele August 1980

_____.__._._.___.___I__

References Bennet, R.G. and Gol ay, M.N. ,

• 979 : " Numerical Modeling of Buoyant Plumes in a Turbulent Stratified Atmosphere," MIT-EL 79-002, M.I.T.

Briggs, G.A., 1969: Plume Rise, U.S. A.E.C.

Csanady, G.T., 1973: Turbulent Diffusion in the Environment, D. Reidel Pub. Co.

Teske , it.E. , Lewellen, W.S . , Segur, H.S. ,1978: " Turbulence Modeling Applied to Buoyant Plumes ," E.P. A. Report 600/4-78-050.

i.^,

2

7 Question 9. 8-28-80 in .tegard to the enviwnmental study conducted by UCLA, what would have been the results if mo.te of the TLD's wste consideted in the analyses. -

As indicated during the site visit, do not utili::e the TLD's that were located at distances appteciably upcind from the dischstge stack tJ1at were iritialty placed to dete".mine .the effects of aff-site and potentially super-imposed fugitive sou,1ces.

The a w... ; value of all 16 TLD locations within 300 feet of the stack ;-

was 8.86 mr/ quarter. Of these 16 locations, ten were in the 90* '

northeast quadrant which contains. the prevailing wind quadrant. The average value for these ten stations was 9.31 mr/ quarter. There was only one other test location in the NE quadrant, TLD #18 was approximately 780' NE of the release point on an elevator structure above the Physics Building. The average value at that station was 8.9 mr/ quarter. If

i. included 'th the ten others in the NE quadrant, the overall average for the N: luadrant is 9.27 mr/ quarter.

The TLD on' the stack (ll . 6 mr/ quarter) is included in all of the averages reported above.

i

+

i i

t 4

1

- .l

Question 10. 8-27-80 Your May 13, 1980 response to question *2 indicated that the caicuiated redudion of a factor of 200 due to instaitation of a hoid-up tank is expected to show a facter of 10 reduction in reality due to " seep" into the teactor room and "some losses" from cpe.tation of the' pneumatic sampie ttans;.urt system. This is a large ioss in effcctiveness. How much do you estimate.

is due to each of these effects? Could a non-activating gas be used in the pneumatic system, whe.te it night provide some ,telief from argon escaping the core exttact line while not invoiving large amounts of pressurized 9as handling?

An improved estimate of the effective capture is about 96%. The loss dis-tribution is estimated to be:

Leakage to room 2.75%

Pneumatic transfer system 1.36%

Delayed release 0.16%

Total Losses 4.27%

The losses by d.1ayed release depend somewhat on the operating cycle.

The loss cited above assumes 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> of operation,15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> of holding, and 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> to release (depressuring).

The largest loss (leakage to the room), is the least certain but ig an upper bound estimate. It is based upon a concentration of 5 x 10-' pCi/ml.

This number is the estimated detection threshold of the Triton model 855

, that has been used on several occasions to attempt argon-41 detection in that rcom. The null result of these measurement attempts indicates that

~ the argon-41 concentration is less than 5 x 10-7 Ci/ml in that room.

Unless the estimated loss to the room proves to be considerably over-estimated, the cost-benefit ratio is not very attractive to inert gas utilization in the pneumatic transfer system.

i i

V ..

L '

5&+

Question 11. 8-27-80 Du,thtg the reactor 2.ifetime and shtce AMF has stopped manufacturing Argonaut reactor components, what has been the UCLA experience ht regard to repait and replacement parts for reactor facility components that have matfunctioned or tequited replacement? 0: hat would be expected b1 the futute?

If parts are no longe.1 available off the shelf, could they be fabricated, manufactured, substLtuted fo.t. A,te as-buit.t and fabrication d,wcings and specifications available for the various components or component parts that may have to be replaced.

Although AMF was the prime contractor for the ULuA reactor, the mechani-cal work was done by Levine and McCann, the electrical work (equipment) largely by Honeywell. The mechanical components (control blades, shafts, bearings, gears, couplings, clutches and drives) are a combination of pur-chased and fabricated parts. The mechanical assembly was never an off-the-shel f item. The system is very simple and although exact replacement of some purchased parts might be difficult todays similar components exist and could be re-engineered into the existing equipment.

The or.ly mechanical components of the control system that have been replaced are bearings.

The most significant repair / replacement work on the UCLA reactor was the replacement of the fuel boxes and related plumbing. This was done in 1971 because of exterior corrosion of the aluminum piping that had been cast into the concrete base. The effort required plumbing and sheet metal crafts, but not sophisticated engineering. Parts procurement problems were non-existent.

Over the years, major components of the control console have gradually been replaced with improved equipment. Because of solid state circuitry, the reliability is probably better now than with the vacuum tube units of the original console.

A reasonably complete set of construction drawings by General Nuclear Engineering and AMF exists. There may have been unrecorded field changes, and in general, core maintenance requires entry to assess the problem, and a part would not be fabricated without examination of that part to determine the appropriate fix.

I I

i t

1 s

Question -12. 8-27-80 What is the ptesent sLwta af the petatnent fix fo.1 the conttot tod with-dmtw event af December 1979? Has the Radiatson use Comittee made a decision? Has installation af the fix been completed?

The Radiation Use Committee approved the modification (by signature on individual letters) and unanimous aporoval was reported to the Committee at the March 20.-1980 Meeting. At the-June 17,.1980 Meeting, the Reactor Supervisor reported the installation and testing had been completed and that the performance was entirely satisfac+wy.

j '.

4 _, 5 - , , , , _ . , _ . ,._,..,__.,,..-...~_m.., . . - . . . , . . .. . . . . . _ ~ . _ _ _ . , _ , _ , , ,-,.m.

Question 13. 8-27-80 Figures 111-1, -2, -3, -6 of de Licen.se renewal application do not clearly show the operation of de conttot rods. Mcw do dey operate in relation to the core in the event of an carthquaket Discuss de effects on conttot blade operability of any seisnic shake tests performed on de reactor core /

conttot system. Has dere been any reactor operating experience durbtg seismic events. If so, describe conttot rod activities.

The four control / shim blades are firmly fixed to horizontal drive shafts which rotate to swing the blade through a maximum arc of about 45 degrees.

The blades are rotated in a vertical plane, When at the lower extremity of the 45* arc they are horizontal and between adjacent fuel bo'es. Con-trol blade withdrawal is effected by rotating the drive shafts so that the blades rotate upward and away from the core center.

b'ade withdrawal is measured / indicated as a percent of the maximum (45 )

rota ti on . It is typically less than 50% indicating an elevation of less tran 22.5 degrees from the horizontal .

The blades rotate within closed magnesium shrouds, the space surrounding the shrouds is stacked graphite. The shrouds define the clear path for the blade rotation as well as the graphite stacking geometry.

Control blade operability under earthquake has not been tested or experienced.

If the earthquake precursor is sensed oy the seismic scram interlock befoia any large effects occur, then the blades will fall (rotate downwards) by gravity. A sudden sharp jolt could conceivably shift the graphite rela-tive to the foundation and shroud. Binding of one or moo Nntrol blades by deformation of the shroud is conceivable. No such esant was observed in connection with the seismic shake tests.

Results of the seismic shake tests were reported at the Winter Meeting of the American Nuclear Society in 1968 [iJ. It was reported that no immediate adverse effects were observed, however a possible delayed effect was observed appn.ximately six months le..er. The delayed effect was the failure of a " drop rod" test during a e,

• start check off. The failure was attributed to shifting of the lead shielding in the rod-drive shaft clearance space. Although not provably related to the seismic test, the coincidence did not go unremarked. More importantly, the lead shot arrangement was recognized as a potential high-friction situation and remedial action was undertaken.

There is no experience with reactor operation during an earthquake.

In regard to the " frozen-olade" scenario with all four blades locked, it may be remarked that the reactor has been operated in a simulation of this  ;

mode for many consecutive hours. Nominally in the manual mode, the power level is stabilized by the negative temperature coefficient of reactivity.

The actual behavior under a quake induced " frozen blade" scenario depends -

upon the propagation of various multiple failure sequences. The seismic )

interlock should demand " full-scram" including the " dump water" command.

_ _ _ _ _ _ . _ _ J

1. 13.- page 2 Execution of the latter command adds approximately 25 dollars of negative reactivity to the core. If the interlocks fail, and the reactor is super-

-critical, an over-power scram would be called, again including the " dump water" command. Failure of the electrical and/or air utility would simi-larly result in opening the dump valve. ,

Because of the variety of possible signals to " dump water", the simplest approach to a worst-case scenario is to hypothesize that the dump-valve is- stuck in the closed position.

In mild cases, the operating staff can intervene. A " stuck" dump valve can be levered open, and a control blade can be torqued down with a pipe wrench. These are not highly sophisticated maneuvers, but the reactor is not a highly sophisticated device.

A most drastic case assumes that the blades are frozen, the dump valve is stuck, and that the entire 8 story structure has collapsed through the roof of the reactor building, burying the reactor in a mass of rubble.

All of the utilities are off but the fuel boxes and dump line remain intact.

If super-critic.al, the core water inventory would be dri,er. to the boiling point. Evaportsion of 10 to 20% (3 to 6 gallons) of co e water would render the core subcr ;tical in 3 to 6 minutes. This time scale is based upon an initial excess reactivity sufficient to drive to reactor to a void limited power level of 150 kw. No fission product release attends the scenario.

References :

1. Smith, C.B. , and Matthiesen, R.B. , " Vibration Testing and Earthquake Response of Nuclear Reactors" Presented at the Winter Meeting of the American Nuclear Society, Washington D.C. , November 1968.

l l

i

)

Question 14. 8-27-80 Paqe V/1-3 of the reneaul application defines "secwted expc.timent". (Chat is the status af a "secwte experiment" in the event af an ea,tthquake?

Secured experiments rest on a spacer supported by a graphite plug at the

. bottom of a vertical port. The spacer elevates the sample to the mid plane of the core, the region of maximum flux. The sample is laterally constrained by the sample hole liner, and is vertically supported by the spacer.

Neither the sample, the core, nor_ the top shielding are secured against verti:31 downward accelerations greater than 1 g. Cohesive forces suf-ficiert to produce downward accelerations greater than 1 g are unlikely, the tuilding and all of its contents would fall at no more than 1 g. The sample wcold not move relative to the core.

l

)

E',C t 05 U o E

1. In order to consider pesting the r:of of the "ath/ Science building as a

" restricted area" it would be necessary to first consider relocating the meteorology licoratory facilities from tha t building. Is is conceivable for the .:e:eorology facilities to be moved from tce "ath/ Science building

7. tc ecother 3uilding.oould the astronomy facilities also , ave to ce ce,ovee

/ or are tr.ay princi.: ally utilized when the reactorsp ic ct : e ra ti nc. ?

Are there advantages to .NEL in having the meteorology facilities located cn '.he roof o f the Fa th/ Science buil ding? Are student progra:ns or Ur.is er-sity (reactor) programs incorporated in the use of this facility?

2. When standing on the roof of the . Math / Science building and looking at the reactor stack, it is not apparent that the ventilatien system is ocerating.

It is further not apparent that the reactor is cperating. As the Math /

Science roof area is an open access zone, it would appear that information attasting to the status of reactor operati:.,n would be helpful . Could a blinking red light be installed together with a ;ermanent sign ;osted .shere g' it can be easily read from the roof of the "ath/ Science building that (to wit) "when the red light is blinking the UCLA research reactor is in ope ra ti on . Please do not loiter in this area".

[

3. To reflect the , measures taken by UCLA, the Technical 5:ecificaticns will speci fy that the roof containing the stack shall be declared a " restricted area".

a. I: was noted during the site visit that ecui; rent for proposed Argen 'I iolcue :anxs are being obtaired and stored, please provide a schedule for c:mpletion o f this facility. q..

ut

5. What is the reliability of emissicns Icnitoring equi; ment and accuracy o' calibration standards and methods? Who checxs calibra tions ? How are rec:rds ma i n '.a i n e a ?

O h

6. What was the reliacility of the instrumentation and sanoling systers used Te gr for the SF6 dispersion tes'.s conducted by UCLA? ,y t

5 ce p b'

7. What is the reliability and accuracy of the film badges at various lesels ~ ~ ~

g and types o f expcsure used by UCLA? g, p ce q d. M . r rw i

8. In tne request for Amendment .'lo.10, UCLA utilized an overali emissions g , _ r .-

reduction of 160 that as calculated on the bases of reac:ce usa, available g s.,c-meteorciogical cilution ocal s . ,winc diraction in forma tion, and :uil ding oc:u;ancy in:orma: ion.

.n tne i n:= rvening y e a rs , me t eoro n.gica s 5:a ti on.

,gg , .

have been ?staclished uch closer tc T A roviding more apolicable in- O* #. iE f:rmatien, acre rigcrous eteorologi:3i mixing and trans:or: rodels are a nilable, TLA has c:nducted meteorological studies and additional in-forma: ice. is possibly available on building occu:arcy. With the use of nis ore :urren: in fcr .a tion , ple ase :recare calcul a tion s to :redic t the c Acantrati n of Ar-41 at the Pa tn/Scienca building fr;m sta k er.issi:ns due to raacice c; era:icn. Utilize the Te;3,.a tt "ours c' ; ara ticn ia. II/3

'C r ria C tCr ;;e ra ti ng t # :2.

. 1 1 l L. 1 1

D** D "DT 3 $ l mer o f m,.9 . - a L

Also please submit any actual release data available such as from sccke or vapor tests, and ccmpare with the calculated results.

Use roof occupancy factors of 10% and 1000, and wind direction factors of 100" and 50%. Also consider infinite immersion, and ncn #nfinite immersion factors indicated ir. the footnote associated with the calculations for Ar-41 concentrations pro,ided in 10 CFR 20 Appendix 8.

9. In regard t; .ne environmental s udy :crducted by UCLA, what .<culd nave been :ne results if more of the I.0's eere considered in the analjses. As indicated during the site visit, co not utilize the TLD's that aere located at distances appre:iably u; wind from the discharge stack that were initally olaced to determine the ef fects of off site and poten ially su erimposed fugitive scurces .

10. Your May 13, 1980 response to cuestion d2 indicated that the calculated reduction of a factor of 200 due to installation of a hold-oc :ank is \ _ ,

expected to shcw a factor of 10 reduction in reality due to ;' seep" in to the reactor room and "scme losses" from oceration of the pneuratic sample p.S rk p.#

transport sys tem. This is a large less in effectiveness. Mcw much do pCe b",

you estimate is due to each of these effects? Could a ncn-activa ting cas be used in the ;neumatic P' A from argon escacing tne cTeD:3'ct r lir.e while not involving largesys g . em, w amounts of pressurized gas handling? D

11. Curing the reac:ce lifetire and since AN has stooced manufacturing Argonaut reac or cc.aponen ts , .vna t has been ta: UCLA experience in re;ard to repair and replacament parts for reactor facility ccaponents : hat ,

have mal functioned or required replacement? 'Jhat would be ex;ected in the fut;re?

If parts are no icnger available off the shelf, could they be fabri:ated, manufactured, substituted for. Are as-buil: and fabrication drawings and specifications available for the various components or compcnent ; arts that may have to be replaced. )

12. 'Jha t is the present status cf the pernanent fix for the control red with-draw event of :e: ember 1979? Has the :adiatico Use Corsnittae made a decision?

Has installation of the fix been :omoleted? I

13. Figures III/5-1, -2, -3, -6 of the license renewal application do not clearly shcw the Operation of the control rods. How do they operate in relation to the core in ine event of an earthquake? Discuss the effects cn centrol blade operability of any seismic shake tests performed on the reactor core /centrol system. Has there been any reactor 3,cera:ing experience dt ring saismic events.

If so, describe control rod activities.

12 E3ge 7/l-3 sf the rare al 3; plica !ca de fices "se:ured u;eri e,t". Mt f?',Y '

is the s tatas of a " secure e <;eri. ent" in ne event o f an sar:h:take ? v

13. Ses:icos ::ncerning "secur'ty" will be ;;se d uncer se; ara te ::ver.

T D* D 'D 6 3 j\, b J\ d A m i

UNIVERSITY OF CALIFORNIA LOS ANGELES Atmospheric Dispersion of Argon 41 from the UCLA Nuclear Reactor A thesis submitted in partial satisfaction of the requirements for the degree Master of Science in Engineering by Mark Phillip Rubin 1976