ML20032D809

Estimated Airborne Release of Pu from Atomics Intl Nuclear Matls Development Facility in Santa Susana Site,Ca,As Result of Postulated Damage from Severe Wind & Earthquake Hazard, Increment of Analysis

ML20032D809
Person / Time
Site:	07000025
Issue date:	09/30/1981
From:	Ayer J, Mckinney M, Mishima J Battelle Memorial Institute, PACIFIC NORTHWEST NATION
To:
Shared Package
ML20032D807	List: ML20032D806 ML20032D808 ML20032D809
References
PNL-3935, UC-11, NUDOCS 8111170521
Download: ML20032D809 (49)

Text

PNL-3935 UC-11 An Increment of Analysis ESTIMATED AIRBORNE RELEASE OF PLUT0NIUM FROM ATOMICS INTERNATIONAL'S NUCLEAR MATERIALS DEVELOPMENT FACILITY IN THE SANTA SUSANA SITE, CALIFORNIA, AS A RESULT OF POSTULATED DAMAGE FROM SEVERE WIND AND EARTHQUAKE HAZARD J.Mishima(a)

J. E. Ayer M. A. McKinney, Editor September 1981 Prepared for Division of Environmental Impact Studies Argonne National Laboratory under Contract DE-AC06-76-RLO 1830 Pacific Northwest Laboratory Richland, Washington 99352 (a) Advanced Fuel and Spent Fuel Licensing Branch Division of Fuel. Cycle and Material Safety U.S. Nuclear Regulatory Commission k o!

~

CONTENTS SLM4ARY AND CONCLUSIONS vii INTRODUCTION 1-BUILDING AND PROCESS DESCRIPTION 3

BUILDING DESCRIPTION (EDAC 1978a) 3 GLOVEB0X ROOM 5

5 PROCESS DESCRIPTION.

ENGINEERED SAFEGUARDS 5

Ventilation and Exhaust (Mishima 1980) 5 Fire Protection System 11 AREAS OF CONCERN 12 DAMAGE SCENARIOS 13 WIND DAMAGE (Mehta, Mcdonald and Alikhanlou 1980, pp. 32-34) 13 Nominal Windspeed of 110 (49.2 m/sec); 3 x 10-6/yr Probability of Occurrence 13 Nominal Windspeed of 130 mph (58.1 m/sec); 8 x 10-7/yr Probability of Occurrence 13 Nominal Windspeed of 150 mph (67.1 m/sec); 4 x 10-7/yr Probability of Occurrence 13 Nominal Windspeed of 170 mps (76 m/sec); 1 x 10-7/yr Probability of Occurrence 14 EARTHQUAKE DAMAGE (EDAC 1978b, pp. 52-53) 14 Linear Acceleration of 0.2 to 0.35 g; 1 x 10-2/yr Probability of Occurrence 14 Linear Acceleration of 0.35 to 0.55 g;.2 x 10-3/yr Probability of Occurrence 14 ar Acceleration in Excess of 0.55 g; 0.6 g has 1.3 x Ling /yr Probability of Occurrence 10-3 14 iii

~

APPROACH AND FACTORS USED IN ESTIMATING SOURCE TERMS 15 INVENTORY AT RISK 15 Process Inventories of Plutonium 15 Relocated Process Materials Inventories.

16 Surface Contamination 16 Filters 16 DAMAGE RATIO 17 FRACTIONAL AIRBORNE RELEASE OF PARTICULATE MATERI AL 17 Crushing of Glovebox 18 Perforation of Glovebox 19 Damage to Exhaust HEPA Filters 20 Aerodynamic Entrainment 20 ATMOSPHERIC EXCHANGE RATE 21 Wind Hazard 21 Nominal Windsreed of 110 mph 21 Nominal Windspeed of 130 mph 21 Nominal Windspeed of 150 mph 22 Nominal Windspeed of 170 mph 22 Earthquake Hazard 22 SOURCE TERM RANGES 22 Upper Bound 23 Average 23 Lower Bound 23 SOURCE TERM ESTIMATES 24 iv l

1 I

SOURCE TERM ESTIMATES FROM WIND HAZARD 24 Nominal Windspeed 110 mph (49.2 m/sec); 3 x 10-6/yr Probability of Occurrence 24 Nominal Windspeed 130 mph (58.1 m/sec); 8 x 10-7/yr Probability of Occurrence 27 Nominal Windspeed 150 mph (67.1 m/sec); 4 x 10-7/yr Probability of Occurrence Occurrence 29 Nominal Windspeed of 170 mph (76 m/sec); 1 x 10-7/yr Probability of Occurrence 31 SOURCE TERM ESTIMATES FROM EARTHQUAKE HAZARDS 34 Linear Acceleration of Less Than 0.55 g; Greater Than 2 x 10-3/yr Probability of Occurrence 34 Linear Acceleration in Excess of 0.55 g; Less Than 10-3/yr Probability of Occurrence 34 REFERENCES 35 ATTACHMENT A.

A.1

=

l' V

FIGURES 1 Plan View of NMDF and Schematic of Exhaust System 4

2 Isometric Drawing of NMDF with Glovebox Arrangement 6

3 Mixed-Carbide-Process Flow Diagram g

4 The Range and Types of Damage Postulated for the NMDF at a Nominal Windspeed of 110 mph 26 5 The Range and Types of Damage Postulated for the NMDF at a l

Nominal Windspeed of 130 mph 28 l

t l

6 The Range and Types of Damage Postulated for the NMDF at a Nominal Windspeed of 150 mph 30 7 The Range and Types of Damage Postulated for the NMDF at a Nominal Wind; peed of 170 mph 32 l

l TABLES i

1 Source Term Estimates for AI's Nuclear Materials Development.

ix 2 Sumary of Data for Glove Boxes 7

3 Inventory-at-Risk 10 4 Amounts of Combustible Materials in Room 127 11 5 Fractional Airborne Release Factors.

18 vi

SUMMARY

AND CONCLUSIONS, The potential mass of airborne releases of plutonium (source term) that could result from wind and seismic damage is estimated for the Atomics Inter-national Company's Nuclear Materials Development Facility (NMDF) at the Santa Susana site in California.

The postulated source terms will be useful as the c

basis for estimating the potential dose to the " maximum" exposed individual by inhalation and to the total population living within a prescribed radius of the site. The respirable fraction of airborne particles is thus the principal concern.

The estimated source terms (Table 1) are based on the damage ratio, i.e.,

the fraction of enclosures crushed or punctured during events of varying sever-ity, and the potential airborne releases if all enclosures suffer particular levels of damage.

In an attempt to provide a realistic range of potential source terms that include most of the normal processing conditions, a "best estimate" bounded by upper and lower limits is provided.

The range of source terms is calculated by combining a high best estimate and a low damage ratio, based on a fraction of enclosures suffering crush or perforation, with the air-borne release from enclosures based upon an upper limit, average, and lower limit inventory of dispersible materials at risk. Two throughput levels are considered. The factors used to evaluate the fractional airborne release of materials and the exchange rates between enclosed and exterior atmospheres are discussed.

The postulated damage and source terms are discussed for wind and earth-quake hazard scenarios in order of their increasing severity.

The largest postulated airborne releases from the NMDF are for the maximum wind hazard (nominal windspeed 170 mph) and for linear acceleration exceeding 0.55 g.

Both scenarios postulate virtual complete loss of the structure excluding the vault. Wind and earthquake hazards using higher windspeeds or o

linear accelerations should not result in substantially greater source terms.

The source terms are expressed as mass of plutonium particles, 10 um aero-dynamic equivalent diameter or less, released up to 4 days after the events.

vii h

x.

i 4

i I

Because of the.small inventories routinely found in the facility and the use of a' batch process, the. instantaneous airborne release for the maximum wind and earthquake scenarios is dominated by the. postulated release from crushing of

' the: glovetox exhaust filters.

~

4 T

q l

?

l' 1

l j

2 j

i 4

4 1

3 a

I

. 4 viii 1

4 e

L__.

TABLE 1.

S0urce Term Estimates for AI's Nuclear Materials Development Facility Due to Wind and Earthquake Hazard MASS RELEASE OF PLUTONIUM IN THE RESPIRABLE SIZE R ANGE(a), g UPPER LOWER EVENT BOUND AVERAGE BOUND WIND HAZARD NOMINAL WINDSPEED 110 mph (49.2 m/sec). 3x10-6 PER YEAR PROBABILITY OF OCCURRENCE INSTANTANEOUS ADDITIONAL MASS RELEASED IN NEXT 2 HOURS 0.003 ADDITIONAL MASS RELEASED IN NEXT 6 HOURS 0.009 ADDITIONAL MASS RELEASED IN NEXT 16 HOURS ADDITIONAL MASS RELEASED IN NEXT 3 D AYS f

NOMINAL WINDSPEED 130 mph (58.1 m/sec). 8x10 7 PER YEAR PROBABILITY OF OCCURRENCE INSTANTANEOUS 0.01 0.005 0.003 ADDITIONAL MASS RELEASED IN NEXT 2 HOURS 0.003 1x10 5 1 x10-5 ADDITIONAL MASS RELEASED IN NEXT 6 HOURS 0.009 3x10-5 3x10-5 ADDITIONAL MASS ret. EASED IN NEXT 16 HOURS 8x10-5 8x10-5 8x10-5 ADDITIONAL MASS RELEASED IN NEXT 3 DAYS 4x10 4 4x10 4 4 x10-4 NOMIN AL WINDSPEED 150 mph (67.1 m/sec),4x10 7 PER YEAR PROBABILITY OF OCCURRENCE INSTANTANEOUS 0.03 0.02 0.02 ADDITIONAL MASS RELEASED IN NEXT 2 HOURS 0.003 4x10-4 4 x10-4 ADDITION AL MASS RELEASED IN NEXT 6 HOURS 0.009 0.001 0.001 ADDITIONAL MASS RELEASED IN NEXT 16 HOURS 0.001 ADDITIONAL MASS RELEASED IN NEXT 3 D AYS 0.02 NOMINAL WINDSPEED 170 mph (76 m/sec),1x10-7 PER YEAR PROBABILITY OF OCCURRENCE INSTANTANEOUS 4.

4.

4.

ADDITIONAL MASS RELEASED IN NEXT 2 HOURS 0.02 0.004 0.003 ADDITIONAL MASS RELEASED IN NEXT 6 HOURS 0.06 0.01 0.01 ADDITIONAL MASS RELEASED IN NEXT 1G HOURS 0.02 0.03 0.03 ADDITIONAL M ASS RELEASED IN NEXT 3 DAYS 0.7 0.2 0.2 EARTHQUAKE HAZARD LINEAR ACCELER ATION LESS THAN 0.55g, GREATER NO SIGNIFICANT DAMAGE THAN 2x10-3 PER YEAR PROBABILITY RESULTING IN AIRBORNE OF OCCURRENCE RELEASE

~

LINEAR ACCELERATION EXCEEDING 0.55g. 0.6 g HAS A 1.3 x 10 3 PER YEAR PROBABILITY OF OCCURRENCE INSTANTANEOUS 4.

4.

4 ADDITIONAL MASS RELEASED IN NEXT 2 HOURS 0.02 4x10-5 4,10-5 ADDITIONAL MASS RELEASED IN NEXT 6 HOURS 0.06 1 x10-4 1xiG 4 ADDITIONAL MASS RELEASED IN NEXT 16 HOURS 0.2 3 x10-4 3>.10-4 ADDITIONAL MASS RELEASED IN NEXT 3 DAYS 0.7 0.002 0.001 (a) PARTICLES 10 pm AND LESS AERODYNAMIC EQUlVALENT DIAMETER

- LESS THAN 10-79 PLUTONIUM iy

INTRODUCTION If the structure and equipment that contain radioactive materials fail because of the stresses imposed by the impact of a natural phenomenon, the downwind population can be subjected to a radiological hazard from the airborne material. The estimated airborne releases of contained radioactive material form the basis for calculating the radiation dose, which is one component of an overall risk analysis.

This report is a part of an interdisciplinary study sponsored by the j

United States Nuclear Regulatory Commission (NRC) and coordinated by the Divi-sion of Environmental Impact Studies of the Argonne National Laboratory ( ANL).

It is one increment in a series dealing with the potential airborne releases of plutonium from licensed plutoaium' fuel-fabrication facilities.

The study estimates the potential release from the Atomics International NMDF at Santa Susana, California, as the result of a severe wind and earthquake hazard.

j The estimates of airborne plutonium releases were developed by identifying the damage sustained by the structure and equipment at varying severities of wind and earthquake. The Pacific Northwest Laboratory (PNL)(a) staff used data developed by other specialists.

The description of the facility was pro-vided by the Engineering Decision Analysis Company (EDAC 1978a) as was data on the potential responses of the structure and equipment to various severities of earthquake hazard (EDAC 1978b).

The Disaster Research Institute at Texas Tech provided sim'lar infonnation for wind hazard (Mehta, Mcdonald and Alikhanlow 1980).

The primary concern in the calculation of downwind dose for this study is inhalation (McPherson and Watson 1978, p. 3).

In this increment of the series the primary emphasis is the release of plutonium particulate material of a size range that can be carried downwind and inhaled.

Particles of 10-um (a) Pacific Northwest Laboratory is operated for the U.S. Department of Energy by Battelle Memorial Institute.

1

aerodynamic equivalent diameter (AED)(a) or less are conservatively assumed to be.the respirable fraction.

Such an assumption overstates the potential effect by a factor of 1.5 to greater than an order of magnitude, depending upon the lung deposition model chosen (Mercer 1977, Figure 1).

The behavior of the structure and equipment in accident situations is not precisely understood.

With such uncertainties, the estimates of airborne releases tend to be con-servative, that is, estimates are probably greater than the releases that would actually be experienced.

(a) Aerodynamic equivalent diameter: Particles exhibiting the aerodynamic behavior of a unit density sphere of the stated size.

2

lf:1 BUILDING AND PROCESS DESCRIPTION The first step in estimating the potential airborne releases of plutonium is to find out what materials and equipment are present and to identify those features and operations that could influence the material released. The infor-mation was gathered from documents issued by Engineering Decision Analysis Com-a pany (EDAC), the NRC, and PNL. This information includes descriptions of the structure and of the process (which influences the types and locations of equipment). The amounts and forms of materials present and information on engineered safety systems are also included.

BUILDING DESCRIPTION (EDAC 1978a)

The NN)F (055 Building) is a one story, windowless, tilt up, concrete building located in the Atomics International Nuclear Development Field Labo-ratory, which is in the southeastern portion of Ventura county, 29 miles north-west of downtown Los Angeles (AI 1976). The site is in a pocket in the Simi Hills, 800 ft to 1000 ft above the populated valley floor (AI 1976) and is in a relatively. remote mountain site (NRC 1980).

The building is rectangular, 202 ft long (north-south direction) by 60 ft wide (east-west direction) by 17 ft high. Figure 1 is a plan of the building.

The floor is a concrete slab on grade.

Steel "I" beams, 20 ft between centers, support the concrete tilt-up panels (20 ft x 17 ft) that are tied together'by concrete inserts welded to the main building columns.

The panels are also sup-ported by the column footings.

Lightweight concrete-filled roof panels (20 ft x 2 ft) are welded to roof beams running in an east-west direction (60 ft), which are supported by the main building column.

The interior parti-tions are of metal-stud construction with gypsum-board faces.

There are sev-eral doors in the exterior walls of the facility.

The two standard size doors in the east and west exterior walls of the glovebox room are protected by con-crete enclosures from wind-generated missiles.

The vault is attached to the southern corner of the west exterior wall of the Glovebox Room and is an ll-ft-square, cast-in-place concrete box.

3

I I

l I

i I

l

.I I l i

O R

=

.I a

r

+

gQ,

,],

MEcnANiCAL

.k-2,,=.

+

s q

gQ,

,C1

,f,1 '

=

i o

i E ouir.

4,

"' e 2

=

i I

I 1

I I

I-l I

I I

I l

_. [

tA. AREA p

,A GLOVE 80X ROOM 5

mf MS OFF6CE AREA sis,,,

us g

I -n

r

• g N

^ 'l"^""

a m

)

i I

I I

l 7

I o

, t 4'

{

}

}

,l

,,, !p n.

M'

=

4 :

'"c'a'c^' 'a'a"-

m l

y l'J _

lif i

e i

n i

i l_

I i=

=

a 1

i vAuti e

FIRST FLOOR in LEGEND

^

~

- BARRIER WALLS amms= mums CONTAINMENT DOORS M BARRIER WALLS I

I ButLDING CONTAINMENT - PRIM ARY ARE A I

I BUILDING CONTAINMENT - SECONDARY AREA L

M MOTORIZED BUTTERFLY VALVE iy MOTORIZEO DAMPER MECHANICAL E nutP.

O*

CEILING REGISTER

[

__t

/

BLOWER l

isr +

FILTERS 1 ABSOLUTE 24 24* 2 ABSOLUTE e

l l

ENCLOSED 24 x24* 3 AUTO. FIBERGLASS ROLL 5 FIBEPGLASS MEZZANINE FLOOR PRE P8,ED j8ERGLASS g

FIGURE 1.

Plan View of HMDF and Schematic of Exhaust System e

GLOVEBOX ROOM The glovebox room occupies the central portion of the NMDF and is rectan-gular.

Its plan dimensions are 60 ft x 90 ft.

Part of the mezzanine (Rm 202) is open to the glovebox room.

Figure 2 shows the arrangement of the two glove-box lines that, along with the process and argon-purification equipment, are the principal items in the room.

Thc gloveboxes are constructed of 3/16-in.

stainless steel with 3/8-in. plexiglas windows. The designation and dimensions are given in Table 2.

PROCESS DESCRIPTION The NMDF is an experimental facility for the development of mixed-carbide fuel. The batch process used in the facility is shown in Figure 3.

Only 10 to 12 batches wera processed during the last year of operation (see Attach-ment A) and, for the purposes of this analysis, each batch is assumed to need I week of 8-h/ day operation to be processed.

The quantity of radionuclides at risk is shown in Table 3.

Due to the differences in toxicities, only the plu-tonium is considered (10 CFR 20 APP B).

The quantity of flammable materials l

that can be located in the facility are shown in Table 4.

ENGINEERED SAFEGUARDS Ventilation and Exhaust (Mishima 1980)

Room air is suppliec

• iffusers set near the ceiling of the Glovebox Room. Room air is exha; 2 ft x 2 ft HEPA filters in the risers near floor level along the ea.

d vest walls. The exhaust is carried by sheet metal ducts at ceiling levels. Most of the inert gas (glovebox atmosphere) recycles through a purification system.

Only enough gas is exhausted to main-tain a preset negative pressure.

The few gloveboxes having an air atmosphere extract room air via HEPA filters and exhaust into the stainless steel exhaust ducts running over each glovebox line.

The room glovebox exhausts are joined in Room 130 (Filter Room).

The system is shown in Figure 1.

5 9

w

Off 8CE ""I*

ctOygs0 g AusT MAN D

5g ($

TTED O

<g

,s pg

,.h{.

/

AtSOTH t

SIDE S)

/

+

p*E CHANICAL EQUIPMENT w

t MEZZANINE y,

}

g V MECHAN 8 CAL Q*. -

(, ',,, g i e

..,.,7, E OUgpME NT 1

../

4

}'~.$ y

~.,,

F g geROOM e'

W GLOyg 80X y

ROOM O

Y d,

8*

7 ' 'h~\\

i

,* ' N

{

},,.

+,

VAULT %

/

d y'

s.

. f i_ f

/l\\

Y' 9

/

/

(

\\

g gfd l

/

[N

/

.i f

[

N

[

Ann LOCIL f '.

/"

ELECTRICALA

/,

EOU3pMENT

/

[

N /

FIGJRE 2.

Isometric Drawing of NP0F with Glovebox Arrangement s

• N y

[

'$~

'r' Sumary of-D'ata for Glove Bak.

s-TABLE 2.

es Station Box Box Size

~

a Glove Box Name No.

No.

(in. x in. x in.)-

Coment -

Air entrance 1

K-1 36 x 30/23 x 33 inpentid side'T y

s e

Air entrance 1A K-1A 36 x 30/23 x 33 23 in.--upper,

s

, dimension' Inert gas entrance 2

K 36 x 30/23 x 33 30 in.--lower dimension

. s Inert gas entrance 2A K-2A 36 x 30/23 x 33 Blend box 3(a)

K-3 84 x 42 x 42 Weight box 3A K-3A 42 x 42 x 48 Unload box--sinter 4(b)

B-1 84 x 30 x 42 furnace

~

Dry and mill box 5(b)

K-11 84 x 42 x 42 Load box--sinter 6(b)

B-2 84 x 30 x 42 furnace Press box 7(b)

K-7 84 x 42 x 48 2-in.-thick steel J

plate installed in the'60x O.D. grind box 8(b)

K-8 84 x 42 x 48 2-in.-thick steel plate installed in the box Granulator and 9

K-6 84 x 42 x 42 screening box Transfer box 10 B-3 42 x 24 x 42 Inspection box 11 B-8 84 x 42 x 48 Weigh box 12 K-12A 42 x 24 x 42 Pin-loading box 13 K-12 84 x 42 x 42 Waste-packaging box 14 K-14 84 x 42 x 42 Dissolution and 15(b)

K-15 84 x 42 x 42 sample preparation 7

N 4

X' L.

.? '

jf',

~.

i:

TABLE 2.

(contd) 3

\\

Station Box Box Size

,j

.y, Glove Box -Nwae.

No'.

~ No.

(in. x in. x in.)

Coment E, Bjlancebox f

15A,

-K-15A-42 x 24 x 42 f;;,

i X-ray fluorescence?

16(b)-

g.16 42 x 42 x 42 N ' box

."fdi?sionspectrograph,'

17(b)

K-17 42 x 42 x 42 half box-

~

I

. Arc spark bcx 17A K-17A 42 x 24 x 42 Inert as fusion box 18(b)

K-18 84 x 42 x 42

,_s.

19(b)

K-19 84 x 42 x 42 Netallograpliy pre-

~

.paration Metallography obser-20(b)

K-20 84 x 42 x 42 vation

. Wet-chemistry box 21(b)

K-21 84 x 42 x 42 TGA equipment box 22(b)

K-5 84 x 42 x 42 23 m

Power-processing box 24 K-13 142 x 45 x 48 Sintering-furnace tox 26 K-4 84 x 42 x 48 Pellet pressing and 27 K-9 84 x 42 x 42 granulator Transfer box 28 B-4 69 x 24 x 42 Arc-casting box 29 K-5 72 x 42 x 42 Fuel pin-weld box 30 B-7 48 x 32 x 33 Glovebox(AEC) 31(a) 84 x 42 x 42 (a)Boxesofprimaryconcern.

(b) Boxes of secondary concern.

8 L

BALL MILL BLEND UO2 + PuO2+C h

,. BRIO:JETTE h

REDUCE IN VACUUM FURNACE h

^

COMMINUTE IN SPEX MILL h

SIEVE BALL MILL h

PRESS COMPACTS h

SINTER '.N FLOWING l

ARGON ATMOSPHERE h

INSPECT AND SAMPLE FOR ANALYSIS FIGURE 3.

Mixed-Carbide-Process Flow Diagram e

G I

l e-l 9

1

o TABLE 3.

Inventory-at-Risk Amount of

% of Time' Glove-Radionuclide in Pro ess Box Nermal Processing Chemical Compounds (a)

Step No.

Process Operatiens Pu. a U-235. 1 Physical Form 3

Batching, Slugging 225 540 Powder A, 8 20 4

Carbothermic 0 to 120 0 to 350 Slugs A, 8, C 10 reduction 5

Size reduction 0 to 120 0 to 350 Powder C

20 '

6 Sintering 0 to 120

.O to 350 Pellets C

10 -

8 Pressing and 0 to 120 0 to 350 Powder and C

40 inspection pellets 7

None 0

0 15 Coulometry 10 40 Dissolved in A, B, C solutions 15 None 17 Emission 1

1 Crushed pellets A, B, C Spectroscopy 18 Carbon analysis 20 80 Powder A, 8, C 19 Meta 11ography 2

8 Pellets C

20 Meta 11ography 20 80 Pellets C

21 0xygen analysis 2

5 U-Pu-?t C

22 Thermogravimetric 100 400 Solidified waste Analytical solutions solidified in ben-tonite clay--stored until disposed of 13 Storage of Fuel 0 to 360 0 to 1050 Pellets C

l Pellets (c) A ' PuC, B = tJ0..C = Mixed (Pu - U) Carbide as U--20% Pu.

2 2

I l-I 10 -

[

d 4

TABLE 4.

Amounts of Combustible Materiais in Room 127 Material Normal Amount Maximum Amount Methanol 500 ml 1000 ml Benzene 0

0 Vacuum oil 20 gal 20 gal Lapping oil 500 ml 1000 ml Carbowa 0

0 Epoxy 1 qt 2 qt Kimwipes Kim towels 10 lb 15 lb Cardboard Paper sacks PVC sheet 100 lb 224 lb PVC bags 30 lb 40 lb Spray paint 30 oz 40 oz Film 0.5 lb 1.0 lb Sodium 50 g 200 g Exhaust air passes through three banks of HEPA filters held in nine sheet metal enclosures.

The volume of the glovebox room is approximately 2.06 x 5

3 10 ft and, at a rated flow of 9,000 cfm, the system produces five air changes per hour in that room.

The exhaust fans and stack are not attached to the building.

Fire Protection System Fire protection is provided by thermal detectors (rate-of-rise detectors) in each glovebox and smoke detectors in th'e NMDF at the ceiling level. Dry extinguishers are located throughout the NMDF. The pipe for a sprinkler system has been installed.

Both thermal and smoke detectors are provided in the vault.

11 e

AREAS OF CONCERN:

The handling of (Pu + U)-C presents some additional potential hazards.

Snall quantities of plutonium can become volatile during high-temperature oper-ations (Burnham, Skardahl and Chika11a 1964; Paffreyman and Potter 1964) and may accumulate in equipment.

Plutonima carbide oxidizes slowly in air at 200 to 300 C and burns brightly at 400 C (Cleveland 1967).

Plutonium carbide has remained in air at ambient temperature for 2 months with no reaction (Cleveland 1967).

The ignition temperature of uranium carbide decreases with increased surface (Snowdon et al. 1964) and the plutonium carbide is more reactive than the uranium (Cleveland 1967). Oxidation, with or without igni-tion, can convert large pieces such as pellets to loose powder.

The quantities and forms of plutonium and uranium as well as the process-ing activities in each glovebox in the NMDF are shown in Table 1.

The greatest I

quantity of plutonium in the most dispersible form present in the NMDF (loose powder) is. listed for glovebox #3, batching and slugging.

Glovebox #5 holds a lesser quantity of. plutonium but in the form of a (Pu + U)-C powder. Glove-box #8 likewise may contain (Pu + U)-C powder before being pressed into pellets. Gloveboxes #4 and #6 hold comparable amounts of (Pu + U)-C but as bulk solids (slugs and pellets). Thus, the gloveboxes are rated in order of decreasing concern:

Glovebox #3--Batching and Slugging e

e Glovebox #5--Size Reduction Glovebox #8--Pressing and Inspection o

e Glovebox #4--Carbothermic Reduction e Glovebox #6--Sintering The remaining boxes in the system are analytical boxes holding lesser quantities of plutonium and uranium.

e 12 1

O A

DAMAGE 3CENARIOS Once the significant features of the facility are defined, the damage leading to loss of containment &nd the stresses on the contained material that can make them airborne must be described. The damage to the structure and equipment from the impact of increasing levels of wind and earthquake hazard has been assessed. The wind damage response developed by Mehta, Mcdonald and Alikhanlou (1980) ranged from minimal damage to a glovebox to failure of the entire building. The only earthquake building damage leading to a significant airborne release of Pu was a failure of the entire building (EDAC 1978b).

WIND DAMAGE (Mehta, Mcdonald and Alikhanlou 1980, pp. 32-34)

Nominal Windspeed of 110 (49.2 m/sec); 3 x 10-6/yr Probability of Occurrence One of the two standard doors in the exterior walls of the glovebox room fails due to a change in air pressure. The enclosure around the outside of the ioor prevents the entry of wind-generated missiles, but wind could circulate through the room resulting in minimal damage to gloveboxes and filters.

There is no significant damage to the filter room or vault.

Nominal Windspeed of 130 mph (58.1 m/sec); 8 x 10-7/yr Probability of Occurrence Failure of the exterior doors could cause the collapse of interior parti-tions in the Filter Room. Debris from the wall could strike the filter enclo-sures. No significant damage to the vault is found.

Nominal Windspeed of 150 mph (67.1 m/sec); 4 x 10-7/yr Probability of Occurrence One or two gloveboxes could be crushed by the failure of the interior partitions in the north or south end of the glovebox room.

The roof and exterior walls in this area remain intact.

Collapsing beams could strita enclosures.

Small portions of the roof deck

~

may collapse along with the beams.

No damage to the vault is found.

13 p-t

-+

m T

y e

q t

-Nominal Windspeed of 170 mps (76 m/sec); 1 x 10-7/yr Probability of Occurrence Collapse of roof causes collapse of exterior walls.

All gloveboxes, fil-ters and filter enclosures are crushed. No significant damage to the vault occurs.

EARTHQUAKE DAMAGE (EDAC 1978b, pp. 52-53)

Linear Acceleration of 0.2 to 0.35 g; 1 x 10-2/yr Probability of Occurrence No significant damage leading to airborne release occurs.

Linear Acceleration of 0.35 to 0.55 g; 2 x 10-3/yr Probability of Occurrence No significant damage leading to airborne release occurs.

Linear f.cceleration in Excess of 0.55 g; 0.6 g has 1.3 x 10-3/yr Probability of Occurrence Exterior walls fail at 0.6 g causing the roof to collapse.

Gloveboxes l

(filters and filter enclosures) are crushed.

The vault remains intact at levels exceeding 2 g.

9 14

APPROACH AND FACTORS USED IN ESTIMATING SOURCE TERMS t

Once the damage to the facility and equipment is assessed, source terms are estimated o provide data for the calculation of potential radiation dose to the general population. A principal concern is the quantity of particulate material in the size range that can be transported downwind, inhaled by humans and deposited in the lungs. The quantity of material made airborne by an event depends on the quantity of hazardous material placed in jeopardy by the event

[

and the type and level of stress imposed.

The fraction of materia? mada air-borne that is released from the facility depends on the number of barriers between the airborne material and the outdoors and the transmission of airborne material through the barriers.

The factors used in estimating the source terms are discussed below.

m INVENTORY AT RISK The quantity of plutonium placed in jeopardy by the impact of severe winds and earthquakes comes from two sources--process materials and process materials that have been relocated and accumulated at other locations.

Plutonium is placed in jeopardy only if the primary barrier (i.e., glovebox, filter enclo-sure, etc.) loses its integrity.

The inventory is considered according to three assumptions--presence of the maximum, the average and no process mate-rials. Relocated process materials are assumed to always be present.

The fraction of time the process material will be found at the locations is also a consideration.

Process Inventories of Plutonium The quantity of plutonium that could be located in the various gloveboxes in the glovebox room of the NMDF is shown in Table 2.

Since the carbide-fuels operation is a batch process, the plutonium used can only be in one location at a time.

(The percent of processing time that quantity would be present at that location is also shown.) Only 10 to 12 batches of fuel were processed during the final year of operation (see Attachment A). A 1-shift, 5-d/wk operation is assumed and a 1-wk period is required to process a batch.

Thus, the process materials are only in jeopardy 40 h/168 h week x 12 wk/52 wk per year, or about 5.5% of the time.

15

Pelocated Process Materials Inventories Surfacu Contamination The level of surface contamination estimated for the principal process 235 gloveboxes (#3, 5, 7 and 9) based upon the quantity of Pu + U picked up g Pu/m ) and 2 x 10-8 2

by wiping the surfaces is 5 x 10-9 g Pu/cm2 (5 x 10-5 g

235,2) (see Attachment A).

The level of contamination U

/cm2 (2 x 10-4 gU 235

~

7 in other gloveboxes is estimated to be a factor of 10 lower and the level of uncertainty was estimated to be up to several hundred percent.

A surface con-tamination level based upon the visibility of deposited powders, 7.5 g 2

powder /m, has been assumed for previous studies of mixed-oxide fuel-fabrica-tion facilities (Mishima, Schwendiman and Ayer 1978, Mishima et al. 1980, and Mishima and Ayer 1980).

A conservative value, 200 times the estimated value, 2

of 1 x 10-4 g Pu/m was applied for this study.

Filters A maximum " average" loading of 1 g Pu for glovebox exhaust filters, with a maximum loading of 4 g Pu at change out, has been reported (Westinghouse 1974, pp. 5.1-5.13).

This value will be applied for all glovebox exhaust fil-ters for this study.

The potential accumulation of Pu in downstream filters are estl W.ed using the assumption that the transmission through the initial filter is 5 t 10-4 and that filters in the final filter banks are not changed as frequently as upstream filters.

If 5 x 10'4 g Pu accumulate in the second filter for each 1 g accumulated on the initial filter, and the second filter remains in the place for 100 filter changes, its loading would be 0.05 g Pu.

A value of 0.1 g Pu is applied for the loading of the initial HEPA filters of the final bank, 0.05 g for the second filter, and 0.01 g Pu for the third filter.

8 In two of the situations above, glovebox surface contamination and exhaust filter, most of the material accumulated would be in the form of carbides.

Because of the fine size of the materials deposited due to the low, intermit-tent exhaust flow and the presence of trace quantii.ies of oxygen in the inert-ing gas, conversion to Pu0 appears to be virtually certain with time.

The 2

16 i

x y

[y.Ag

,a

, ~, g }; ;v 3.},. -

c 1

e } } g ; g. -

q'

cirborne contamination released'to the exhaust system is a fine particle in an air atmoghere and is probably rapidly converted to oxide. Thus, both surface contamination and material accumulated on filters are considered oxides.

DAMAGE RATIO Damage ratio is a term used to denote the fraction of enclosures damaged to the two levels of loss of integrity used in this study--perforation and crush. Perforation is defined as:

" Perforation of the glove box:

Pieces of timber, concrete blocks, loose pieces of pipe or equipment could strike a glove box, causing an opening in the glove box window.

Plutonium stored in canisters is not likely to be released in this case, but loose material in powder form could possibly escape the confines of the glove box.

Failure of an exterior wall could allow the wind to circulate throughout the building, causing loose objects to be thrown against the glove boxes. Windborne debris could cause missile impact on the glove box and may cause perforation of the glove box."

(Mehta, i

Mcdonald and Alikhanlou 1980, p. 25)

Crush is defined as:

" Crushing of glove box:

If a heavy object falls on the glove box, structural members of the box may collapse resulting in the glove box being crushed.

This event could occur if a load-bearing wall or building frame should collapse, thus allowing the roof structure to fall downward.

In this case, the roof integrity of the glove box would be violated. The material inside the glove box would be exposed to the atmosphere."

(Meata, Mcdonald and Alikhanlou 1980,

p. 24)

FRACTIONAL AIRBORNE RELEASE OF PARTICULATE MATERIAL The various factors applied to estimate the airoorne release of plutonium as a result of the postulated damage scenarios are listed in Table 5.

Some 4

17

TABLE 5.

Fractional Airborae Release Factors Event Factor e Crushing of glovebox containing powder Volume of glove box x 300 mg powder /m3 containing surface contamination 10-2 m/

only e Perforation of glovebox containing powder Fraction of volume of glove box affected x 100 mg powder /m3 with surface contamination only 10-4 m/

Damaged to filters e

crushing 10-1 of accumulated material airborne perforation 10-2 of accumulated material airborne e Aerodynamic entrainment 10-10 sec of exposed material powder, less than 5 mph

/

10-8 sec of exposed material powder, greater than 5 mph

/

considerations that influence the applicability of these factors for the damage scenarios described are noted in the following paragraphs.

Crushing of Glovebox Crushing is defined as a complete loss of containment such as rupture of

^

the steel shell or loss of one or morc of the large viewing windows.

The event is assumed to generate sufficient excess force to inject uncontained powders into the air and to break glass or thin-wall rigid-plastic containers.

1.

Glovebox containing powder--Vibrating powder resting on a surface does not appear to provide as much dispersion of the powder as tum-

~

bling.

The mass airborne concentration is indicated by experimental 18

data measuring the value within secorids after tumbling of a powder assumed with a density.and size distribution similar to the Pu02 for this study-(Mishima, Schwendiman and Ayer 1978, p. 30).

2.

Glovebox containing surface contamination only--Surface contamination can range from particles settled on the surface to material mixed into the surface. 'The resuspension factor applied is a value mea-sured for a combination of mechanical and aerodynamic stresses (Mishima, Schwendiman and Ayer 1979, p. 44).

Perforation of Glovebox Perforation is defined as a partial loss of containment (loss of one or more glove ports, loss of a portion of a viewing window, etc.) that allows air to circulate through the glove box. The rate of release from the break will depend upon the size of the opening, whether the exhaust system continues to function, and the velocity of air entering the opening.

The particulate mate-rials airborne within the volume are released from the glove box with time.

If the exhaust flow is zero,' the release is exponential with time.

Release of greater than 99% of the airborne material within 30 min is considered inst.ntaneous.

1.

Glovebox containing powders--The force transmitted to the glove box during perforation is assumed to be considerably less than during crushing.

Since a finite period of time is required to release the airborne particulcte material, a mass concentration measured approxi-mately 1 minute after tumbling a fine powder and considered quasi-stable, was selected. Furthermore, the disturbance due to perfora-tion can be more localized than in crushing and it was judged that the powder only occupied a fraction of the volume of the glovebox (Mishima, Schewendiman and Ayer 1979, p. 39).

2 Glovebox containing surface contamination only--A reduced resuspen-sion factor of 10-10/m was selected to reflect the reduced force and area involved in perforation (Mishima, Schwendiman and Ayer 1980,

p. 18).

l n

19

a 1

Damage to Exhaust HEPA Filters Unless otherwise stated, the filters attached to glove boxes are assumed-to suffer the same damage as the glove box to which they are attached.

Secon-dary filters within the building will also suffer the same type of damage.

1.

Crushing of HEPA filters--Although the filter material (glass-fiber mats) is fragile, the plutonium particulate material accumulated can be embedded in the filter and associated with other materials such as dust, condensed organic vapors, etc. These materials may not be readily dispersed in a respirable, transportable size-range. A con-servative airborne fractional value of 10% of the accumulated mate-rials released are assumed in the absence of experimental data (Mishima, Schwendiman and Ayer 1979, p. 46).

1 2.

Perforation of HEPA filters--A reduced fractional-airborne-release factor of 1% is applied to reflect the reduced level of stress required for this level of damage (Mishima, Schwendiman and Ayer 1979, p. 47).

Aerodynamic Entrainment Powders and liquids can be entrained in the air passing over their sur-faces.

Particle suspension resuits from initiation of movement in larger par-ticles that subsequently transfer momentum to particles-in a size-range that allows suspension. Under similar conditions of air flow over a liquid film, droplets are each less likely than particles to become airborne because of the higher energy required to break up the film and form droplets.

Powder by Air Velocities Greater than 5 mph (2.2 m/sec).

A conserva-e tive suspension rate (10-8/sec), measured for a homogeneous bed and with wind velocity variations over a year's duration, is applied (Mishima, Schwendiman and Ayer 1978, p. 39).

O 20

Powder by Air.-Velocities Less than 5 mph (2.2 m/sec). A suspension e

rate (10-8/sec) measured from a homogeneous bed at these velocities is applied (Sehmel and Lloyd 1974, p. 853, Figure 3).

ATMOSPHERIC EXCHANGE RATE After the particulate material is injected into'the air, it requires the airflow to move it from its starting point to the ambient atmosphere. Diffu-sion is only a serious consideration for particles less than 1 um (Dennis 1976,

p. 52). The two areas in the NMDF in which plutonium could become airborne are the glovebox room and the filter room / airlock.

The glovebox room is located in the central portion of the NMDF and is approximately 90 ft long by 60 ft wide by 17 ft high. The filter room is located in the center of the south end of the facility and is 40 ft long by 30 ft wide by 17 ft high. The airlock connects the glovebox room to the double doors in the south wall of the facil-ity and runs along the west border of the Filter Room.

Its dimensions are approximately 30 ft long by 10 ft wide _by 17 ft high.

The estimated airflows through these areas and the equipment contained therein are described below for the various postulated accident scenarios.

Wind Hazard Nominal Windspeed of 110 mph It is postulated that if one of the two standard-sized doors, (approximate dimensions 2.5 ft by 7 ft) in the exterior walls. fails, air at its existing velocity could enter the facility and the initial volumetric flow would be 170,000 cfm. The nominal airspeed in an east-west direction is 13 mph.

Since flow via the room exhaust system is approximately.9,000 cfm, the inflow via the door is soon reduced due to backpressure. Only minor damage, a crack, is pos-tulated for a single glovebox.

Airspeed through the equipment is. low ~(only a fraction of the airspeed in the room).

Nominal Windspeed of 130 mph All exterior doors are assumed to fail and the air circulating into the airlock causes failure of the interior partition in the filter room. Debris 21'

from the failure of the wall damages a portion of filter enclosures.

The sur-2 f ace area of the double door is 60 ft and the airspeed is 11,440 fpm.

5 Assuming no losses, air enters the facility at 6.9 x 10 cfm.

The nominal airspeed in a north-south direction through the combined filter-room airlock is approximately 18 mph, which reduces to 7.6 mph in the glovebox room.

The air is assumed to exit the facility via another unfiltered opening (door).

Nominal Windspeed of 150 mph All exterior doors are assumed to fail and air circulating into the air-lock causes failure of the interior partition of the filter room and the north and south walls of the glovebox room. Debris from the interior partition dam-ages filter enclosures and debris from failure of the walls could damage the gloveboxes nearest the walls. -As in the previous scenario, the area of the 2

opening by which the air can enter the facility is 60 ft but the airspeed is increased to 13,200 fpm. The nominal airspeed in a north-south direction through the combined filter room-airlock is approximately 20 mph and 9 mph through the glovebox room. The air is assumed to exit the facility via the failed door at the same rate as it enters.

Nominal Windspeed of 170 mph Collapse of the roof and exterior walls exposes all the debris from the failed equipment to the existing windfield.

Earthquake Hazar_d Approximately 70% of the measured airspeeds at the site are less than 7 mph (NRC 1980, Table 3, p. 7). The exterior walls and roof are assumed to fail at linear accelerations exceeding 0.55 g.

Thus, a windspeed in excess of 5 mph is assumed for the upper-bound calculations and less than 5 mph for the average and lower-bound calculations.

SOURCE TERM RANGES 1.1 order to provide some quasi-realistic bounds to the quantity of plu-tonium estimated to be released from the damage scenarios, three estimates 22 l

(

are provided--upper-bound, average, and lower bound.

The assumptions under

(

which the estimites are made are:

l Upper Bound the upper-bound damage ratio occurs e

e the stated' inventory that.can be present is found at each location all areas have a maximum loading, on the average, of surface e

contamination e all exhaust filters are fully loaded maximum anticipated airspeeds are assumed for earthquake scenarios e

Average the average ("best estimate") damage ratio occurs e

the stated inventory at each location is reduced by the fraction of e

time it is normally found at that location all locations have a maximum loading, cn the average, of surface e

contamination e all exhaust filters are fully loaded average wind velocities are assumed for earthquake scenarios e

Lower Bound the lower-bound damage ratio occurs e

no process material is present and the maximum loading, on the aver-e age, of surface contamination is present e all exhaust filters are fully loaded average wind velocities are assumed for earthquake scenarios e

9.

23

+

SOURCE TERM ESTIMATES In the previous sections of this document, inventories of dispersible materials in various areas, damage levels, fractional airborne releases, and

' atmospheric exchange rates required to estimate the source terms for the pos-tulated damage scenarios'were described.

These components are combined in this section with the specific conditions postulated for each hazard to. arrive at three. source term estimates for each scenario--an upper limit, a best estimate, and a lower limit.

The estimates are divided into the mass of airborne plutonium particulate material in the respirable size fraction released during five time intervals-

. covering a four-day period. The quantity designated as instantaneous is the mass released from the facility within a few minutes following the hazardous event. The mass estimated for the remaining four time periods comes from two sources--the delayed release of material airborne in enclosures and the resus-

' pension of the dispersable materials exposed to the ambient windfield.

Drawings are used to illustrate the type and range of damage that could result-in key areas from the scenarios described. The illustrations are not an attempt to show what actually happens--the' data available and the state-of.

the-art are not sufficient to predict the precise levels of damage that would be inflicted upon each item.

Certain details of the facility have been omitted for clarity in the' drawings.

The discussion is divided into wind and earthquake hazards in order of their increasing severity.

SOURCE TERM ESTIMATES FROM WIND HAZARD

-6 Nominal Windspeed 110 mph (49.2 m/sec); 3 x 10 /yr Probability of Occurrence It is postulated that if one of the standard-sized doors in the exterior walls'of the glovebox room fails, the window of one of the gloveboxes nearest the door would be cracked by the wind circulating around the room.

The enclo-sure around the exterior door prevents the entry of wind-generated missiles.

The ventilation and exhaust system continues to function.

24

The'two gloveboxes nearest the doors are #8 (nearest the door in the west

' exterior wall) and #16.(nearest the door in the east exterior wall).

Glovebox

1. 8, 00 Grind-Box, is 84 in. x 48 in. x 48 in, and can contain 120 g Pu + 350 g

~

235 2

-U It is assumed to h' ave an internal contaminated surface area of 266 ft 2

(24.7 m ).

G1ovebox #16-is an analytical enclosure (x-ray' fluorescence) and has no routine inventory.. Its dimensions are 42 in. x 42 in. x 42 in. and it 2

is assumed to have an internal contaminated surface area of 147 ft2 (13.7 m ),

Because of its higher inventory and larger assumed contaminated surface, glove-box'#8'was used for this ar, jsis.

It is postulated that if a significant crack occurs in the viewing window of glovebox #8, the mixed carbide present would be completely oxidized by the air drawn into the enclosure (see Figure 4). The exhaust system is functional and 99% of the material airborne is drawn into the exhaust filter.

It is assumed that as much as 1% of the inventory in the glovebox (1.2 g Pu) as an upper' bound can back-diffuse through the crack and be released to the glovebox room atmosphere. The room exhaust system also remains functional and most of the particles released into the glovebox room are carried to the room filters or are deposited on various surfaces.

It is assumed that the 1% of the mate-rial released to the glovebox room diffuses from the facilities at a linear rate over the next 8 h.

The assumption for the average case is that 0.06 x 120 g of Pu are avail-able in the glovebox and only the finely divided carbide oxidizes.

Essentially all of the material is carried to the exhaust filters and the material released to the room behaves as-described above. Thus, the quantity released to the atmosphere around the building is less than 10-7 g Pu.

Surface contamination at the level of 10-4 g Pu/m is present in all 2

2 cases. The assumed area contaminated is 24.7 m and the postulated inventory is 2.5 mg Pu.

The release factor for a perforated glovebox is 10-4 and 2.5 x 10-7 g Pu become airborne from this source.

If 10-2 back-diffuses from the glovebox and 10-2 of that quantity is released to the ambient atmo-sphere _around the facility, the calculated airborne release would be much less than 10-7.g Pu.

25

______m_______.__._.________.___.___.m___._________.______.____._m___m_.__a

-_w 4

93## p ->

goS

,Q "W"

. g

+

,s p

~

Y

's

..y

~

c

~

' ^/

2

/

k~g,-

\\

g

~_/:K')~.

f

..g

. /

s

/

)

I

/:

ak.,-

.I

~..

~-

.v,.

S

?.

3,*$ **"# u9** os93****9ost*

(OT es SM st

• ss* 4M

1. 9*, '

e

[1 Airborne release of Pu by aerodyn wic entrainment is not considered to be of concern in this scenario since no new contaminated surfaces were exposed to the glovebox room atmosphere.

The estimated airborne releases are listed in Table 1.

All particles air-borne from this scenario are assumed to be less than 10 m AED.

Nominal Windspeed 130 mph'(58.1 m/sec); 8 x 10-7/yr Probability of Occurrence In addition to the damage postulated in the previous scenario, if the interior ptrtition between filter the room and airlock fails at this windspeed, debris from the failure of the partition could damage 1/3 of the filter enclo-sure with upper and lower bounds of 2/3 and 1/6, respectively.

Damage to any of the filter enclosures would likely make the exhaust system not able to func-tion. The situation is illustrated in Figure 5.

Each enclosure holds three standard-size HEPA filters that sequentially filter the room air and periodically filter the glovebox exhaust. The filters were assigned inventories of 0.1, 0.05 and 0.01 g Pu, respectively.

The esti-mated total inventory for all nine enclosure:; is 0.151 g Pu.

The release fac-tor for crushed filters is 10-1 and the estimated releases to the ambient atmosphere around the facility are:

upper bound--2/3 (0.151 g Pu x 10-1) = 0.01 g Pu e

average--1/3 (0.151 g Pu x 10-1) = 0.005 g Pu e

lower bound--1/6 (0.151 g Pu x 10-1) = 0.0025 g Pu.

e The Pu not instantaneously airborne is exposed to the air velocity in the area (approximately 18 mph) and is subject to entrainment due to aerodynamic l

stress. A entrainment rate of 10-8/sec is applied for powders at this air-speed. The estimated airborne releases are:

l upper. bound--0.141 g Pu (3.6 x 10-5/h) = 5 x 10-6 g Pu/h e

average--0.146 g Pu (3.6 x 10-5/h) = 5 x 10-6 g Pu/h e

lower bound--0.148 g Pu (3.6 x 10-5/h) = 5 x 10-6 g Pu/h.

e The combined estimated airborne releases from this scenario (including the airborne releases from the previous scenario) are shown in Table 1.

I 27

e,.J W ed Y goSe # **

,yN n

. i c.

y-s p

Qs A s

N

'y Ns

/ 's e

j 4

/, 3, s

'.Y. %e:,Q j

?

y

• $c 4

L

~

Of+

, 'y 1[.

..- [

r o

-d r

%,K'

• N s

.y x

3 x

?y ::

,.' b

\\

N

, ;~. _ N w.

A

^

)/,,

[

2

^ f y\\,

V K,,,.

N p

N'N

~.

t

/

FIGRE 5.

The Range and Types of Damage Postulated for the NK)F at a Nominal Windspeed of 130 mph j

w Nominal Windspeed 150 mph (67.1 m/sec); 4 x 10-7/yr Probability of Occurrence h

In addition to the damage described in the previous two scenarios, the north and south interior walls in the Glovebox Room may collapse at 170 mph, A few of the roof panels could be dislodged by the failure of the wall be ns and could damage one or two gloveboxes near the walls.

The gloveboxes nearest the north interior wall are #1a, 2a, 13, and 14; and #26, 27 and 28 nearest the south interior wall.

Gloveboxes #1a and 2a are air and inert-gas entry boxes, 36 in. x 30 in. x 33 in., and do not routinely contain any radioactive inventory. Glovebox #13 is a pin-loading box, 84 in. x 42 in. x 42 in., and normally only contains Pu during manual opera-tions. Glovebox #14 is a waste-loading box, 84 in. x 42 in. x 42 in., and nor-mally holds small que.ctities of highly diluted radioactive materials. Glove-boxes #26, 27 and 28 form a cluster and are designated as pellet-pressing, granulating, and sintering boxes. They appear to constitute a greater hazard than the other gloveboxes and, for this reason, these boxes were used for this analysis. The situation is illustrated in Figure 6.

It is postulated that the damage to the gloveboxes maybe sufficiently severe to constitute a complete loss of integrity--crush.

None of the boxes has a' stated inventory and therefore the airborne release is based upon surface contamination and exhaust-filter inventory.

For this reason, the two larger gloveboxes (#26 and 27) are used in this analysis.

The airborne release for crushed gloveboxes containing surface contamina-tion only is estimated by applying a resuspension factor of 10-2/m. The dimensions of glovebox #26 are 84 in. x 42 in. x 48 in. and its voluire is 3

2.78 m.

The dimensions of glovebox #27 are 84 in. x 42 in. x 42 in and its 3

volume is 2.43 m. Using a maximum average surface contamination level of 9 IU/d, the estimated instmtaneous releases are:

g Pu/m )(2.78 m ) = 3 x 10-6 g Pu 2

3 glovebox #26--(10-2/m)(10-4 g Pu/m )(2.43 m ) = 2 x 10-6 g Pu.

2 3

glovebox #27--(10-2/m)(10-4 e

29

DR AWING CUT AWAY FOR CLARITY b.

.'[.

+~

. s

-y p

x

. N c

\\

~

q Q?$

phj

$s e

's IB" r

s, p,

..,.%._'~

~

s N',j

)

8 e3

"~

. h. N'.

\\

I

))$

, z.

y 3

s w

\\

'I

/

' 11

^

h FIG. IRE 6.

The Range and Types of Damage Postulated for the NWF at a Nominal Windspeed of 150 mph i

e e

e

~

Both gloveboxes'contain equipment that has been used in processing carbide fuels and the equipment is assumed to have some residual, finely divided car-bide present.

Because the carbide can ignite,-the resultant oxide can be more readily entrained.

It is estimated that the. quantities of Pu which become air-borne from this source are':

e upperbound--0.08 g Pu o average--0.04 g Pu e lower bound--0.02 g Pu.-

The instantaneous airborne release estimate for the crushed exhaust fil-ters-is 0.2 g Pu, based upon a maximum inventory of 1 g Pu per filter and a

-release fraction of 0.1.

All Pu not instantaneously suspended is potentially exposed to the exist-ing windfield. The nominal air velocities in the glovebox room and airlock filter Room exceed 5 mph and the higher resuspension flux applies. The esti-mated time-dependent airborne release rates are:

upper bound--5.72 g Pu x 3.6 x 10-5/h = 2 x 10-4 g Pu/h e. average--5.76 g Pu-x 3.6 x 10-5/h = 2 x 10-4 g Pu/h lo er' bound--5.79 x 3.6 x 10-5/h = 2 x 10-4 g Pu/h.

e The total estimated airborne release of Pu, including the contribution from the two previous scenarios', is listed in Table 1.

All airborne particles are assumed to be less than 10 m AED.

Nominal Windspeed of 170 mph (76 m/sec); 1 x 10 7/yr Probability of Occurrence A total loss of the interior and exterior walls and roof of the NM3F is postulated at this windspeed, resulting in complete loss of contained equip-ment. The situation is illustrated in Figure 7.

The instantaneous airborne release estimate is calculated from the quan-

.tity of Pu, the type of Pu material, the fraction of time the Pu is present, and the degree of damage suffered by the container. Only five gloveboxes (#3,

.4, 5, 6.and 8) have significant inventories.

Since the process used is a batch

. process, the Pu would only be found at a single location. Thus, the largest l

quantity of Pu in its most dispersible form (powder) was used in the l

t 31

l I

DRAWING CUT AWAY FOR CLARITY

-Q

.?

' \\

\\

~. ?

M, s

w.

,^

s

'; qq

)>

" 1'

.',/

_ fp '

5 '*

Qf]

S' N:

y]lk'N' O

p fx

,e r

r, m

s.

N (pp

.,'e

/

~

A g! '

d pD.

% S s

, f 4'"

m

)/ '

os

\\..-

'(47 ' o. <

j l-; / 8 1

E h y

s

~.,

4

/-

A N

~

g,,. q x

3,N..'s 1

o f.'

s FIGJRE 7.

The Range and Types of Damage Postulated for the NM)F at a Nominal Windspeed i

of 170 mph l

,{

=

g

.h,,.

' 'N

.,p p '.

'e

I t

l l

l upper-bound calculation. The fraction of time the various quantities of Pu might be present at various locations was usr~ to estimate the average contri-bution from this source.

The estimates are:

j.

upper bound--300 mg mixed carbide / oxide /m x 2.43.n3 = 0.7 g 3

powder (0.3 g Pu) 3 3

average--0.06 (300 mg Pu/m x 2.43 m ) = 0.01 g Pu e

lower bound--no inventory.

e Consistent with previous analyses, it is assumed that 10% of the particles released are less than 10 pm AED.

The instantaneous airborne release from the glovebox and building exhaust

~

filters is based on their inventory and the severity of damage inflicted upon them.

There are 33 gloveboxes in the glovebox room, each assumed to hold a maximum Pu content of 1 g.

There are 9 enclosures in the filter room, each containing 0.15 g of Pu. The total inventory in the filters is 34.4 g Pu and, applying the release factor for filter crush (0.1), the instantaneous airborne release contributed from this source is:

34.4 g Pu x 0.1 = 3 g Pu.

l All of the Pu that becomes airborne from this source is assumed to be less than 10 pm AED.

All of the gloveboxes are assumed to have all internal surfaces contami-I nated to the maximum average level (1 x 10-4 g Pu/m ).

Airborne releases 2

are estimated by a resuspension factor that relates the surface contamination to airborne concentration.

It is assumed that resuspension occurs during the crush of the box and is limited to the volume of' box. The total volume of the 3

33 gloveboxes is 59.7 m and the instantaneous release contributed from this source is 6 x 10-5 g Pu.

All of the airborne materials are assumed to be less than 10 um.

One last potential source for the instar.taneous airborne release of Pu is the residual traces-of carbide in the process equipment.

Since carbide can 33

1

\\

ignite when exposed to air, it may be more readily suspended during oxidation.

For the purposes of this analysis, it was assumed that 20 g Pu was present and that 0.01 fraction was made airborne due to its exposure to air, All airborne particles are assumed to be less than 10 um AED.

All material not instantaneously airborne is assumed to be exposed to aerodynamic entrainment. Although much of the material is clearly buried by the debris, the time-dependent airborne release is based on all materials using the higher resuspension flux. The estimates are:

upper bound--275 g Pu x 3.6 x 10-5/h = 0.01 g Pu/h average--58.3 g Pu x 3.6 x 10-5/h = 0.002 g Pu/h lower bound--49.9 g Pu x 3.6 x 10-5/h = 0.002 g Pu/h.

All of the materials that become airborne from this source are assumed to be less than 10 m AED.

The estimated airborne releases from this scenario are listed in Table 1.

SOURCE TERM ESTIMATES FROM EARTHQUAKE HAZARDS Linear Acceleration of Less Than 0.55 g; Greater Than 2 x 10~ /yr Probability of Occurrence No significant damage leading'to airborne release of Pu occurs.

Linear Acceleration in Excess of 0.55 g; 0.6 g has 1.3 x 10-3/yr Probability of Occurrence Although the details and manner of collapse may vary, the consequences of the collapse of the interior and exterior walls and roof are considered identi-cal. All instantaneous airborne releases and the upper-bound time-dependent releases are as shown for the 170 mph wind-hazard scenario.

Due to the data on prevailing winds at this site, the average and lower-bound time-dependent releases were estimated using the resuspension flux for windspeeds of 5 mph and less. The estimates are:

average--58.3 g Pu x 3.6 x 10-7/h = 2 x 10-5 g Pu/h lower bound--49.9 g Pu x 3.6 x 10-7/h = 2 x 10-5 g Pu/h.

e The estimated airborne releases for this scenario are listed in Table 1.

34

REFERENCES AI.

1975. Technical Information in Support of the Atomics International Application for Broad Nuclear Materials License.

AI-16-46, Rockwell Inter-national, Atomics International Division, Canoga Park, California.

AI.

1976.

Environmental Impact Assessment of Operations at Atomic Inter-national Under Special Nuclear Materials License No. SNM-21.

AI-76-21, Rockwell International Atomics International Division, Canoga Park, California.

Burnham, J. B., R. E. Skavdahl and T. D. Chikalla.

1964.

" Plutonium Bearing Refractory Carbides," In: Carbides in Nuclear Energy, Vol. 1.

(L. E.

Russell, et al., Editor), Macmillian and Company, Ltd., London.

Cleveland, J. M.

1967.

" Compounds of Plutonium," Chapter 12 of Plutonium Handbook, Vol. 1.

(0. J. Wick, Editor), Gordon and Beach, New York.

Dennis, Richard, ed. 1916. Handbook on Aerosols.

TID-26608, Technical Information Center, U.S. Energy Research and Development Administration, Washington, D.C.

EDAC.

1978a.

Structural Condition Documentation and Structural Capacity Evaluation of the Atomics International Nuclear Materials Development Facility at Santa Susana, California. Task I--Structural Condition.

EDAC 175-070.01, Engineering Decision Analysis Company, Inc., Irvine, California.

EDAC.

1978b.

Structural Condition Documentation and Structural Capacity Evaluation of the Atomics International Nuclear Materials Development Facility at Santa Susana, California.

Task II--Structural Condition.

EDAC 175-070.02, Engineering Decision Analysis Company, Inc., Irvine, California.

McPherson, R. B. and E. C. Watson.

1979.

Environmental Consequences of Postu-lated Plutonium Releases from the Babcock and Wilcox Plant, Leechburg, Penn-sylvania, as a Result of Severe Natural Phenomena.

PNL-2833, Pacific North-west Laboratory, Richland, Washington.

Mehta, K. C., J. R. Mcdonald and F. Alikhanlou.

1980. Response of Structures to Wind Hazard at the Atomics International Nuclear Development Facility, Santa Susana, California, Volume 1.

Institute for Disaster Research, Texas Tech University, Lubbock, Texas.

Mercer, T. T.

1977.

" Matching Sampler Penetraticq Curves to Definition of Respirable Fraction," Health Physics 33(3):259-264.

Mishima, J.

1980.

Identification of Features Within Plutonium Fabrication Facilities Whose Failure May Have a Significant Effect on the Source Term.

Features Observed in Atomics International-Nuclear Material Development Facility at Santa Susana, California.

Pacific Northwest Laboratory, Richland, Washington.

35

Mishima, J. and J. E. Ayer.

1980.

Estimated Airborne Release from the 102 Building at the General Electric Vallecitos Nuclear Center, Vallecitos.

California, as a Result of Sev'ere Wind and Earthquake Hazard.

PNL-3601, Pacific Northwest Laboratory, Richland, Washington.

Mishima, J., L. C. Schwendiman and J. E. Ayer.

1978.

An Estimate of Airborne Release of Plutonium from Babcock and Wilcox Plant as a Result of Severe Wind Hazard and Earthquake. PNL-2812, Pacific Northwest Laboratory, Richland, Washington.

Mishima, J., L. C. Schwendiman and J. E. Ayer.

1979.

Estimated Airborne Release of Plutonium From Westinghouse Cheswick Site as Result of Postulated Damage from Severe Wind and Seismic Hazard. PNL-2965, Pacific Northwest Laboratory, Richland, Washington.

Mishima, J., L. C. Schwendiman, and J. E. Ayer.

1980. Estimated Airborne Release of Plutonium from the Exxon Nuclear Mixed 0xide Fuel Plant at Rich-land, Washington, as a Result of Postulated Damage from Severe Wind and Earthquake Hazard. PNL-3340, Pacific Northwest Laboratory, Rich.and, Washington.

NRC.

1980.

Environmental Character Around Atomics International's Chatsworth California Pla2t.

L. E. Rouse letter to Dr. M. E. Remley, dated 5/7/80, Docket No. 70-25, Project M-3.

Paffreyman, M. and P. E. Potter.

1964.

"The Volatility of Plutonium Carbide," In: Carbides in Nuclear Energy, Vol.1 (L. E. Russel et al.,

Editor) Macmillian and Company, Ltd., London.

Sehmel, G. A. and F. A. Lloyd.

1974.

" Particle Resuspension Rate," In:

Proceedings of the Symposuim on Atmosphere Surface Exchange of Particulate and Gaseous Pollutant _s_, eds. R. J. Engelmann and G. A. Sehmel, CONF-740921, U.S. Energy Research :nd Development Administration, Technical Information Center, Washington, D.C.

Snowdon, R. G.,

et al.

1964.

"The Behavior of Carbides in Hydrogen and Oxygen," In: Carbides in Nuclear Energy, Vol. 1, (L. E. Russel, et al.,

Editor) Macmillian and Company, Ltd., London.

Westinghouse.

19/4. Westinghouse Cheswick Site, Fuels Development Laboratories, Environmental Report. Westinghouse Nuclear Fuels Division, Box 355, Pittsburgh, Pennsylvania.

G 36

Attachment A. Page 1

,m.

i.:

b Atomics internationat Division E90c Ce kro Avenue Rockwell Canega Park Ca:. forma 91304 m aj 34i.,cco International February 17, 1978 In reply refer to 7.8A:T.yl#E M r. Richard W. Starostecki, Chief Fuel Reprocessing and Recycle Branch Division of Fuel Cycle and Material Safety U. S. Nuclear Regulatory Cornmission Washington, DC 20555

Subject:

Evaluation of Natural Phenomena Effects, Docket No. 70-25 Your letter of January 27 requested additional information necessary for the evaluation of the effects of natural phenomena on our Nuclear Materials Development Facility (NMDF) near Santa Susana, California.

The informa-tion on " inventory-at-risk" has been summarized in Table I, included as.

I should also indicate that 10 - 12 batches consisting of 120 grams of Pu and 350 grams of U-235 were processed during the last year of the last operational period at the NMDF.

In response to your second question, the one-inch natural gas supply line has been moved outside the building. One other line, which supplies a pre-mixed gas consisting of 94 percent argon and 6 percent hydrogen for regeneration, comes into the glovebox room (Room 127) and is attached to each of 13 inert gas atmosphere pur:fier units.

The surface contamination in the powder handling boxes (Boxes 3, 5, 7, and

9) has been estimated on the basis of recent smear survey results. These estimates result in contamination levels of 5 x 10-9 tram of Pu/cm2 and

~

2 x 10-8 gram of U-23 5/,.m2 The surface contamination in the other glove-boxes is estimated at a factor of 10 less than the above amounts.

The amounts of Pu and U-235 tied-up in the equipment have been estimated to be about 20 grams of Pu and 50 grams of U-235.

• I should point out that these estimates are subject to a rather large uncertainty, perhaps as much as several hundred percent.

e The amounts of combustible materials found in Room 127 durinc normal operations haie been summarized in Table II, included as Enclosure 2.

The up-to-date grading plan for the site of the NMD? consists of a group of five drawiny (No. 's 30 3-GEN-C40, -C41, -C42

-C--4 and -C4 5) together A.1 L

Attachment A. Page 2 with the storm drain master plan (Drawing No. 303-GEN-C92) which are included here as Enclosure 3.

4 If you have any questions or desire further information, please call me at (213) 341-1000, Extension 2238.

Sir.cer 1v A

M. E.

y, Mayager Health, Safety & Radiation Services Atomics International Division a

D d

9 e

e E

9 A.2

4 PNL-3935 UC-ll DISTRIBUTION No. of No. of Copies Copies 0FFSITE W. Burkhardt

'J.S. Nuclear Regulatory A. A. Chunt.

Commission DOE Patent Division Washington, DC 20555 9800 S. Cass Avenue Argonne, IL _60439 U.S. Nuclear Regulatory Commission 27 DOE Technical Information Center Division of Technica' Information and J. E. Carson Document Control Division of Environmental 7920 Norfolk Avenue Impact Studies Bethesda, MD 20014

~

Argonne National Laboratory 9700 S. Cass Avenue J. W. Johnson Argonne, IL 60439 U.S. Nuclear Regulatory Commission W. J. Hall Washington, DC 20555 Newmark & Associates Civil Engineering Building L. C. Rouse University of Illinois U.S. Nuclear Regulatory i

Urbana, IL 61801 Commission Washington, DC 20555 i

N. M. Newmark Newmark & Associates W. E. Vesely 6

Civil Engineering Building U.S. Nuclear Regulatory University of Illinois Connission Urbana, IL 61801 Washington, DC 20555 D. W. Pepper 0NSITE Savannah River Laboratory Environmental Transport Division DOE Richland Operations Office E. I. duPont deNemour Company Aiken, SC 29801 H. E. Ransom J. R. Mcdonald 38 Pacific Northwest Laboratory Texas Tech University Institute for Disaster Research C. E. Elderkin P. O. Box 4089 J. D. Jamison Lubbock, TX 79409 M. A. McKinney J. Mishima (25) 15 J. E. Ayer L. C. Schwendiman U.S. Nuclear Regulatory E. C. Watson Commission R. K. Woodruff Washington, DC 20555 Technical Information (5)

Pubiishing Coordination (2)(Ro)

]

l Distr-1