ML20106A339

NRC Staff Exhibit S-29,consisting of Fes

ML20106A339
Person / Time
Site:	Limerick
Issue date:	05/22/1984
From:	PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC
To:
References
OL, OL-S-29, S-29, NUDOCS 8408170167
Download: ML20106A339 (125)

Text

gg M g l ^

This early program has been updated and expanded; it is pre-%

K l 3

in the FES-CP.

fj sented in Section 6.1.5 of the applicant's ER-OL and is summarized'here'in 8

1 H

t.

Tables 5.8 through 5.11.

d h.i

. /h :

2 The applicant states that the preoperational program will have been implementeda fl L'

at least 2 years before initial criticality of Unit 1 to document backgroundC w

^

levels of direct radiation and concentrations of radionuclides that exist in

~

~

I ~

The preoperational program will continue up to initial..

the environment.

criticality of Unit 1, at which time the operational radiological monitoring' program will commence.

The staff has reviewed the preoperational environmental monitoring plan of the The current NRC staff applicant and finds that it is acceptable as presented.

position is that a total of about 40 dosimetry stations (or continuously an inner ring of recording dose-rate instruments) should be placed as follows:

stations in the general area of the site boundary and an outer ring in the 6 to 8 km (4 to 5 mile) range from the site with a station in each sector of each ring (16 sectors x 2 rings = 32 stations).

The remaining eight stations should 1

be placed in special. interest areas such as population centers, nearby residences The station and schools, and in two or three areas to serve as control stations.

I'.

locations have been reviewed by the NRC staff and are specified in Table 5.9.

5.9.3.4.2 Operational. -

The operational offsite radiological-monitoring program is conoucted to provide l

data on measurable levels of radiation and radioactive materials in the site environs in accordance with 10 CFR 20 and 50.

It assists and provides backup support to the effluent-monitoring program recommended in RG 1.21, " Measuring, Evaluating and Reporting Radioactivity in Solid Wastes and Releases of Radio-active Materials in Liquid and Gaseous Effluents from Light-Water Cooled Nuclear l

Power Plants."

l The applicant states that the operational program will.in essence be a j

continuation of the preoperational program described above, with some periodic adjustment of sampling frequencies in expected critical exposure pathways--such l'

as increasing milk sampling frequency and deletion of fruit, vegetable, soil, and gamma radiation survey samples.

The proposed operational program will be j

reviewed prior to plant operation.

Modification will be based upon anomalies and/or exposure pathway variations observed during the preoperational program.

I The final operational-monitoring program proposed by the applicant will be I

reviewed in detail by the NRC staff, and the specifics of the required monitor-l ing program will be incorporated into the operating license Radiological l

Technical Specifications.

i 5.9.4 Environmental Impacts.of Postulated'Acciden'ts.

i 5.9.4.1 Plant Accidents '

}

The staff has considered the potential radiological impacts on the environment l

i. '

of possible accidents at the Limerick Generating Station, Units 1 and 2, in accordance with a Statement'of Interim Policy published by the Nuclear Regula-

[L tory Commission on June ~ 13, 1980 (45~FR 40101-40104).

The following discussion reflects the staff's considerations and conclusions.

I 8408170167 840522 i

Limerick FES.

PDR ADIX;K 05000352 PDR G

4

p..

KUCLE AR REGuttJ03Y CCM*JNS'0?l Dodst tio.N-35n MT cir,:;2: tun te.

M e

t, u:, vx.: a

%, or

- _~

1 m,

_F0 c_,rt_r_, r, Staff tZ 13:.'!i:>i:0.

V A c.i::ct.

_,EG'T.TO _

?h!Kt

_ _. _ _ _._. 4?>.* :T*:0.

& O:'s Dt.TE _

,d@

k: le ::w Cst.;f.

~,~

%.RL T

l l

l t

N h

A

\\',

Tcble 5.8 Prsp;rztienal radiolegical envir:nmental monit ring pr: gram summary

.C a

No. of Frequency of Year Sample type stations Analysis analysis E

F 1982 Direct radiation 48 Gamma dose A

(partial)

Monthly m

Air (particulate & iodine) 17 Radioiodine (I-131)

Gross beta Weekly Gamma isotopic composite Monthly Surface water 5

Gamma isotopic Monthly Tritium composite Quarterly

\\

Gross ~ beta (soluble & insoluble)

Monthly Drinking water 5

Gamma isotopic Monthly Tritium composite Quarterly Gross beta (soluble & insoluble)

Monthly Groundwater 2

Gamma isotopic Semi-annually Tritium Semi-annually T

g Sediment 3

Gamma isotopic Semi-annually Fish 3

Gamma isotopic Semi-annually Vegetation 1

Radiofodine Monthly when available

' Milk I'2 Radiciodine (I-131)

Quarterly Gamma isotopic Quarterly Small game 1

Gamma isotopic Annually 1983 Direct radiation 48 Ganma dose Monthly (partisi)

Air (particulate & iodine) 17 Gross beta Weekly Gairuna isotopic composite Monthly 4

)

Table 5.8 (centinued) r-if No. of Frequency of Year Sample type stations Analysis analysis W

1983 Surface water 5

Gamma isotopic Monthly

?!

Tritium composite Quarterly Gross beta (soluble & insoluble)

Monthly Drinking water 5

Gamma isotopic Monthly Tritium composite Quarterly Gross beta (soluble & insolube)

Monthly s

Groundwater 2

Gamma isotopic Semi-annually Tritium Semi-annually Sediment 3

Gamma isotopic Semi-annually Fish 3

Gamma isotopic Semi-annually Vegetation 1

Radioiodine Monthly during 4'

growing season IS Milk 12 Radioiodine (I-131)

Quarterly

' mall game 1

Gamma isotopic Annually S

1984 Direct radiation 48 Gamma dose Monthly

\\

Air (particulate & iodine) 17 Radiciodine (I-131)

Weekly (7 stations)

Gross beta Weekly Gamma isotopic composite Monthly t

e h

Tcble 5.8 (centinu d) r-y No. of Frequency of y

Year Sample type stations Analysis analysis X-7e 1984 Surface water 5

Gamma isotopic Monthly y

Tritium composite Quarterly v'

Gross beta (soluble & insoluble)

Monthly Drinking water 5

Gamma isotopic Monthly Tritium composite Quarterly Gross beta (soluble & insoluble)

Monthly Groundwater 2

Gamma isotopic Semi-annually Tritium Semi-annually Sediment 3

Gamma isotopic Semi-annually Fish 3

Gamma isotopic Semi-annually Vegetation 1

Radiofodine Monthly during us growing season Milk 13 Radioiodine (I-131)

Bi-weekly during grazing season, monthly at other times (4 stations)

Monthly analysis

\\

only (9 stations)

Gamma isotopic Quarterly Small Game 1

Gamma isotopic Annually Source:

ER-OL Table 6.1-45, through Revision 17, February 1984

- - - - - - - - - - - - - - - - -^

e

/

Table 5.9 Preoperational radiological environmental monitoring program station locations Distance Location description Code Sector (km)

TLD (inner ring)

Evergreen & Sanatoga Rd.,

3651 N

0.97 N sector site boundary Sanatoga Rd., NNE sector 351 NNE 0.97 site boundary Possum Hollow Rd.

551 NE 0.64 Limerick Training Center 751 ENE 0.60 Keen Rd.

1051 E

0.80 Limerick Information Center 1151 ESE 0.80 Longview Rd., SE sector 14S1 SE 0.97 site boundary Longview Rd., SSE sector 16S2 SSE 0.97 site boundary Railroad tracks along 1851 S

0.48 Longview Rd.

Impounding basin, SSW sector 2151 SSW 0.80 site boundary Transmission tower, SW se-tor 2352 SW 0.80

.l site boundary f

WSW sector site boundary 25S1 WSW 0.80 Met' tower 2 site 26S3 W

0.64 WNW sector site boundary 2951 WNW 0.80 NW sector site boundary 32S1 NW 0.97 Met tower 1 site 34S2 NNW 0.97 TLD (outer ring)

Ringing Rock substation 35F1 N

6.8 Laughing Waters GSC 2E1 NNE 8.2 Neiffer Rd.

4E1 NE 7.4 Pheasant Rd. Game Farm site 7El ENE 6.8 Transmission corrider, 10E1 E

6.3 Royersford Rd.

Trappe substation 10F3 ESE 8.8 Vaughn substation 13E1 SE 6.9 Pikeland substation 16F1 SSE 7.9 Limerick FES 5-55 c.

w Tabic 5.9 (continued)

Distance Location description Code Sector (km)

Snowden substation 1901 S

5.8 Sheeder substation 20F1 SSW 8.4 Porters Mill substation 2401 SW 6.3 Transmission corrider, 25D1 WSW 6.4 Hoffecker & Keim Sts.

Transmission corrider, 28D2 W

6.1 W. Cedarville Rd.

Prince St.

29El WNW 7.9 Poplar substation 3102 NW 6.3 Yarnell Rd.

34E1 NNW 7.4 TLD (control stations and other selected locations)

Sanatoga substation 2B1 NNE 2.4 Birch substation 5H1 NE 42 Pottstown landing field 6C1 ENE 3.4 Reed Rd.

9C1 E

3.5 King Rd.

13C1 SE 4.7 3508 Market St., Philadelphia 13H3 SE 45 Spring City substation 1501 SE 5.1 Linfield substation 1781 S

2.6 Planebrook substation 18G1 S

21 Ellis Woods Rd.

2001 SSW 5

Manor substation 22G1 SW 28 Old Schuylkill Rd.

2681 W

2.7 Yost Rd.

2981 WNW 2.9 Lincoln substation 3101

NW 4.8 Friedensburg substation 32G1 NW 25 Pleasantview Rd.

3581 NNW 3.1 Dairy farms SC1 NE 4.2 9El E

6.6 9G1 E

18 1081 ESE

1. 8 10C1 ESE 4.5 Limerick FES 5-56

x

'b Table 5.9 (continued)

Distance Location description Code Sector (km) 11El ESE 7.9 17C2 S

4.0 17D1 S

5.8 18C1 S

3.1 2181 SW 2.7 22F1 SW 16 25B1 WSW 2.1 36El N

7.6 Air particulate and iodine Sanatoga substation 2B1 NNE 2.4 Pottstown landing field 6C1 ENE 3.4 Reed Rd.

9C1 E

3.5 Keen Rd.

1053 E

0.80 Limerick Information Center 1151 ESE 0.80 King Rd.

13C1 SE 4.7 2301 Market St., Philadelphia 13H4 SE 46 Longview Rd., SE sector 14S1 SE 0.97 site boundary Spring City substation 1501 SE' 5.1

~

Linfield substation 1781 S

2.6 Ellis Woods Rd.

2001 SSW 5

Manor substation 22G1 SW 28 Old Schuylkill Rd.

26B1 W

2.7 Yost Rd.

29B1 WNW 2.9 Lincoln substation 3101 NW 4.8 Met tower 1 3452 NNW 0.97 Pleasantview Rd.

3581 NNW 3.1 Vegetation Limerick Information Center 1151 ESE 0.80 garden Fish Upstream of Limerick (Kein St.

29Cl*

bridge to Hanover St. bridge)

Limerick FES 5-57 t

O Table 5.9 (continued)

Distance Location description Code Sector (km)

Downstream of Limerick 20S1*

discharge Middle of Vincent pool upstream 16C5*

to Pigeon Creek Game Fricks Lock, Limerick vicinity 26S5*

Sediment Upriver from Limerick discharge 33A2*

Linfield bridge area 1682*

Vincent Dam pool area 16C4*

Water sampling stations j

Surface water:

I Limerick intake 2451*

Fricks Lock boat house 2452*

Linfield bridge 16B2*

Philadelphia Suburban Water 15F5*

Company Perkiomen pumping station 10F2*

Drin' king water

~

Philadelphia Suburban Water 15F4*

Company Phoenixville Water Works 15F7*

Citizens Home Water Company 16C2*

Pottstown Water Authority 28F3*

Belmont Water Works 13H2*

(Philadelphia)

Well Water Limerick Information Center 1151*

Well Water S sector farm near site 18Al*

• See ER-OL Figures 6.1-23 through 6.1-29 for details.

Source:

ER-OL Table 6.1-46, through Revision 17, February 1984 Limerick FES 5-58 6

~

Table 5.10 Detection capabilities for environmental sample analyses Sensitivity Nonroutine Sample type Analysis LLD*

reporting levels Units Surface water Gross beta 4

200 pCi/l

-(insol)

Gross beta 4

200 (sol)

Tritium 2000 20000 Gamma Mn-54 15 1000 Fe-59 30 400 Co-58 15 1000 Co-60 15 300 Zn-65 30 300 Zr-95 30 400 Nb-95 15 400 Cs-134 15 30 Cs-137 18 50 Ba-140 60 200 La-140 15 200 Drinking water Gross beta 4

200 pCi/1 (insol)

Gross beta 4

200 (sol)

Tritium 2000 20000 Gamma Mn-54 15 1000 Fe-59 30 400 Co-58 15 1000 Co-60 15 300

~

Zn-65 30 300 Zr-95 30 400 Nb-95 15 400 Cs-134 15 30 Cs-137 18 50 Ba-140 60 200 La-140 15 200 Well water Tritium 2000 20000 pCi/1 Gamma Mn-54 15 1000 Fe-59 30 400 Co-58 15 1000 300 Co-60 15

- 300 -

Zn-65 30

~'

Zr-95 30 400 Nb-95 15 400 Cs-134 15 30 Cs-137 18 50 Ba-140 60 200 Limerick FES 5-59 L

Table 5.10 (continued)

Sensitivity-Nonroutine Sample type' Analysis LLD*

reporting levels Units Milk

.I-131 1

3 pCi/1 Gama Cs-134 15 60 Cs-137 18 70 Ba-140 60 300 La-140 15 300 Food products Gama i

I-131 0.06 0.1 pCi/g(wet)

Cs-134 0.06 1.0 Cs-137 0.08 2.0 Game Gama Cs-134 0.06 pCi/g(wet)

Cs-137 0.08 i

Fish Gama Mn-54 0.130 30 pCi/g(wet) 4 Fe-59 0.260 10 Co-58 0.130 30 Co-60 0.130 10 Zn-65 0.260 20 Cs-134 0.130 1

Cs-137 0.150 2

1 1

Sediment Gama l

Cs-134 0.150 pCi/g(dry)

Cs-137 0.180 j

3 i

Air particu-Gross beta 0.01 pCi/m lates Gama Cs-134 0.05 10 Cs-137 0.06 20-

~

l Air iodine I-131 0.07 0.9 pCi/m3 Direct radia-TLD RG 4.15

' mrad /std 4

tion month

• LLD is the "a priori" lower limit of detection, defined as the smallest concen-tration of radioactive material in a sample (picocuries per unit of mass or volume) that will yield a net count, above system background, that will be detected with 05% probability, with only 5% probability of falsely concluding

~

that a blank observation represents a "real" signal.

Source:

ER-OL Table 6'.1-47, through Revision 17, February 1984 i

Limerick FES 5-60

Table 5.11 Environmental sampling and measuring. equipment Sample type of measurement Equipment Airborne particulate &

Continuous air pump that radiciodine passes approximately 1 cfm through filter paper and charcoal cartridge Surface water (composite)

Automatic composite sampler

. Drinking water (composite)

Automatic composite sampler Direct radiation Thermoluminescent dosimeter i

Fish Trap net, seine, hook and line, electro fishing apparatus and/

or equivalent equipment Source:

ER-OL Table 6.1-48, Revision 17, February 1984 Section 5.9.4.2 deals with general characteristics of nuclear power plant acci-dents, including a brief summary of safety measures provided to minimize the probability of their occurrence and to mitigate their consequences if they should occur. Also described are the important properties of radioactive mate-rials and the pathways by which they could be transported to become environ-mental hazards.

Potential adverse health effects and impacts on society asso-ciated with actions to avoid such health effects also are identified.

Next, actual experience with nuclear power plant accidents and their observed health effects and other societal impacts are described.

This is followed by a summary review of safety features of the Limerick station and of-the site that act to mitigate the consequences of accidents.

i The results of-calculations of the potential consequences of accidents that have been postulated in the design basis are then given.

Also described are the results of calculations for the Limerick site using contemporary probabil-istic methods and their inherent uncertainties to estimate the possible impacts and the risks associated with severe accident sequences of low probability of

[

occurrence.

5.9.4.2 General Characteristics of Accidents The term " accident," as used in this section, refers to any unintentional event not addressed in Section 5.9.3 that results in a release of radioactive mate-I rials into the environment.

The predominant focus, therefore, is on events that can lead to releases substanttally in excess of permissible limits for normal operation.

Normal release limits are specified in the Commission's i

regulations at 10 CFR 20, and 10 CFR 50, Appendix I.

I There are several features that combine to reduce the risk associated with accidents at nuclear power plants.

Safety features provided for in design, Limerick FES 5-61

.m m;

Q-

'/.

~

' construction,.and operation comprise the first line of defense and are to a very large extent devoted to the prevention of the release of radioactive materials from their normal places of confinement within the plant.

There are also a number of additional lines of defense that are designed to mitigate the conse-quences of. failures in the 'first line.

These safety features are designed tak-ing into consideration the specific locations of radioactive materials within the plant; their amounts; their nuclear, physical, and chemical properties; and their relativ'e tendency to be transported into and for creating biological Descriptions of these features for Limerick Units 1

. hazards in the environment.

and 2 may be found in the applicant's FSAR and in the staff's Safety Evaluation Report (SER, NUREG-0991).

The most important mitigative features are described in Section 5.9.4.4(1) below.

(1) Fission Pro' duct Characteristics By far the largest inventory of radioactive material in a nuclear power plant is' produced as a byproduct of the fission process and is located in the uranium During oxide fuel pellets in the reactor core in the form of fission products.'

' periodic refueling shutdowns, the assemblies containing these fuel pellets are i

transferred to a spent-fuel storage pool so that the second largest inventory of radioactive material is located in this storage area.

Much smaller inven-l tories of radioactive materials also are normally present in the water that i

l circulates in the reactor coolant system and in the systems used to process gaseous and liquid radioactive wastes in the plant.

i i

1 All these radioactive materials exist in a variety of physical and chemical I

forms.

Their potential for dispersion into the environment depends not only l

on mechanical forces that might physically transport them, but also upon their i-inherent properties, particularly their volatility.

The majority of these materials exist as nonvolatile solids over a wide range of temperatures.

Some, j

Such however, are relatively volatile solids and a few are gaseous.in nature.

j-characteristics have a significant bearing upon the_ assessment of the environ-

~

mental radiological impact of accidents.

i The gaseous materials include radioactive forms of the chemically inert noble l

j gases krypton and xenon.

These have the highest potential for release into the atmosphere.

If a reactor accident were to occur involving degradation of the fuel cladding, the release of substantial quantities of these radioactive gases i

from the fuel is a virtual certainty.

Such accidents are of low frequency, but j

are considered credible events (see Section 5.9.4.3).

It is for this reason that the safety analysis of each nuclear power plant incorporates a hypothetical l

l design-basis accident that postulates the release of the entire contained inven-l tory of radioactive noble gases from the fuel in the reactor vessel into the containment structure.

If these gases were further released to the environment as a possible result of failure of safety features, the hazard to individuals from these noble gases would arise predominantly through the external gamma 4

I radiation from the airborne plume.

The reactor containment structure and other 1

features are designed to minimize this type.of release.

i Radioactive forms of iodine are formed in substantial quantities in the fuel by I

I the fission process and in some chemical forms may be quite volatile.

For these l

reasons, they have traditionally been regarded as having a relatively high po-tential for release (1) from the fuel at higher than normal temperatures, or (2) from defects in fuel pins.

If radiciodines are released to the environment,

-Limerick FES 5-62 e

the principal radiological hazard associated with the radioiodines 13 incor-poration into the human body and subsequent concentration in the thyroid gland.

Because of this, the potential for release of radiciodines to the atmosphere is reduced by the use of special structures, components, and systems designed to retain the iodine.

The chemical forms in which the fission product radioiodines are found are generally solid materials at room temperatures, so they have a strong tendency to condense (or " plate out") upon cooler surfaces.

In addition, most of the iodine compounds are quite soluble in or chemically reactive with water.

Although these properties do not inhibit the release of radioiodines from degraded fuel, they do act to mitigate the release both to and from con-tainment structures that have large internal surface areas and that contain large quantities of water as a result of an accident.

The same properties affect the behavior of radiciodines that may " escape" into the atmosphere.

Thus, if rainfall occurs during a relea;e, or if there is moisture on exposed surfaces (for example, dew), the radioiodines will show a strong tendency to be absorbed by the moisture.

Although less volatile than many iodine compounds, virtually all cesium and rubidium (alkali metals) compounds are soluble in or react strongly with water, and would behave similarly in the presence of mois-ture.

In addition, the more volatile iodine compounds are capable of reacting with vegetation and traces of organic gases and pollen normally present in air, while many alkali metal compounds are capable of reacting with siliceous materials such as concrete, glass and soil.

Other radioactive materials formed during the operation of a nuclear power plant have lower volatilities and by comparison with the noble gases, iodine and alkali metals have a much smaller tendency to escape from degraded fuel unless the temperature of the fuel becomes very high.

By the same token, if such mate-rials escape by volatilization from the fuel, they tend (1) to condense quite rapidly to solid form again when they are transported to a region of lower temperature and/or (2) to dissolve in water when it is present.

The former mechanism can have the result of producing some solid particles of sufficiently small size to be carried some distance by a moving-stream of gas or air.

If such particulate materials are dispersed into the atmosphere as a result of failure of the containment barrier, they will tend to be carried downwind and deposit on surfaces by gravitational settling or by precipitation (fallout),

where they will become " contamination" hazards in the environment.

All of these radioactive materials exhibit the property of radioactive decay with characteristic half-lives ranging from fractions of a Second to many days or years (see Table 5.11a).

Many of them decay through a sequence or chain of decay processes, and all eventually become stable (nonradioactive) materials.

The radiation emitted during these decay processes is the reason that they are hazardous materials. As a result of radioactive decay, most fission product elements transmute into other elements.

Iodines transmute into noble gases, for example, while the noble gases transmute into alkali metals.

Because of this property, fissicn products which escape into the environment as one ele-ment may later become a contamination hazard as a different element.

(2)

Exposure Pathways The radiation exposure (hazard) to individuals is determined by their proximity to the radioactive materials, the duration of exposure, and factors that act to 5-63 Limerick FES

\\

/

~

[.

Table 5.11a Activity of radionuclides in a Limerick reactor core

[-

at 3458 MWt (WASH-1400 basis)

Radioactive inventory Group /radionuclide (millions of Ci)

Half-life (days)

A.

NOBLE GASES Krypton-85 0.6 3,950 Krypton-85m 30 0.183 Krypton-87 50 0.0528 Krypton-88 70 0.117 Xenon-133 200 5.28 Xenon-135 40 0.384 B.

10 DINES j;

Iodine-131 90 8.05 Iodine-132 100 0.0958 Iodine-133 200 0.875 Iodine-134 200 0.0366 Iodine-135 200 0.280 C.

ALKALI METALS Rubidium-86 0.03 18.7 Cesium-134 8

750 Cesium-136 3

13.0 Cesium-137 5

11,000 D.

TELLURIUM-ANTIMONY Tellurium-127 6

0.391 Tellurium-127m 1

109 Tellurim-129 30 0.048 Tellurim-129m 6

34.0 Tellurium-131m 10 1.25 Tellurium-132 100 3.25 Antimony-127 7

3.88 Antimony-129 40 0.179 E.

ALKALINE EARTHS Strontium-89 100 52.1 1

Strontium-90 4

11,030 Strontium-91 100 0.403 Barium-140 200 12.8 F.

COBALT AND NOBLE METALS Cobalt-58 0.8 71.0 Coba1t-60 0.3 1,920 Molybdenum-99 200 2.8 Technetium-99m 200 0.25

(

Limerick FES.

5-64 c

s 4

Table 5.11a (Continued)

Radioactive inventory Group /radionuclide (millions of Ci)

Half-life (days)

F.

COBALT AND NOBLE METALS (Continued)

Ruthenium-103 100 39.5 Ruthenium-105 100 0.185 Ruthenium-106 30 366 Rhodium-105 50 1.50 G.

RARE EARTHS, REFRACTORY OXIDES AND TRANSURANICS Yttrium-90 4

2.67 Yttrium-91 100 59.0 Zirconium-95 200 65.2 Zirconium-97 200 0.71 Niobium-95 200 35.0 Lanthanum-140 200 1.67 Cerium-141 200 32.3 Cerium-143 100 1.38 Cerium-144 100 284 Praseodymium-la3 100 13.7 Neodymium-147 60 11.1 Neptunium-239 2000 2.35 Plutonium-238 0.06 32,500 Plutonium-239 0.02 8.9 x 108 Plutonium-240 0.02 2.4 x 108 Plutonium-241 4

5,350 1.5 x 105 Americium-241 0.002 Curium-242 0.5 163

~

Curium-244 0.03 6,630 Note:

The above grouping of radionuclides corresponds to that in Table 5.11c.

The listed inventory has been rounded to one significant digit to reflect its accuracy in describing the Limerick core.

All calculations, however, were done using the CRAC data file at much higher precision.

shield the individual from the radiation.

Pathways that lead to radiation ex-posure nazards to humans are generally the same for accidental as for " normal" releases.

These are depicted in Figure 5.4.

There are two additional possible pathways that could be significant for accident releases that are not shown in Figure 5.4.

One of these is the fallout onto open bodies of water of radioactiv-ity initially carried in the air.

The second would be unique to an accident J

that results in temperatures inside the reactor core sufficiently high to cause uncontrolled or unmitigated melting and subsequent penetration of.the basemat underlying the reactor by the molten core debris.

This situation could create the potential for the release of radioactive material into the hydrosphere Limerick FES 5-65 i

through contact with groundwater, and may lead to external exposure to radiation and to internal exposures if radioactive material is inhaled or ingested from contaminated food or water.

It is characteristic of the transport of radioactive material by wind or by water that the material tends to spread and disperse, like a plume of smoke from a smokestack, becoming less concentrated in larger volumes of air or water.

The results of these natural processes are to lessen the intensity of exposure to individuals downwind or downstream of the point of release, but to increase the number who may be exposed.

The bulk of radioactive releases is more likely to reach the atmosphere than to reach streams or groundwater.

For a release into the atmosphere, the degree to which dispersion reduces the concentration in the plume at any downwind point is governed by the turbulence characteristics of the atmosphere, which vary considerably with time and from place to place.

This fact, taken in conjunction with the variability of wind direction and the presence or absence of precipitation, means that accident consequences are very much dependent upon the weather conditions existing at the time of the accident.

(3) Health Effects The cause-and-effect relationships between radiation exposure and adverse health effects are quite complex (National Research Council, 1979; Land, 1980),

but they have been studied exhaustively in comparison to many other environ-mental contaminants.

Whole-body radiation exposure resulting in a dose greater than about 10 rems for a few persons and about 25 rems for nearly all people over a short period of time (hours) is necessary before any physiological effects to an individual are clinically detectable.

Dosen about 7 or more times larger than the latter dose also received over a relatively short period of time (hours to a few days),

can be expected to cause some fatal injuries.

At the severe but extremely low probability end of the accident spectrum, exposures _of these magnitudes are theoretically,possible for persons in close proximity to such accidents if mea-sures are not or cannot be taken to provide protection, such as sheltering or evacuation.

Lower levels of exposuras also may constitute a health risk, bu.t the ability to define a direct cause-and-effect relationship between any given health effect and a known exposure to radiation is difficult, given the backdrop of the many other possible reasons why a particular effect is observed in a specific indi-vidual.

For this reason, it is necesary to assess such effects on a statistical basis.

Such effects include randomly occurring cancer in the exposed population and genetic changes in future generations after exposure of a prospective parent.

The occurrence of cancer itself is not necessarily indicative of fatality, how-ever. Occurrences of cancer in the exposed population may begin to develop only after a lapse of 1 to 15 years (latent period) from the time of exposure and then continue over a period of about 30 years (plateau period).

However, in the case of exposure to fetuses (in utero.), occurrences of cancer may begin to develop at birth (no latent period) and end at 5ge 10 (that is, the plateau period is 10 years).

The health consequences model used was based on the 1972 BEIR I Report of the National Academy of Sciences (NAS,1972). -

Most autharities agree that a reasonable, and probably conservative, estimate of the randomly occurring number of health effects of low levels of radiation Limerick FES 5-66

up o

exposure to a large number of people is within the range of about 10 to 500 potential cancer deaths per million person-rems (although zero is not excluded by the data).

The range comes from the latest NAS BEIR III Report (1980),

J which also indicates a probable value of about 150.

This value is virtually identical to the value of about 140 used in the NRC health-effects models.

In l

addition, approximately 220 genetic changes per million person-rems would be projected over succeeding generations by models suggested in the BEIR III report.

This also compares well with the value of about 260 per million person-rems l

used by the NRC' staff, which was computed as the sum of the risk of specific genetic defects and the risk of defects with complex etiology.

(4) Health Effects Avoidance Radiation hazards in the environment tend to disappear by the natural processes of radioactive decay and weathering.

However, where the decay process is slow, and where the material becomes relatively fixed in its location as an environ-mental contaminant (such as in soil), the hazard can continue to exist for a relatively long period of time--months, years, or even decades.

Thus, a pos-sible consequential environmental societal impact of severe accidents is the avoidance of the health hazard rather than the health. hazard itself, by re-strictions on the use of the contaminated property or contaminated foodstuffs, milk, and drinking water.

The potential economic impacts that this avoidance can cause are discussed below.

5.9.4.3 Accident Experience and Observed Impacts As of February 1983, there were 76 commercial nuclear power reactor units licensed for operation in the United States at 52 sites, with power generating capacities ranging from 50 to 1180 megawatt electric (MWe).

(Limerick Units 1 and 2 are designed for 1055 MWe per unit).

The combined experience with all these units represents approximately 500 reactor years of operation over an elapsed tim of about 20 years. Accidents have occurred at several of these e

facilities (0ak Ridge National Laboratory,1980; NUREG-0651).

Some of these have resulted in releases of radioactive material to the environment ranging from very small fractions of a curie to a few million curies.

None is known to hava caused any radiation injury or fatality to any specific member of the public, nor any significant individual or collective public radiation exposure, nor any significant contamination of the environment.

This experience base is not large enough to permit a reliable quantitative statistical inference for predicting accident probabilities.

It does, however, suggest that significant environmental impacts caused by accidents are very unlikely to occur over time periods of a few decades.

Melting or severe degradation of reactor fuel has occurred in only one of these units, during the accident at Three Mile Island Unit 2 (TMI-2) on fiarch 28, 1979.

In addition to the release to the environment of a few million curies of noble gases, mostly xenon-133, it has been estimated that approximately 15 curies of radiotodine also were released to the environment at TMI-2 (NRC Special Inquiry Group, 1980).

This amount represents an extremely minute frac-~~

tion of the total radioiodine inventory present in the reactor at the time of the accident.

No other radioactive fission products were release'd to the environment in measurable quantity.

It has been estimated that the maximum cumulative offsite radiation dose to an individual was less then 100 mrems (NRC Special Inquiry Group,1980; President's Commission on the Accident at Three Limerick FES 5-67

Mile Island, 1979).

The total population exposure has been estimated to be in the range from about 1000 to 5300 person rems.

This exposure could produce between none and one additional fatal cancer over the lifetime of the population.'

~

.The same population receives each year from natural background radiation about 240,000 person-rems.

Approximately a half-million. cancers are expected to develop'in this group over~their lifetimes (NRC Special Inquiry Group, 1980; President's. Commission on the Accident at Three Mile Island, 1979), primarily from causes other than radiation.

Trace quantities (barely above the limit of detectability) of radioiodine were found in a few samples of milk produced in the area.

No other food or water supplies were impacted.

Accidents at nuclear power plants also have caused occupational injuries and a few fatalities, but none attributed to radiation exposure.

Individual worker exposures have ranged up to about 5 rems as a direct consequence of reactor accidents (although there have been higher exposures to individual workers as a result of other unusual occurrences).

However, the collective worker exposure

-levels (person rem) are a small fraction of the exposures experienced during normal routine operations that average about 440 to 1300 on-rems in a PWR and 790 to 1660. person-rems in a BWR per reactor year.

Accidents also have occurred at other nuclear reactor facilities in the United States and in other countries (0ak Ridge National Laboratory,1980; NUREG-0651).

Because of inherent differences in design, construction, operation, and purpose I

of most of these other facilities, their accident record has only indirect relevance to current nuclear power plants.

Melting of reactor fuel occurred in at least seven of these accidents, including the one in 1966 at the Enrico Fermi Atomic Power Plant, Unit 1.

Fermi Unit I was a sodium cooled fast breeder 1

J demonstration reactor designed to generate 61 MWe.

This accident did not release sny radioactivity to the environment.

The damages were repaired and the reactor reached full power 4 years following the accident.

It operated successfully and completed its mission in 1973.

A reactor accident in 1957 at Windscale, England, released a significant quan-ti.ty of radioiodine, apprcximately 20,000 curies, to the environment (United Kingdom Atomic Energy Office, " Accident at Windscale," 1957).

This reactor, which was not operated to generate electricity, used air rather than water to

. cool the uranium fuel.

During a special operation to heat the large amount of graphite in this reactor (characteristic of a graphite-moderated reactor), the fuel overheated and radioiodine and noble gases were released directly to the atmosphere from a 123 m (405-foot) stack.

Milk produced in a 518-km2 (200-mi2) area around tne facility was impounded for up to 44 days.

The United Kingdom National Radiological Protection Board estimated that the releases may have caused about 260 cases of thyroid cancer, about 13 of them fatal, and about 7 deaths from other cancers or hereditary diseases (NRPB-R135, Crick and Linsley, 1982).

This kind of accident cannot occur in a water moderated and cooled reactor like Limerick, however.

5.9.4.4 Mitigation of Accident Consequences Pursuant to the Atomic. Energy Act of 1954, the NRC conducted a safety evaluation of the application to operate Limerick Units 1 and 2 (NUREG-0991).

Although NUREG-0991 contains more detailed information on plant design, the principal design features are addressed in the following section.

a Limerick FES 5-68

6 l

I

)

l (1) Design Features Limerick Units 1 and 2 are essentially identical.

Each unit contains features designed to prevent accidental release of fission products from the fuel and to lessen the consequences should such a release occur.

These accident preventive and mitigative features are referred to collectively as engineered safety fea-tures (ESF).

To establish design and operating specifications for ESF, postu-lated events referred to as design-basis accidents are analyzed.

An emergency core cooling system (ECCS) is provided to supply cooling. water to the reactor core during an accident to prevent or minimize fuel damage.

Means of removing heat energy from the containment to mitigate its overpressurization following an accident are also provided.

The containment system itself 's a passive ESF, designed to prevent direct escape of released fission products to the environment.

The Limerick contain-nent structures consist of an inner primary containment and an outer secondary containment.

The primary containment is designed to withstand internal pres-sures resulting from reactor accidents.

The secondary containment surrounds the primary containment and includes all equipment outside primary containment that could handle fission products in the event of an accident.

The secondary containment is designed to collect, delay, and filter any leakaga from the primary containment before its release to the environment for all events up to and including those of design basis severity, and for some events of greater severity.

The secondary containment encloses plant areas that are accessible and, there-fore, ventilated during normal operation.

When a release of radioactivity is detected, normal ventilation is automatically isolated, and two ESFs--standby gas treatment system (SGTS) and reactor enclosure recirculation system (RERS)--

assume control of air flow within and from the secondary containment.

The SGTS and RERS filter the secondary containment atmospheFe and exhaust sufficient filtered air to establish and maintain an internal pressure less than the out-side atmospheric pressure.

This negative precsure is to be sufficient to pre-vent unfiltered air leakage from the building.

Radioactive iodine and particu-late fission products would be substantially removed from the SGTS and RERS flow by safety graoe activated charcoal and high-efficiency particulate air filters.

A filtered exhaust system als; encloses the spent fuel pool.

The main steamlines pass through the secondary containment in going from the reactor to the turbine building.

Any leakage of the main steamline isolation valves, therefore, could pass through those lines without being intercepted by-the SGTS and RERS.

To prevent this passage, a leakage control system is designed to collect main steamline isolation valve leakage and direct it into the secondary containment atmosphere and sumps, so that any airborne emissions are processed by the SGTS and RERS.

All mechanical systems mentioned above are' designed to perform their functions given single failures, are qualified for their anticipated accident environments, and are supplied with emergency power from onsite diesel generators if normal offsite and station power is interrupted.

~

Limerick FES 5-69

l:

1 Much more extensive discussion of.these design features may be found in the applicant's FSAR and the staff's SER (NUREG-0991).

In addition, the implementa-tion of the lessons learned from the TMI-2 accident--in the form of improvements in design, procedures, and operator training--will significantly reduce the likelihood of a degraded core accident that could result in large releases of

-fission products to the containment.

The applicant will be required to meet the TMI related requirements specified in NUREG-0737.' As noted in Section 5.9.4.5(7), the relative improvement in safety from these actions has not been quantified in this statement.

(2) Site Features The NRC's reactor site criteria,10 CFR 100, require that the site for every power reactor have certain characteristics that tend to reduce the risk and potential impact of accidents.

The discussion that follows briefly describes the Limerick site characteristics and how they meet these requirements.

L First, the site has an exclusion area, as required by 10 CFR 100.

The total site area is about 241 ha (595 acres).

The exclusion area, located within the site boundary, is a circular area with a minimum distance of 762 meters (2500 feet) from the center of Unit 1 and Unit 2 to the exclusion area boundary.

There are no residents within the exclusion area.

The applicant owns all sur-face and mineral rights in the exclusion area and has the authority, as re-quired by 10 CFR 100, to determine all activities in this area.

Several state-maintained roads traverse the area, allowing access to the plant and to the Schuylkill River.

One railroad and the Schuylkill River traverse the exclusion area.

The Schuylkill River, including that section within the exclusion area, 4

is used for recreational activities such as boating and fishing.

In the event of an emergency, the applicant has made arrangements with Pennsylvania State Police to control access to and activities on the Schuylkill River and the roads 4

traversing the exclusion area..The applicant also has made arrangements with Conrail for authority to control activities on the-railroad traversing the exclusion area.

Second, beyond and surrounding the exclusion area is a low population zone (LPZ),

also required by 10 CFR 100.

The LPZ for the Limerick site is a circular area with a 1.27-mile (2.04-km) radius. Within this zone, the applicant must ensure that there is a reasonable probability that appropriate protective measures could be taken on behalf of the residents in the event of a serious accident.

The ap-plicant has indicated that 1177 persons lived within a 1.27-mile (2.04-km) radius in 1980.

The major source of seasonal transients within the same 1.27-mile (2.04-km) radius of the site are the patrons of the Countryside Swim Club, which.

is located 1.2 miles west-southwest.

The 1980 industrial employee population within the LPZ was 87 persons.

In case of a radiological emergency, the applicant has made arrangements to carry out protective actions, including evacuation of personnel in the vicinity uf the plant (see also the following section on emergency preparedness).

Third, 10 CFR 100 also, requires that the distance from the reactor,to the near-est boundary of a densely populated area containing more than about 25,000 residents be at least one and one-third times the dista~nce from the reactor to the outer boundary of the LPZ.

Because accidents of greater potential hazards I

Limerick FES 5-70 v

=

x than those commonly _ postulated are highly improbable, although conceivable, it was considered desirable to add the population center distance requirement in 10 CFR 100 to provide for protection against excessive doses.to people in large-centers.

Pottstown borough, with a 1980 population of 22,729, located 1.7 miles northwest of the site, is the nearest population center.

This population center distance is at least one and one-third times the LPZ distance.

The population density within a 30-mile (48.2-km) radius of the site was 1215 people /mi2

-(3147 people /km )'in 1980 and is projected to increase to about 1966 people /mi2 2

2 (5092 people /km ) by the year 2020.

.The safety evaluation of the Limerick site has also included a review of poten-tial external hazards, that is, activities offsite that might adversely affect.

th_e operation of the nuclear. plant and cause an accident.

The review encompassed i

. nearby industrial and transportation facilities that might create explosive, fire, inissile or toxic gas hazards.

The risk to the Limerick station from such hazards has been found to be negligible.

A more detailed discussion of the

" compliance with the Commissic,..'s siting criteria and the consideration of external hazards is in the Limerick SER (NUREG-0991).

(3) Emergency Preparedness.

1he emergency preparedness plans, including protective action measures for Limerick station and environs,' are in an advanced, but not yet fully completed stage.

In_accordance with the provisions of 10 CFR 50.47, effective November 3, 1980, no operating license will be issued to the applicant unless a finding is made by the NRC that the state of onsite and offsite emergency preparedness provides reasonable assurance that adequate protective measures can and will be taken in the event of a radiological emergency..Among the standards that must be met by these plants are provisions for two emergency planning zones (EPZs);

a plume exposure pathway EPZ of about 10 miles (16 km) in radius and an inges-tion exposure pathway EPZ of about :;0 miles (80 km) in radius. Other standards include appropriate ranges of protective actions..for each of these zones, aro-visions for dissemination-to the public of basic emergency planning inforration, provisions for rapid notification of the public during a serious reactor emer-gency, and methods, systems, and equipment for assessing and monitoring actual or potential offsite consequences in the EPZs of a-radiological emergency condition.

NRC and the Federal Emergency Management Agency (FEMA) have agreed that FEMA will make a finding and determination as to the adequacy of state and local government emergency response plans. NRC will determine the adequacy of the applicant's Emergency Response Plans with respect to the standards listed in 10 CFR 50.47(b), the requirements of Appendix E to 10 CFR 50, and the guidance contained in NUREG-0654/ FEMA-REP-1, Revision 1, " Criteria for Preparation and Evaluation of Radiological Emergency Response Plans and Preparedness in Support of Nuclear Power Plants," dated November 1980.

After the above determinations by NRC and FEMA, the NRC will make a finding in the licensing process as to the state of preparedness.

The NRC staff findings will be reported in a supplement,

to the SER. Although the presence of adequate and tested emergency plans canno~t prevent an accident,.it is the staff's judgment that such plans when implemented can mitigate the consequences to the public if an accident should" occur.

l bl r

i i.

. Limerick FES 5 -

ii

'1

~

y 3

5.9.4.5 Accident Risk 'and Impact Assessment (1) Design-Basis' Accidents-As a means of ensuring that certain features of the Limerick facility meet acceptable design and performance criteria, both the applicant and the staff

have~ analyzed the potential consequences of a number of postulated accidents.

-Some of these could lead to significant releases of. radioactive materials.to the environment, and calculations have been performed to estimate the potential i[

radiological consequences to persons off site.

For.each postulated initiating event, the potential __ radiological consequences cover a considerable range of values,. depending upon the particular course taken by the accident and related conditions, including,tind direction ~ and weather prevalent during the accident.

In.the Limerick safety analysis and evaluation, three categories of accidents have been considered by the applicant and the staff.

These categories are-

. based on probability of occurrence and include (1) incidents of moderate fre-quency (events'that can reasonably be expected to occur during any year of operation); (2) infrequent accidents (events that might occur once during the lifetime of the plant); and (3) limiting faults (accidents not expected to occur but that have the potential for significant releases of radioactivity).

The radiological consequences of incidents in the first category, also called anticipated operational occurrences, are discussed in Section 5.9.3.

Some of-the initiating events postulated.in the second and third categories for the Limerick units are shown in Table 5.11b.

These events are designated design-basis accidents in that specific design and operating features such as described in Section 5.9.4.4(1) are provided to limit their potential radiological conse-quences.

Approximate radiation doses that might be received by a person at the Table 5.11b Approximate doses during_a 2-hour exposure at the exclusion _ area boundary

• Duration Whole-body Thyroid' Accidents and faults of release dose (rems) dose (rems)

INFREQUENT ACCIDENTS Category 2 Fuel-handling accident

<2 hours

0. 5 1

LIMITING FAULTS Category 3 Main steamline break

<2 hours 1

80 Control rod drop hours-days 0.1 '

O.7 Large-break LOCA.

hours-days.

5 300 i

• 2500 feet (762 m) from centers of Unit 1 or 2.

All numbers have been rounded to one significant digit.

I i

Limerick FES 5-72 u

i

\\

~

exclusion area boundary are also shown in the table, along with a characteriza-tion of the duration of the releases.

The results shown in the table reflect a conservative estimate of the potential upper bound of individual radiation exposures from the initiating accidents in Table 5.11b for the purpose of imple-menting the provisions of 10 CFR 100 and are reported in the staff's Safety Evaluation Report (SER, NUREG-0991).

For these calculations, pessimistic (con-servative) assumptions are made as to the course taken by the accident and the prevailing conditions.

These assumptions' include conservatively large amounts of radioactive material released by the initiating events, additional single failures in equipment, operation of ESFs in a degraded mode,* and very poor meteorological dispersion conditions. The results of these calculations show that radioiodine releases have the potential for offsite exposures ranging up to about 300 rems to the thyroid.

For such an exposure to occur, an individual would have to be located at a point on the site boundary where the radioiodine concentration in the plume has its highest value'and inhale at a breathing rate characteristic of jogging for a period of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> during very poor atmospheric dispersion conditions.

The health risk to an individual receiving such a thyroid exposure is the potential appearance of benign or malignant thyroid nodules in about 1 out of 10 cases, and the development of a fatal cancer in about 4 out of 1000 cases.

The staff experience has been that realistic dose estimates for a spectrum of accidents up to and including those as severe as design-basis accidents would result in values considerably lower th'n the design-basis accidents established a

for the purpose of implementing the provisions of 10 CFR Parts 50 and 100 as re-viewed in the staff's SER.

None of the calculations cf the impacts of design-basis accidents described in this section take into consideration possible reductions in individual or popu-lation exposures as a result of any protective actions.

(2) Probabilistic Assessment of Severe Accidents In this and the following three sections, there is a discussion of the proba-bilities and' consequences of accidents of greater severity than the design-basis accidents discussed in the previous section.

As a class, they are considered less likely to occur, but their consequences could be more severe for both the plant itself and for the environment.

These severe accidents (heretofore fre-quently called Class 9 accidents) can be distinguished from design-basis acci-dents in two primary respects:

they all involve sub'stantial physical deteriora-tion of the fuel in the reactor core to the point of melting, and they involve deterioration of the capability of the containment structure to perform its intended function of limiting the release of radioactive materials to the envi-ronment.

It should be understood that even the very severe reactor accidents, unlike weapons, would not result in blast and in high pressure-and high temperature-related consequences to the offsite public or to the environment.

The assessment methodology employed is essentially as described in the reactor,

safety study (RSS, WASH-1400) which was published in 1975 (NUREG-75/014), but includes improvements in the assessment methodology that were developed after

• The containment system, however, is assumed to prevent leakage in excess of that which can be demonstrated by testing, as provided in 10 CFR 100.11(a).

Limerick FES 5-73 N

v

~

N i

1 1

publication of the RSS* (such as better thermal-hydraulic models, more precise core melt phenomenology and containment response analysis).

The assessment is

.also plant and site specific.

1 In the Limerick Environmental Report--Operating License stage (ER-OL) Revi-sion 12, April 1983, the applicant'has presented a plant-and site-specific probablistic assessment of severe accidents, iacluding the effects of external events such as fires and earthquakes.

The details of the applicant's analysis are contained in a supporting document, " Limerick Generating Station Severe j

Accident Risk Analysis (LGS-SARA)," which also includes information from the applicant's earlier submittal " Limerick Generating Station Probabilistic Risk l

i Assessment (LGS-PRA)."

As a direct result of the applicant's efforts in per-forming th'e probabilistic assessment, several risk reduction modifications to the plant design were implemented during its construction.

These modifications have been reviewed by the staff and are incorporated into the staff's analysis.

The NRC staff contracted with the Brookhaven National Laboratory (BNL) to review

- portions of the LGS-SARA.

The results of BNL's review of LGS-PRA is reported in NUREG/CR-3028, and that of the earthquake and fire hazards from the SARA is sum-marized in the draft report attached to the staff's letter to the applicant dated August 31, 1983.

By letter dated March 13, 1984 the applicant informed the staff that errors in the LGS-SARA consequence analysis had been discovered.

The staff

[

has datermined that correction of the applicant's errors will not change the con-j clusions contained herein.

The results of an independent staff analysis of 1

severe accidents are summarized below..Neither the applicant's analysis nor the I

staff's analysis includes the potential effects of sabotage; such an analysis is 4

considered to be beyond the state of the art of probabilistic risk assessment.

However, the staff judges that the additional risks from severe accidents ini-tiated by sabotage are within the uncertainties of risks presented for the

(

severe accidents considered here.

Accident sequences initiated by both internal and external causes that are used in the staff analysis are described in Appendix H to this report, based on in-formation provided by BNL.

Accident sequences are grouped into " release cate-gories" based upon similarities of the sequences regarding core-melt accident progression, containment failure characteristics, and the parameters of atmos-pheric release of radionuclides required for consequence analysis.

Included in the list of potential accident initiators that are called external events are fires and earthquakes.

The staff concurs with the SARA findings 1

that the hazards due to other external events such as floods, tornadoes, trans-portation accidents, industrial accidents, and turbine missiles do not contri-bute significantly to the risk from severe accidents.

• However, there are large uncertainties in the assessment methodology and the results derived from its application.

A discussion of the uncertainties is provided in section 5.9.4.5(7).

Large uncertainties in event frequencies and other areas of risk analysis arise, in part, fro'm similar causes in all plant and site assessments; hence the results are better used in carefully constructed comparisons rather than as absolute values.

External event freqdencies used here are, however, more representative of the Limerick site than those used in the RSS.

Limerick FES 5-74 L

Table 5.11c provides information used in the staff's consequence assessment for-

.each specific release category and summarizes the BNL analysis described in The information includes time estimates from termination of the

Appendix-H.

fission process during the accident until the beginning of release to the envi-t ronment (release time), duration of the_ atmospheric release, warning time for offsite evacuation, and estimates of the energy associated with the release, height of the release location above the ground level, and fractions of the core

-inventory (see Table 5.11a) of seven groups of radionuclides in the release.

The radionuclide release fractions shown in Table 5.11c were derived using WASH-1400 radiochemistry assumptions of fission product releases from fuel and their attenuation through various elements of the primary system and contain-ment such as the suppression pool and aerosol transport in the con _tainment building as described in Appendix H.

The number in parentheses following the designation of each release category in Table 5.11c indicates its relative rank in terms of the magnitude of the core-fraction of cesium estimated to be in the release.

Cesium was chosen because of its biological significance.

Th[BNL-calculated mean value (i.e., the point estimate or the best estimate) of probability associated with each release category used in the staff ar,alysis, is shown in Table 5-11d (see Appendix H and 'Section 5.9.4.5(7)).

In this table, the probability of each accident sequence or release category is shown in two separate parts based on the cause of the accident.

One contribution to the probability is ascribed to the accident-initiating events that include plant.

internal causes, fires, and earthquakes of low to medium severity (effective peak ground acceleration less than 0.4 g; that is, Modified Mercalli (MM) intensity scale VIII or lower) (see Aooendix H).-[In Table 5.11c of the DES ^ t fsupplementreleasefractionsforfourreleasecategorieswerefoundtobein error (IV-T/DW, IV-T/W, IV-T/W and IV-A/DW) and these have been corrected.nal Tne secono contrioution to tne probability is ascribed to very severe re earthquakes (effective peak ground acceleration equal to or greater than 0.4 g; that is, MM intensity scale IX or higher) (see Appendix H) as potential cause of reactor accidents, which would also alter offsite conditions adversely to seriously hamp,er emergency responses that would mitigate the consequences of such accidents.

(Appendix I provides a description of potential offsite damages from earthquakes of various intensities.) As in the RSS, there are substantial uncertainties in these probabilities.

This is due, in part, to difficulties associated with the quantification of human error and to inadequacies (1) in the data base on failure rates of individual plant components (NUREG/CR-0400),

and (2) in the data base on external events and their effects on plant systems and components that are used to calculate the probabilities.

Analyses of risks have indicated that reactor accidents having mean likelihoods of less than 10 8 per reactor year (i.e., less than once in a billion reactor years), even considering the uncertainties of such estimates, are unlikely to -

contribute substantially to estimated risks.

For this reason, and because of the low prababilities of occurrence of these accidents, the staff has omitted from any further discussion the Table 5-11c accidents and release categories for which the mean probability in Table 5-11d is estimated to be less than 10 8 per reactor year.

The magnitudes (curies)- of radioactivity release to the atmosphere for each acci-dent sequence or release category are obtained by multiplying the release Limerick FES 5-75

s i

8 Table 5.11c Summary of the atmospheric release specifications used in consequence analysis for Limerick Units 1 and 2 Warning Fractions of Core Inventory P.eleased-r-

Release Release time for Energy Release 3

Release time duration evacuation release height Inorgan-

'd a

b g

1 category (hr)

(hr)

(hr)

(10s 8tu/hr) (a)

Xe-Kr Organic I ic I Cs-Rb Te-Sb Ba-Sr Ru La -

n

[

I-T/DW(22)*

5 0.5 4

100 30 1

7(-3)**

2(-3) 2(-2) 8(-2) 1(-3) 5(-3)'

1(-3) g I-T/W(25) 5 0.5 4

100 30 1

7(-3) 1(-4) 3(-4) 1(-3) 2(-5) 7(-5) 1(-5) t I-T/W(24) 5 0.5 4

100 30 1

7(-3) 2(-4) 9(-4) 2(-3) 8(-5) 1(-4) 3(-5)

I-T/SE(14) 2 0.5 1

100 30 1

1(-1) 1(-1) 4(-1) 1(-2) 4(-1) 2(-3)

I-T/H8(20) 2 0.5 1

100 30 1

2(-1) 6(-2) 1(-1) 7(-3) 8(-2) 1(-5)

I-T/LGT(26)*** 2 3

0 1

30 0.7 3(-3) 1(-4) 5(-4) 2(-5) 3(-5) 6(-6)

I-T/IET(18) 2 3

0 1

30 0.7 2(-2) 1(-1) 5(-2) 2(-3) 3(-3)-

6(-4)

II-T/W(8) 20 4

5 1

30 1

7(-3) 7(-1) 3(-1) 2(-1) 4(-2) 4(-2) 3(-3)

II-T/SE(14) 30 0.5 7

100 30 1

1(-1) 1(-1) 4(-1) 1(-2) 4(-1) 2(-3)

III-T/W(10) 3 1

2 100 30 1

7(-3) 8(-2) 2(-1):

6(-1) 2(-2) 4(-2) ~

7(-3)

III-T/SE(5) 2 0.5 1

100 30 1

4(-1) 5(-1) 5(-1) 5(-2)-

5(-1) 3(-3)

III-T/H8(20) 2 0.5 1

100 30 1

2(-1) 6(-2) 1(-1) 7(-3)

'8(-2)

- 1(-5)

III-T/LGT(26) 0.5 4

0 1

30 0.7 3(-3) 1(-4) 5(-4) 2(-5) 3(-5) 6(-6)

III-T/IGT(18) 0.5 4

0 1

30 0.7 2(-2) 1(-1) 5(-2) 2(-3) 3(-3) 6(-4)

IV-T/DW(2) 1 3

0.5 1

30 1

7(-3) 5(-1) 5(-1) 5(-1) 6(-2) 9(-2)-

7(-3).

v.L IV-T/W(4) 1 3

0.5 1

30 1

7(-3) 5(-1) 5(-1) 5(-1) 6(-2) 8(-2) 6(-3).

IV-T/W(3) 1 3

0.5 1

30 1

7(-3) 5(-1)-

5(-1) 5(-1) 6(-2) 9(-2) 7(-3)'

IV-T/SE(5) 2 0.5 2

100 30 1

4(-1) 4(-1) 5(-1) 5(-2) 5(-1) 3(-3)

I-S/DW(23) 5 0.5 4

100 30 1

7(-3) 3(-3) 5(-3) 3(-3) 6(-4)

'3(-4).

4(-4)

IV-A/DW(1) 1 3

0. 5 1

30 1

7(-3) 5(-1) 5(-1) 5(-1) 6(-2) 9(-2) 7(-3)

IS-C/DW(13) 0 3

0.4 1-30 1

7(-3) 8(-2) 1(-1) 6(-1) 7(-3) 8(-2)~

7(-3)

IS-C/SE(14) 1 0.5 1

100 30 1

1(-1) 1(-1) 4(-1) 1(-2) 4(-1) 2(-3)

IS-C/DW(12) 1 3

1 1

30 1

7(-3) 8(-2) 1(-1) 6(-1) 8(-3) 1(-1) 7(-3)

IS-C/SE.(14) 2-0.5 2

100

.30 1

1(-1) 1(-1) 4(-1) 1(-2) 4(-1) 2(-3)-

S-H20/W(11) 3 5

3 1

130 1

7(-3) 1(-1)

- 2(-1)

.3(-1)'

1(-2) 5(-2)'

4(-3) 5-H20/SE(5) 4 0.5 4

100 30 L

4(-1) 4(-1) 5(-1) 5(-2) 5(-1) 3(-3)-

5-H20/W(9) 3 4

3 1

30 1

7(-3) 3(-1) 3(-1) 4(-1) 3(-2) 6(-2) 5(-3)-

"See Section 5.9.4.5(7) for discussion of uncertainties. Estimated numbers were rounded to one significant digit only for the purpose of this table, bSee Appendix H for designations and descriptions of the release categories.

1. rganic iodine is added to inorganic lodine for consequence calculations because organic iodine is likely to be converted to inorganic or particulate--

0 forms during environmental transport.

dIncludes Ru, Rh, Co, No. Tc.

' Includes *Y, La, Zr, Mb, Ce, Pr, Md. NP, Pu, Am. Co.

• teumber in parentheses indicates relative ranking of the release category according to cesium fraction.

• • 7(-3) = 7 x 10 3 = o,007,

• • • This release category is combined with III-T/LGT in consequence analysis.

)

)

Table 5.11d ' Summary of the calculated mean (point estimate) probabilities of atmospheric release categories Probability.of.the release category initiated by internal Probability of the release causes, fires, and low to category initiated by Release moderately severe earthquakes severe earthquakes-category (per. reactor year)

(per reactor year)

I-T/DW 2(-5)*

6(-7)

I-T/WW 2(-5) 5(-7)

I-T/9W 2(-6) 6(-8)

I-T/SE 8(-9) 2(-10)***

'I-T/HB 8(-7) 2(-8)

'I-T/LGT**

2(-5) 5(-7)

-I-T/Off 2(-5) 6(-7)

II-T/WW 2(-6) 2(-8)

II-T/SE 4(-10)***

4(-10)***

III-T/WW 2(-6) 4(-7)

III-T/SE 3(-10)***

7(-11)***

III-T/HB 3(-8) 7(-9)

III-T/LGT 7(-7) 2(-7)

III-T/GT 9(-7) 2(-7)

IV-T/0W 2(-7) 5(-8)

IV-T/WW 2(-7) 4(-8)

IV-T/WW 2(-8) 5(-9)

IV-T/SE 3(-11)***

1(-11)***

I-S/DW 4(-8) 0 IV-A/0W 5(-9) 0 IS-C/DW 1(-8) 1(-7)

IS-C/SE 1(-12)***

1(-11)***

JS-f/0W 1(-7) 9(-7)

.IS-f/SE 1(-11)***

9(-11)***

S-H20/WW 1(-8) 4(-8)

S-H20/SE 1(-12)***

4(-12)***

S-iPl6/99 1(-8) 4(-7)

Total prob-ability per reactor-year 9(-5) 5(-6)

• 2(-5) = 2 x 10 5 =.00002

• • This release category is combined with III-T/LGT in consequence analysis.

• • • Any release category with probability less than 10 9 per reactor year is omitted from consequence analysis because of its low probability and-insignificant contribution to risks.

NOTEi Please see S'ection 5.9.4.5(7) for discussion of uncertainties.

Estimated numbers were rounded to one significant digit only for the purpose of this table.

Limerick FES 5-77

l:

fractions shown in Table 5-11c by the amounts that would be present in the core at the time of the hypothetical accident and by depletion factors as a result of inplant radioactive decay during the release time.

The core inventory of radio-nuclides are shown in Table 5.11a for Units 1 and 2 at a core thermal power level of 3458 MWt. This is the power level used in the FSAR for analysis of radiological consequences and is used here instead of the 3293 MWt expected maximum power to correct for power density variations and instrument error in measurement of power levels normally present in operating reactors.

The 54 nuclides shown in the table represent those (of the hundreds actually expected to be present in the operating plant) that are potentially major contributors to the heal'.h and economic effects of severe accidents.

They were selected on the i

basis of the half life of the nuclide, consideration of the health effects of daughter products, and the approximate relative offsite dose contribution.

The potential radiological consequences of these releases have been calculated by the computer code CRAC, based on the consequence model used in the RSS (see

._NUREG-0340), adapted and modified as described below to apply to a specific site.

The essential elements are shown in schematic form in Figure 5.4a.

Environmental parameters specific to the site of Limerick station have been used and include (1) meteorological data for the site representing a full year (1976) of con-o 1

secutive hourly measurements and seasonal variations with good data recovery characteristics (annual average probabilities of wind blowing into 16 directions of the compass are shown in Table 5.11e) j (2) projected population for the year 2000 extending throughout regions of p

80-km (50-mile) and 563-km (350 mile) radius from the site i

j (3) the habitable land fraction within a 563-km (350-mile) radius I

WAtet DATA 9

DU31PficM AMDIC OF RA113ACT!vt

ggi, truAst upt.TM u

EFFECT3 Cuun ncstMTrf PeruulT!m Nim o

PROPDTY awst anus

~

CMTAM!antia NII" Figure 5.4a Schematic outline of consequence model Limerick FES 5-78

/

Table 5.11e Annual ~ average wind-direction probabilities for the Limerick site based on data for the year.1976 Wind blowing toward Probability (fraction-the. direction of the year)

N 0.07 NNE 0.07-NE 0.06 ENE 0.05 E

0.10 i

ESE 0.16 SE 0.11

.SSE 0.04 S

0.04 j

^

SSW 0.03 SW 0.03 i

WSW 0.04 W

0.07 WNW 0.03 NW 0.04 NNW 0.06 Total 1.00 l

(4) land-use statistics on a countywide basis within and statewide basis out-side of a 80-km (50-mile) region, including farm land values, farm pro-duct values including dairy production, and growing season information, i

j for the counties, the State of Pennsylvania and each surrounding state

.within the 563-km (350-mile) region for the region beyond 563 km (350 miles), the U.S. average population density 4

)

was assumed.

The calculation was extended out to 3200 km (2000 miles) from the site, to ac-count for the residual radionuclides that would remain in the atmosphere at large distances, with rain assumeo in the interval between 563 km and 3200 km to

{

deplete the plume of all non-noble gas inventory. To obtain a probability dis-tribution of consequences, calculations were performed assuming the occurrence of each release category at each of 91 different " start" times distributed i

throughout a 1 year period.

Each calculation utilized site-specific hourly j

meteorological data and seasonal information for the period following each

" start" time.

The consequence model was also used to evaluate the consequence reduction bene-i fits of offsite emergen.cy response such as evacuation, relocation, and other protective actions.

Early evacuation and relocation of people would consider-ably reduce the exposure from the radioactive cloud ind the contaminated ground l

Limerick FES 5-79

in the' wake of the cloud passage.

The evacuation model used (see Appendix J) xhas been revised from that used in the RSS for better site-specific application.

In the staff calculation, three sets of assumptions were made about the short-term emergency response that would likely be_ undertaken to minimize the severe accident health effects from early or short-term radiological exposure.

Table 5.11f lists the assumptions and parameters for each emergency response scenario evaluated.

The first set of parameters assumes evacuation of the population within 10 miles (16 km).

The effective evacuation speed in Table 5.11f is based on an evalua-tion made by thetapplicant's contractor, NUS Corporation, in an evacuation time estimate study (NUS, 1980).

The estimate of the delay time before evacuation in the same study has been rejected by the applicant in LGS-SARA and, therefore, is t

not used in the staff analysis.

Instead, the value of delay time in Table 5.11f is a staff assumption and is based partly on considerations of the NRC require-ment regarding prompt notification of the public of the emergency, and partly on the staff judgment regarding the time people would take preparing for evacuation after being notified of the emergency, for a high population density site, dur-ing normal to moderately adverse conditions such as snow, ice, hurricane, low to moderately severe earthquakes (up through M intensity scale VIII), etc.

The values of delay time before evacuation and effective evacuation speed used in the staff analysis are assumed only to be average values.

Within the 10-mile emergency planning zone there normally would be some facilities (such as nursing homes, hospitals, prisons, schools, etc.) where special equipment or personnel may be required to effect evacuation, and there may be some people who choose not to evacuate.

Therefore, actual effectiveness could be greater or less than i

that characterized by the average values.

Because special consideration will be given in emergency planning for Limerick to any unique aspects of dealing with special facilities, it is not expected that actual evacuation effectiveness would be very much less than that'modeled by the average values used here.

For areas beyond 10 miles (16 km), however, the parameters selected reflect the assumptions that an extension of emergency response would occur during a large accident and people would be advised to leave areas that would be considered to be-highly contaminated (see below for criterion), i.e., people would relocate.

i Relocation of the public from the highly contaminated areas beyond 10 miles (16 km) is assumed to take place 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> after plume passage.

The criterion j

for this relocation is whether the projected 7-day-ground dose to the total bone marrow, as projected by field measurements, would exceed 200 rems (which is only s110htly above the average threshold exposure for potential early fatality with l

minimal medical treatment); otherwise people in highly contaminated areas are assumed to be relocated within 7 days.

The offsite emergency response mode e

characterized by.these assumptions is designated Evac-Reloc.

t The second set of parameters reflects the hypothesis that the planned evacuation may not take place in a real situation for one or more reasons'such as'short warning time, indecision regarding whether to evacuate or not because of uncer-l tain plant conditions, or adverse site conditions that would cause long delay before evacuation.

In lieu of evacuation, it was assumed that people in the footprint of the plume within 10 miles (16 km) would leave the area (i.e.,

relocate) 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> after plume passage.

This 6-hour relocation f,im,e is similar to the time for evacuation assumed in the first set based on 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> delay and about 2.5 miles per hour evacuation speed.

Beyond 10 miles (16 km), relocation Limerick FES 5-80

c Tcble 5.11f Emergency rssponsa assumptiens fer cach rateter unit

'l Shielding protection-e 3.

Zone 8 factor (fraction) 0-Eff4ctive Relocation relocation dose R

_ Effective downwind zone size Zone B criterion (bone During' Other.

m Emergency Evacuation Delay evacuation distance (mi)

W M ae marrow dose evacuation,

times, response distance time speed moved ***

Zone relocation projected for plume /

. plume /

s::t no.* '

(mi)**

(hr)

(mph)

(mi)

At Bt

. time (hr) 7 days) (rems) ground ground 1

10 2

2.5 15 0

>10 12 200 11/0.51 0.7515/0.3311 2

N/Att N/A

.N/A N/A 10ttt >10 12 200 N/A 0.7511/0.3311 3

N/A N/A N/A N/A 0

>0 24 200 N/A 1.0111/0.5111-aSets 1, 2, and 3 are also identified as Evac-Reloc, Early Reloc, and Late Reloc, respectively, in text, tables, and figures.

- **To change miles to km, multiply the values shown by 1.609.

• • • An artificial parameter used only to represent a realistic path-length for each evacuee over which radiation exposure T

to the evacuee is calculated in the CRAC code.

tZone A is the 10-mile plume exposure pathway emergency planning zone; Zone B is the area outside Zone A.

ftN/A - Not Applicable.

tttRelocation takes place 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> after ground contamination.

1During evacuation, automobiles are assumed to provide essentially no shielding to gamma rays from the plume and some shielding to gamma rays from the contaminated ground.

The selected values of sh'ielding protection factors for the plume and the ground during evacuation are taken from Table VI 11-13 of Appendix VI of WASH-1400.

11At other times than during evaucation, shielding protection factors are the average values representative of normal' The selected values of the activities of the people during which some people are indoors and some are outdoors.

shielding protecticn factors for the plume and the ground for this situation are taken from Table VI 11-13 of Appen-dix VI of WASH-1400.

111During an abnormal situation in the site region caused by a external event such as a severe earthquake, it is assumed that many,of the buildings may not remain habitable to provide shielding protection to the people against. gamma rays So, the shielding factor for the plume is taken to be 1.

However, the nature of the ground surface from the plume.

So, the is assumed to become altered by debris and possibly mud / slush / water generated from a severe earthquake.

ground shielding factor (provided by the altered ground and whatever building struct situation and 0.7 for an ordinary and uncovered ground surface.

NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

H

i was assumed as in the previous set of assumptions.

The offsite emergency re-sponse mode characterized by these assumptions is designated Early Reloc and was used for an alternative risk analysis.

The third set of parameters reflects a radiological emergency response situation hampered by a severe type of external event, such as a severe regional earth-quake, which would seriously limit the ability to evacuate, and would also eliminate or reduce the shielding protection that the public would otherwise experience.

However, relocation of the public from highly contaminated areas

'24 hou s after plume passage was assumed.

The criterion for this relocation was the same as in the first set of assumptions, but relocation was assumed to extend outward from the site exclusion area boundary (762 meters, as opposed to the 10-mile (16-km) EPZ boundary); otherwise people are assumed to be relocated within 7 days.

The offrite emergency response mode characterized by this third set of assumptions is designated Late Reloc.

l l

The environmental protective actions considered as part of relatively long-j term offsite emergency response to reduce health effects from chronic exposure include:

(1) either complete denial of use (interdiction), or permitting use only at a later time after appropriate decontamination, of food stuffs such as crops and milk; (2) decontamination of severely contaminated land and property when it is considered to be economically feasible to lower the levels of con-tamination to protective action guide (PAG) levels *; and (3) denial of use (interdiction) of severely contaminated land and property for varying periods of time until the contamination levels are reduced by radioactive decay and weathering to such values that land and property can be economically decontami-nated as in (2) above.

These actions would reduce radiological exposures and health effects to the people from immediate and/or subsequent use of or living in the contaminated environment, but would also result in economic costs to implement them.

Lowering the PAG levels would lower the delayed health effects but would increase costs.

Estimates of m.eteorology-averaged societal consequerices of several types condi-tional upon occurrence of each release category in Table 5.11c are tabulated in Appendix K.

For each release category, separate estimates are provided using each of the offsite emergency response modes in. Table 5.11f.

These conditional mean values are of use only in judging the relative severity of each release category and they cannot be used directly for risk. assessment without simulta-neous association with the probability of the release category to which the consequences are due.

Therefore, in the following paragraphs, the impacts of severe accidents in the Limerick reactors are appropriately weighted by their probabilities.

• PAG 1evels used in CRAC analyses are not to be confused with those drafted by the U.S. Environmental Protection Agency (EPA-520/1-75-001, September 1975), or by the U.S. Department of Health and Human Services (47 FR 47073, October 22, 1982), for reactor accidents.

PAG 1evels used in CRAC are defined in Table VI 11-6 of WASH-1400, and were based on the recommendations of the former U.S. Federal Radiation Council and the British Medical Research l

Council.

However, for control of long-term external irradiation, the PAG level for urban areas in WASH-1400 Table VI.11-6 was used in CRAC for all areas (urban and rural).

Limerick FES 5-82

~~

The consequences and risks

• of severe accidents in the Limerick reactors ini-tiated by plant internal causes, fires, and low to moderately seves'e earthquakes were evaluated using the release categories in Table 5.11c, the corresponding probabilities in Table 5.11d, and the parameters of the Evac-Reloc mode of off-site emergency response in Table 5.11f. The consequences and risks of accidents initiated.by very severe regional earthquakes that could also affect the offsite conditions so as to seriously hamper evacuation or early relocation were eval-uated using the accident parameters in Table 5.11c, the corresponding probabil-ities in Table 5.11d, and the parameters of the Late Reloc mode of offsite emer-Finally, the overall evaluation of consequences gency response in Table 5.11f.

and risks of reactor accidents at Limerick from internal causes, fires, and low to high severity earthquakes is made by combining the results for Evac-Reloc and Late Reloc offsite emergency response modes.

The results of the staff' calculations using the consequence model are radio-logical doses to individuals and to populations, health effects that might result from these exposures, costs of implementing protective actions.and costs associated with property damage by radioactive contamination, and land area that would be subject to long-term interdiction.

These results are presented and Breakdowns for each type of consequence in terms of contribu-discussed below.

tions from accidents initiated by severe earthquakes and from accidents initiated

~

by other causes considered in the analysis are presented in Appar. dix L.

+

An alternative overall evaluation of consequences and risk in which the Evac-Reloc mode of offsite emergency response is replaced by the Early Reloc mode is presented in Appendix M.

The staff critique of the principal aspects of the applicant's consequence analysis in the Environmental Report-Operating License stage (ER-OL), which is identified to be the same as in LGS-SARA, is provided in Appendix N.

There are large uncertainties in each facet of the estimates of consequences both in the staff analysis and the applicant's analysis (see Section 5.9.4.5(7)).

(3) Dose and Health Impacts of Atmospheric Releases The results of the staff calculations of the. environmental dispersion of radio-active releases to the atmosphere and the radiological' dose to people and health impacts performed for the Limerick station and site are presented in the form of probability distributions in Figures 5.4b through 5.4f and are included in the impact summary Table 5.11g.

The graphs in Figures 5.4b through 5.4f (and in similar Figures 5.4g and 5.4h introduced later) display a type of proba-bility distribution called a complementary cummulative-distribution function (CCDF).

CCDFs are intended'to show the relationship between the probability of a particular type of consequence being equaled or exceeded and the magnitude These graphs are useful in visualizing the degree to which of the consequence.

the probability of occurrence of consequences decreases _ as the magnitude of the Probability per reactor-yearh81s the chance that a given /

consequence increases.

event would occur or a given consequence magnitude would be exceeded in 1 year

• Risk of a particular kind of consequence is to be understood as "the average value of several estimates of the product of magnitude of the particular consequence and its associated probability.

+'

H h re 5-91

,l' Limerick FES 5-83

a F

Qy1,8 1d Id

..... 10'........,8

.......8

........d Idr 1

1 I

,o

.......,o

$o z

LEGEND

-o i

o = WHOLE BODY DOSE 7 25 REM

/

y-o = THYROID DOSE 7/ 300 REM a

m7 A = BONE MARROW DOSE 7 200 REM 7 f

Ag 0

g

.- =

E*

a i

m E-.

OT T

g m

o

%-=.

s-A

[

?

T ~

o-

~

11 a

i*!

N N

N

\\

t t

o o

r

-i 2-co N

O b-t o

o u -e i-A :

>T i

-T S

o o

U #5

\\

5*

x,Y,

-'o

$'o m 18 16 18 10' id

,,,,,,, ;6 16

,,..... ;8-1 1

X= NUMBER OF AFFECTED PERSONS Figure 5.4b Probability distributions of individual dose impacts NOTE:

See Section 5.9.4.5(7) for a discussion,of uncertainties.

i

_.$.1 d

'o o

o*

0w o4 oe o-n T

Y _*! :

_I 7

._Y

$(o" u

i:

=:

o

! _: ~

rE_ _

r S

g 1.

EN c

k LO ba I

I MT l

A a

0L W

ru f,

5U t

. 1 I..

N a

P n

O I

s IP e

m r

o I

. DT D u

r NI s

f E

o

. EWS d

p g

x n

G O

1 e

i d, ENP t

1 LOX Y

n l

o u

I TE D

i s

t e

O AE B

la r

LR u

s UI E

p e

s PT d

o l

e L

p i i

1 6,

ON l

i PE O

m t

I f

n 1

I o

0 i

==

5 a

W X

s t

ao n

n r

x M

o i

e i

h c

E t

t n

(

6 R u

i u

b w

1 f

6 i

N N

r n

o 1

O t

o n

s i.

S ts o

i R

d am i

l e s

p E

y ur s

P t

p-u on c

.d L i

l po s

i s

i 6,

._ i 1A b

er d

m.

1 T

a h e O

b t p a

o T

r o0 r

=

P t0 o

f

^

0, X

e

~

c s0

)

o0 7

6 4

d8

(

1 5

5 6

l t 1

e au 4

r uo 9

u nb g

na i

a 5

F s

ei n

g o

t an i

.i T

.l ro t

(.

ei c

vt e

1..

aa S

i ed e

ha e

Tr S

_d 1

2 6

.F:

1 E: - - -

E: -

E5 -

_E : : -

o o

?a-QW oW o-

'a-s 1 !: :

M A X m g f

• o U m N A N o x 4 k m a > H j. m < m k n.

iD 1*

o S

E

)

TON i,ugu riii E-,O n

e q

9 11,

....i,d 16

.....01'

.....01'

.....01' 16 Idr r-

......o

'd

" E LEGEND

~L

~

a = WITIIIN 50 MILES-EXCLUDING THYROID Q-p-

o = WITIIIN 50 MILES-TIIYROID ONLY I'S A = ENTIRE EXPOSED POPULATION-EXCL.TIIYROID 7

0 5

+ = ENTIRE EXPOSED POPULATION-TIIYROID ONLY rS bg:

Cr:T -

.T doep-~

mEC lo w

N!

bo o

x-e e-EQ p

4

• -h

~

g To l

y' O

5

=

x:

W'o.

l b

_ 'o A :

e-e 2

4 y

OT T

%o

'o ea a

e a s a sus e

a s a s s su a

s e asssig '

i a e s ssu s

a a a issj u

a s s a sis a

a a ssTTT e

A 16 Id 16 10 10' 16 16 16 X= LATENT CANCER FATALITIES l

j Figure 5.4d Probability distributions of cancer fatalities NOTE:

See Section 5.9.4.5(7) for a discussion of uncertainties, i

.......,a'

,......,id

.......i,d

.......i..d

.......Idr

.o i d.......i d

.......i d l

E.

b-LEGEND 2.

"i o = W/ SUPPORTIVE MEDICAL TREATMENT :

g o = W/ MINIMAL MEDICAL TREATMENT

.,.o As, -

r -

-m wo

!S -5 s

_L 3

Eb_

N *!

-N

>J o

x-4'o n I

m -r A

i; b

N o

9A

!~

xb_

m a,

s,

~

._'o o

N E d I

i po g-E s_

To m

E Ao_

to - =

O E

.?o Z

4 8

id 16 10' 16 16........d 16

.'o i

1 X=EARLY FATALITIES Probability distribution of early fatalities Figure 5.4e See Section 5.9.4.5(7) for a discussion of uncertainties.

NOTE:

C i

Id id Id 16 16 16 16 16

....... r

_3,

,o o

s w

x k~

~7 g

,a

,a_,_

ki "

'f g

z; g

-e

. -i E

>I%

a o

o m -e 2-

~

T m

g g

4:

0 b

3 ^5 5#

E*a:

N T

o g

m -e x

O g

b fl.

a T

T o

o

-,d Id 1d

.......,d 16........,6 1d 16-I 1

1 X=EARLY INJURIES Figure 5.4f Probability distribution of early injuries NOTE:

See Section 5.9.4.5(7) for a discussion of uncertainties.

m

e T

9

. b[

io

,o1 ow o4 o_-

o-

.?o4 E: :

. ~

T:..

e:

g~

. ~.

d.!:

i.

s 1 u

s a

s a

=

a

)f u I s

1 s

s s

s u

e a

r s

u s

a a

e d

m d.

k s 1 I.

u n

I s

o a

i a

t s

S a

R g

s i

A t

s L

im ldL 6

s IO f

u 1

D o

s a

a t

0 s

s 8

o s

a 9

c e

i 1

s f

t o

n dN 6

s iI n

a i

1 s

o t

I i

r T

s t

e s

S u

c a

O b

n i

u C

r t

f s

L s

o l0' A i

6 d

n u' 1T o

1 v

O y i t

s a

T t

s a

i s

=

l u

s X

i c

a b

s a

i a

b d

o d

r a

6 s

r 1

u I P

s o

a g

f a

4 a

)

a 5

7

(

s e

5 r

1 si;6 u

4 6

u 1 g

9 i

F s

a 5

s a

no a

i t

a ce

'0 E: :

'_0 S

2'

=:: -

z:: _

5_-

g 1

1 e

o-o-

o-bW O'o-. hL Na

>p 3 e> * =

mOGA b

'o-kg e

S gA suO D

ETON

,Y; w 'mu U.m*

n n

C o

a us Table 5.11g Sammary of environmental impacts and probahfittles Population Latent cancer fatalities (persons)

Early fatalities Land exposure.

Persons exposed over uhele body (perscas)

Cost of area for (million Excluding offsite tene-ters Probability-200 rees25 rems person rees)*

thyreld Thyroid With With aftigation inter-of impact 300 rems total whole supportive minimal Early measures diction per reacter-thyreld marrow body 50 elles Total 50 elles Total 50 miles Total medical medical injtries (alliions (elllions year dose hse dose (90 km)

(80 km)

(80 km) treatment treatment (persons) of 1980 8) of m yas a

10

• 8 0

0 0

0 0

0 0

0 0

0 0

0 0

10

• 2(3) 3(1) 2(4)***

2(1) 3(1) 1(3) 2(3) 3(2) 3(2) 0 1(0) 9(1) 1(3) 4(1) 5 s le

• 7(3) 2(2) 5(4) 4(1) 5(1) 2(3) 3(3) 4(2) 6(2) 0 2(1) 2(2) 3(3) 7(1) 10

• 4(4) 5(3) 3(5) 7(1) 1(2) 5(3) 7(3) 2(3) 2(3) 1(3) 2(3) 4(3) 6(3) 1(2)

T IS '

2(5) 3(4) 1(6) 1(2) 3(2) 1(4) 2(4) 4(3) 4(3) 9(3) 1(4) 3(4) 2(4) 3(2) to 10 s 5(5) 2(5) 3(6) 2(2) 5(2) 2(4) 3(4) 6(3) 6(3) 2(4) 3(4) 2(5) 3(4) 7(2)

C l

See Flysre

5. 4 5.46 5.46 5.4c 5.4c 5.44 5.4d 5.4d 5.44 5.4e 5.4e 5.4f 5.4g 5.4h a

• About 260 cases of genetic effects may occur in the s'acceeding generattens per slillen persen ree to the exposed generation.

l

• • About 2.6 ellifen aguare meters equais I square alle.

• • • 2(4) = 2 x 10* = 20000.

ISTE: Please see Section 5.S.4.5(7) for discussion of uncertainties. Estimated niabers were rounded to one significant digit only for the purgese of this table.

i 9

i.

of operation for one reactor.

Different accident releases and atmospheric dis-persion conditions, source-term magnitudes, and dose effects result in wide ranges of calculated magnitudes of consequences.

Similarly, probabilities of equaling or exceeding a given consequence magnitude would also vary over a wide range because of varying probabilities of accidents and dispersion conditions.**

Therefore, the CCDFs are presented as logarithmic plots in which numbers varying over a large range can be conveniently shown on a graph scaled in powers of 10.

I For example, a consequence magnitude of los means a consequence magnitude of one million (1 followed by six zeroes); a probability of 10 s per reactor year means a chance of 1 in one million or one millionth (0.000001)'per reactor year.

All release categories shown-in Table 5.11c contribute to the results; the conse-quences from each are weighted by its associated probability (Table 5.11d).

For these calculations, the Evac Reloc mode of offsite emergency response was assumed for accidents initiated by causes internal to the plant, by fires and by low to moderately severe earthquakes; and Late Reloc mode of offsite emergency response Wds assumed for SCCidents initiated by very severe regional earthquakes (see' Table 5.11f).

Figure 5.4b.shows the probability distribution for the number of persons who might receive whole-body doses equal to or greater than 25 rems, total bone marrow doses equal to or greater than.200. rems, and thyroid-doses equal to or greater than 300 rems from early exposure;*** al1~on a per reactor-year basis.

The 200 rem total bone marrow dose figure corresponds, approximately, to a threshold value for which hospitalization would be indicated for the treatment of-radiation injury.

The 25-rem whole-body dose (which has been identified earlier as the lower limit for a clinically observable physiological effect in nearly all people) and the 300-rem thyroid dose figures correspond to the Commission's guideline values for reactor siting in 10 CFR 100.

Figure 5.4b shows in the left-hand portion.that there are, approximately, 60 chances in 1 million (6 x 10 s) per reactor year that one or more persons may receive doses equal to or greater than any of the doses specified.

The fact that the three curves run almost parallel in horizontal lines initially shows th'at if one person were to receive'such doses, the chances are about the same that up to 10 would be so exposed.

The chances of larger numbers of persons being exposed at those levels are seen to be considerably smaller.

For example, the chances are less than 1 in 1 million (10.s) that 10,000 or more people might receive doses of 200 rems or greater. 'A majority of the exposures reflected in this figure would be expected to occur to persons within a 40-km (25-mile) ra--

dius of the plant.

Virtually all would occur within a 160-km (100-mile) radius.

Figure 5.4c shows the probability distribution for the total population exposure in person-rems; that is, the probability per reactor year that the total popula-Most of the population-tion exposure will equal or exceed the values given.

l-

• ry in the plots means reactor-year.

• • See (7) below for further discussion of areas of uncertainty.

• • • Early exposure to' an individual includes external doses from the radioactive cloud and the contaminated ground, and the dose from internally deposited l

radionuclides from inhalation of contaminated air during the cloud passage.

i Other pathways of exposures are excluded.

i Limerick FES 5-91 i j l

..-._ _ _ _.,_. _., _ d

~_.

/

exposure _up_to 100 million person-rems would occur within 80-km (50 miles) but very severe releases would result in exposure to persons beyond the 80-km (50-mile) range, as shown.

For perspective, population doses shown in Figure 5.4c may be compared with the annual average dose to the population within 80 km (50 miles) of the Limerick site.resulting from natural background radiation of about 800,000 person-rems,

- and to the anticipated annual population dose to the general public (total U.S.)

from normal plant operation of about 80 person-rems (both units, excluding plant workers) (Appendix D of the environmental statement, Tables D.7 and D.9).

' Figure 5.4d represents the statistical relationship between population exposure and the induction of fatal cancers that might appear over a period of many years

'following exposure.

The impacts on the total population and the population within 80 km (50 miles) are shown separately..Further, the fatal latent cancer i

estimates have been subdivided into those attributable to exposures of the thyroid and all other organs.

The majority of latent cancer (including thyroid) fatalities would occur within 80 km (50' miles) of the plant.

Figure 5.4e shows probability distributions of early fatalities.

Two curves are minimal; see Appendix J of this q, g g f medical treatment (supportive and shown representing benefits of two and Appendix F of Appendix VI of 1

l WASH-1400) that would likely be given to individuals receiving excessive doses to the total bone narrow from early exposure.

One curve shows the results con-l sidering the benefit of the supportive medical treatment.

The early fatalities i

with supportive medical treatment are predicted to be essentially all within 32 km (20 miles)~of the site.

The other curve shows the results including the benefit of minimal medical treatment.

The early fatalities with minimal medical l

treatments are predicted to be essentially all within 80 km (50 miles) of the site.

As discussed in Appendix J, because it is conceivable that for very severe but low probability accidents, some of the people requiring supportive medical treatment may not actually receive it, the likely probability distribution of the early fatalities would be between the two curves shown in Figure 5.4e.

Figure 5.4f shows the probability distributions of. early injuries that may result from acute radiation exposure.

The cases of early injuries are predicted to be all within 160 km (100 miles) of the site.

An additional potential pathway for doses resulting from atmospheric release is from fallout onto open bodies of water.

This pathway has been investigated in the NRC analysis'of the Fermi Unit 2 plant, which is located on Lake Erie, and for which appreciable fractions of radionuclides in the plume could be deposited in the Great Lakes (NUREG-0769).

It was found that for the Fermi site, the indicated individual and societal doses from this pathway were smaller than the interdicted doses from other pathways.

Further, the individual and societal liquid pathway doses could be substantially eliminated by the interdiction of L;

the aquatic food pathway in a manner comparable to interdiction of the terres-L trial food pathway i.n the present analysis.

Becane Limerick is not on a large surface water body, the fraction of radioactive material that could fall out in p

nearby rivers, streams,.or lakes would be correspondingly reduced.

The staff L

has also considered fallout onto and runoff and leaching into water bodies in 4

connection with a study of severe accidents at the Indian Point reactors in Limerick FES 5-92 e

1 l.

southeastern New York (Written staff testimony on Commission Question 1, Sec-tion III.D by Richard Codell on Liquid Pathway Considerations for the Indian Point ASLB Special Hearing, June 1982-April 1983).

In this study empirical-

'models were developed based upon considerations of radionuclide data collected in the New York City water supply system as a result of fallout from atmospheric

. weapons tests.

As with.the Fermi study, the Indian Point evaluation indicated that the uninterdicted risks from this pathway were fractions of the interdicted risks from other pathways.

Further, if interdicted in a manner similar to inter-diction assumed for other pathways, the liquid pathway risk from fallout would be a very small fraction of the risks from other pathways.

Considering the LGS and the regional meteoroing and hydrology, the staff sees nothing to indicate that the liquid pathway ct... ibution to the total accident risk would be signifi-i i

cantly greater than found for Fermi 2 and Indian Point.

This water pathway would be of-small importance compared to the results presented here for fallout onto. land.

(4) Economic and Societal Impacts As noted in Section 5.9.4.2, the various measures for avoiding adverse health i

effects, including those resulting from residual radioactive contamination in-the environment, are possible consequential impacts of severe accidents.

Calcu-lations of the probabilities and magnitudes of such impacts for Limerick station t

i and environs also have been made.

(NUREG-0340 describes the model used.) Unlike the radiation exposure and health effect impacts discussed above, impacts asso-ciated with avoiding adverse health effects are more readily transformed into i

economic impacts.

1 The results are shown as the probability distribution for cost of offsite.niti-gating actions in Figure 5.4g and are included in the impact summary Table 5.11g.

]

The factors contributing to these estimated costs include the following:

i l

evacuation costs value of crops contaminated and condemned

{

. value of milk contaminated and condemned costs of decontamination of property where practical indirect costs resulting from the loss of use of property and incomes derived therefrom The last-named costs would derive from the necessity for interdiction to pre-vent the use of property until it is either free of contamination or can be economically decontaminated.

t i

Figure 5.4g shows that at the extreme end of the accident spectrum these costs l

could exceed tens of billions of dollars, but that the probability that this l

would occur is exceedingly small (less than one chance in 10 million per

]

reactor year).

l Additional economic impacts that can be monetized include costs of related health effects, cost of regional industrial impacts, costs of decontamination 4

of the facility itself,,and the costs of replacement power.

Probabi,11ty dis-tributions for these impacts have not been calculated, but they are included j

in the discussion of risk considerations in Section 5.9.4.5(6) below.

i -

Limerick FES 5-93 3,

t 4,

_ _ _,. ~ _ _, _

As an additional impact of environmental contamination, Figure 5.4h shows the probability distribution of severely contaminated land area in square meters (about 2.6 million square meters equals 1 square mile) that would not be returned to use by decontamination, because decontamination procedures would not be very effective.

Such areas would be marked for long-term interdiction (more than 30 years).

At the extreme end of the accident spectrum, Figure 5.4h shows that such areas could be as large as several hundreds of square miles, but the probability that this could occur is extremely small (less than 1 chance in 10 million per reactor year).

This impact is also included in Table 5.11g.

The geographical extent of the kinds of impacts discussed above, as well as many other types of impacts, is a function of several factors.

For example, the dispersion conditions and wind direction following a reactor accident: the type of accident, and the magnitude of the release of radioactive material are all important in determining the gcegraphical extent of such impacts.

Because of these large inherent uncertainties, the values presented herein are mean values of the important types of risk based upon the methodology employed in the accident consequence model (NUREG-0340) and do not indicate specific geo-graphical areas.

(5) Releases to Groundwater A groundwater pathway for radiation exposure to the public and environmental contamination that would be unique for severe reactor accidents was identified in Section 5.9.4.2(2) above.

Consideration has been given to potential environ-mental impacts of this pathway for the Limerick station.

The penetration of the basemat of the containment building can release molten core debris to the strata beneath the plant. The soluble radionuclides in the debris can be leached and transported with groundwater to downgradient domestic wells used for drinking water or the surface water bodies used for drinking water, aquatic food, and recreation.

Peleases of radioactivity to the groundwater underlying the site could also occur via depressurization of the containment atmosphere and releases of radioactive.ECCS and suppression pool water through the failed containment.

An analysis of the potential consequences of a liquid pathway release of radio-activity for generic sites was presented in the " Liquid Pathway Generic Study" (LPGS) (NUREG-0440).

The LPGS compares the risk of accidents involving the liquid pathway (drinking water, irrigation, aquatic food, swimming, and shore-line usage) for four conventional, generic, land-based nuclear plants and for a floating nuclear plant for which the nuclear reactor would be mounted on a barge and moored in a water body.

Parameters for each generic lend-based site were chosen to represent averages for a wide range of real sites and were thus

" typical", but represented no real sites in particular.

The discussion in this section is a summary of an analysis performed to compare the liquid pathway consequences of a postulated accident at the Limerick site with that of the generic small-river land-based site considered in the LPGS.

The comparison is made on the basis of population doses from drinking contaminated water, eating contaminated fish, and ench shoreline uses as recreation.

The parameters that were evaluated include the amounts and rate of release of radioactive materials to the ground, ground water travel time, sorption on geological media, surface water transport, drinking water usage, aquatic food consumption, and recreation area usage.

Limerick FES 5-94

7 Yo _:

}7o _ : :To*

oe ow o-om

_T,o{

E E- :

-:i : ' T

_T T

03

=i_ _

i :

5_: _

1 3

1 a

a a

s f.

jf I.

v 1 a

s a

s a

n e

o s

i e

t i

c t

8 d

n i

1 d

i a 1 r

a

=

s e

t s

t r

s s

n i

e s

cn S

a u

e s

R r

f 8

(\\

d T E

a o

1

I d

n 3

E n

3 s

o a

M a

i e

l s

N a

s E

f q

u s

R o

cs a

A n

i 8

U o

d 6

i 1

Q t

r m

1 s

S u

v o

b f

s i

N r

s

)

a I

t 7

s a

(

i A

d 5

e E

8 dR y

4 1

s 1A t

9 i

a a

l

=

5 a

i X

b n

s a

o e

b a

i o

t r

c a

P f

e S

1 s6

(

1 h

e v

4 e

a S

s 5

s s

e a

r E

u T

a g

O 8

,d F

i N

1 e, I s

e y

s s

a s

d F

i :

=i- - _

i:.

i:

,d I o-=. :. ' o 5_ :

a" oe

~ _

i ow o

oe oe' xA'$r4ADN4ae

>smaHW*gCu1 y

i oD s

s5 r[,.nwyO n

a$

,lll1!ll'

i All of the' reactors considered in the LPGS were Westinghouse pressurized water

~ reactors (PWRs) with ice condenser containments.

There are likely to be signi-ficantly different mechanisms-and probabilities of releases of radioactivity for the Limerick boiling water reactor (BWR).

The staff is not aware of any studies I

which indicate the' probabilities or magnitudes of liquid releases for BWRs.

The source term used for Limerick in this comparison is assumed to be equal to that used in the LPGS.

Doses to individuals and populations were calculated in the LPGS without con-sideration of interdiction methods such as isolating the contaminated ground-water or denying use of the water.

In the event of surface water contamination, alternative sources of water for drinking, irrigation, and industrial uses would be expected to be found, if necessary.

Commercial and sports fishing, as well as many other related activities, could be restricted.

The consequences would, therefore, be largely economic and social rather than radiological.

In any event, the individual and population doses for the liquid pathway range from fractions to very small fractions of those that can arise from the airborne pathways.

]

The Limerick site is about 244 meters (800 feet) from and 33.5 meters (110 feet) above the Schuylkill River.

The aquifer underlying the site is composed of red shale, sandstone, and siltstone. Most of the grouns ater movement in the i

aquifer follows secondary openings that have developed following the deposition I

of the beds.

The most important openings are nearly vertical joint planes; they l

cross each other at various angles throughout the beds.

Where these joints are present, they provide an interconnected series of channels through which ground-water can flow, giving the material a low to moderate permeability.

y h

i The weathered upper bedrock in the power block area has been removed and the small fracture zones in the remaining rock have been filled with concrete.

Should a core melt accident occur at the Limerick site and the leached radio-nuclides find a path through the concrete basemat, the tight bedrock beneath the basemat would tend to confine the effluent and greatly limit its transport g

downgradient.

For the purposes of this analysis, however, the radioactive j

effluent was conservatively assumed to travel immediately through the under-

[

lying rock and move downgradient toward the river.

The applicant performed an analysis of the liquid pathway release following a postulated core melt accident and determined a groundwater travel time of 3.28 years from the reactor building to the Schuylkill River The groundwater travel time calculated for the LPGS generic site was 0.61 years.

1i.

The staff has evaluated the applicant's groundwater travel time calculation and E

the data used to choose the pertinent parameters and considers the applicant's J

analyses to be conservative.

The average bedrock permeability, estimated from

{p site permeability tests, is 65 m (214 feet) per year, and the effective porosity l4 is estimated to be 0.05.

The groundwater gradient likely to exist after plant l

4 construction is estimated to be no greater than 0.025, based on well hydrographs 4

at the site.

From these values, the staff estimates a groundwater travel time of 7.5 years for.the 244 meters to the river.

j It was demonstrated in the LPGS that for holdup times on the order of years, t

virtually all the liquid pathway population dose results from Sr-90 and Cs-137.

i Limerick FES 5-96 H

Therefore, only these two radionuclides are considered in the remainder of this analysis.

The radionuclides Sr-90 and Cs-137 usually move much slower than groundwater because of the effects of sorption (ion-exchange) on the geologic media.

How-ever, most of the measured values of the retardation effects of sorption are applicable only to soil or pulverized rock.

There is only limited data avail-able on retardation in fractured geologic media.

At the Limerick site, however, the fractures in the siltstone and sandstone are partially filled with calcite, sand, and clay.

Hence, part of the flow path would be through porous media, and ion exchange can be expected to retard the movement of radionuclides to the Schuylkill River.

Based on measured retardation related distribution coeffi-cients (Kd) for similar rock types and soil (Isherwood,1981), a Kd of 2 was selected for Sr-90 and a Kd of 20 for Cs-137.

Both Kd values selected are on the low side of representative values and are, therefore, considered to be con-servative.

A total porosity of 25% was selected as representative of the frac-tured and filled media through which the radioactive effluent would travel.

From these values, retardation coefficients of 20 for Sr-90 and 193 for Cs-137 were determined as being reasonably censervative for.the transport media.

The calculated radionuclide travel time is then 150 years for Sr-90 and 1447.5 years for Cs-137.

The radionuclide travel times for Sr-90 and Cs-137 in the LPGS are 5.7 years and 51 years, respectively.

As a result of radioactive decay, the estimated amount of Sr-90 entering the Schuylkill River would be reduced to about 3% of the amount determined in the LPGS.

The amount of Cs-137 would be about 14 orders of magnitude less than that in the LPGS, and its contribution to population dose via the various pathways (drinking water, fish consumption, and racreation activities) need not be considered further.

The primary pathway for Sr-90 to humans is through drinking water.

Comparison of drinking water population doses will be based upon the ratio of population served to river flow, which takes into account the effects of dilution.

Down-stream of the Limerick site, there are approximately 1.9 million people using the Schuylkill River as a drinking water supply.

The average flow in Schuylkill River.is about 1900 ft /sec resulting in a population to flow ratio of 1000 3

3 people /ft /sec.

The corresponding ratio in the LPGS for a small river site is about 32 people /ft3sec.

Hence, for a similar release to a river, the total drinking water dose at Limerick without a change i.n drinking water supply, would be about 30 times worse.

However, since the concentration of Sr-90 entering the water would be only 3% of that of the LPGS, the total drinking water dose is roughly equivalent to that determined in the LPGS.

The staff concludes that population dose as a result of the liquid pathway contribution at the Limerick site would be about the same as that from the generic site.

The staff recognizes that, because of the differences in design of the Limerick reactor as compared to the reactor design analyzed in the LPGS, a different inventory of radionuclides could be released following a core melt accident and postulated breach of the basemat.

This uncertainty, along with uncertainties j

in the amount of radionuclides that could be released, could result in a dif-ferent dose comparison than the one presented.

However, the staff also con-siders the potential for.a release through the basemat at the Limerick site i

following a core melt accident to be i,ignificantly less than that for the design considered in the LPGS.

Therefore, the total risk from the liquid pathway is still estimated to be less than or about the same order as that in the LPGS.

j Limerick FES 5-97

F 1

In conclusion, Limerick should be considered about equal in regard to risk from the liquid pathway (groundwater) in comparison to other land-based sites.

In addition, the long groundwater travel time ensures that mitigation measures such as slurry walls, grouting, dewatering, and other measures can be completed in time to protect downstream drinking water and fisheries.

A comprehensive dis-cussion of accident mitigation measures has been presented by V. A. Harris (Harris,1982).

(6) Risk Considerations The foregoing discussions have dealt with both the frequency (or likelihood of occurrence) of accidents and their impacts (or consequences).

Because the ranges of both factors are quite broad and uncertain (see (7) below), it also is useful to combine them to obtain average measurce of environmental risks.

Such averages can be particularly instructive as an aid to the comparison of radiological risks associated with accident releases with risks associated with normal operational releases and with other forms of risks.

A common way in which this combination of factors is used to estimate risk is to multiply probabilities by the consequences. The resultant risk is then expressed as a measure of consequences per unit of time.

Such a cuantification of risk does not mean that there is universal agreement that peoples' attitudes about risks, or what constitutes an acceptable risk, can or should be governed solely by such a measure.

However, it can be a contributing factor to a risk judgment, although not necessarily a decisive factor.

Table 5.11h shows average values of societal risk estimates associated with population dose, early fatalities with two types of medical treatment (minimal and supportive), early injuries, latent cancer fatalities, costs for evacuation and other protective actions, and land area for long-term interdiction.

These average values are obtained by summing the probabilities multiplied by the con-sequences over the entire range of the distributions. Because the probabilities are on a per-reactor year basis, the averages shown also are on a per-reactor-year. basis.

Incremental risks per reactor year of early fatality (with two types of medical treatment) and latent cancer fatality associated with spatial intervals up to 50 miles (80 km) from the Limerick reactors are shown in Appendix L.

The population exposures and latent cancer fatality risks may be compared with those from normal operation shown in Appendix 0 and Section 5.9.3.2 of this statement.

The comparison (excluding exposure to station personnel) shows that the accident risks are up to 30 times higher.

For a different perspective, the latent cancer (including thyroid) fatality risks of 3 x 10 4 persons per reactor-year within 1 mile (1.6 km) of the site exclusion area boundary (EAB) (based on data in Table L.4 in Appendix L) and 5 x 10 2 persons per reactor year within the 50-mile (80-km) region (from Table 5.11h) may be compared with such risks from causes other than reactor accidents.

Approximately 3000 persons are pro-jected to live within 1 mile (1.6 km) from the EAB and 7 million persons are

_[

projected to ifve within,the 50-mile (80-km) region in the year 2000.

The back-ground cancer mortality rate is 1.9 x 10 3 cancer fatality per persori per year Limerick FES 5-98 u

F:

f

.f*

Table 5.11h Est.imated values of societal risks from severe accidents, per reactor year Estimated risk within Estimated risk within Consequence type ~

the 50-mile-region the entire region 1.

Early fatalities with 5(-3)*

5(-3)

Supportive medical treatment.(persons) 2.

Early fatalities with 8(-3) 8(-3) minimal medical treat-ment (persons) 3.

Early injuries (persons) 2(-2) 2(-2)

-4. ' Latent cancer fatalities ~ 4(-2) 7(-2)

(excluding thyroid)

(persons) 5.

Latent thyroid cancer 1(-2) 1(-2) fatalities (persons) 6.

Total person-rems

.7(2)

(3) 7a.

Cost of-offsite mitiga-5(4) 5(4) tion measures (1980 $)

7b.

Regional industrial 5(4)***

impact costs (1980 $)

1 7c.

Plant costs (1980 $)

1(5) 8.

Land area for long-term 1(3) 1(3) interdicti,on (m )=*

2

• 5(-3) = 5 x 10 3 =.005

• • About 2.6.million m2 equals to 1 mi2,

• • • Excludes costs of crop and milk interdiction, which are' included in 7a.

NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

Estimated numbers were rounded to one significant digit only for the purpose of this table, in the U.S (American Cancer Society, 1981).

Therefore, at this rate, about-6 background cancer' fatalities per year are expected in the population within 1 mile (1.6 km) of the EAB, and 10,000 background cancer fatalities in the population within the 50-mile ~(80-km) region in the year 2000.

Thus, the risk of cancer fatality from reactor accidents at Limerick is small compared to the risk of normal occurrence of such fatality.

The ratio of latent cancer fatality risk from reactor accidents at Limerick to l

the population living within 50 miles of the plant in the year 2000 to the can-car fatality risk in the same population from all other causes is 5 x 10.s

[

(5 x 10 2/10,000) on a per reactor-unit basis.

Limerick FES 5-99

0 i

There are no early fatality, early, injury, long-term land interdiction, or economic risks associated with protective actions and decontamination for normal For releases; but these risks can be associated with large accidental releases.

perspective and understanding of the meaning of the early fatality risk of 5 x 10 s persons per reactor-year with supportive medical treatment and 8 x 10 s persons per reactor year with minimal medical treatment (from Table 5.11h), the staff notes that occurrences of early fatalities with supportive and minimal medical treatments would be contained, approximately, within the 20-mile (32-km) and 50-mile (80-km) regions, respectively.

The number of persons projected to live within these regions in the year 2000 are 0.8 million and 7 million, respectively.

The background risk for the average individual in the U.S. is 5 x 10 4 accidental death per year (NUREG/CR-1916).

Therefore, the expected number of non-Limerick accidental fatalities per year within the 20-mile (32-km) and 50-mile (80-km) regions are 400 and 4000, respectively, in the year 2000.

Thus, the risk of early fatality with supportive or minimal medical treatment from reactor accidents at Limerick is extremely small compared with that from non-Limerick accidents.

For an added perspective, the risk of early fatality within 1 mile (1.6 km) of the exclusion area boundary (EA8) from reactor acci-2 dents may be compared with early fatality risk from nonnuclear accidents in the

]

From Tables L.2 and L.3 in Appendix L, the Limerick risks of early same region.

L fatality with supportive or minimal medical treatments are 5 x 10 4 persons per c

reactor year and 6 x 10 4 persons per reactor-year, respectively, in this region.

At the average rate of 5 x 10 4 -nonnuclear accidental death per individual per year in the U.S., the number of nonnuclear accidental fatalities in the popula-

[

tion of 3000 projected to live within 1 mile (1.6 km) from the EA8 in the year y'

2000.would be 2 per year.

This also shows that the early fatality risk from reactor accidents at Limerick is expected to be small compared with risk of non-y U

nuclear accidental deaths.

3 The ratio of (1) risk of early fatality with minimal medical treatment from y

reacter accidents at Limerick to an average individual living within a mile of E

ll the site exclu.sion area boundary to (2) the risk to the same individual of acci-U dental death from all other causes, is 3 x 10 4 (6 x 10 4/3000 + 2/3000) on a y

per ieactor-unit basis, f

To provide a reasonable bound to the role of evacuation in risk estimates from 0

the release categories not initiated by severe earthquakes, as.well as to as-sess the sensitivity of risks from these release categories with respect to un-certainties in executing an evacuation, an analysis of these release categories was made by assuming the Early Reloc mode of offsite emergency response (see

, b..

Table 5.11f).

Results of the analysis are provided in Appendix M.

These results, when combined with those previously calculated for the release cate-gories initiated by severe eathquakes, show only slight increases in the risks of latert cancer and early fatalities and also corroborate the preceding con-y Ij clusions that these risks from Limerick reactor accidents are small compared with the background risks from nonnuclear causes.

Figure 5.41 shows the calculated risk of whole-body dose to an individual from The early exposure as a function of the downwind distance from the plant.

values are on a per-reactor year basis and~all release categories contributed l :!

to the dose, weighted by their associated probabilities.

For purpoies of com-i i

p!'{

parison the risk of receiving a whole body dose of 99 mrems per year from 1

I

'j Limerick FES 5-100 b

-. r-

~._g=**

.1 n

8o x

=

n m

N 5

'c.

s w

T 3

i 8

13 N

5

.7

\\

ac O.

h "i

1 z

w

_ft c

..n m

e-8 80.

us 0

!s

~

~

o oas w

E' so

'b.

.E

~3

~:

69 ~

'O e.e s'.e d.o d.o 20.o 2s.o so.s 35.e so.o ts.e 98.#

DISTANCC INILCl Figure 5.41 Risk of downwind individual dose versus distance NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties

7 P

's l

d natural background is a virtual certainty for any individual living in the M} l Limerick site region (see Table D.7 in Appendix D).

M' Figures 5.4j, 5.4k, and 5.41, respectively, display risk to an individual of Y

early _ fatality, early injury, and latent cancer fatality, all from early expo-d

-sure, as functions of distance from the Limerick reactors and on a per-reactor-q year basis. The curves in these figures were generated without regard to the y

differences in the likelihood of wind blowing in different directions (the staff H

used 16 direction sectors of the compass). To'obtain risk curves for a specific i

direction (1 out of the 16), all values on the curves along the vertical axis must be multiplied by 16P, where P is the annual average probability of the wind i

blowing toward the direction of interest.

The values of P for the Limerick site derived from 1976 meteorological data are shown in Table 5.11e.

For comparison to early fatality risk to an individual from Limerick reactor accidents, the following nonnuclear risks, per year, of accidental fatality to an individual living in the United States may be noted (National Research Council, 1979,

p. 577):

automobile accident 2.2 x 10 4, falls 7.7 x 10 5, drowning 3.1 x 10 5, F

burning 2.9 x 10 5, and firearms 1.2 x 10.s.

For comparison to the estimated latent cancer fatality risk to an individual from the Limerick reactor accidents, i

it should be noted that the risk of cancer fatality to an-individual in the U.S.

e from nonnuclear causes is 1.9 x 10 3 per year (American Cancer Society, 1981).

The economic risk associated with evacuation and other protective actions could be compared with property damage costs associated with alternative energy gene-i c,

ration technologies. The use of fossil fuels, coal, or oil, for example, would emit substantial quantities of sulfur dioxide and nitrogen oxides into the at-mosphere and, among other things, lead to environmental and ecological damage through the phenomenon of acid rain (National Research Council, 1979, pp. 559-560).

In the judgment of the staff, this effect has not been sufficiently quantified to draw a useful comparison at.this time.

The staff has also considered the health care costs-resulting from hypothetical accidents in'a generic model developed by the Pacific Northwest Laboratory (Nieves, 1982).

Based upon this generic model, the staff concludes that such costs may be a fraction af the offsite costs evaluated herein, but that the model is not sufficiently constituted for application to a specific reactor site.

A severe accident that requires the interdiction and/or decontamination of land n';

areas is likely to force numerous businesses to temporarily or permanently close.

These closures would have additional economic effects beyond the contaminated areas through the disruption of regional markets and sources of supplies.

Esti-mates of these risks were made using:

(1) the RSS consequence model (Appen-t b

dix VI, WASH-1400) and (2) the regional input-output modeling system (RIMS II),

l; developed by the Bureau of Economic Analysis (BEA).

lLlI' The industrial impact model developed by BEA is based on contamination levels of l,'

a physically affected area defined by the RSS consequence model.

Contamination

(

levels define an interdicted area immediately surrounding the plant, followed by fj' an area of decontamination, an area of crop interdiction, and finally an area of milk interdiction.

I*

h WF M

Limerick FES 5-102 e

$qa,

1 Q.

'o se ;

5 5

y y

"g y

y D,

i LEGEM i

u o -W/ MINIMAL MEDICAL TfiEATMENT a -W/ SUPPORTIVE MEDICAL TREATMENT

~

t'-'o _

i "3

l be' j

'o_

-i d

e 12

~

5 m

et

'o.

Y if,

~i i

g g

k9

~

o_

E

-i 9

'o s.e s'.e 10.8 15.e ao.e.

2s.e so.e 35.e m.e 45.s sea.e

~-

DISTf1NCC INILCl

,1 4

Figure 5.4j Individual risk of early fatality versus distance NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties 1

i

 ;':::._ ;K, _ _

TQ': a:7,7--'

s-_,,_

4

. _ __ :=. z'w._:-.

.-~

l C

S-

-s,,

nw m'o 8

3 3

5 5

3 s

y y

n I.

'o_

5 1:'

u*-

o.

-g g

-g e-sC w

E

e. o..

u, u

.m

]

g

".[

O saf

+

.a a

it E;

e o

• o.

?

i A

g -

so o.s s.e is.e is.e an.o as.s m.a n.e f,e so,.

DISTf1NCC li1lLCl Figure 5.4k Individual risk of early injury versus distance NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties

l I

C se 2.

n X

Yo.

mn LEGEND n -EXCLUDING TilVROID o - TilVROID

~

e*

S o.

~

e*

l b

'O-i.

g g

i g

g

~

a.

ui w

~:

a o.'

4 8

7'o, E

ai

~

~

[

g To

~

y y

e.e s.e so.e es.s ao.o 25.o 3n.e 3s.e e.e e.e ss.e DISTilNCE IMILCl Figure 5.41 Individual risk of latent cancer fatality versus distance NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties

Specific assumptions used in the analysis are (1) ~In the interdicted area, all industries would lose total production for more than a year.

(2) -In the decontamination zone, there would be a 3-month loss in nonagri-cultural output; a 1 year loss in all crop output (except there would be no loss in greenhouse, nursery, and forestry output); a 3-month loss in dairy output; and a 6-month loss in livestock and poultry output.

In the crop interdiction area, there would be no loss in nonagricultural (3) output; a one year loss in agricultural output (except there would be no loss in greenhouse, nursery, and forestry output); no loss in livestock and poultry output; and a 2-month loss of dairy output.

(4)_ In the milk interdiction zone, there would be only a 2-month loss in dairy output.

The estimates of industrial impacts are made for an economic study area that An consists of a physically affected area and a physically unaffected area.

accident that causes an adverse impact in the physically affected area (for example, the loss of agricultural output) could also adversely affect output in In addition to the physically unaffected area (for example, food processing).t impacts could occur in the physically unaffected area:

decreased demand (in the physically affected area) for output produced in (1) the physically unaffected area decreased availability of production inputs purchased from the physically (2) affected area Only the impacts occuring during the first year following an accident are con-The longer term consequences are not considered because they will vary widely depending on the level and nature of efforts to mitigate the acci-sidered.

dent consequences and to decontaminate the physically affected ater.s.

The estimates assume no compensating effects, such as the use of unused capacity in the physically unaffected area to offset the initial lost production in the income payments to individuals displaced from their physically affected area on These compen-jobs that would enable them to maintain their spending habits.

Realistically, these com-sating effects would reduce the industrial impacts.

The estimates using no pensating effects would occur over a lengthy period. compensating i-The output loss risk can be estimated by mutiplying the probabilities of the j

release categories representative of those in Table 5.11c by the probability of l

}

The the wind blowing in various directions and the associated consequences.

overall risk associated with these release categories was then estimated as the

(

The estimated overall risk values using output sum of the individual products.

losses as the measure of accident consequences, expressed in a per r'eactor year This includes $2000 as the 1,

basis, is $50,000 (1980 dollars) per reactor year.

]'F cost of crop and milk interdictions calculated in CRAC runs for consequence d

i 5-106 Limerick FES it 6

analysis. 'The corresponding expected employment loss is between two and three jobs per reactor year.

Half of the total risk per reactor year is accounted for by the cases of wind blowing toward the east-southeast.

The risk is least severe with the wind blowing toward the east-southwest.

Because of the economic mix of the entire region, the composition of impacts consists of 85% nonagri-cultural impacts, 4% agricultural impacts, and 11% indirect impacts of decreased exports and supply constraints.

There are other economic impacts and risks that can be monetized but that are not included in the cost calculations discussed earlier.

These are accident impacts on the facility itself that res, ult in added costs to the public-(rate-payers, taxpayers, and/or shareholders).

These costs would be for decontami-nation and repair or replacement of the facility, and replacement power.

Ex-perience with such costs is currently being accumulated as a result of the Three Mile Island accident.

If an accident occurs during the first full year of Limerick Unit 1 operation (1985), the economic penalty associated with the ini-tial year of the unit's operation is estimated at $1500 million for decontami-nation and restoration, including replacement of the damaged nuclear fuel.

This is based on a conservative (high) 10% escalation of the $950 million cost in 1980 dollars estimated for Three Mile Island (EMD-81-106).

Although insurance would cover $300 million or more of the $1500 million, the insurance is not credited against the $1500 million because the $300 million times the risk prob-ability should theoretically balance the insurance premium.

In addition, staff estimates additional fuel costs of $50 million (1985 dollars) for replacement power during each year Limerick Unit I was being restored.

This estimate as-sumes conservatively (high cost) that two-thirds of the energy that would have been forthcoming from the unit (assuming 55% capacity factor) would be replaced by coal-fired generation and one-third by oil-fired generation.

Assuming the nuclear unit does not operate for 8 years, the total additional replacement power costs would be approximately $400 million in 1985 dollars.

The probability of a core melt or severe reactor damage is assumed to be as high as 10 4 per reactor year (this accident probability is intended to account for.all severe core damage accidents leading to large economic consequences for the owner, not just those leading to significant offsite consequences).

Multiplying the previously estimated costs of $1900 million for an accident to Limerick Unit 1 during the initial year of its cperation by the above 10 4 prob-ability results in an economic risk of approximately $190,000 (in 1985 dollars l

or $120,000 in 1980 dollars) applicable to Limerick Unit 1 during its first year j

of operation.

This is also aproximately the economic risk (in 1985 dollars) to Limerick Unit 1 during the second and each subsequent year of its operation, i

l Although nuclear units depreciate in value and may operate at reduced capacity factors so that the economic consequences of an accident become less as the units become older, this is conservatively (high cost) considered to be offset by a slightly higher escalation rate than discount rate.

The economic risk to Limerick Unit 2 (in 1985 dollars) is also approximately

$190,000 (or $120,000 in 1980 dollars) during the first year and each subse-quent year of operation because of the balancing effect of escalation and the i

presen'.-worth discount factor.

Limerick FES 5-107

t (7) Uncertainties i

_The probabilistic risk assessment discussed above has been based mostly on the methodology in the RSS, which was published in 1975 (NUREG-75/014).

Although substantial improvements have bien made in various facets of the RSS methodology since this publication was issued, there are still large uncertainties in the results of the analysis presented above because of the uncertainties associated with the likelihoods of the accident sequences and containment failure modes leading to the release categories, the source terms for the release categories, and the estimates of environmental consequences.

Relatively more important contributors to uncertainties in the results ' presented in this supplement are as follows:

Probability of Occurrence of Accident If the probability of a release category were to be changed by a certain factor, the probabilities of various types of consequences from that re-lease category would also change exactly by the same factor.

Thus, an order of magnitude uncertainty in the probability of a release category would result in an order of magnitude uncertainty in both societal and individual risks stemming from the release category.

As in the RSS, there are substantial uncertainties in the probabilities of the release categories.

This is due, in part, to difficulties associated with the quantification of human error and to inadequacies in (1) the data base on failure rates of individual plant components, and (2) the data base on external events and their effects on plant systems and components that are used to calculate the probabilities.

Severa earthquakes are one cause of accidents.

Uncertainties in the esti-mates of probabilities of severe earthquake induced core melt sequences are judged to be very large because of (1) the relatively sparse data base on severe earthquakes in the eastern U.'. and-(2) the unavailability of a

an acceptably precise and definite procedure to quantify seismically

- induced accident sequences.

In LGS-SARA, the spectrum of probabilities of seismically induced core melt sequences varied over a wide range (several orders) of magnitudes.

However, the mean'(point or best estimate) proba-bilities of seismically induced core melt accident sequences used in the staff analysis (which essentially came from LGS-SARA) are within the range of probabilities developed 4n LGS-SARA, and are within a factor of about 6 of the upper end of the spectrum of probabilities in LGS-SARA.

Thus, the point estimates of seismic probabilities used to evaluate risks are more representative of Limerick than WASH-1400 values, and consider the applicant's estimate of the range of seismic frequency uncertainty.

The staff has concluded that the high and low values of the range should not l

be characterized as 95% and 5% limits, but rather as a representative range P

of the seismic sequence frequencies, which incorporates a large part (but not necessarily all) of the uncertainties with such events.

This statement reflects the staff's view that the rigorous definition of seismic hazard and 1'

its uncertainty at low probabilities is beyond the state-of-the-art at this T

time and should be recognized as such.

Different studies would not neces-sarily yield equivalent results.

For example, an interium report to be pub-

[

lished " Seismic Hazard Characterization of the Eastern U.S." of an ongoing i

^]

!:[

Limerick FES 5-108 u

E

~ - - +

-,, - _ ~-

l study being carried out by Lawrence Livermore National Laboratory (LLNL) for the NRC shows seismic. hazard calculations for the Limerick site which overlap, ~ but are not necessarily coincident with, the range of seismic hazard assumed.in LGS-SARA.

The median (50%) hazard calculated in the. interim LLNL report is within, but near the high.end of, the_ range of hazard curves utilized in LGS-SARA.

Additional studies of seismic hazard in the eastern U.S. are being carried out by such groups as the Electric Power Research Institute.

Given the

~

highly judgmental nature of seismic hazard calculations, there is not reason to believe that these studies or the final;LLNL report would not show differences in estimated seismic hazard and uncertainty between them-selves and the LGS-SARA, particularly at the low probabilities being calculated for Limerick.-

The staff believes that only the use of a full range of seismic probabilities in' risk analysis-would be appropriate.

_ However, to keep the risk analysis manageable, the staff has used the point estimates of probabilities of seismically induced release categories in the risk analysis, and has provided below a discussion of uncertainty in the risk estimates arising from the use of point estimates of probabilities.

Inspection of the results shown in Tables L-la and b and M-la and b indi-cates that with the use of the mean values of probabilities of the severe earthquake initiated release categories, these release categories contri-bute:

(1) dominantly (about 4 to 30 times higher) to the risks of early fatality; (2) about equally to the risk of early injury; and (3) much less to the other types of risks--all compared to the contributions.from the release categories initiated by causes other than severe earthquakes.

If, j.

instead of using tne mean probabilities, the staff had used the values of probabilities of earthquake-initiated release categories from the high estimates, theni (1) the total risks of early fatality would be increased by a factor of about 6 (because the high estima~ ties of probabilities of the earthquake-initiated release categories are about 6 times higher than the lnean values); (2) the total risk of early injury would be increased by a factor of about 4; and (3) the other types of risks would be increased by factors of about 2.

On the other hand, if th,e staff had used the low estimates of probabilities of the earthquake-initiated release categories (which are lower than the mean values by several orders of magnitudes),

1 then the contributions to the risks from these release categories would be negligible compared to.those from the release categories initiated by causes other than severe earthquakes.

Therefore, use of the full = range of probabilities of earthquake-initiated release' categories would result in spreads in the staff's risk estimates; values of the risks would fall within ranges of about one-thirtieth to about 6 times the values depicted-in Tables 5-11h, L-la and b, and M-la and b.

We do not mean to imply that higher risk estimates are more appropriate than the median, mean or lower estimates.

Indeed the most significant earthquake damage anywhere within the vicinity of the Limerick Site, in the two to three-hundred years during which we have records, are fallen chimneys 50 kilometers away.

i during an earthquake:at Wilmington, Delaware in 1871 whose magnitude can l

be estimated to have been less than 5.0.

We certainly cannot exclude from the range of reasonable assumptions the judgment that'there essentially is Limerick FES 5-109

- ~ -.. --.

f ls.

L no risk to the public resulting from earthquake-induced damage at the seismically engineered nuclear power plant at Limerick during its operating j j life.

i h[

Overall,. accident prob' abilities may be expressed in terms of the probabil-ity of core melt, and considered an important measure of the likelihood of 4

4 environmental and human impacts from severe reactor accidents.

To provide i

gd some perspective on the uncertainty in such estimates, Figure 5.4m compares the estimate of core melt probabilities and their uncertainties based on f

contemporary PRA-based estimates for several different reactors.

Except a

for Limerick, the results presented on Figure 5.4m are taken directly from published PRAs without modification (Rowsome and Blond, 1982).

The results for Limerick.are based on staff contractor estimates for Limerick (NUREG-3028).

The PRAs were not necessarily performed using consistent

^'

methodologies or assumptions, and some of the PRAs evaluate designs that have subsequently been altered.

Caution should be exercised when using these results because there are very large uncertainties in these analyses.

n No attempt has been made to adjust the results to compensate for inconsis-tency of approach or methods.

Therefore, the appropriateness of the com-

]

parison may be in question.

However, all of the studies have analyzed, in roughly the same manner, the so-called " internally" initiated events.

e 1

Quantity and Chemical Form of Radioactivity Released r

The models used in'these calculations contain approximations to describe the physical behavior of the radionuclides which affects the transport within 9

the reactor vessel and other plant structures and the amounts of release.

This relates to the quantity and chemical form of each radionuclide species

+

j that would be released from a reactor. unit during a particular accident sequence.

Such releases would originate in the fuel and would be attenu-ated by physical and chemical processes in route to being released to the environment.

Depending on the accident sequence, attenuation in the reactor vessel, the pri.aary cooling system, th~e containment, and adjacent buildin,gs would influence both the magnitude and chemical form of radio-7

- active releases.

The releases of radionuclides to the environent, called i

source terms, used in the staff analysis were determined using the RSS methodology applicable to a BWR of Peach Bottom design; therefore, the RSS methodology may not have been fully appropriate for the Limeri::k BWRs.

Information available in NUREG-0772 and from the latest research activi-L ties sponsored by the Commission and the industry indicates that source b

terms used in the staff analysis cannot be much higher in the maximum, but could be substantially lower.

Some lower source term values could be higher also, primarily because of the manner in which the source term was evaluated for early releases using the.RSS methodology.

The impact of lesser values of source term, would be substantially lower estimates of health effects, particularly early fatalities and injuries.

The source terms resulting from the applicants PRA would, for example, yield significantly lower L

estimates of risk than those used by the staff in this report.

The NRC S

staff anticipates better source term information at the end of 1984 when U,

tht: staff's Accident Source Term Program Office and the American Physical

' y;

'ociety complete their studies.

m ud W

OW ip

-Limerick FES 5-110

. hj-n

1-02 o

internal events only internal plus external events -

l

~~

4 1-03 a

m mo>-

O<w 2

1-04 T

=

w a

p t

a

~

m<m oa L

1-06 7

l Y

1-06 n

n E

N n

E E

E E

3 3

I 3

o a

A E

a j

E g

e e

c e

z z

g.,

D o

e 3

3 e

3 3

o e

e 3

E 3

3 z

.!P 3

]

]

3 2

5 a.

3 n

en a a.

n Figure 5.4m Core melt probability uncertainty bounds for internal i

events and internal plus external events Limerick FES 5-111

.(

+ 4,.

Atmospheric Dispersion Modeling' for the Radioactive Plume Transport, m%

Including the Physical and Chemical Behavior of Radionuclides in Particu-f late Form in the Atmosphere ih

[.

This uncertainty is due to differences between the modeling of the atmo-t y

spheric transport of radioactivity in gaseous and particulate. states in the 4[

CRAC code and the actual transport, diffusion and deposition or fallout W[j i -

that would occur during an accident (including the effects of precipitation).

The phenomenon of plume rise because of heat that is associated with the.

H, atmospheric release, effects of precipitation on the plume, and fallout of 4

particulate matter from the plume all have considerable impact on both the dj magnitude of early health consequences and the distance from the reactor to which these consequences would occur.

The staff judgment is that these O

factors can result in substantial overestimates or underestimates of both H

early and later effects-(health and economic).

I:

Errors of Completeness, Modeling, Arithmetic, and Omission This area of lumped uncertainty includes such topics as the omission of a model of sabotage, modeling errors in event trees, common cause failures other than those originating in external events or fires, improvements in 9;

design or operating criteria undertaken or to be undertaken by the appli-cant, potential errors in the different models used to assess risks, it statistical errors, and arithmetic errors.

The impact on risk estimates Q

of this class of uncertainty could be large, but is unknown and virtually j

impossible to quantify accurately (Rowsome,1982).

Because of the depth to which the applicant and the staff have considered risks for Limerick, however, uncertainties of this type are not expected to be as large as for other reactors for which less comprehensive probabilistic risk assessments

}

have been performed.

i E

Other areas that.have substantial but relatively less effect on uncertainty U.

than the preceeding items are

~

u

- Duration and Energy of Release, Warning Time, and Inplant Radionuclide Decay Time The assumed release duration, energy of release, and the warning and the inplant radioactivity decay times may differ from those that would actually occur during a real accident.

t For a relatively long duration (greater than a half-hour) of.an atmospheric release, the actual cross-wind spread (the width) of the radioactive plume that would develop would likely be larger than the width calculated by the t'

dispersion ~model in CRAC.

However, the effective width of the plume is calculated in the code using a plume expansion factor that is determined M,

by the release duration.

For a given quantity of radionuclides in a re-

[]4i lease, the plume and, therefore, the area that would come under its cover Q

would become wider if the release duration were made longer.

In effect,

y this would result in lower air and ground concentration's of radioactivity

2 but a greater area of contamination.

fO w

9

$$U Limerick FES 5-112

(

.Q c

.,=

c--__.

n The thermal' energy associated with the release affects the plume rise phenomenon, which results in relatively lower air and ground concentrations

.in the closer-in regions and relatively higher concentrations as a' result of fallout in the more distant regions.

Therefore, if a large amount of thermal energy were associated with a release containing large fractions of core -inventory of radionuclides, the distance from the reactor over which

{

early health effects may occur is-likely to be increased.

Warning time before evacuation has considerable impact on the effective-ness of offsite emergency response.

Longer warning times would improve the effectiveness of the response.

The time from reactor shutdown until the beginning of the release to the environment (atmosphere), known as the time of release, is used to calcu-late the depletion.of radionuclides by radioactive decay within the plant before release.

The depletion factor for each radionuclide (determined by

~ the radioactive decay constant and the time of release) multiplied by the release fraction of the radionuclide and its core inventory. determines the l

actual quantity of the radionuclide released.to the environment.

Longer l

release times would result in release of fewer curies to the environment l

for given values of release fractions.

i i

The first three of the parameters discussed above can have significant impacts on accident consequences, particularly early consequences.

The staff judgment is that the estimates of early consequences and risks could be substantially exceeded, or could be substantial overestimates, because of uncertainties in the first three parameters.

Meteorological Sampling Scheme Used The meteorological sequences used with the selected 91 start times (sampling) in the CRAC code may not adequately-represent all meteorological variation's that may occur over the life of the plant.

This factor is

-judged to produce greater uncertainties for early effects and less for latent effects.

Emergency Response Effectiveness The modeling assumptions of the emergency response of the people residing around the Limerick site may not correspond to what would happen during an actual' severe reactor accident.

Included in these considerations are such subjects as evacuation effectiveness under different ci.rcumstances, possi-ble sheltering and its effectiveness, and the effectiveness of population relocation. The staff judgment is that the uncertainties associated with emergency response effectiveness could cause large-uncertainties in esti-mates of early health consequences.

The uncertainties in estimates of latent health consequences and' costs are considered smaller than those of early health consequences.

A limited sensitivity analysis in this area is presented in Appendix M.

It indicates that-for release categories initi-ated by causes other than severe earthquakes, the risk of early, fatality with supportive'or minimal medical treatment would be increased by factors

.of less than 5, if people from within the plume exposure pathway EPZ would

-not evacuate to evade the plume but would wait for the plume to leave the

. area and then relocate from the contaminated ground after a time interval Limerick FES 5-113

D-t.

equal to the evacuation time assumed for the Limerick site.

Under the same assumptions, increases in risks of other health effects would be less.

However, the increase in risks of all health effects from release cate-gories initiated by all causes (severe earthquakes and other causes) taken together would be within about 20%.

Dose Conversion Factors and Dose Response Relationships for Early Health Consequences, Including Benefits of Medical Treatment There are many uncertainties associated with estimates of dose and early health effects on individuals exposed to high levels of radiation.

Included are the uncertainties associated with the conversion of contamination levels to doses, relationships of doses to health effects, and considerations of the availability of what was described in the RSS as supportive medical treatment (a specialized medical treatment program of limited reso0rces that would minimize the early health effect consequences of high levels of

- - radiation exposure following a severe reactor accident).

The staff analysis shows that the variation in estimates of early fatality risks stemming from considerations of supportive medical treatment alone is less than a factor of 3 for the Limerick site.

Dose Conversion Factors and Dose Response Relationships for Latent Health Consequences In comparison to early health effects, there are even larger uncertainties associated with dose estimates and latent (delayed and long-term) health effects on individuals exposed to lower levels of radiation and on their succeeding generations.

Included are the uncertainties associated with conversion of contamination levels to doses and doses to health effects.

The staff judgment is that this category has a large uncertainty.

The un-certainty could result in relatively small underestimates of consequences, but it also could result in substantial overes.timates of consequences.

(Note:

radiobiological evidence on this subject does not rule out the

. possibility that low level radiation could produce zero consequences.)

Chronic Exposure Pathways, Including Environmental Decontamination and the Fate of Deposited Radionuclides Uncertainties are associated with chronic exposure pathways to people from long-term use of the contaminated environment.

Uncertainty also arises from the possibility that the protective action guide levels that may actually be used for interdiction or decontamination of the exposure path-ways may differ from those assumed in the staff analysis.

Further, uncer-tainty arises as a result of the lack of precise knowledge about the fate of the radionuclides in the environment as influenced by such natural pro-cesses as runoff, weathering, etc.

The staff's qualitative judgment is 1

that the uncertainty from these considerations is substantial.

4 Economic Data and Modeling

{

There are uncerta'inties in the economic parameters and economrc modeling, l1 such as costs of evacuation, relocation, medical treatment, cost of decon-l1 tamination of properties, and other costs of property damage.

Uncertainty in this area could be substantial.

I l

Limerick FES 5-114 u la,E,,6N w...

4 t-Fission Product Inventory The fission product inventory presented in Table 5.11a is an approximation of that which would be present after extended operation at maximum power.

'The amount of each isotope listed will, in fact, vary with time in a manner dependent upon_the fuel management scheme and the power history of the_ core.

The actual inventory at the time of an accident could r.ot be much larger for any isotope than the amount in Table 5.11a, but, especially for long-lived fission products, could be substantially smaller.

The means for quantitative evaluation of the uncertainties in a probabilistic risk analysis such as the type presented here are not well developed.

T_he.

staff, however, has attempted to identify all sources of uncertainty, and to e.ssess the net effect upon the uncertainty of the risk estimates.

Based upon the insight gained from the review of similar PRAs for Indian Point and Zion, it.is the judgment of the staff that the risk estimates for Limerick could be too low by _a factor of about 40. or too high by a factor of about 400.

The risk estimates are equal to the= integrals of the corresponding probability distribu-tions of the consequences (CC0Fs).

As a result, errors in probabilities and consequences are partially offset.

Because of the magnitude of uncertainties, the staff has concluded that estimates of-the absolute magnitudes of probabili-i ties, consequences, and -risks do not provide an accident perspective unless the uncertainties are also considered.

When the accident at Three Mile Island occurred in March 1979, the accumulated experience record was about 400 reactor years.

It is of interest to note that i

this was within the range of frequencies estimated by the RSS for an accident of this severi_ty (National Research Council,1979, p. 553).

It should also be noted that the Three Mile Island accident has resulted in a very comprehensive evaluation of similar reactor accidents by a number of investigative groups both within and outside of the NRC.

Actions to improve the safety of nuclear power plants have resulted from these investigations, including those from the Presi' dent's Commission on the Accident at Three Mile Island and from NRC staff investigations and task forces.

A comprehensive "NRC Action Plan Developed as 4

l a Result of the TMI-2 Accident" (NUREG-0660, Vol I) collects the various recom-mendations of these groups and describes them under the subject areas of:

i Operational Safety; Siting and Design; Emergency Preparedness and Radiation Effects; Practices and Procedures; and NRC Policy, Organization, and Management.

NUREG-0737, " Clarification of TMI Action Plan Requirements," and Supplement 1 to NUREG-0737 identified those requirements that were approved for implementa-1-

tion.

The action plan presents a sequence of actions, some already taken, that results in a gradually increasing improvement in safety as individual actions l

are completed.

The Limerick units are receiving and will receive the benefit of these actions on the schedule discussed in the SER.

The improvement in safety from these actions has not been quantified, however.

P

-(8) Comparison of Limerick Risks with Other Plants To provide a perspective as to how the Limerick reactors compare in terms of risks from severe accidents with some of the other nuclear power plants that are either operating or that are being reviewed by the staff for possible issuance of a license to operate, the estimated risks from severe accidents for several nuclear power plants (including those for Limerick) are shown in Figures 5.4n through 5.4v for three important categories of risk.

The values for individual Limerick FES-5-115

tl

)1i 1

lr),,

}

o:

t.

3 r 2E t. I<GE >. 3. _

F L

i i

g m

t t

1 i

,9 o

o m

m 4

u e

1 r

r e

i

- ~ _ "..:

c k

5 i3

<t.& g.

i

+

4 F

n E

_g n

S t

a op(E l

1 a

f lp s n

.l aet lnri itsm a

I csoa

. t e

e nt nese

!i=

.i n

s i

si)d i

et

,jl tera l

l hfe oror i8 l

ml oo y

pp s ajt I

eef rrva 1

eaet a

1 I

ttra oiel n

i ai.

f grt ey Soa aif !=

5nrcr I

ti 1

ros ain l

1 ferk 6 ecseaw el I

tici nvct ii 1

oiih 8

I tnd eges iii I

s nu ctp fosp

s t

l on o

lsfr liot

F t

I odri irse 4T an l

4.a we v

LE gtvm v Uel G

nae ANP E

Ford 8

e N

iee 32I r R n e E - o t

g s

. Gp

< inai g

lc a,"

t I

o ci IC.

uf a

f R f

ronl aru a!*

I s

1 i c u 6 r 6 P ct e

ro 9 R

5il r see aT p

I g a A e p 4saa

.t pl

.v u r t sif,

I pl dei a

m a

npe n

to con ai t

I

.t s

ewte

.ti r

aI.<

I

.. - -. :-~

- ~

ll lll

I 141.

~

~

~

=

II 14 -

143 7

Ij g V a

{ }a 7 8

}{ t J E l 1 7 { s p

d o

=

u n

- = E e

1 E 1 3 J e 3 e 3 3 *

- e E !

E e

a.u u u o a o z 3 2 2 3 j j g g j j j Figure 5.p Estimated early fatality risk with supportive medical treatment i

(persons) from severe reactor accidents for several nuclear power plants either operating or receiving consideration for issuance of license to operate for which site-specific applicatforts of HUREG/CR-1695 accident releases have been used to calculate off-site consequences. Bars are drawn to illustrate effect of

~;

uncertainty range discussed in text.

See footnotes following Figure 5.4v.

Limerick FES 5-118 v

k;

H e

8 E$=

E go I gn>

• 2gmEO3 "{

o,*e k >i8* o* 25o$.e uF3G g

i

=

i g

9 E

3 E

A R

Y e

E E

e e

e

.+

RO T

e CA e

E e

R l

R 8

i E

e e

P e

S e

e e

e N

e e

O o

S R

E P

i

.8 I.

a e

a a

a

.\\

e

.t e

l n

i n

e e

a e

- - a

.A in.

C

- " e-a a

e P..

e, k

m a

3 3

e oi t

P 2.

en 3

n n

e e

n k

c.

io io f

e e

lu d

f k

a d

e r

2 h

e o o e

e t

y i.

.e n

n 2

n n o ie r

h r

e P

k n

a c

G d

e o

r s

P.

ic o

w.

o e

i d

is V

O c

f C

2 n

r u

m u

r n

r f

s e

a

i. i n e l

o l

i n

m m n r

ie y

k L

m q

t f

P t

e m

le o

a r

r a

r d

o r

n s

io N

r n

e w

d.

d io i

o y

la e

N o

e r

a i le e e.

t u

u a

I I

Z L

S B

C C C C F

G H M P

P S

S S

S S

W W W e

. ^Vesmo3*"

', %D',D'" a *M s a m 7

@xn* coy 0.,rMsO*O w,;,ed a>,ma 4>ee3e n2g.S t

v Ocs*

t o

r

,sQ *.<ese 1eantrO1,nO,_ a@*r" %Os m e < a s e,. 3cn72' oO(65 "U* e3*re s

s D

,.Q 13O OO3m a a&sb+r.O3 %O ;mCn3n4 O%

0VessMe.3Q O5

, "7 5

e e

,-Q3m. *rO Oves >trS*

v $ @ O*r3O* rem >t

• 7m e3O o%. *., Ocs8 W

• A<

• i r

r-

-r Es g ;v w8 ppd

i; L

i i

I:

t is 1

. j.

i:

1..

r.

(

1.o IP 1*

o

<i y

3 b

3

= i.

=

E 3

g

=

5m

.)

$3 Im l:

.o i

14 I

i f

f i I 11 8I ie i

r e

e a

l l

,l

+

~

a

) f' '

y y

2 2

5 y

_1 a

c

.i i

[. t '

Figure 5.4r Estimated latent cancer fatality risk, excluding thyroid (persons)

, f-from severe reactor accidents for nuclear power plants having "j;

plant-specific PRAs, showing estimated range of uncertainties.

I'i; See footnotes at the end of Figure 5.4v.

l Limerick FES 5-120 i. !!!. '

e l

IA 2

m

O s n

a s

a s

i a

a a

s a

a a

1 s

e

.O l

I s

a a

3 a

s I

a s

a 8

a u

a u

e

.I l

l s

e a

a a

a I

a a

A s

e R

E u

n Y

R

=

O e

T 2

CA E

l R

.k l

R s

^-

a e

u a

E u

P s

2 a

S a

=

a NO e

a S

e R

a EP a

e

.h g

l i

e e

I s

e a

a s

e a

a s

e a

e

't

• I g

l a

a i

W e

a s

H e

s a

a a

a e

s

~

~

.h k

3, ae P

2 en 3

e e

f h

r 2

n k

e e

lu d

f k

e d

e t

y N

e r

h r

e r

o o c

G d

e n o ic n

e o

r e

e n

n 2

o f

n w

• o e

i d

ir l

o r

n f

C 2

s V

a r

u m

u r

n m m n y O e

o e

t b

L q

e f

P

' n a

. m H

r N

• i o

e r

a i

le e

e t

u u

a r

d t

lo N

s o

y e

l r

a B

B C C C C F

G H M P

P S

S S

S S

W W W m* gCh * **

rec 3O,,ma a 1

48+4a +< 5;N* NMO *Ca*3G +r7<5*=0a ^VD5hO3n n+

s o +rf 3 +c OO3O4D5 s

D s

~

0r r

(-

t E

i 5

45O3 t f < # -D54 5 MOO +rO5 OOO*aD36e O5 *

<D5sO *- 3CO* #D"

=

s s

4 VO1D5 V

• O3*re oD

(

n De5

(-

s n

(-

• =

2

• M"rD 5

=

5DOD*=<*3e O03n [@1s+*O3 O1 OVm.1a+*3Q O5

• nn a

4 eiCp30e O-pr-(

- (

(

i

• Ar~

(-

s O* =

T

• D sf 05 T7*=O7 i r(

• Oe3nm +rO O3m1a+o %

n * + D4 & DO * % '* O OV"a +* C. a +

~

Dr*O3i Og n

e 2CArONO:aNmoW OOO*au3e 3i 3

s rs& &DD3 Cif 8*

a s D

0 = %I

- O nD r0 O0 0C _ A+rf 5N f4nDn D

v x

t r

r i#i

((

D Dn De5n p m a1sE3,,0

= **

6*+re OO3efJ 4

f% D 0 +r O-D~4(

cer 0+rfD n+58 nD3Cf3Oii*

CD i s5

• O 4

a*iOCthoDa *3 +rmX*(

C3Om1+rO 3+cM 5s3e(

Mf( %O O +r3O +r f i O +c,2D (3a n

t(

s O

D DD Dn s

f D

O=

m*=CC5e m*&<*

=

• =l 1 W mmM W8YH e

L,,.!I j i j

l,l l1

]l I

1lllljl I1l6 lf 1!

,lI,k

)a

'lii k

i\\.

!{,

L F

im

.a =gE

• 5t8 5(,

ig m

m m

m 4

1 e

u i

r 1

r i

e

...:.=

c 1

0 k

5 k*EI~ jeeE a.! t

...::r

.. ~

b-

. ~ ". -. ~,

F 4

E

}2Ii5 l'

t t

S m

oosE

}2Ii0 A'

t B

ppes eevt 4e rrei l'

I u

aarm m

ttea

$*5 I'

i m

.e i t

nre m

ged 32}

e a

Socl erta' Eb e

ot e

rre fe n

33I, ocat I

oec tict 5ei nvih t

oidy tner 52!

t egno s

ti 0E: 2 $i t

csd 5

ao tnfc 5N" 1

soa I

2 tirn 2

hd c

5*5i eese t

rer eav 5 1.

ntef t

dira oat EEI onl a t

f l

fni 2o $4 Fout t

ircy 2U i

g l

uier rsai L

esrs 3 $.! A M i

y.5 u

k E

5ap 3^$,

i v

no(

l G

ANP E

Ua N

4cwp r R n e

a E -

D

.v e e e id$a t

g G >s t

rr e

I o

s o C.

Rci fp o aE!

t s

1 i f

l n fc u 6 las n6 P in) 3*liI i

g a A o 9 h

ct esf a p t

i2i1n I

epl n

r -

p ei seo eim l

d a

t i5Gj n

t th t

o s

s oe i

r IL a e

t b-

___ _ i 6

'1.e 141 7 t>

Ih p

<l IM l

K

,- =

=

.g I,

144

=

14

=

I I

f f

I I

I I

I t

f I

f f

I t

I i

it i

t t

e I.I I.1 ll l1 N

M l

}

=

c 3

5 I

~

a Figure 5.4u Estimated latent thyroid cancer fatality risk (persons) from severe reactor accidents for nuclear power plants having i

plant-specific PRAs, showing estimated range of uncertainties.

See footnotes at the end of Figure 5.4v.

Limerick FES 5-123

~

.e

/ j o

[-

+

a

~

!j 1.s 4j

?til

y mit

[p g/(

ig.

PJ

'# r T

jij

+

p I

!)

im r

^

E o

m g

E.

o m

.s 3

o 5

m

• Se r u

r I

~

~

.rJ to :-

o o

im I

i i

t t

t i

I t

t I

t t

t t

t t

t t

t g

t t

i 1

a d

I"j 2

'i 4

I g

a o

l I

h Figure 5.4v Estimated latent tnyroid cancer fatality risk (persons) from severe reactor accidents from several nuclear power plante either i

operating or receiving consideration for issuance of ifcense to l3 operate for which site-specific applications of NUREG-1695 acci-dent releases have been used to calculate off-site consequences.

Bars are drawn to illustrate effect of uncertainty range dis-

.il,,a cussed in text.

See footnotes on following page.

o la Limerick FES 5-124 Il b a

I Notes for Figures 5.4n through 5.4v Except for Indian Point, Zien, and Limerick, risk analyses for other plants in these figures are based on WASH-1400 generic source terms and probabilities for severe accidents and do not include external event analyses.

Ary or all of the values could be under or over-estimates of the true risks.

1-01 = 1 x 10 1 tAssumes evacuation to 25 miles.

ttWith evacuation within 10 miles and relocation from 10-25 miles, aExcluding severe earthquakes and hurricanes.

NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

plants are based upon three types of estimates:

from the RSS (labeled WASH-1400 Average Plant), from independent staff reviews of contemporary probabilistic risk assessments (Indian Point 2 and 3, Zion and Limerick), and from generic applica-tions of RSSMAP accident sequences to reactor sites for environn ntal statements by the staff (for 21 nuclear power plants).

The RSS risk estimates were intended to illustrate the general level of risk from a variety of plant designs at a variety of sites, and these estimates appear in Figures 5.4n, q and t as point estimates along with the corresponding point estimates obtained oy the other types of analysis.

Figures 5.4o, r and u show the range of uncertainty that is estimated for those four plants for which a plant-specific probabilistic risk assessment has been parformed.

Figures 5.4p, s and v are included to illustrate the effect uncertainties of a factor of 100 would have upon comparison emongst risk estimates using a fixed set of accident sequences, but site-specific mete-orology and population.

The display of risk in three sets of figures is intended to allow comparison of risks similarly evaluated, and to allow an over-all comparison ~of risks to be made among all types of risk evaluations available.

Figures 5.4n through 5.4v indicate that the estimated Limerick risks may be higher than those for some plants, and lower than those for severcl other plants but, except for early fatalities at the Wolf Creek site, not by a margin that would exceed the uncertainties in the estimates themselves.

Similarly, Figure 5.4m, which comparcs core melt probabilities for Limerick with several other reactors, indicates that the estimated likelihood of a core melt accident i

at Limerick is roughly the same es for sevaral opersting reactors.

Furthermore, i

any or all of the estimates of risk could be under or overestimates.

5.9.4.6 Conclusions The foregoing sectinns consider the potential environmental impacts from acci-dents at Limerick station.

These have covered a broad spectrum of possible l

accidental releases of radioactive materials into the environment by atmospheric j

and liquid pathways.

Included in the considerations are postulated design-basis accidents and more severe accident sequences that lead to a severely damaged reactor core or core melt. The applicant also considered similar accidents in the ER-OL.

The staff has considered the technical merits of the applicant's Limerick FES 5-125

~-

7 y 7 bl Lassessment and the uncertainties involved, and agrees in several areas and dis-agrees in several other areas (see Appendix N).

Notgabledisagreementsarein j g the area of. source i.erms.and offsite emergency response modeling.

For several

~

W.

sequences the staff's source terms are considerably higher; the offsite emer-gency response modeling is site specific and more pessimistic for severe earth-M@h

' quake conditions in the site region than that modeled by the applicant.

As a result, the applicant's risk estimates are substantially lower than the staff f

estimates.

In both the applicant's and the staff's analyses of accident risk, d:

however, there~are very large uncertainties, m;

.This section documents the staff's use of PRA in its inquiry into the environ-i, mental impacts of reacto.r accidents.

The staff's inquiry into the implications of the risk assessments for reactor design and operation; to wit, questions of y

compliance with the reactor safety regulations and the questions of whether plant-specific vulnerabilities to severe accidents warrant requirements more; stringent than the norm, will be documented elsewhere.

,f The environmental impacts that have been considered include potential radiation exposures to individuals and to the population as a whole, the estimated likelf-hood of core melt accidents, the risk of near-and long-term adverse health effects that such exposures could entail, and the potential economic and societal consequences of accidental contamination of the environment. These impacts could be severe, but the likelihood of their occurrence is judged to be small and comparable to that of other reactors.

This conclusion is based on (1) the fact that considerable experience has been gained with the operation of 2_

similar facilities without significant degradation of the environment, (2) the fact that, to obtain a license to operate, the Limerick station must comply with the applicable Commission regulations and requirements, (3) a comparison with the estimated core melt probabilities of other reactors, and-(4) a proba-bilistic assessment of the risk based upon the methodology developed in the l

RSS, improvements on the RSS methodology including external event analysis, and a sensitivity analysis of offsite emergency response modeling.

The overall assessment of environmental risk of accidents, assuming protective actions, shpws that the risks of population exposure and latent cancer fatality are within a factor of 30 of those from normal operation.

Accidents have a poten--

tial for early fatalities and economic costs that cannot arise from normal operations; however, the risks of early fatality from potential accidents at the site are small in comparison with risks of early fatality from other human activities in a comparably sized population, and the accident risk will not add significantly to population exposure and cancer risks.

Accident risks from i

Limerick are expected to be a small fraction of the risks the general public l

incurs from other sources.

Further, the best estimate calculations show that the risks of potential reactor accidents at Limerick are within the range of such risks from other nuclear power plants.

Based on the foregoing considerations of environmental impacts of accidents, i f4 which have not been found to be significant, the staff has concluded that there

' y are no special or unique circumstances about the Limerick site and environs that l >l would warrant consideration of alternatives for Limerick Units 1 and 2.

L 5.10 Impacts from the Uranium Fuel Cycle

,4 pi The Uranium Fuel Cycle rule, 10 CFR 51.20 (44 FR 45362), reflects the latest-information relative to the reprocessing of spent fuel and to radioactive waste

)

p y

Limerick FES 5-126 F

w e.

-w-.

r--,

,-.,.-,,m,_.-

,-,,.--m--

4

_,me.-

.s ce 6

EVALUATION OF THE PROPOSED ACTION 6.1 _ Unavoidable Adverse Impacts that can be attributed to the operation of the Lime These impacts are summarized in Table 6.1.

The applicant is required to adhere to the following conditions for th tection of the environment:

(1) that may result in any significant adverse enviro

~

not evaluated or that is significantly greater than that evaluated in activities to the Director of the Office of Nuclear and will receive written approval from that office before proceeding w such activities.

(2)

The applicant will carry out the environmental monitoring programs ou lined in Section 5 of this statement, as modified and approved by the staff and implemented in the Environmental Protection Plan and Techn Specifications that will be incorporated in the operating licenses (3)

If an adverse environmental effect or evidence of irreve mental damage is detected during the operating life of the plant, the applicant will provide the staff with an analysis of the problem and a proposed course of action to alleviate it.

6.2 Irreversible and Irretrievable Commitments of Resources There has been no change in the staff's assessment of this impact sin earlier review except that the continuing escalation of costs has increase the dollar values of the materials used for constructing and fueling the plant.

6.3 Relationship Between Short-Term Use and Long-Term Productivity There have been no significant changes in the st review.

t 6.4 Benefit-Cost Summary 6.4.1 Summary and ecsts that are associated with the operation station.

They are summarized in Table 6.1.

I Limerick FES 6-1

\\

t'

98 Table 6.1 Benefit-cost summary for Limerick Primary impact and effect Quantity on population or resources (Section)*

Impacts **

BENEFITS Direct Electrical energy 10 billion kWh/yr Large Additional generating capacity 2110 MWe Large (design rating)

(Sec. 6.4.1)

COSTS Envircnmental

~"'

Damages suffered by other water users Surface water consumption (Sec. 5.3.2)

Small Surface water contamination (Sec. 5.3.2)

Small Groundwater consumption (Sec. 4.3.2)

None Groundwater contamination (Sec. 4.3.2)

None Damage to aquatic resources Impingement and entrainment (Sec. 5.5.2)

Small Thermal effects (Secs. 5.3.2 & 5.5.2)

Small Chemical discharges (Sec. 5.3.2)

Small Diversion flow effects (East Branch)

(Sec. 5.5.2.3)

Moderate Damage to terrestrial resources Station operations (Secc~5.5)

Small Transmission line maintenance (Sec. 5.5.1)

Small Adverse socioeconomic effects Loss of historic or archeological resources (Sec. 5.7)

Moderate Increased demands on public facilities and services (Sec. 5.8)

Small Increased demands on private facilities and services (Sec. 5.8)

Small Noise (Sec. 5.12)

Moderate-Small Adverse nonradiological health effects Water quality changes (Sec. 5.3.2)

None Air quality changes (Sec. 5.4)

• See footnotes at end of table.

Limerick'FES 6-2 p

4P r -

Table 6.1 (Cqntinued)

Primary-impact and effect Quantity on population or resources (Section)*

Impacts **

Adverse radiological health effects Routine operation (Sec. 5.9.3)

Small Design basis accidents (Sec. 5.9.4)

Small Severe accident risks (Sec. 5.9.4)

Small Uranium fuel cycle (Sec. 5.10)

Small

• Where a particular unit of measure for a benefit / cost category has not been specified in this statement or where an estimate of the magnitude of the benefit / cost under consideration has not been made, the reader is directed to the appropriate section of this report for further

.information.

• • Subjective measure of costs and benefits is assigned by reviewers, where quantification is not possible:

"Small" = impacts that in the reviewers' judgments, are of such minor nature, based on currently available infor-mation, that they do not warrant detailed investigations or considera-tions of mitigative. actions; " Moderate" = impacts that in the reviewers' judgments are likely to be clearly evident (mitigation alternatives are usually considered for moderate impacts); "Large" = impacts that-in the reviewers' judgments, represent either a severe penalty or a major benefit.

Acceptance requires that large negative impacts should be more than offset by.other overriding project considerations.

6.4.2 Benefits A major benefit to be derived from the operation of the Limerick station is the approximately 10 billion kWh of baseload electrical energy that will be produced annually (this projection assumes that both units will operate at an annual average capacity factur of 55%).

The addition of the plant will also improve the applicant's ability to supply system load requirements by contributing 2110 MW of generating capacity to the Philadelphia Electric Company sys_ tem (1055 MW from Unit 1 in 1985 and 1055 MW from Unit 2 in 1989).

6.4.3 Costs i

No significant socioeconomic costs are expected from either the ope' ration of the Limerick generating station or from the number of station personnel and their families living in the area.

The socioeconomic impacts of a severe acci-dent could be large; however, the probability of such an accident is small.

6.5 Conclusion As a result of it; analysis and review of potential environmental, technical, and social impacts, the NRC staff has prepared an updated forecas,t of the-effects of operation of the Limerick generating station.

The NRC staff has Limerick FES 6-3

o-

..e I

i determined that the Limerick' generating station can be operated with minimal--

I envircnmental impact.

To date, no new information has been obtained that alters the overall favorable balancing of the benefits of station. operation versus the environmental costs that resulted from evaluations made at the construction permit stage.

6.6 Reference U.S. Nuclear Regulatory Commission, NUREG-0586, " Draft Generic Environmental Impact Statement on Decommissioning Nuclear Fac.ilities," January 1981.

j-

+

j 4

[

e i

i f

)

I

[

)

i

(

i 1

l Limerick' FES 6-4

-e

j;f ',c y "

APPENDIX H LIMLRICK ACCIDENT SEQUENCES AND RELEASE CATEGORIES USED IN CONSEQUENCE' AMLYSIS s

For_the purpose of performing accident consequence analyses for the Limerick-

DES and FES,- the-staff requested Brookhaven National Laboratory (BNL) to help develop specifications of atmospheric release of radionuclides from severe accidents in the Limerick reactors based on the applicant's two probabilistic risk analyses (PRAs), Limerick Generating Station P obabilistic Risk Assessment

.(LGS-PRA)F and the Limerick Generating Station Severe Accident Risk Analysis (LGS-SARA).2 The specifications included (1) identification of core-melt acci-dent sequences leading to atmospheric release initiated by internal causes, fires, and earthquakes; (2) probabilities of the wquences; and (3) quantities-and forms of radionuclides (source terms) and the other parameters necessary for appropriate characterization of atmospheric release frorr. these sequences.

The ground rules recommended by the staff for the BNL analysis relate to the method of estimating source terms.

There has been significant research activity in this area sponsored by both industry and the Commission since the publication of the Reactor Safety Study (RSS)3 in 1975.

Updated fission product source term assessment methods are currently being developed.and are receiving exten-sive peer review.

However, it is the judgment of the staff that the applica-tion of the evolving methodologies for assessment of source terms in licensing activities before they are thoroughly and carefully appraised would be premature.

Therefore, the staff requested that BNL use the RSS prescriptions of fission c

product release from the damaged fuel, primary system holdup, credit for decon-tamination by suppression pool scrubbing, and fallout, plateout, and transport of radionuclides in the containment leading to atmospheric release.

These RSS prescriptions are explained below.

In the RSS methodology, quantities of fission products released from the core f

material were based on four release components:

gap, melt, oxidation, and

-vaporization.

The gap release is modeled as a single event and is assumed to occur at accident initiation as the result of rupture of fuel cladding.

It con-sists mostly of activity that would be released to void spaces within the fuel' i

rods during normal reactor operatior., and rapid depressurization of contained gases provides the driving force for escape.

The melt release occurs from the-4 fuel while it first heats to melting and becomes molten.

High gasflows in the j

i

' core during this period sweep the activity out of the core region. _ The melt release is divided into'10 equally sized releases evenly spaced between.the time l

of core melt and the time of core slump.

The oxidation release is modeled as a.

~

single release that occurs when the-reactor pressure vessel d@V) head fails and is the result of oxidation of that ' fraction of the core. debris that is assumed to interact with water on the diaphragm floor or to fall into the suppression pool.

Finely divided fuel material'is scattered into an oxygen atmosphere,and under-

,goes extensive' oxidation, which liberates specific fission products.

The vaporization release _is assumed to start after vessel failure when core-concrete Limerick FES H-1 o

r interactions begin.

Turbulence caused by internal convection and melt sparging by gaseous decomposition products of :oncrete produce the driving forces for escape.

The vaporization release is divided into 20 parts,10 releases of exponentially decreasing magnitude in tne first half hour followed by 10 more release,s, also of exponentially de:reasing magnitude, during the next I hours.

Also in the RSS methodology, no specific credit for attenuation of fission products released from the RPV to containment building is allowed in the pri-mary system.

Thus, all the fission products released during the gap and melt release phases are assumed to enter the containment building.

) ;-

For fission product attenuation as a result of scrubbing by water in the sup-gl pression pool, a decontamination factor (DF) of 100 is used for the subcooled

~

pools and e DF of 1 is used for the saturated pools.

(Noble gases and organic iodine are not subject to pool scrubbing.)

f-In the RSS methodology, the fission product transport within the containment building volumes is predicted using the CORRAL-II code.

This code is used in conjunction with the fission product release model, pool scrubbing model, and the MARCH code.

l As stated earlier, in the source term assessment made by BNL for use in the Limerick DES, only the RSS methodolgy was used.

Use of the RSS methodology for Limerick may have resulted in over-estimates of source terms for some accident sequences and underestimates of source terms for others.

However, because the evolving methodologies have not been fully appraised, the staff used its current practice of following the RSS source term assessment methodology in licensing evaluations.

On balance, however, the staff has concluded that the risks esti-mated using the RSS source term methodology are reasonable, particularly when considered within the overall numerical uncertainties discussed in Section 5.9.4.5(7).

The staff worked with BNL during the analysis, and-the final results have been reviewed by the staff and fcund adequate.

Following the staff's guidelines, BNL.develtped 27 release categories for use in the Limerick DES.

The same 27 release categories have also been used in the staff analysis in the FES.

Char-acteristics of these releare categor'es are shown in Table 5.11c and their likelihoods (point estimates of mean annual probabilities) in Table 5.11d.

As noted in Section 5.9.4.5(2), source terms associated with four of the release categories.in Table 5.11c, and probabilities of some of the release categories in Table 5.11d include revisions made after pubitcation of the DES, For iden-tification and quantification of these release categories, BNL considered (1) the sequence of events and conditions that could lead to core melt (acci-dent damage states); (2) the containment building failure modes and radionuclide.

release paths; and (3) the actual characterization of radionuclide releases to the environment.

Procedures used for identification of these release categories and their brief descriptions are summarized below.

Initially 67 plant damage states were identified for the Limerick reactors.

Subsequently, however,10 surrogate damage states were 'ound to encompass these original 67 damage states.

This was possible because many of the original damage states were found to be very similar in terms of the core-melt accident progres: ion and containment failure characteristics.

Table H.1 gives a brief description of each of the surrogate damage states and uses simple designators Limerick FES -

H-2

c do identify -the damage states for easy reference.

The first six of the surro-gate' damage states given in Table H.1 include damage states discussed in LGS-PRA and NUREG/CR-3028,4. but they also include the dam?ge states initiated by fires and low to moderately severe earthquakes discussed in LGS-SARA.

The last four

~

of the suYrogate damage states in Table H.1 include damage states discussed exclusively in LGS-SARA.

Mean probabilities per reactor year assigned to the 10 surrogate damage-states are shown in Table H.P.

- Using the 10 surrogate damage states, BNL performed analyses to determine the

. Limerick containment failure modes and radionuclide release characteristics using the MARCH / CORRAL computer code system *.

Seven containment failure modes and release paths were identified (sec. Table H.3) and analyzed.

They can be subdivided into leakage. failures and structural failures.

The leakage failures prevent the more catastrophic structural failure and, in some of the cases,

make effective use~of the standby gas treatment system (see Section 5.9.4.4(1)).

The structural failures result in release pathways that either (1) bypass the suppression pool by failing the drywell or by causing the suppression pool to 4

drain or (2) pass through the suppression pool.

The mechanisms for developing these release pathways ata overpressure from steam or noncondensibles, over-pressure from hydrogen burns (for the containment deinerted cases), seismic (earthquake) failure,of structures and systems, and steam explosion-induced failures.

Analyses showed that there could be only 40 combinations of the 10 surrogate damage states and the 7 containment failure modes (and release paths) with non-zero probabilities (having any possibility of occurrence).

The other 30 combinations were considered as essentially impossible.

The 40 combinations.of surrogate damage states and containment failure modes (and leakage paths) were further reduced because the accident progressions resulting in radionuclude release to the atmosphere associated with a number of them are very similar.

This resulted in 27 release categories for conse-quence analysis.

These release categories are described in Table H.4.

It should be noted that the labeling of each release cate' gory has been made both in terms of the surrogate damage state and the matching containment failure mode,or leakage path.

As stated earlier, specifications (including the source terms) of each of the 27 release categories developed by BNL are shown in Table 5-11c. The timing of the radionuclide release, energy of release, duration of release, and warning time for evacuation shown in Table 5.11c were based on the MARCH analysis.

The time of release 'is defined as the time of containment failure for those cases in which the meltdown would take place in an intact containment building.

For those cases, when the containment building would fail prior to core damage, the l

time of release is defined as'the start of core melting.

The duration of l

release is defined as the time for the containment building to blowdown to-

?

l

• The MARCH computer code used includes a new decay heat model based on the

!~

ANS-5.1-1979 standard.

The 1979 standard produces an integrated decay heat.

over the first hour after the reactor shutdown about 20% greater than the 1971 standard used in the previous BNL review (NUREG/CR-3028)4 of the LGS-PRA.

The j

main effect 'of the new decay heat model has been the change in timing of major events during the progression of the accidents.

The time to core meltdown, core slump, reactor pressure vessel failure, and containment failure predicted using the new decay heat model are significantly earlier than in NUREG/CR-3028.

~

J Limerick FES H-3

atmospheric pressure. ' However, if the building fails first (meltdown into a failed containment building), the duration of release was defined to be from the start of core melting to the completion of vaporization release.

The warn-ing time.is defined as.the time period between the start of the core melt and the tim,e of containment failure.

If the containment builoing fails first, the

warning. time'was defined as the time from the time of containment failure to the start of core melt. The energy of release is the energy release rate asso-ciated with the release at the time of containment failure. 'In those cases j.

where'the release could be spread out over many hours, the energy of release J

would be low.

The height 'of release was chosen to be 25 m (82 ft) in all cases.

Following the guidelines provided by the staff, BNL subdivided the mean proba-bility of each release category initiated by earthquakes into two parts.

One part was associated with the release category that would be initiated by very severe earthquakes (effective peak ground acceleration equal to or in excess

. of 0.4g*), and the other part was associated with the same release category initiated by low to moderately severe earthquakes (effective peak ground acceleration less than 0.4g).

The latter part was added to the mean probability of the same release category initiated by internal causes and fires.

The re-arranged mean probability for each release category is shown in Table 5.11d.

h The purpose of such breakdown was to aid in making appropriate assumption k

regarding offsite emergency response in the consequence analysis.

It was the

[

judgment of the staff that earthquakes resulting in effective peak ground

}

acceleration equal to or greater than about 0.4g would be of severity of Modi-fied Mercalli (MM) intensity ccale IX or worse.**

Earthquakes of MM intensity s_cale IX or higher would be likely to seriously hamper the offsite emergency response efforts.

(See Appendix I for description of offsite damages likely to be caused by earthquakes of various MM intensity scales.)

There are substantial uncertainties in the estimated mean probabilities shown in Table 5.11d.

Further, the mean probability of a release category is not neces-l, sarily the representative of the full spectrum of values of its probability.

['

Particularly for seismically induced release categories, values of probabilities span several orders of magnitudes between low and high estimates.

However, it is

[

the judgment of the staff that the use of the mean probabilities in consequence analysis, supplemented by discussion of uncertainties resulting from this use, provides a reasonable risk perspective.

For discussion of uncertainties see Section 5.9.4.5(7).

• g stands for acceleration due to gravity and is numerically about 32 feet per second per second.

• • The lack of actual recording associated with this intensity and the controversy surrouriding the definition of effective peak ground acceleration made the choice of 0.4g imprecise.

A sensitivity analysis performed with a range of values of effective peak ground acceleration such as 0.35g to 0.5g would have been more appropriate.

However, it wa the staff's judgment that breakdown of probabil-ities of seismically induced release cattgories using several values from the range 0.35g to 0.5g of effective peak ground acceleration would no't have resulted in probability sets very different from those obtained by using 0.4g.

Limerick FES.

H-4

' ~

--__--._-_---______________._-___.__d

s.

};

l REFERENCES a

1.~

Letter, from PEco to NRC, submitting operating license application and a report, " Limerick Generating Station, Probabilistic Risk Assessment,"

March 17, 1981.

2.

Letter, fr=. C. J. Bradley, PECo, to a. denwencer, NRC, submittirg report

" Limerick Generating Station, Severe Accident Risk Assessme.nt," April 21, 1983.

3.

U.S. Nuclear Regulatory Commission, NUREG-75/014, " Reactor Safety Study--

An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants,"

October 1975 (formerly WASH-1400).

4.

I. A. Papazoglou, et al., " Review of Limerick Generating Station Probabil-istic Risk Assessment," U.S. Nuclear Regulatory Commission, NUREG/CR-3028, February 1983.

I l

Limerick FES H-5 8

~

ye:

Table H.1 Description of surrogate damage states Designator Description I-S These~are LOCA (loss-of-coolant accident)-initiated sequences (medium and small pipe breaks only) involving loss-of-coolant c

inventory makeup.

They would result in a relatively fast core melt, with the containment intact at the time of core melt.

I-T These are sequences initiated by transient events

• involving loss of-coolant inventory makeup.

Core melt is expected to be relatively fast and the containment to be intact at the time of core melt.

II-T These are transient or LOCA-intiatsd sequences involving loss of~

containment heat removal or inadvertent steam relief valve opening accidents with inadequate heat removal capability.

Core melt is expected to be relatively slow as a result of the lower decay power level, with the containment failing before core melt.

III-T These sequences are transients ' involving loss of scram (fas't shut-down of reactor) function and inacility to provide coolant makeup, large LOCAs with insufficient coolant makeup, transients with loss of heat removal, and long-term loss of-coolant inventory makeup.

Core melt is expected to be relatively fast, and the containment intact at core melt.

IV-T These sequences are transients that involve loss of scram function and a loss of containment heat removal or all reactivity control, but with coolant makeup capability.

Core melt is expected to be relatively fast with the containment fail.ing before core melt because of overpressure.

IV-A As above but initiated by large LOCAs IS-C These sequences are seismically (earthquake) induced sequences that lead to failure of the coolant inventory / makeup systems and a breach of wetwell integrity with the reactor scrammed.

Core melt is expected to be fast, with the containment failing before core melt because the residual heat removal (RHR) system suction lines I

are severed.

L IS-C As above, but coupled with a loss of the scram function.

S-H2O These sequences are seismically induced reactor vessel failures (plus random reactor-vessel failure), coupled with immediate con-tainment failure.

Core melt is fast, with the vessel and contain-ment both failed at the time of core melt.

This sequence assumes the vessel break is high, which would allow water to be, retained in the bottom of the vessel before core slump.

• See next page for footnote.

I Limerick FES H-6

.\\

i

.e Table H.1 (Continued)

Designator Des /ription T

~ ' '

S-H2O As above, but with a vessel failure location that results -in complete draining of the water from the vessel.

"In general, the term reactor transient applies to any significant deviation from the normal operating values of any of the key reactor operating param-i eters.

More specifically, transient events can be assumed to include all those situations (except for the LOCA, which is treated separately) that could lead to fuel heat imbalances.

When viewed in this way, transients cover the reactor in its shutdown condition as well as in its various operat-ing conditions.

The shutdown condition is important in the consideration of transients because many transient conditions result in shutdown of the reactor, and decay heat removal systems are needed to prevent fuel heat imbalances as a result of core decay heat.

Transients may occur as a consequence of an operator error or the malfunc-tion or failure of equipment.

Many transients are handled by the reactor control system, which would return the reactor to its normal operating condi-tion.

Others would be b~eyond the capability of the reactor control system and would require reactor shutdown by the reactor protection system to avoid damage to the reactor fuel.

In safety analyses, the principal areas of interest are increases in reactor core power (heat generation), decreases in coolant flow (heat removal), and increases in reactor coolant system (RCS) pressure.

Any of these could potentially result from a malfunction or failure, and they represent a poten-tial for damage to the reactor core and/or the pressure boundary of the RCS.

The analysis of reactor transients has been directed-at identifying those malfunctions or failures tbt can cause core melting or rupture of the RCS pressure boundary.

Regardless of the way in which transients might cause core melting, the consequences are essentially the same; that is, the molten core would be inside the containment and would follow the same course of events as a molten core that might result from a LOCA.

Each potential transient is assessed to fall into either one of two general categories, the anticipated (likely) transients and the unanticipated (unlikely) transients.

The large majority of potential transients are those l

that have become commonly known as anticipated transients.

All other trans-i ients are considered to fall into the unanticipated transients category.

The i

relatively low probability (unanticipated) transients can be eliminated from l

the risk determination because their potential contribution to risk is small i

compared to that of the more likely (anticipated) transients that would pro-duce the same consequences.

The anticipated transient initiators for which successful reactor scram could l

be accomplished have been divided into five groups for analysis of the Limerick reactors.

These groups are Limerick FES H-7 e

b.*

Table H.1 (Continued)

(1) transients resulting in turbine trip (2) transients leading to isolation of the reactor vessel from the main condenser, a main steamline isolation valve (MSIV) closure, and loss of feedwater (3) transients resulting from loss of offsite power (4) transients resulting from inadvertent open relief valve (IORV)

(5) orderly and controlled manual shutdown Thirty-seven BWR transients identified from operating experience data are listed in Table 2.9 of NUREG/CR-30284 and are included in the first four of the above groups.

If the reactor protection system fails to scram the reactor after an initiating event in any of the first four transient groups, then an anticipated transient without scram (ATWS) condition results.

four groups of ATWS initiators were, therefore, considered:

The following

-(1) turbine trip ATWS (2) MSIV closure ATWS (3) loss of offsite power ATWS (4) IORV ATWS AW Limerick FES H-8

T l

1 Table H.2 Mean (point estimate) probabilities of surrogate damage states by initiating events Probability per reactor year Surrogate Low to moderately Severe damage Internal severe earthquakes earthquakes state causes Fires (EPA * ~< 0. 4g)**

(EPA * > 0.4g)**

I-S 8(-8)***

I-T 8(-5) 3(-6) 9(-7) 2(-6)

II-T 4(-6) 1(-8) 4(-8)

III-T 3(-6) 8(-8) 7(-7)

IV-T 3(-7) 2(-8) 1(-7)

IV-A 5(-9)

IS-C 1(-7) 9(-7)

IS-C 1(-8) 1(-7)

S-H2O 1(-8) 4(-8) 5-H20 1(-8) 4(-7) 1 TOTAL 9(-5) 3(-6) 1(-6) 4(-6)

• EPA stands for effective peak ground acceleration.

• • g stands for an acceleration equal that due to gravity and is numerically equal to 32 feet per second per second

• • • 8(-8) = 8 x 10 s NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

Estimated numbers were rounded to one significant digit only fcr the purpose of this table.

l l

Limerick FES H-9

j

~

Table H.3 Containment failure mode and release path notation i

)p Designator Description i

DW Containment failure via overpressurization.

Failure location in the drywell.

l WW Containment failure via overpressurization.

Failure location in the wetwell above the suppression pool.

99 Containment failure via overpressurization.

Failure location in t-the wetwell below the suppression pool resulting in loss of suppression pool water.

SE Failure via in-vessel steam explosion generated missiles. '

HB Failure via hydrogen burning during the periods when the contain-ment atmosphere is de-inerted. This failure mode also includes hydrogen detonation and ex-vessel steam explosion failure modes, which are of very low frequency.

LGT Containment leakage rates sufficiently low to allow the standby gas treatment system (SGTS) to operate effectively.

l LGT Containment leakage rates so high that the SGTS is ineffective.

i i

l I 1

l 4

i l

i Limerick FES,

H-10

A.

,e

' Table H.4 Description of the release categories Category Description 1.

RELEASE CATEGORIES ASSOCIATED The damage state I-T is defined in WITH SURROGATE DAMAGE STATE I-T Table H.1 and basically consists of transients with loss-of-coolant inventory makeup.

Core melt in such situations is expected to be relatively fast and occurs within an intact containment.

After ves-sel failure, the majority of the core materials are retained on the diaphragm floor below the reactor vessel.

Contain-ment failure occurs via gradual overpres-surization (except for SE, HB, LGT and LGT release -- see Table _H.3) several hours after vessel failure as a result of core / concrete interactions.

1 I-T/DW This release category assumes an over-pressure failure in the drywell wall.

The gap and melt releases would be directed to the suppression pool and subjected to a DF of 100 (water is sub-cooled) before they reach the wetwell airspace.

The vaporization release would be directed to the drywell without any pool' scrubbing.

All fission products in i

the drywell and wetwell would be subjected to agglomeration and settling as predicted by the CORRAL.. code before vessel failure, several hours after the pressure vessel failure.

I-T/WW This release category assumes a failure in the wetwell above the suppression pool.

The gap, melt, and vaporization releases would be released to the drywell and wet-well as described above.

The only dif-forence is that when the containment fails, fission products in the drywell must pass through the downcomers and suppression pool before they are released to the atmosphere.

i H-11 Limerick FES

,,8 Table H.4 (Continued)

Category Description I-T/5N This release category assumes a failure in the wetwell below the suppression-pool, which drains the water.

The gap,

. melt, and vaporization releases would be released to the containment as described above.

The only difference is that at containment. failure the suppression pool would be drained so that fission products in the drywell no longer have to pass through the suppression pool (as'in the I-T/WW release path) before they are released to atmosphere.

I-T/SE This release category results from an in-vessel steam explosion generated mis-

sile, BNL assumed this occurs at~ core slump and opens a direct path from the primary system to atmosphere.

In the LGS-PRA, this failure mode was similar to RSS release category BWR-1.

The release corresponds to an anticipated transient without scram sequence analyzed in Appendix V of the RSS, in which the steam explosion was assumed to occur after only 13% of the core had melted.

Consequently, most of the melt release would be released to containment without pool scrubbing.

However, BNL U, sed a steam explosion release that more appropriately reflects BNL's analysis of the sequence.

I-T/HB This release category could result from hydrogen, burn failures during the time when the containment atmosphere is de-inerted.

BNL used the same release cate-gory as in the LGS-PRA, but reduced the core fraction associated with the oxida-tion releases in a manner consistent with WASH-1400.

(Note in the LGS-PRA, this release category was representative of ex-vessel steam explosions.)

i l

r l

l Limerick FES H-12 l

[6:

l j.

Table H.4 Description of the release categories Category Description I-T/LGT and I-T/LGT These release categories result from containment leakage and assume that the SGTS operates (LGT), or that'it does not operate (LGT).

BNL_used the LGS-PRA releases, but changed the timing to cor-respond to the BNL MARCH analysis.

2..: RELEASE CATEGORIES ASSOCIATED The damage state II-T is defined in WITH SURROGATE DAMAGE STATE II-T Table H.1 and basically assumes loss of containment heat removal.

Eventually, the containment would fail and csuse the

~

loss of inventory makeup.

As the con-tainment would fail prior to core melt and the suppression pool is saturated (DF of 1), the location of containment. failure (DW, W or W -- see Table H.3) is of rather less importance than it is for the I-T damage states.

t II-T/WW This release category assumes a failure in the wetwell above the suppression pool.

1 The melt release would be directed to the ~

.(

suppression pool, but would not be sub-jected to pool decontamination because the water would be saturated.

The vapor-ization release would be directed to the drywell, then through the downcomers to the wetwell air space, and finally to the atmosphere.

This one failure location was also used to represent failures in the drywell (DW) and wetwell below the suppression pool (W).

This assumption is reasonable because, as the pool is' i

saturated, the different flow paths would not retult in significant differences in calculated release fractions (see IV-T below).

II-T/SE This release category results from an in-vessel steam explosior) generated mis-sile.

The release path used in the LGS-PRA, which was taken from Appendix V l

of the RSS, was considered appropriate L

and was used.

Differences relate only to l

the timing, which now corresportds to the j

j present analysis of a II-T damage state.

t i

t Limerick FES H-13

. )

J Table H.4 Description of the release categories Category Description 3.

RELEASE CATEGORIES ASSOCIATED The damage state III-T corresponds to WITH SURROGATE DAMAGE STATE a transient event coupled with loss of III-T scram function (see Table H.1).

Core melt would be rapid and into an intact containment.

Containment failure is predicted to occur after vessel failure as a result of overpressurization.

How-ever, the suppression pool would be saturated so that the gap, melt, and vaporization releases would not be sub-jected to decontamination by the pool.

Consequently, again (as for the II-T damage state) one failure location was used to represent the three potential locations.

III-T/W This release category is similar to the I-T/W sequence; however, because the pool is saturated, the melt release would not be subjected to pool scrubbing.

III-T/SE The steam explosion release category used in the LGS-PRA was considered appropriate and was used.

Differences in conditions postulated were related only to timing, which was made consistent with a MARCH thermal-hydraulics analysis.

III-T/HB_t III-T/LGT and These release categories are also consid-III-T/LGT dered as possible and would be similar to I-T/HB, I-T/LGT and I-T/LGT, respectively.

4.

RELEASE CATEGORIES ASSOCIATED The damage state IV-T is defined in WITH SURROGATE DAMAGE STATE IV-T Table H.1 and essentially consists if ATWS sequences in which continued coolant makeup was postulated to result in over-pressurization failure of the containment before core melt.

The suppression pool would be saturated for these sequences and hence the DF would be unity.

IV-T/0W, IV-T/W and IV-T/W For these release categories, the impacts of the three Lotential failure locations (DW, W, and W) were analyzed.

Because of the saturated pool, similar release fractions were estimated.

The,sc calcula-tions support the use of only one failure location for the II-T and III-T damage states.

The release paths (DW, W, and

~ ) for the three locations are discussed W

in detail above.

Limerick FES

• H-14

O Table H.4 Description of the release categories Category Description IV-T/SE The steam explosion release category used in the LGS-PRA for Class III (damage state III-T) was considered appropriate to this damage state.

Consequently, this release category is used, with the timing changed to be consistent with the BNL MARCH analysis.

5.

RELEASE CATEGORIES ASSOCIATED The damage states I-S and IV-A are defined WITH SURROGATE DAMAGE STATES in Table H.1 and correspond to LOCA-I-S AND IV-A initiated sequences.

They were calculated to have a low frequency but, because of differences in flow paths relative to transients, were analyzed separately.

I-S/0W This release category would result in the release of the melt and vaporization releases to the drywell, thus bypassing pool scrubbing.

However, because the containment would fail several hours after vessel failure, the release fractions are not significantly different from the I-T/0W flow path (in which the gap and melt releases were subjected to suppression pool scrubbing.)

IV-A/DW This release category is similar to IV-T/DW except'that the initiating event is a large LOCA.

6.

RELEASE CATEGORIES ASSOCIATED The damage states IS-C and IS-C are WITH SURROGATE DAMAGE STATES defined in Table H.1 and could be induced IS-C AND IS-C by earthquakes.

The RHR suction lines could be severed, resulting in partial loss of the suppression pool.

The gap and melt releases would be directed to the suppression pool and subjected to decontamination (the water would be sub-cooled and the DF = 100) before release via the severed RHR suction lines.

The vaporization release would be directed to the drywell and then flow through the downcomers into the wetwell.

However, as the suppression pool would be drained below the downcomer outlet, the vaporiza-tion release would not be subject to pool scrubbing.

The difference between IS-C and IS-C relates to the scram function and does not influence the flow paths; only the timing of the sequence is affected.

Limerick FES H-15

o

' Table H.4 Description of the release categories Category

'1 Description

$ j

,i JS-C/DW and IS-E/DW t.

The failure mode for these release cate- -

ri gories was considered to be similar to a hI DW mode in LPG-SARA.

However, this should ei not be interpreted as a failure location J d in the drywell..Rather, for release anal-(

ysis purposes, a containment failure of the type DW is postulated.

s i

IS-C/SE and IS-l/SE g!

For these release categories, the f!

in-vessel steam explosion failures were assumed to be similar to the I-T/SE

j release.

Only the timing was altered to reflect the MARCH analysis.

"l 7.

RELEASE CATEGORIES ASSOCIATED i1:

WITH SURROGATE _ DAMAGE STATES The damage states 5-H2O and S-H2O are S-H20 AND S-H2O defined in Table H.1; they also would-i be earthquake induced.

The RHR suction lines would be severed, but the vessel also could fail at the start of the acci-dent.-

Thus, the core would melt into a failed containment and none of the releases would be subjected to pool scrubbing.

The _on_1y differences between the S-H20 and 5-H2O sequences relate to the location of possible failure in the vessel.

For the S-H2O sequence, water would remain in the vessel and be avail-able for interact slumping occurs. ing with core debris as This would affect move-ment of the fission products and allow the potential for an in-vessel steam explosion.

I The S-Iif6 damage state involves a failure-of the vessel so that the water would be complete'ly drained at the start of the accident.

Thus, there would be no in-vessel debris / water interaction and no potential for an in-vessel steam explosion.

5-tt20/~W, S-H20/SE and S-H20/W These release categories ge considered possible.

Assignment of W failure mode to damage states S-H2O and S-IIf3 relates only to similarity of fission product release path and lack of suppression pool scrubbing, rather than the actual failure location.

Limerick FES.

H-16 7 ** h, - -.

e

\\;

=

l L

l APPENDIX I DESCRIPTION OF POTENTIAL OFFSITE DAMAGES FROM EARTHQUAKES OF VARIOUS INTENSITIES, ACCORDING TO THE MODIFIED MERCALLI INTENSITY SCALE OF 1931

[ Adapted from Seiberg's Mercalli-Cancani scale, modified and condensed.]

I.

a.

Not felt, except rarely under especially favorable circumstances.

Under certain conditions, at and outside the boundary of the area in which a great shock is felt.

b.

Sometimes birds or animals reported uneasy or disturbed.

Sometimes dizziness or nausea experienced.

c.

d.

Sometimes trees, structures, liquids, bodies of water may sway, doors swing very slowly.

II.

a.

Felt indoors by few, especially on upper floors, or by sensitive, or nervous persons.

b.

Sometimes hanging objects may swing, especially when delicately suspended.

Sometimes trees, structures, liquids, bodies of water may sway, c.

doors swing very slowly.

d.

Sometimes birds or animals reported uneasy or disturbed.

e.

Sometimes dizziness or nausea experienced.

III.

a.

Felt indoors by several persons.

b.

Motion, usually rapid vibration.

Sometimes not recognized to be an earthquake at first.

c.

d.

Duration estimated in some cases, Vibration like that due to passing of light or lightly loaded e.

trucks or heavy trucks some distance away.

f.

Hanging objects may swing slightly.

g.

Movements may be appreciable on upper level of tall structures. --

h.

Standing motorcars rocked slightly.

IV.

a.

Felt indoors by many, outdoors by few.

b.

Awakened few, especially light sleepers, Frightened no one, unless apprehensive from previous experience.

c.

d.

Vibration like that due to passing of heavy or heavily loaded trucks.

Sensation like heavy body striking building, or falling of heavy e.

objects inside, f.

Rattling of dishes, windows, doors; glassware and crockery clink and clash.

Creaking of wa11s, frame, especially in the upper range'of,this grade.

g.

Limerick FES*

I-l l.

h.

Hanging objects swing in numerous instances.

i.

Liquids in open vessels slightly,di'sturbed.

j.

Standing motorcars rocked noticeably.

V.

a. Felt indoors by practically all; outdoors by many or most.

b.

Outdoors direction estimated.

c.

Awakened many or most.

d.

Frightened few, slight excitment, a few ran outdoors.

e.

Buildings trembled throughout.

f.

Dishes, glassware broken to some extent.

g.

Windows cracked in some cases, but not generally.

h.

Vases, small or unstable objects overturned, in many instances, with

' occasional falls.

i.

Hanging. objects, doors,-swing generally or considerably.

j.

Pictures knocked against walls or swung out of place.

k.

Doors, shutters opened or closed abruptly.

1.

Pendulum clocks stopped, started, or ran fast, or slow.

Small objects, furnishings moved, the latter to a slight extent.

m.

'n.

Liquids spilled in small amounts from well-filled open containers.

Trees, bushes shaken slightly.

o.

VI.

a.

Felt by all, indoors and outdoors, b.

Frightened many; excitement general; some alarm; many ran outdoors.

c.

Awakened all.

d.

Persons made to move unsteadily.

4 Trees, bushes shaken slightly to moderately.

e.

f.

Liquid set in strong motion.

g.

Small bells rang-church, chapel, school, etc.

h.

Dama0e slight in poorly built buildings.

1.

Fall of plaster in small amount.

1 j.

Plaster cracked somewhat, especially fine cracks (in) chimneys in some instances.

k.

Dishes, glassware broken in considerable quantity, also some windows.

1.

Knickknacks, books, pictures fall.

Furniture overturned in many instances.

m.

Moderately heavy furnishings moved.

n.

VII.

a.

Frightened all; general alarm, all ran outdoors.

b.

Some, or many, found it difficult to stand.

c.

Noticed by persons driving motorcars.

d.

Trees and bushes shaken moderately to strongly.

e.

Waves on ponds, lakes, and running water.

f.

Water turbid from stirred-up mud.

g.

Incaving to some extent of sand or gravel stream banks.

h.

Large church bells, etc. rang.

i.

Suspended objects quiver.

j.

Damage negligible in buildings of good design and construction.

b)

Limerick FES I-2

Nu k.

Damage slight to moderate in well-built ordinary buildings; considerable in poorly built or badly designed buildings,' adobe houses, old walls (especially without mortar), spires, etc.

1.

Chimneys cracked to considerable extent, walls to some extent.

m.

Fall of plaster in considerable to large amounts; also some stucco

falls, n.

Numerous windows broken; furniture to some extent.

o.

Loosened brickwork and tiles shaken down.

p.

Weak chimneys broken at the roofline (sometimes damaging roofs).

q.

Cornices fall from tower's and high buildings, r.

Bricks and stones dislodged.

s.

Heavy furniture overturned, with damage from breaking.

t.

Considerable damage to concrete irrigation ditches.

VIII. a.

Fright general; alarm approaches panic.

b.

Persons driving motorcars disturbed.

c.

Trees shaken strongly; branches, trunks broken off, especially palm trees.

d.

Sand and mud ejected in small amounts.

e.

Temporary and permanent changes in flow of springs and wells; dry wells renewed flow, temperature changes in spring and well waters.

f.

Damage slight in structures (brick) built especially to withstand earthquakes.

g.

Damage considerable in ordinary substantial buildings:

partial collapse, racked; tumbled down wooden houses in some cases; threw out panel walls in frame structures; decayed piling broken off.

h.. Walls fall.

i.

Cracked, broke solid stone wallt seriously; wet ground to some extent, also ground on steep sicpes, j.

Chimneys, columns, monuments, factory stacks, towers twist, fall.

k.

Very heavy furniture moved conspicuously,, overturned.

IX*.

a.

Panic general b.

Ground cracked conspicuously, c.

Damage considerable in (masonry) structures built especially to withstand earthquakes, d.

Some wood frame houses built especially to withstand earthquakes, thrown out of plumb, Damage great in substantial (masonry) buildings, some collapse in e.

large part; wholly shifted frame buildings off foundations, racked frames.

f.

Damage serious to reservoirs.

g.

Underground pipes sometimes broken.

• It is the staff's judgment that MM Intensity Scale of IX and higher.would be associated with effective peak ground acceleration of about or greater than 0.4g.

Limerick iES I-3

e t

t 0

X, c.

Ground cracked, csp:cially when loose and wet, up to widths of several inches; fissures up to a yard in width parallel to canal l

and stream. banks.

b.

Landslides considerable from river banks and steep coasts.

Sand and mud shifted horizontally on beaches and flat land.

c.

],

. d.

j Level of water in wells changed.

e.

Water thrown on banks of canals, lakes, rivers, etc.

j f.

Damage serious to dams, dikes, embankments, c'

Damage severe to well-built wooden structures and bridges, g.

some destroyed.

h.

Dangerous cracks developed in excellent brick walls.

i.

Most masonry and frame structures destroyed, also their foundations, J.

Railroad rails bent slightly.

k.

Pipelines buried in earth torn apart or crushed endwise.

1.

Open cracks and broad wavy folds in cement pavements and asphalt road surfaces.

i j

~ XI.

a.

Many and widespread disturbances in ground, varying with ground I

material, b.

Broad fissures, earth slumps, and land slips in soft, wet ground.

Water ejected in large amounts charged with sand and mud.

c.

d.

Sea-waves Damage seve(tidal waves) of significant magnitude, e.

re to wood frame structures, especially near shock centers.

f.

Damage great to dams, dikes, embankments, often for long distances, Few, if any, masonry structures remained standing.

g.

h.

piers, or pillars,Large, well-built bridges destroyed by the wrecking o i.

Yielding wooden bridges affected less, j.

Railroad rails bent greatly and thrust endwise.

k.

Pipelines buried in earth put_ completely out of service.

t f

XII.

a.

Damagetotal-practicallyallworksofIonstructiondamaged greatly or destroyed.

b.

Disturbances in ground great and varied, numerous shearing

cracks, Landslides, falls of rock of significant charact'er, slumping c.

of river banks, etc., numerous and extensive.

d.

Large rock masses wrenched loose, torn off.

Fault slips in firm rock, with notable horizontal and vertical g.

offset displacements.

f.

Water channels, surface and underground, disturbed and modified greately.

Lakes dammed g.

Waves seen on, waterfalls produced, rivers deflected, etc.

h.

ground surfaces (actually seen, probably, in some cases).

1.

Lines of sight and level distorted.

j.

Objects thrown upward into the air.

i Limerick FES I-4 9

J APPENDIX J CONSEQUENCE MODELING CONSIDERATIONS J.1 Evacuation Model

" Evacuation," used in the context of offsite emergency response in the event of substantial amount of radioactivity release to the atmosphere in a reactor acci-dent, denotes an early and expeditious movement of people to avoid exposure to the passing radioactive cloud and/or to acute ground contamination in the wake of the cloud passage.

It should be distinguished from " relocation" which denotes a post-accident response to reduce exposure from long-term ground contamination.

The Reactor Safety Study (RSS) (WASH-1400, NUREG-75/014) consequence model con-tains provision for incorporating radiological consequence reduction benefits of public evacuation. The benefits of a properly planned and expeditiously carried out public evacuation would be manifested in a reduction of early health effects associated with early exposure; namely, in the riumber of cases of early fatality (see Section J-2) and acute radiation sicknesa that would require hos-pitalization.

The evacuation model originally used in the RSS consequence model is described in WASH-1400 as well as in NUREG-0340.

However, the evacuation model that has been used herein is a modified version (SAND 78-0092) of the RSS model and is, to a certain extent, oriented toward site emergency planning by inclusion of site-specific delay time before evacuation and effective evacu-ation speed as model parameters.

The modified version is incorporated into the current version of the CRAC code (and the CRAC2 code which is a modified ver-sion of CRAC) and is briefly outlined below.

The model assumes that people living within portions of a circular area with a specified radius (such as the 10-mile (16-km) plume exposure pathway Emergency Planning Zone (EPZ)), with the reactor at the center, would acuate if an accident should occur involving imminent or actual release significant quantities of radioactivity to the atmosphere.

Significant atmospheric releases of radioactivity would in general be preceded by one or more hours of warning time (postulated as the time interval between the awareness of impending core melt and the beginning of the release of radio-activity from the containment building)--although for some specific release categories the warning time could be less than an hour.

For the purpose of calculation of radiological exposure, the model assumes that those people who would potentially be under the radioactive cloud that would develop following the release would leave their residences after a specific amount of delay time

• and then evacuate.

The delay time is reckoned from the beginning of the warning time and is recognized as the sum of the time required by the reactor operators to notify the responsible authorities; the time required by the authorities to interpret the data, decide to evacuate, and direct the people to evacuate; and the time required for the people to mobilize and get underway.

i l

• Assumed to be of constant value which would be the same for all evacuees.

Limerick FES J-1 m

l P

s i

l l The model assumes that while leaving the area each evacuee would move radially out and in the downwind direction

• with an average effective speed ** (obtained by dividing the zone radius by the average time taken to clear the zone after the delay time) over a fixed distance ** from the evacuee's starting point.

The fixed distance used in the analysis discussed in Section 5.9.4.5(2) was selected to be 15 miles plume exposure p(athway EPZ radius).24 km) (which is 5 miles (8 km) more After reaching the end of the travel dis-tance, the evacuee is assumed to receive no further radiation exposure.

In a real evacuation, paths of evacuees would be dictated by the site road net-work.

However, each segment of actual trajectory of an evacuee would project a com-ponent in the downwind direction which, in the consequence model, is assumed to be radial.

Therefore, each evacuee's actual motion would have a component of motion along the radial downwind direction.

The evacuation model assumption that evacuees originating from areas that would come under the radioactive cloud would move radially out over a certain distance amounts to only an artifice for dose calculation:

as if the evacuee's radiological exposure is due to their component motion along the radial downwind direction (over a component path length which is assumed to be 15 miles).

The model incorporates a finite length uf the radioactive cloud in the downwind direction; this would be determined by the product of the duration over which the atmospheric release would take place and the average windspeed during the release.

It is assumed that the front and the back of the cloud formed would move with an equal speed, which would be the same as the prevailing windspeed; therefore, its length would remain constant.

At any time after the release, the concentration of radioactivity is assumed to be uniform over the length of the cloud.

If the delay time would be less than the warning time, then all evacuees would have a head start, i.e., the cloud would be trailing behind the i

evacuees initially.

On the other hand, if the delay time would be more than the warning time, then, depending on initial locations of the evacuees there are possibilities that (1) an evacuee would still have a head. start, (2) the t

cloud would already be overhead when an evacuee starts out to leave, or (3) an l

evacuee would be initially trailing behind the cloud.

However, this initial picture of cloud people disposition would change a^s the evacuees travel, depend-l ing on the relati,ve speeds and positions between the cloud and people.

It is possible that the cloud and an evacuee would overtake one another one or more j

times before the evaucee would reach his or her destination.

In the model, the radial position of an evacuating person, while stationary or in transit, is compared to the front and the back of the cloud as a function of time to deter-mine a period of exposure to airborne radionuclides.

The model calculates the time periods during which people are exposed to radionuclides on the ground while they are stationary and while they are evacuating.

Because radionuclides would be deposited continually from the cloud as it passed a given location, a person while under the cloud would be exposed to ground contamination less con-centrated than if the cload had completely passed.

To account for this reason-ably, the revised model assumes that persons are exposed to the total ground contamination when completely passed by the cloud; to one half the calculated concentration when they are anywhere under the cloud; and to no concentration when they are in front of the cloud.

• In the RSS consequence model and the CRAC and CRAC2 codes, the radioactive cloud is assumed to travel radially outward only.

• • Assumed to be a constant value for all evacuees.

Limerick FES J-2

~-

8 The model provides for use of different values of the shielding. protection factors for exposure from airborne radioactivity and contaminated ground for stationary and moving evacuees during delay and transit periods.

The model has the same provision for calculation of the economic cnst asso-ciated with implementation of evacuation as in the original RSS model.

For this purpose, the model assumes that for atmospheric' releases of durations 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> or less, all people living within a circular area of 5-mile (8-km) radius centered at the reactor, plus all people within a 90* angular sector within the plume exposure pathway EPZ and centered on the the downwind direction, will evacuate and temporarily relocate.

However, if the duration of release exceeds 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />, the cost of evacuation is based on the assumption that all people within that entire plume exposure pathway CPZ would evacuate and te pnrarily i

relocate.

For either of these situations, the cost of evacuation and reloca-tion is assumed to be $225 (1980 dollar) per person, which includes cost'of food and temporary sheltering for a period of 1 week.

J.2 Early Health Effects Model The medical advisers to the RSS (WASH-1400, Appendix IV, Section 9.2.2, and Appendix F) proposed three alternative dose-mortality relationships that can be used to estimate the number of early fatalities that might result in an exposed population.

These alternatives characterize different degrees of postexposure medical treatment from " minimal," to " supportive," to " heroic"; they. are more fully described in NUREG-0340.

There is uncertainty associated with both the mortality relationships (NUREG/CR-3185), and the availability and efficacy of different classes of medical treatment (Elliot, 1982).

Estimates of the early fatility risks using the dose-mortality relationship that is based upon the supportive treatment alternative are presented in the texts of Section 5.9.4.5.

This implies the availability of medical care facilities and services for those exposed in excess of 175 rems, the approximate level that the medical advisors to the RSS indicated would be indicative of the potential need for more than minimum services to reduce early fatality risks.

At the extreme low probability end of the spectrum (i.e., at the 1 chance in 100 million per reactor year i

level), the number of persons involved might exceed the capacity of facilities for such services, in which case the number of early fatalities might have been underestimated.

To gain perspective on this element of uncertainty, the staff has also performed calculations using the most pessimistic dose-mortality rela-tionship based upon WASH-1400 medical experts' estimated dose-mortality rela-tionship for minimal medical treatment and using identical assumptions regarding offsite emergency response as made in Section 5.9.4.5.

These results are also presented in Section 5.9.4.5.

The staff has also considered the uncertainties associated with the WASH-1400 dose-mortality relationship for minimal medical treatment and has concluded that early fatality risk estimates as bounded by the uncertainties discussed in Section 5.9.4.5(7) are reasonable.

This is because it is inconceivable that a major reactor accident at Limerick would not be fol-lowed by a mobilization of medical services, services which can be expected to reduce mortality risks to less than those indicated by the WASH-1400 description of minimal medical treatment.

J.3 References I

Elliot, D.A., Task 5 letter report from Dr. D. A. Elliot of Andrulis Research Corp. to Ms. A. Chu, NRC Project Officer, on Technical Assistance Contract No. NRC-03-82-128, December 13. 1982.

Limerick FES J-3 r

r, l

h P

Sandia Laboratories, "A Model of Public Evacuation for Atmospheric Radiological Releases," SAND-78-0092, June 1978.

r b

U.S. Nuclear Regulatory Commission, NUREG-75/014, " Reactor Safety Study--An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants,"

i

. formerly WASH-1400, October 1975.

'l f

--, NUREG-0340, Overview of the Reactor Safety Study Consequences Model'"

October 1977 l.,

--, NUREG/CR-3185, " Critical Review of the Reactor Safety Study Radiological Health Effects Model," Sandia Laboratories (also SAND-82-7081) March 1983.

.Il!

j l i

l l

i i

e*-

I i

P i

l e

jt I

f-i

)i Limerick FES J-4 e

,?

j.-

APPENDIX K CONDITIONAL MEAN VALUES OF ACCIDENT CONSEQUENCES The conditional mean values of potential societal consequences of several kinds from each release category in Table 5.11c are shown in Table K.1.

These means were calculated by the CRAC code and represent averages of each kind of con-sequence for each release category over the spectrum of the Limerick site meterological conditions.

" Conditional" mean values are so called because these mean values are conditional upon the occurrence of the accidents repre-sented by the release categories.

Probabilities of release categories have not been factored into these mean value estimates.

The conditional mean values are provided for a perspective only; they are devoid of much importance without simultaneous association of probabilities of the release categories to which the mean values are due. They are useful, however, in' judging the relative importance of different sequences.

Table K.1 is useful for risk calculations.

It can be used to calculate the risk of any particular kind of consequence (shown in the table) from any of the listed release categories by simply multiplying the conditional mean value of the given consequence by the probability per reactor year (Table 5.11d) of the release category to which the mean value is due.

It can also be used to cal-culate the risk of any particular kind of consequence from a group of release categories by calculating the sum of the products of the conditional mean values of the consequence and the probabilities of the respective release categories in the group; the group may include some or all of the release categories.

4 Limerick FES.

K-1

.~.

-.=

)

q-

==w

=== = = =-- -- = - - -

=

- =-

- :===~--=--=

~

~~ - ~. =-~

+

--- > = m =~

a o

i Table K.1 Conditional mean values of societal consequences from individual release categories for three alternative offsite emergency response modes Offsite

' I " * '8'#I

1 Consequence Emergency

?r Category Response Mode I-T/DW I-T/W 1-T/W I-T/SE" I-T/HB I-T/L6f II-T/W III-T/W III-T/HB III-T/LGT h

1. Early fatalities Evac-Reloc 0

0 0

2(2)"*

1(1) 5(-1) 0 0

1(1) 0' with supportive Early Reloc 1(0) 0 0

7(1) 1(1) 1(0) 2(2) 3(1) 1(1) 0 medical treatment Late Reloc 3(1) 5(-1) 5(-1) 1(2) 5(1) 2(3) 4(2) 2(2)-

2(-2)

(persons)

2. Population receiving Evac-Reloc 0

0 0

2(3) 4(2) 4(1) 5(2) 2(3) 4(2) 3(0) in excess of 200 Rees Early Reloc 1(1) 0 0

1(3) 3(2) 2(1)

'2(3) 2(3) 3(2)

O total marrow dose Lata Reloc 1(2) 3(0) 1(0) 1(3) 9(2) 5(3) 7(3)

.1(3) 5(0) from early exposure (persons)

3. Early injuries Evac-Reloc 4(1) 0 0

3(3) 5(2) 5(1).

2(3) 3(3) 4(2) 8(-1) 6(2) 3(3) 5(2) 5(0)

(persons)

Early Reloc 5(1) 1(-2) 2(-2) 3(3) 4(2) 4(1)

Late Reloc 2(2) 2(0) 1(0) 1(3) 6(2) 3(3) 6(3) 1(3) 9(0) b

4. Delayed cancer fatal-Evac-Reloc 6(2) 1(1) 4(1)

.6(3) 2(3) 1(3) 4(3) 4(3) 2(3) 2(1) ities (excluding Early Reloc 6(2) 3(1) 5(1) 6(3) 2(3) 1(3) 4(3) 4(3) 2(3) 3(1) thyroid) (persons)

Late Reloc 7(2) 3(1) 5(1) 2(3) 1(3) 4(3) 4(3) 2(3) 3(1)

5. Delayed thyroid Evac-Reloc 1(2) 2(1) 2(1) 8(2) 6(2) 2(2) 1(3) 9(2) 6(2) 1(1) cancer fatalities Early Reloc 1(2) 2(1) 2(1) 8(2) 6(2) 2(2) 1(3) 1(3) 6(2) 2(1)

(persons)

Late Reloc 2(2) 2(1) 2(1) 7(2) 2(2) 1(3) 1(3) 7(2) 2(1)

6. Total person-rees Evac-Reloc 1(7) 5(5) 8(5) 4(7) 2(7) 2(7) 6(7)

'6(7) 2(7) 4(5)

Early Reloc 1(7) 5(5) 9(5) 4(7) 2(7) 2(7) 6(7) 6(7) 2(7) 5(5)

Late Reloc 1(7) 5(5) 1(6) 2(7) 3(7) 7(7) 7(7) 3(7) 6(5)

7. Cost of offsite Evac-Reloc 3(8) 5(7) 6(7) 2(9) 1(9) 1(9) 4(9) 3(9) 1(9) 1(6) mitigation measures Early Reloc, 2(8) 2(6) 3(6) 2(9) 1(9) 1(9) 4(9) 3(9) 1(9) 1(6)

(1980 dollars)

Late Reloc 2(8) 2(6) 3(6) 1(9) 1(9) 4(9) 3(9)'

1(9) 1(6)

8. Land area for Evac-Reloc 1(6) 2(4) 3(4) 7(7) 2(7) 3(7) 1(8) 6(7) 2(7) 0 long-ters interdic-Early Reloc 1(6) 2(4) 3(4) 7(7) 2(7) 3(7) 1(8) 6(7) 2(7) 0 2

tion (n )

Late Reloc 1(6) 2(4) 3(4) 2(7) 3(7) 1(8) 6(7) 2(7) 0

• This release category has a probability less than 10

• per reactor year to be initiated by severe earthquakes; it is not analyzed with Late Reloc mode for its insignificant contribution to risks due to its low probability.

• • 2(2) = 2 x 10 = 200.

• • • These release categories are initiated by plant internal causes; therefore, the Late Reloc mode does not apply.

NOTE: Please see Section 5.9.4.5(7) for discussion of uncertainties. Estimated numbers were rounded to one significant digit only for the purpose of tMs table.

tes s )

w.

+ ~

r.

i Table K.1 (Continued)

Release Categories Offsite g.

Category Response Mode III-T/IET IV-T/DW IV-T/W IV-T/IE I-S/DWa** IV-A/DW*** IS-C/DW IS-C/DW S-H20/5E S-Ii20M 1

Consequence Emergency h

1. Early fatalities Evac-Reloc 6(-1) 6(2) 5(2) 6(2) 0 7(2) 3(2) 1(2) 0 0

with supportive Early Reloc 1(0) 1(3) 1(3) 1(3) 0 M3) 7(2) 7(2) 2(2) 6(2) 3(3) 3(3) 2(3) 3(3) medical treatment Late Reloc 7(1) 4(3) 4(3) 4(3)

(persons)

2. Population receiving Evac-Reloc 5(1) 5(3) 4(3) 4(3) 0 4(3) 2(3) 2(3) 4(2) 4(2) in excess of 200 Rees Early Reloc 3(1) 6(3) 5(3) 4(3) 5(-1) 5(3) 3(3) 3(3) 1(3) 2(3) total marrow dose Late Reloc 1(3) 1(4) 1(4) 1(4) 9(3) 9(3) 5(3) 8(3) from early exposure (persons)

3. Early injuries Evac-Reloc 6(1) 5(3) 4(3) 3(3) 0 3(3) 2(3) 2(3) 5(2) 6(2)

(persons)

Early Reloc 4(1) 5(3) 4(3) 4(3) 5(-1) 3(3) 3(3) 3(3) 2(3) 2(3) 6(3) 6(3) 3(3) 5(3)

Late Reloc 7(2) 7(3) 6(3) 7(3)

[

4. Delayed cancer fatal-Evac-Reloc 1(3) 5(3) 5(3) 5(3) 2(2).

5(3) 4(3) 4(3) 3(3) 4(3) ities (excluding Early Reloc 1(3) 5(3) 5(3)'

5(3) 2(2) 5(3) 4(3) 4(3) 3(3) 4(3) 4(3) 4(3) 3(3) 4(3) thyroid) (persons)

Late Reloc 1(3) 6(3) 6(3) 6(3)

5. Delayed thyroid Evac-Reloc 2(2) 2(3) 2(3) 2(3) 3(1) 2(3) 9(2) 9(2) 7(2) 1(3) cancer fatalities Early Reloc 2(2) 2(3) 2(3)

- 2(3) 3(1) 2(3) 9(2) 1(3) 8(2) 1(3) 1(3) 1(3)-

8(2) 1(3)

(persons)

Late Reloc 2(2) 2(3)-

2(3) 2(3)

6. Total person-ress Evac 9eloc 2(7) 8(7) 7(7) 6(7) 3(6) 8(7) 5(7) 5(7) 4(7) 6(7)

Early Reloc 2(7) 8(7) 8(7) 8(7) 3(6) 8(7) 5(7) 5(7) 5(7) 6(7) 6(7) 6(7) 5(7) 7(7)-

Late Reloc 3(7) 9(7) 8(7) 9(8)

7. Cost of offsite Evac-Reloc 1(9) 5(9) 5(9) 5(9) 9(7) 5(9) 2(9) 2(9) 2(9) 3(9) mitigation measures Early Reloc 1(9) 5(9) 5(9) 5(9) 4(7) 5(9) 2(9) 2(9) 2(9) 3(9) 2(9) 2(9)

.2(9) 3(9)

(1980 dollars)

Late Reloc 1(9) 5(9) 5(9) 5(9)

8. Land area for Evac-Reloc 3(7) 1(8) 1(8) 2(8) 3(5) 1(8) 5(7) 6(7) 5(7) 8(7) long-term interdic-Early Reloc 3(7) 1(8) 1(8) 2(8) 3(5) 1(8) 5(7) 6(7) 5(7) 8(7) 5(7) 6(7) 5(7) 8(7) tion (m )

Late Reloc 3(7) 1(8) 1(8) 2)8) 2 e

T

f

,e APPENDIX L CONSEQUENCES AND RISKS OF RELEASE CATEGORIES INITIATED BY SEVERE EARTHQUAK AND THOSE OF RELEASE CATEGORIES INITIATED BY OTHER CAUSES Probability distributions of accident consequences and probability-weighted values of these consequences (i.e., risks) are presented and dis::ussed in Sections 5.9.4.5(3), 5.9.4.5(4), and 5.9.4.5(6).

The results presented in those sections were the combined results from release categories initiated by. internal causes, firer and low to moderately severe earthquakes, and from release cate-gories initiated by severe earthquakes.

The severe earthquake initiated release categories were analyzed with the assumption of late relocation (Late Reloc) mode of offsite emergency response (see Section 5.9.4.5(2) and Table 5.11f).

Release categories initiated by causes other than severe earthquakes '?ere ana-lyzed with the assumption of. evacuation and relocation (Evac-Reloc) mode of-offsite emergency response (see Section 5.9.4.5(2) and Table 5.11f).

A separate display of radiological contributions to the overall results (presented in sec-tions cited above) from release categories initiated by severe earthquakes and from release categories-initiated by causes other than severe earthquakes is provided here.

Additionally, breakdowns of societal consequences of early fatalities and latent cancer fatalities in terms of contributions from spatial 1

intervals up to 50 miles (80 km) from the Limerick reactors are also presented.

Figures L.1 through L.20 display the breakdowns of each of the graphical plots presented in Figures 5.4b throug'n 5.41 in the sections cited above~into two components--one ascribed to the. severe earthquakes and the other ascribed to j

the other causes.

In Figures L.1 through L.20, the graphical plots of Figures 5.4b through 5.41 are reproduced for easy reference.

i i

Tables L.la and b provide a breakdown of each category of risk shown in Table 5.11h into the two components as stated above.

From these tables it is apparent that the release categories initiated by severe earthquakes are the dominant contributors to the risk of early fatality'(with supportive or minimal medical treatment).

These release categories contribute almost equally as the release categories initiated by other causes to the risk of early injury.

How-ever, the release categories initiated by causes other than severe earthquakes 4

are the dominant contributors to the other types of risk in Tables L.la and b.

1 Table L.2 shows the contributions to the risk of early fatality with suppor-tive medical treatment from the spatial intervals within 50 miles (80 km) of the plant.

Contributions from each spatial interval is also broken down into component contributions ascribed to severe earthquakes and the other causes.

Table L.3 shows similar results for arly fatality as in Table L.2, but with minimal medical treatment.

Table L.4 shows the risk of. latent cancer fatality in similar fashion as in i

Table L.2 for early fatality.

Latent cancer fatality risks shown in Table L.4 include risks of both thyroid and nonthyroid cancer fatalities.

Limerick FES L-1 i

,,.m

,-.- - m D 1d 16 1d

...... 1,6 10 10

..,d Idt 1

g g

=

o k

D = LATE RELOC o = EVAC-RELOC x

r47 y

w3 a = EVAC-RELOC + LATE RELOC 7 0

0 vi o

3e n

+

01 ~

o_

"E To L

in N

o g"'E Io

~

m E

EW 5

f N

r;To_

k i

Jo w:

w O

r-i e

g-to o

u ri 6

_o k _5 i~

p 3

bi _

o y el t

t

- o i

w m

l l

9 4

ho_

T u 1d

........,d

........ 8 16 10'

....... ;6 1d' 16-

_o m-1 1

1 X= NUMBER OF AFFECTED PERSONS Figure' L.1 ' Probability distribution of population with whole body dose greater than or equal to 25 rems NOTE:, Please see Section 5.9.4.5(7) for a discussion

'of uncertainty.

16......16 16 16 i d'

,,.....i d 1,6

,....... t id o

q g:

a = LATE! RELOC

p g

o = EVAC-RELOC x

N7 A = EVAC-RELOC + LATE RELOC 7

c 0

M

~5 E~

m a

y g

St ~

~L o-w5 i*

Og b

m -5 i-m m

4:

m i

i O

r-E g.

O-lO p WE 1

EH

& _E i

p

-b sb-

\\

ywg Q:

S,o

... \\m..

.b

.......,I 8

Id 18 18 10' id 16 16-1 A

X= NUMBER OF AFFECTED PERSONS Figure' L.2 Probability distribution of population with thyroid dose Jreater than or equal to 300 rems NOTE:

Please see Section 5.9.4.5(7) for a discussion of uncertainty.

i

~

>^Qi1.d id 1d 10'

..... 0'

........d

..d 10b 1

1 I

o

....... _ o 2

o o = LATE RELOC 3.

o =k EVAC-RELOC 9-g7a a = EVAC-RELOC + LATE RELOC 70.'

n E

=

a iw m

g Mt

~

% o-.

't x

o m

o n Ri ~!

N i

o m

2 I:

I

~

r; o_

\\7 8!

S T

$7 o

t-o-

\\

L di o

E!

EW H'

3

_; o-T o

i p

4

$'o m 1d 1d 1d

,,,,,,,,,d

,.......,d

....... ;6 1d

.......,6-

-o t1, 1

1 1

1 X= NUMBER OF AFFECTED PERSONS

. Figure L.3 Probability distribution of population with total bone marrow dose greater than or equal to 200 rems NOTE:

Please see Section S.9.4.5(7) for a discussion of uncertainty.

T 1 6,

.......d........ d 1d 16..... 1,d.......u. f......1.&......1,di 1

i o

.o c

LEGEND Er o = LATE RELOC 4

o = EVAC-RELOC

~

g a}

a = EVAC-RELOC + LATE RELOC ~7 7

a 4\\"!

5#

n0

E y

M'

• o Zo i

@w:E, Tw i

m N

N ME

_\\

_t.

To._

o o

EW I.

C'o -

.. __a ue mi im A :

p.,

E-*I Z o_

k

-Y

_o dE

e. e EH m

m Sf.

{

i o.

\\

_To m WE Ew To_

to w

Id 10*

Id,,......Id

.16,,,,,,,,,d 16 10P,,,,,,,, f I

lo X= TOTAL PERSON REM WHOLE BODY Figure L.4 Probability distribution of population exposure within 50 miles (80 km)

NOTE:

Please see Section 5.9.4.5(7) for a discussion of u'ncertainty.

b

'......10' id Id 16......,,d

........d.........f idi 10 i

i i

o R

LEGEND

- 5l o = LATE RELOC o = EVAC-RELOC g?a_

a = EVAC-RELOC + LATE RELOC 7a m

ANw 5 ia m

x m

m I

-T o

o x

o

^ % __

A!

D

'o

'.o N ME Ed O

b:

Yo.

To v a5 Ea 9.

r;-

g m

$g.

o o

e-o

• E i*

m A

0; O __

a -5 o

s-

'O.

o

........i 1,6 1d 1d 1d 1d........,d

.......,6 kW Ilr

........,,.4 I

1 X= TOTAL PERSON REM WHOLE BODY Figure L.5 Probability distribution of population exposure, entire region NOTE:

Please see Section 5.9.4.5(7) for a discussion of uncertainty.

1

........'........!10' 10'

,........d........6 dE

%T 16 1d 10 i

1 o

o LEGEND i;-

m i;

m 0 = LATE RELOC p

o = EVAC-RELOC 37 A = EVAC-RELOC + LATE RELOC 7 w

a 0

M b< ~ :E En m

4 N

ho o

w.

Z

=~

o o

]o Ho Z WE EH m

g I.

T

(

T

$g~1 r-4 h.

4Yo l

'o h WE id A.

H

)

D'o.

To

.-l. H :

e m

f m

(

ot t

cr:a,6 o

.......i' a'

1 1d 16 10 16 16 16 16~

X= LATENT CANCER FATALITIES Figure L.6 Probability distribution of latent cancer fatalities, excluding thyroid within 50 miles (80 km)

NOTE:

Please see Section 5.9.4.5(7) for a discussion of uncertainty.

_.. -, u: mm.. __._ m. _..... - 2=...e m

xy 18 Id 18 Id 10'

.......,d

........d i

I 1

E go, 4

i LEGEND o = LATE RELOC E

p o = EVAC-RELOC

[

37 b< w ::

a = EVAC-RELOC + LATE RELOC 7 a_

m 0

'^

.-i d.

i kI.

O*p.

$o o.

Z L) 8o Z WE

_o ra iH

.g I..

4 "Tg T

g O

g r-A N :

T o-Y

,h "i

_o E

A iM 3

N'o T

1w:

_o

w m

A m

6 OT o

T (4 w 8

......i y

O A 1

.......,d 18 16

.......10' 16 16 16 I

X= LATENT CANCER FATALITIES Figure'L.7 Probability distribution of latent thyroid cancer fatalities, 50 miles (80 km)

NOTE:

Please see Section 5.9.4.5(7) for a discussion of uncertainty.

i

id Id 16 10' 10'

.......d

........d 10}

i i

xb.

.......:_o

......a 3-h"!

LEGEND i"

2.

m o = LATEi RELOC a

p o = EVAC-RELOC a = EVAC-RELOC + LATE RELOC T

37 g;

o g

E<-

3-l d

m

-t

_o N

g o

r R

d.

FYO

.O z; e E i

m 5*

E-I.

T T

o o

x ~1 o

E E

.q

=

"b

_b.4 4 wg 3

3. E E

s M C_

t

_ O. 4 1W=

m b

m

-T OTg o

m,d

% I Id 1d 1d '

Id

.......,6 1d 16~

1 X= LATENT CANCER FATALITIES Figure L.8 Probability distribution of latent cancer fatalities, excluding thyroid, entire population NOTE:

PleaselseeSection5.9.4.S(7)foradiscussion~ofuncertainty.

4 an

yr Id 10f Id 10' 10'

........d

.g.,d Id I

I

.......r 4

-g U

$i o = LATE RELOC 1

p

~-

R-37

~

o = EVAC-RELOC g

A = EVAC-RELOC + LATE RELOC 70 n

4e=

0 F

E iw

~

$T -

-I o

o o* r m

I" Z

s'o

_'o x ei rq 5*

I:

E<-.

MC

'-3 m

7

~T 0

g

o o

i 7

x-

"T b ~_!

_'o o

!~

A :

~

N

~

D'o

,_1 w.

.'o g --

m tu 4

gn O'

gx; o

_o a,d

.......d

.......d

.......d 1d 16

......,d

.......,6

........i-a 1

1 I

I I

1 X= LATENT CANCER FATALITIES Figure L.9 Probability distribution of latent thyroid cancer fatality, entire population NOTE:

Please see Section 5.9.4.5(7) for a discussion of uncertainty.

am a

w wa a x s an

id

......id id.......I d 16

,......i.d

.......Id

.......Idr b

........e

.o C

LEGEND k

g a = LATE RELOC gY :

h o = EVAC-RELOC x

a = EVAC-RELOC + LATE RELOC L o

A

" :E i~

m E-a:

I.

.T o-

- - - 5W g

• 5 b:

ho bo N

_E 2*

i N

E lN

~

xo F,,__,,,

,,, _,, _,, ja t c m

p r-ho

$o"!

h 1

i*

p p

ko o<

grl w :

g O

E i

(2; g

.g tg 8

Id 16 10 10' 16 16 16-1 X=EARLY FATALITIES Figure L.10 Probability distribution of early fatality with supportive medical treatment NOTE:

Please see Section 5.9.4.5(7) for a discussion of uncertainty.

W w

= ~ -

=--.

7 16 id 18 16

......., '........ d 1,6........ r -

10 i

Id g

o 4

x a = L' ATE RELOC R

A ~-

o = EVAC-RELOC

[

gi a

A = EVAC-RELOC + LATE RELOC 7

=

9 g

3 i

hT ~

da~1

~flg

~

A>

m-

< 'o

[

'_o O ' tt

i w

• fo.

M u-!

N o

t-4 L

a) i ~

go.

(

\\

lb s-!

3 i

!~

m'o -

A (Q e E Yo O

N E

• 4 3

'o d........,d Id

........ d 1d 16 1d 16

........i

-o 1

I I

........i, X=EARLY FATALITIES Figure L.11 Probability distribution of early fatalities with minimal medical treatment NOTE:

Plea.se~see Section 5.9.4.5(7)'for a discussion of uncertainty.

~

)

1 Id

..1d Id Id 10'

.......d

.,,d

......6 1

I -

1 7

g.

g n

a = LATE RELOC M

[

47- '

o = EVAC-RELOC

~

a A = EVAC-RELOC + LATE RELOC '7 m

- < =

0

=m 5

b.

@T

.T

,:z,; o_

r

-=

o

~

=w b

$b-%

~

\\

1 N*E N

i*

m v.

O

~

g'O.

M Y.o n *!

p e

e 0

5%. -

(

p 'o W

-1 E-.*!

g o

m

<T o

N To

\\

f as m

'o.

.'o m,

e 16 Id 1d Id 10' 16 Id 16 X=EARLY INJURIES Figure L.12 Probability distribution of'early injuries NOTE:

PleaselseeSection5.9.4.5(7)foradiscussionofuncertainty.

9 2

m i 10' 10' 16 Id 16 1d........ d 1W 10Y 1

r-o'

......!.i o

y LEGEND a = LATE RELOC

~

o = EVAC-RELOC n

T

^ = EVAC-RELOC + LATE RIjLOC 7 o

0 m

e.:

w

e g

A\\

- s'

&Io.

1. o o" D E*

U x

N i

O

~

s T

o h.-t :

e

_ 'o s.,

N

W s.

v.

T o

(

_c

• E E*

'T 3

5 5

w

~

4

%Y

.e.

Ao

'o m -E

\\

5-o

)

54 A

'o -

o

.T eE Ea T.

T o_ '

o

.....m

.......i

.......n

.......y 10 IW

.......,6 Id 16 Id 1d SW 10^

_,g 1

l X= TOTAL COST IN 1980 DOLLARS Figure L.13 < Probability distribution of mitigation measures cost NOTE:

Please see Section 5.9.4.5(7) for a discussion of uncertainty.

l l

1 10' 10*

1d 1d 16...... !..d Id........d*'

Idj I

1 i.

'o

.......a o

T LEGEND o = L ATE : RELOC o = EVAC-RELOC T

a = EVAC-RELOC + LATE RELOC 7 a

_0

?

-s 5-

. m x

Al :

AT T

N o?

o p; w:

2

w A

~

.Y

'o o

hws EH 4

hu Af ~

N k

'T o

a M-5 id b.

'T

,-t 0

Et -

~t

<o o

m "!

o a:

A o_

o

)

'8

'o _-

T.o w

10' Ad 16 Id 16........,d

........ ;d p Id Id*

I 1

X= AREA IN SQUARE METERS Figure L.14 Probability distribution of land area interdiction q

NOTE:

Please see Section 5.9.4.5(7) for a discussion of uncertainty.

\\

e

^ '., ' '2 a

4 n

10.i w

g y

y y

y g

i i

g y

~

m

~

LEGEND m

D = EVAC RELOC 10-2 =.

O = WE RELOC

=

x A = EVAC RELOC+ LATE RELOC 5

O s

Q w

6 10'3 E-7

,s:

E I

g o

4 O 10 E-o m

o b

.g O

5 I 10 g-5 3

3 t

10-8 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0

'50.0~

DISTANCE (MILE) l Figure L.15 Individual risk of downwind dose versus distance 1

i NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties

C R

i 3.

nx m.

O v

LEGEMI o - EVAC RELOC o - LilTE RELOC r

a - EVAC RELOC

• LATC RELOC

~

23 i

5 2

i, n

i g

i.

7 w

o g

e:,

n

~i l

=

i u

i

a' E

)a o.

's

~:

E 2:

T*

e.s s'.e 35.s is.s S.e S.e S.e S.e S.e d.e sa.e DISTiteCC tillLCl Figure L.16 Individual risk of early fatality with supportive medical treatment versus distance NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties s

s e

4 e.

.y.

x

's

- ?)

C H

a-s g-W a

2 m

LEGOg3 a-tync RCLoc q,

e - LHTC RCLOC

~

p*.

s a - EvHC NELDC

• LATC art oc

~<

j M

g

\\

T<

i o-1 1

M

~

7

'c.

4

-^g

~.3 ce 5

e, 5Q'

~ s o

~

~l i

g ~

s a

,o.e 1

E 5

3 5

y s

a.s 5.e so.e d.e as.e 25.e 3s.e 35.s to.e es.g so.a 3

.~

DISTfWCC It1lLCl Figure L.17 Individual risk of early fatality with minimal medical treatment versus distance NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties i

i

.I

?

e B

CD 3.

n 7

M 8 O m

i m

M:

5 5

I I

5 5

5 5

5 LEGEM o - EVAC RELOC o - LnTC RELOC e :

A - EVOC HELOC

• IJ1TC RELOC

~

'Si i

k

\\.

~

o

-1 i

3 5

na

~

'O.

d b

8 g

A

-3 1

i

=

9~

~

4 so, g

y 3

3 y

3 5

3 n 3 e.s 5.s 10.0 15.s 20.e 25.e M8 35.8 to.e d.e so.e DISTANCC INILCl

~

Figure L.18 Individual risk of early injury versus distance NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties s

e

,\\

o i

E C

I

{

x 4

M so re:

5 5

5 4

E y

t 3

a y

3 3

-[

LEGD0 i

i n - CVOC RELOC

(

9 o - LOTC RELOC b.

A - EVHC RELOC

• LATC RELOC

~

f

~i s

M i

g i

i e

8*e*

\\

b. i ac 2

)

n oc 7

'$2

=

e 2

9 o

s

?o_

~

r i

.t i

?

a,o, ;

i s.s s.s so.e is.s Jo.o 25.s 3s.a 35.s to.

d.e DISTANCC JHILCl sn.a Figure L.19 Individual risk of latent cancer fatality (excluding thyroid) versus distance NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties i

.. _m,._

C 0

3.

n w

8O-5 i

LEGEM i

=

a-EVfC RELOC o - LilTE RELOC 4 - EvitC RELOC

• UITC RELOC

~

p' o_:

3 flE

.a

. t

~

e.

' Si.

1 E

y, g

e

'21 r-i m

o E

M Mt

.o,

'o_

".~

i i

s.e s'.e le.e is.e as.e 25.s so.e 35.s so.e es.e me u

5 5

3 3

8 8

8

~

DISTilNCC lillLCl Figuro L.20 Individual risk of latent thyroid cancer fatality versus distance NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties i

a e

1,8 Ttblo L.la Societal risks within 50 miles (80 km) of Limerick i-site with Evac-Reloc* and Late Reloc* offsite emergency response modes Risk per reactor year From causes other than From severe Consequence type severe earthquakes earthquakes (Evac-Reloc)

(Late Reloc)

Total 1.

Early fatalities with 2(-4)**

5(-3) 5(-3) supportive medical treatment (persons) 2.

Early fatalities with 7(-4) 8(-3) 8(-3) minimal medical treatment

- -(persons) 3.

Early injuries (persons) 1(-2) 1(-2) 2(-2) 4.

Latent cancer fatalities 4(-3) 7(-3) 4(-2)

(excludin (persons)g thyroid) 5.

Latent thyroid cancer 9(-3) 2(-3) 1(-2) fatalities (persons) 6.

Total person-rems 6(2) 9(1) 7(2) 7.

Cost of offsite 4(4) 5(3) 5(4) mitigation measures (1980 dollars) 8.

Land area for long-term 1(3) 1(2) 1(3) interdiction (square meters)

• See Section 5.9.4.5(2).

• • 2(-4) = 2 x 10 4 =.0002 NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.Estimated numbers were rounded to one significant digit only for the purpose of 4

this table.

Limerick FES L-22 w

,e -

Table L.lb Societal risks within the entire region of Limerick site with Evac-Reloc*.and Late Reloc* offsite emergency response modes Risk per reactor year From causes other than From severe Consequence severe earthquakes earthquakes type (Evac-Reloc)

(Late Reloc)

Total 1.

Early fatalities with 2(-4)**

5(-3) 5(-3) supportive medical treatment (persons) 2.

Early fatalities with 7(-4) 8(-3) 8(-3) minimal medical treatment (persons) 3.

Early injuries (persons) 1(-2) 1(-2) 2(-2) 4.

Latent cancer fatalities 6(-2) 1(-2) 7(-2)

(excluding thyroid)

(persons)

~'

5.

Latent thyroid cancer 1(-2) 2(-3) 1(-2) fatalities (persons) 6.

Total person-rems 1(3) 1(2) 1(3) 7.

Cost of offsite 5(4) 6(3) 5(4) mitigation measures (1980 dollars) 8.

Land area for long-term 1(3) 2(2) 1(3) interdiction (square meters)

• See Section 5.9.4.5(2).

• • 2(-4) = 2 x 10 4 =.0002 NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

Estimated numbers were rounded to one significant digit only for the purpose of this table.

l Limerick FES L-23

P.

!, ie K

Table L.2 Contributions to risk of early fatality with I

supportive medical treatment from spatial intervals within 50 miles (80 km) of Limerick site with Evac-Reloc* and Late Reloc* offsite emergency response modes g

Risk per reactor year From causes other than '

From servere severe earthquakes earthquakes Spatial interval (Evac-Reloc)

(Late Reloc)

Total from (mi) - to (mi)t (persons)

(persons)

(persons)

L 0.0 - 0.5**

2(-5)***

4(-5) 6(-5) 0.5 - 1.0 1(-5) 6(-5) 8(-5) 1.0 - 1.5****

4(-5) 3(-4) 3(-4)

I(

1.5 - 2.0 4(-5) 3(-4) 4(-4) 2.0 - 2.5 4(-5) 4(-4) 4(-4) 2.5 - 3.0 2(-5) 3(-4) 4(-4) 3.0 - 3.5 3(-5) 6(-4) 6(-4) 3.5 - 4.0 2(-5) 5(-4) 6(-4) 4.0 - 4.5

- 6(-6) 3(-4) 3(-4) 4.5 - 5.0 2(-6) 3(-4) 3(-4) 5.0 - 6.0 9(-7) 3(-4) 3(-4) 6.0 - 7.0 4(-7) 2(-4) 2(-4) 7.0 - 8.5 1(-6) 3(-4) 3(-4) 8.5 - 10.0 6(-7) 2(-4) 2(-4) 10.0 - 12.5 2(-6) 3(-4) 3(-4) 12.5 - 15.0 2(-8) 2(-6) 2(-6) 15.0 - 17.5 3(-8) 5(-8) 8(-8) 17.5 - 20.0 4(-8) 0 4(-8) 20.0 - 25.0 0

0 0

25.0 - 30.0 0

7(-7) 7(-7) 30.0 - 35.0 0

0 0

35.0 - 40.0 0

0 0

40.0 - 45.0 0

0 0

45.0 - 50.0 0

0 0

Total 2(-4) 5(-3) 5(-3) tTo change miles to km, multiply the values shown by 1.609.

• See Section 5.9.4.5(2).

• • This circular zone includes the Site Exclusion Area.

• • • 2(-5) = 2 x 10 5 =.00002

• • • • 93% of the area of this annulus is included within an annulus 1 mile wide outside of the site exclusion area boundary.

NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

Estimated numbers were rounded to one significant digit only for the purpose of this table.

s Limerick FES L-24 J

r Table L.3 Contributions to risk of early fatality with minimal medical treatment from spatial i

intervals within 50 miles (80 km) of Limerick and Late Reloc* offsite a

site with Evac-Reloc emergency response modes

.L -

Risk per reactor year From causes other than From servere severe earthquakes earthquakes Spatial interval (Evac-Reloc)

(Late Reloc)

Total from (mi) - to (mi)t (persons)

(persons)

(persons) 0.0 - 0.5**

5(-5)***

4(-5) 1(-4) 0.5 - 1.0 4(-5) 7(-5) 1(-4) 1.0 - 1.5****

8(-5) 3(-4) 4(-4) 1.5 - 2.0 6(-5) 4(-4) 5(-4) 2.0 - 2.5 7_(-5) 5(-4) 6(-4) 2.5 - 3.0 5(-5) 4(-4) 5(-4) 3.0 - 3.5 6(-5) 8(-4) 8(-4)

^

3.5 - 4.0

.5(-5) 7(-4) 8(-4) 4.0 - 4.5 2(-5) 4(-4) 4(-4) 4.5 - 5.0 2(-5) 4(-4) 4(-4) 5.0 - 6.0 1(-5) 4(-4) 5(-4) 6.0 - 7.0 3(-6) 4(-4) 4(-4) 1 7.0 - 8.5 3(-6) 5(-4) 5(-4) 8.5 - 10.0 7(-7) 5(-4) 5(-4) 10.0 - 12.5 9(-5) 1(-3) 1(-3) 12.5 - 15.0 9(-6}

1(-4) 1(-4) 15.0 --17.5 1(-5) 5(-5)-

7(-5) 17.5 - 20.0 1(-5) 2(-5) 3(-5) 20.0 - 25.0 1(-5) 1(-5) 3(-5) 25.0 - 30.0 2(-5) 2(-4) 2(-4) 2(-5) 30.0 - 35.0 1(-5) 9(-6)

~1(-6) 35.0 - 40.0 7(-8) 1(-6) 40.0 - 45.0 3(-8) 8(-7).

9(-7) 45.0 - 50.0 3(-6) 3(-7) 3(-6)

~

Total 7(-4) 8(-3) 8(-3) tTo change miles to km, multiply the values shown by 1.609.

• See Section 5.9.4.5(2).

• • This circular zone includes the Site Exclusion Area.

• • • 5(-5) = 5 x 10 5 =.00005

• • • • 93% of the area of this annulus is included within an annulus 1-mile wide outside of the site exclusion area boundary.

i NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

Estimated numbers were rounded to one significant digit,only for the purpose of this table, a

i Limerick FES L-25

f' o !;

Table L.4

'l' Contributions to risk of latent cancer (including J-!

thyroid) fatality from spatial intervals within 7

50 miles (80 km) of Limerick site with Evac-Reloc*

Il ;

and Late Reloc* offsite emergency response modes f

Risk per reactor year From causes other than From severe

!j severe earthquakes earthquakes Spatial interval (Evac-Reloc)

(Late Reloc)

Total l

from (mi) - to (mi)t (persons)

(persons)

(persons) 0.0 - 0.5**

3(-5)***

1(-6) 3(-5)

O.5 - 1.0-

1. 0 - 1. 5 * * * *..

.5(-5) 3(-6) 5(-5) 2(-4) 2(-5) 2(-4)

.1.5 - 2.0 2(-4) 3(-5) 2(-4) 4 i

2.0 - 2.5 2(-4)

~4(-5) 3(-4) 2.5 --3.0 2(-4) 4(-5) 2(-4) 3.0 - 3.5 3(-4) 9(-5).

4(-4) 3.5 - 4.0 3(-4) 9(-5) 4(-4) 1g 4.0 - 4.5 1(-4) 5(-5) 2(-4) 4.5 - 5.0 1(-4) 5(-5) 2(-4) 5.0 - 6.0 2(-4) 8(-5) 2(-4) q 6.0 - 7.0 2(-4) 8(-5) 2(-4) 7.0 - 8.5 2(-4) 1(-4) 4(-4) 8.5 - 10.0 2(-4) 1(-4) 4(-4) 10.0 - 12.5 3(-3) 8(-4) 4(-3) 12.5 - 15.0 1(-3) 2(-4) 1(-3) 2 15.0 - 17.5 2(-3) 4(-4) 3(-3) 17.5 - 20.0 2(-3) 4(-4) 2(-3) 20.0 - 25.0 7(-3) 1(-3) 8(-3) 25.0 - 30.0 1(-2) 2(-3) 2(-2) 30.0 - 35.0 6(-3) 1(-3) 7(-3) 35.0 - 40.0 5(-3) 8(-4) 6(-3) 40.0 - 45.0 2(-3) 3(-4) 2(-3) 45.0 - 50.0 2(-3) 3 (-2,.)

2(-3)

Total 5(-2) 9(-3) 5(-2) tTo change miles to km, multiply the values shown by 1.609.

• See Section 5.9.4.5(2).

• • This circular zone includes the Site Exclusion Area.

• • • 3(-5) = 3 x 10 5 =.00003

• • • • 93%oftheareaofthisannulusisincludedwithinanannu1bs1 mile wide outside of the site exclusion area boundary.

NOTE:

Please'see Section 5.9.4.5(7) for discussion of uncertainties.

Estimated numbers were rounded to one significant digit only for the purpose of this table.

l l

Limerick'FES L-26

!.}

1.i -

^

e

L

! ~

~

APPENDIX M AN ALTERNATIVE EVALUATION OF THE RELEASE CATEGORIES INITIATED BY CAUSES OTHER THAN SEVERE EARTHQUAKES 4

The results presented in Sections 5.9.4.5(3), 5.9.4.5(4), and 5.9.4.5(6) and in Appendix L include contributions from the release categories. initiated by severe earthquakes, and fron the release categories initiated.by internal causes, fires, and low to moderately severe earthquakes..The release categories not initiated by severe earthquakes were analyzed with the assumption of Evac-Reloc offsite emergency response mode-(see Section 5.9.4.5(2) and Table 5.11f).

To provide a reasonable bound to the role of evacuation in risk estimates from the latter release categories, as well as to display sensitivity of risks fron these release categories with respect to pertubations in evacuation, an analysis'of the,se release categories was made assuming the Early Reloc mode of offsite emergency response described in Section 5.9.4.5(2).

The results of this analy-sis are provided in this appendix.

Only the probability-weighted societal con-sequences (i.e., the societal risks) resulting from this alternative evaluation

~

i are presented below.

l Tables M.la and b are similar to Tables L.la and b, respectively,- in Appendix L.

The numbers in the second columns of Tables M.la and b are the estimates of risks of various kinds from the release categories initiated by causes other than severe earthquakes evaluated with the Early Reloc mode of offsite emer-The numbers in the third columns are reproduced from the third gency response.

columns of Tables L.la and b and are the estimates of risks ascribed to the severe earthquake-induced release categories as before. The numbers in the fourth columns represent alternative estimates of,overall risks (for comparison with those shown in Table 5.11h) from release categories initiated by all causes, and are the sums of the numbers in the preceding columns for each risk type.

Number in parentheses in Tables M.la and b below the entry for each type of risk (health effects and population exposure only) is the ratio of the-risk

, estimate in these tables and the corresponding risk estimate in Tables L.la.and b.

This ratio is indicative of the sensitivity of each type of risk to the choice between the Evac-Reloc and Early Reloc modes of offsite emergency response for the release categories initiated by causes other than severe earthquakes.

From inspection of the ratios (see above), it is apparent that the risk of early fatality (with supportive or minimal medical treatment) is most sensitive to the choice of emergency response mode.

The risk of early fatality is about 3 to 4 times as large for the Early Reloc mode as that for the Evac-Reloc mode for release categories not initiated by severe earthquakes.

However, because r

the risk of early fatality is dominated by the release categories initiated by severe earthquakes, the overall risk of early fatality with supportive or mini-mal medical treatment is only about 20% higher for the choice of the Early l

Reloc over the Evac-Reloc mode.

The other types of risks in

' Limerick FES M-1

l-

~

i Tables M.la and b are less sensitive to the choice between the Early Reloc i

and Evac-Reloc modes.

i,'!;

' Tables M.2, M.3, and M.4, respectively, display the contributions to' the risks of early fatality with supportive medical treatment and with minimal medical treat

• ment, and latent cancer (including thyroid). fatality from the spatial

.,1, intervals within 50 miles (80 km) of the plant.

_1 a

?

u

p

s I

~

+

+

.5

',ll

'd

.i

)

.i n

.o p

r Limerick FES M,

e

l i e.

.t'

.c Table M.la Societal risks within 50 miles (80-km) of Limerick site with Early Reloc* and Late Reloc* offsite emergency response modes-Risk per reactor year From causes other than

'From severe-Consequence severe _ earthquakes earthquakes type (Early Reloc)

(Late Reloc)

Total 1.

Early fatalities with 1(-3)**

5(-3) 6(-3) supportive medical (4)

(1) treatment (persons) 2.

Early fatalities with 2(-3) 8(-3) 1(-2) minimal medical treatment (3)

(1)

(persons) 3.

Early injuries (persons) 1(-2) 1(-2) 2(-2)

(1)

(1) 4.

Latent cancer fatalities, 4(-2) 7(-3) 4(-2) excluding thyroid (1)

(1)

(persons) 5.

Latent thyroid cancer 1(-2) 2(-3) 1(-2) fatalities (persons)

(1)

(1) 6.

Total person-rems 6(2) 9(1) 7(2)

(1)

(1) 7.

C'ost of offsite 4(4) 5(3) 4(4) mitigation measures (1980 dollars) 8.

Land area for long-term 1(3) 1(2) 1(3) interdiction

.(square meters)

• See Section 5.9.4.5(2).

• • 1(-3) = 1 x 10 8 =.001 NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

Estimated numbers were rounded to one significant digit only for the purpose of this table.

Limerick FES M-3 r.

Table M.lb Societal risks within the entire region of Limerick site with Early Reloc* and Late Reloc* offsite emergency response modes Risk per reactor year From causes other than From severe Consequence type severe earthquakes earthquakes (Early Reloc)

(Late Reloc)

Total 1.

Early fatalities with 1(-3)**

5(-3) 6(-3) supportive medical (4) treatment (persons)

(1) 2.

Early fatalities with 2(-3) 8(-3) 1(-2) minimal medical treatment (3)

(persons)

(1) 3.

Early injuries (persons) 1(-2) 1(-2) 2(-2)

(1)

(1) 4.

Latent cancer fatalities, 6(-2) 1(-2) 7(-2) excluding thyroid (1)

(persons)

(1) 5.

Latent thyroid cancer 1(-2) 2(-3) 2(-2) fatalities (persons)

(1)

(1) 6.

Total person rems 1(3) 1(2) 1(3)

(1)

(1) 7.

Cost of offsite 5(4) 6(3) 5(4) mitigation measures (1980 dol'.rs) 8.

Land area for long-term 1(3) 2(2) 1(3) interdiction (square meters)

• See Section 5.9.4.5(2).

• • 1(-3) = 1 x 10 3 =.001 NOTE:

Please see Section 5.9.4.5 7) for discussion of uncertainties.

Estimated numbers were roun(ded to one significant digit only purpose of this table.

I i

1 Limerick FES M-4

,e Table M.2 Contributions to risk of early fatality with supportive medical treatment from spatial intervals within 50 miles (80 km) of the Limerick site with Early Reloc* and Late Reloc*

offsite emergency response modes Risk per reactor year From causes other than From severe severe earthquakes earthquakes Spatial interval (Early Reloc)

(Late Reloc)

Total from (mi) - to (mi)t (persons)

(persons)

(persons) 0.0 - 0.5**

6(-5)***

4(-5) 1(-4) 0.5 - 1.0 e

6(-5) 6(-5) 1(-4) 1.0 - 1.5****

2(-4) 3(-4) 5(-4) 1.5 - 2.0 2(-4) 3(-4) 5(-4) 2.0 - 2.5 1(-4) 4(-4) 5(-4) 2.5 - 3.0 1(-4) 3(-4) 4(-4) 3.0 - 3.5 1(-4) 6(-4) 7(-4) 3.5 - 4.0 9(-5) 5(-4) 6(-4) 4.0 - 4.5 3(-5) 3(-4) 3(-4) 4.5 - 5.0 3(-5) 3(-4) 3(-4) 5.0 - 6.0 2(-5) 3(-4) 3(-4) 6.0 - 7.0 6(-6) 2(-4) 3(-4) 7.0 - 8.5 2(-6) 3(-4) 3(-4) 8.5 - 10.0 6(-7) 2(-4) 2(-4) 10.0 - 12.5 2(-6) 3(-4) 3(-^)

12.5 - 15.0 2(-8) 2(-6) 2(-6) 15.0 - 17.5 3(-8) 5(-8) 8(-8) 17.5 - 20.0 4(-8) 0 4(-8) 20.0 - 25.0 0

0 0

25.0 - 30.0 0

7(-7) 7(-7) 30.0 - 35.0 0

0 0

35.0 - 40.0 0

0 0

40.0 - 45.0 0

0 0

45.0 - 50.0 0

0 0

Total 1(-3) 5(-3) 6(-3) tTo change miles to km, multiply the values shown by 1.609.

• See Section 5.9.4.5(2).

• • This circular zone includes the Site Exclusion Area.

• • • 6(-5) = 6 x 10 5 =.00006

• • • • 93% of the area of this annulus is included within an annulus 1-mile wide outside of the site exclusion area boundary.

NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

Estimated numbers were rounded to one significant digit only for the purpose of this table.

~

Limerick FES M-5

7 s

Ta m H.3 Contributions to risk of early fatality with jy minimal medical treatment from spatial

.c intervals within 50 miles (80,km) of the c

Limerick site with Early Reloc* and Late Reloc*

offsite emergency response modes

{

n Risk per reactor year d

From causes F

other than From severe s

severe earthquakes earthquakes l

Spatial interval (Early Reloc)

(Late Reloc)

Total from (mi) - to (mi)t (persons)

(persons)

(persons) 1 d

0.0 - 0.5**

8(-5)***

4(-5) 1(-4) 0.5 - 1.0 1(-4) 7(-5) 2(-4) 1.0 - 1.5****

3(-4) 3(-4) 7(-4) 1.5 - 2.0 3(-4) 4(-4) 7(-4) 2.0 - 2.5 3(-4) 5(-4) 8(-4) i 2.5 - 3.0 2(-4) 4(-4) 6(-4) 3.0 - 3.5 3(-4) 8(-4) 1(-3) 3.5 - 4.0 3(-4) 7(-4) 1(-3) 4.0 - 4.5 1(-4) 4(-4) 5(-4) 1

~

}

4.5 - 5.0 3(-5) 4(-4) 4(-4)

,1

5.0 - 6.0 6(-5) 4(-4) 5(-4) 6.0 - 7.0 3(-5) 4(-4) 4(-4) 7.0 - 8.5 2(-5) 5(-4) 6(-4) 8.5 - 10.0 2(-5) 5(-4) 5(-4) 10.0 - 12.5 9(-5) 1(-3)

-1(-3) 12.5 - 15.0 9(-6) 1(-4) 1(-4) 15.0 - 17.5 1(-5) 5(-5) 7(-5) 17.5 - 20.0 1(-5)

-2(-5) 3(-5) 20.0 - 25.0 1(-5) 1(-5) 3(-5) 25.0

,30.0 2(-5) 2(-4) 2(-4) 30.0 - 35.0 1(-5) 9(-6) 2(-5) 35.0 - 40.0 7(-8) 1(-6)-

1(-6) 40.0 - 45.0 3(-8) 8(-7) 9(-7)..

45.0 - 50.0 3(-6) 3(-7) 3(-6) i 4

Total 2(-3) 8(-3) 1(-2) y tTo change miles to km, multiply the values shown by 1.609.

• See Section 5.9.4.5(2).

,j

• • This circriar zone includes the Site Exclusion Area.

• • • 8(-F ) = 8 x.' 0 5 =. 00008

• • • • 93% of the area of this annulus is included within an annulus 1 mile wide outside of the site exclusion area boundary.

I NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

t Estimated numbers were rounded to one significant digit only a

for the purpose of this table.

j

. Limerick FES M-6 l

+.

y

' Table M.4 -Contributions to risk of latent cancer (including thyroid) fatality from spatial intervals within 50 miles (80 km) of the Limerick site with Early Reloc* and Late Reloc* offsite emergency-response modes Risk per reactor year From causes other than From' severe-severe earthquakes earthquakes Spatial interval (Early Reloc)

(Late Reloc)

Total from (mi) - to (mi)t (persons)

(persons)

(persons)

\\

0.0 - 0.5**

4(-5)***

1(-6) 4(-5) 0.5 - 1.0 7(-5) 3(-6) 7(-5) 1.0 - 1.5****

3(-4) 2(-5) 3(-4) 1.5 - 2.0 3(-4)'

3(-5) 3(-4) 2.0 - 2.5 4(-4) 4(-5) 4(-4) 2.5 - 3.0 4(-4) 4(-5) 4(-4) 3.0 - 3.5 6(-4) 9(-5) 7(-4) 3.5,- 4.0 6(-4) 9(-5) 7(-4) 4.0 - 4.5

' 3(-4) 5(-5) 3(-4) 4.5 - 5.0 3(-4) 5(-5)~

3(-4) 5.0 - 6.0 4(-4) 8(-5) 4(-4) 6.0 - 7.0 3(-4) 8(-5) 4(-4) 7.0 - 8.5 5(-4) 1(-4) 6(-4)-

t 8.5 - 10.0 5(-4) 1(-4) 6(-4) 10.0 - 12.5 3(-3) 8(-4) 4(-3)

{

12.5 - 15.0 1(-3) 2(-4) 1(-3) i 15.0 - 17.5 2(-3) 4(-4) 3(-3) 17.5 - 20.0 2(-3) 4(-4) 2(-3) 20.0 - 25.0 7(-3) 1(-3) 8(-3) l 25.0 - 30.0 1(-2) 2(-3) 2(-2) 30.0 - 35.0 6(-3) 1(-3) 7(-3) 35.0 - 40.0 5(-3)

'8(-4)

-6(-3) 40.0 - 45.0 2(-3) 3(-4) 2(-3) 45.0 - 50.0 2(-3) 3(-4) 2(-3)

Total 5(-2) 9(-3) 6(-2) tTo change miles to km, multiply the values shown by 1.609.

• See Section 5.9.4.5(2).

• • This circular zone includes the Site Exclusion Area.

• • • 4(-5) = 4 x 10 s =.00004

• • • • 93% of the area of this annulus is included within an annulus 1-mile wide outside of the site exclusion area boundary.

NOTE:

Please see Section 5.9.4.5(7) for discussion of uncertainties.

I Estimated numbers were rounded to one significant digit only for the purpose of this table.

Limerick FES M-7'

~+

e i..

s C

7 l

APPENDIX N CRITIQUE OF APPLICANT'S CONSEQUENCE ANALYSIS IN LIMERICK GENERATING

{

STATION ENVIRONMENTAL REPORT-OPERATING LICENSE (ER-OL) 1 l

I l

i.:

y

.p -

APPENDIX N.

' CRITIQUE OF APPLICANT'S CONSEQUENCE ANALYSIS'IN:L'IMERICK GENERATING STATION ENVIRONMENTAL REPORT-OPERATING LICENSE (ER-OL)

~

a In the ER-OL, a total.of 11 source terms (or release ' categories) were used.

Some of-these release categories-were the result of binning (or grouping) of

.several individual source terms.

In some of the bins, the member source terms had very dissimilar release characteristics and release fractions, and the

- source terms selected to represent the bins were considered by the staff to be unrepresentative of the bins.

For this reason, the staff did not use the ER-OL binning of the source terms and chose to use a greater ~ number and more consis-tent set of release categories in its consequence analysis.

However, the 11' different sets of release fractions (source terms) used in the ER-OL and the 27 release categories used in the staff analysis are intended to encompass an equivalent number of combinations of the plant damage states and containment o

failure modes.

The point estimates of radionuclide release fractions for the 11 source terms in the ER-OL are generally lower and warning times for evacuation associated with some of these source terms are longer than those for the release categories

~

used in the staff analysis.

However, exact comparison of source term specifi-cations between those in ER-M and in the staff analysis is difficult because of the different numbers of s iurce terms used in the two analyses.

.The point estimates of probabilities of the source terms in the ER-OL add up to 6 x 10 8 per reactor year for seismic causes and 4 x 10 s per reactor year for -

non-seismic causes.

The staff analysis uses the same total value for the point estimates of the probabilities of the seismically induced' release categories; however, the staff's total of the point estimates of the probabilities of non-seismically induced release categories is 9 x 10 s per reactor year.

The consequence analysis in the ER-OL used the CRAC2 computer code, which is a modified version of the CRAC code used in the Reactor Safety Study (WASH-1400 NUREG-75/014).

Both CRAC2 and the staff version of CRAC (1980) incorporate the same evacuation model which is revised from that used in WASH-1400.

The revised evacuation model is capable of incorporating people's delay time before evacua-tion in addition to their speed during evacuation.

Both the codes are also capable of_modeling a variety of offsite emergency response options--such as shelter and relocation--in addition to evacuation separately or in combination.

-CRAC2 incorporates a modified scheme for sampling the weather data in addition j

to the usual sampling schemes of CRAC.

However, using the modified weather sampling scheme of CRAC2 and the stratified sampling scheme of CRAC, both the codes produced almost identical results, within likely uncertainty bands, in international benchmark exercises for comparison of codes' used in consequence analysis.

Therefore, the use of CRAC2 in the ER-OL is acceptable to the staff.

However, the staff chose to use CRAC for its independent consequence analysis for two reasons:

Limerick FES N-1 s

1

_...,,,r_m-

,.,.m.

1(

(-

(1) L Although CRAC and CR'AC2 produced almost equal results, within likely uncertainty bands, for benchmark problems, there are some differences in' results produced by the' two-codes for other problems which have yet to be properly explained. A detailed comparison between CRAC and CRAC-2 has been sponsored by the staff at Oak Ridge National. Laboratory. 'After the dif-S ferences between the two codes'are understood, the staff may use CRAC2,

.with or without any-additional modifications, in future applications.

(2) The other reason for using CRAC'in the staff analysis is that the_ staff has'used thel1980 version of.CRAC in severe accident consequence analyses in the environmental statements issued after July 1,1980, pursuant to the Commission Statement of Interim Policy, June 13, 1980 (45 FR 40101-40104).

The ' staff has 5 rovided a comparison of risk estimates for Limerick with those made using CRAC in environmental statements for other plants and the use of CRAC2 could prove inconsistent.

Five years worth of meteorological data (from 1972 to 1976) was used in -the ER-OL consequence analysis after some modifications were made to CRAC2, which normally uses only 1 year of meteorological data.

In response to the staff question as to the degree of improvement achieved by using 5 years of data, the applicant provided a comparison of CRAC2 runs for sample problems using each of the five 1 year data periods separately with those using data for the entire 5 year period. The comparison did not show much difference between these runs.

Further, in response to 'the staff. question regarding the adequacy of. the number of weather sequences sampled from 5 years of data, the applicant presented a comparison of CRAC2 runs for sample problems with increased weather sequence samples. No appreciable difference as a result of the increased sampling was noticed.

Therefore, the use of 5 years' worth of meteorological data and the sampling scheme in the ER-OL are acceptable to the staff.

The ER-OL analysis used a core inventory of radionuclides (excluding activation products) calculated for a BWR at a power level of 3293 MWt.

However, the staff analysis used 105% of this power le: vel (3458 MWt), and calculated the core inventory based upon WASH-1400 estimates of fission and activation product distributions.

The use of a lower power level would result in lower offsite consequences.

The ER-OL analysis used an estimated population distribution for the year 2000 up to 500 miles (800 km) from the plant, and economic data related to land use on county-wide basis up to 50 miles (80 km) and on a state-wide basis outside 50 miles (80 km). These are acceptable to the staff, although staff used its own estimates of inputs to the CRAC code.

The other economic data in the ER-OL are not site specific, but they are site specific in the staff analysis.

~

For releases not caused by severe earthquakes, the ER-OL analysis used a generic set of parameters for evacuation within the 10-mile (16-km) Emergency Planning Zone (EPZ):

1, 3, and 5-hour delay times with probabilities of 0.3,'0.4, and 0.3, respectively, and 10 mph (16 km per hour) for effective evacuation speed.

Because this is not site specific, it-is unacceptable to the staff.

A study prepared by the NUS Corporation for the applicant in 1980 provides a basis for the estimate of effective evacuation speed of about 2.5 mph (4 km per hour),

considering the road network and the expected traffic loading for evacuation from the 10-mile (16-km)~EPZ during emergency.

The estimate of the site-3-

specific delay time of about 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> made in the NUS study was rejected by the

' Limerick FES N-2 3

-[

E applicant because the study did not take into account the early warning system that would be required for notification of emergency before the plant would be licensed for operation. 'The staff recognizes the applicant's position. However, s

in lieu of any available estimate of delay time for the site, the staff assumed a delay time of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />, which is consistent with similar estimates for other high population density sites.

The ER-OL assumed a maximum distance of 20 miles

(32 km) traveled by the evacuees; however staff used 15 miles (24 km) for this distance, as it has for other plants.

The staff assumptions of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> for 1

delay time, 2.5. mph (4 km per hour) of evacuation speed, and a travel distance

.of 15 miles ~(24'km) are applied to the situations of releases as a result of plant-internal causes, fires, and low to moderately severe earthquakes (see Section 5.9.4.5(2) for an alterative to the assumption of evacuation from the i

10-mile (16-km) EPZ). - For these situations, the ER-OL also assumes relocation of people from the 10- to 25 mile (16-to 40-km) region 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> after passage of radioactive plume.

Although a similar assumption has been made by the staff in the consequence analyses in the environmental statements for several other plants, the staff judgment is that this assumption for a site with high popula-i tion density would not be appropriate because the large number of people that I#

would be involved in the 10- to 25-mile (16-to 40-km) region would make this scenario unrepresentative.

Instead, the staff analysis assumes that outside of the 10-mile (16-km) EPZ, only people from the highly contaminated areas (see i

Section 5.9.4.5(2)) would be relocated 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> after plume passage.

Shielding i

factors used in the ER-OL are:

(1) the same as in the staff analysis during evacuation, (2) higher than the staff's during delay before evacuation, and (3) lower than the staff's during waiting before relocation.

The values used by the staff are the same as those*used in WASH-1400.

The impact of differences in shielding factors used in the ER-OL from those in WASH-1400 is difficult to 3

j assess, although it is not likely to be substantial.

For releaser, caused by severe earthquakes, the ER-OL assumes evacuation from the 10-i!se (16-km) EPZ after a 3-hour delay with an effective speed of'0.5 meter /

sec, and relocation frc:.. 10- tu 25 mile (16-to 40-km) region 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after plume passage.

How ver, the severity of earthquakes assumed is Modified 1

Mercalli intensity scale of IX or higher, and it is the judgment of the staff that earthquakes of such severity would cause very extensive damage in the site region that would seriously hamper the evacuation.

Therefore, the staff assumed i

no evacuation for these situations but, instead, assumed relocation of people from highly contaminated areas 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after plume passage.

Shielding factors used by the staff are also more pessimistic.

The ER-OL analysis assumed an effective peak ground acceleration of 0.61g or more to be associated with' Modified Mercalli intensity scales of IX of higher.

However, the staff used 0.4g as the dividing line, although it recognizes that there is lack of actual recordings of effective peak ground accelerations asso'ciated with the intensity I

scales.

It is the staff's judgment that although a range of effective peak ground acceleration of 0.35g to 0.5g would be more appropriate, the results of consequence analysis are net sensitive to the choice of values within a range of 0.35g to 0.5g.

Therefore, the staff used only the single value of 0.4g.

The 3

ER-OL assumptions regarding the offsite emergency response during severe earth-quake conditions'as well as the assumption of 0.61 g as the dividing line for classification of less severe and very severe earthquakes result in lower esti-mates of risks from seismically induced source terms.

O m ee,, _ 5 H.,

(

_,_.,___m

_-,--m__,_.,-...-..,.,__--_,..-_.,.o

[

b The ER-OL point estimates of risk from the 11 source terms and the staff's point estimates of risks from 27 release categories are as follows:

s Risk per reactor year Type of risk ER-OL*

Staff 1.

Early fatalities with supportive 3(-4)**

5(-3) medical treatment (persons) 2.

Latent cancer fatalities excluding thyroid (persons) 50-mile (80-km) region 2(-2) 4(-2)

Entire region 3(-2) 7(-2) 3.

Latent thyroid cancer fatalitier (persons) 50-mile (80-km) region 5(-3) 1(-2)

Entire region 6(-3) 1(-2) 4.

Whole body person-rems 50-mile (80-km) region 300 700 Entire region 500 1000 5.

Cost of offsite mitigation 20,000 50,000 measures (1990 dollars)

• 0n March 13, 1984, the applicant informed the staff that the ER-OL consequence calculations are being revised and that the revised calculations will not result in significant changes in the results currently presented. in the ER-OL.

Based upon the applicant's explanation of the source of the error, the staff judges that the impact of these revisions will be relatively small.

• • 3(-4) x 10 4 =.0003.

Estimated numbers were rounded to one significant digit only for the purpose of this table.

In the ER-OL, an uncertainty analysis on risks is provided with respect to four major parameters (1) probability of each source term i,

(2) magnitude and other release characteristics of some of the dominant source terms (3) evacuation and sheltering parameters l

l Limerick FES N-4

(

ff._'

) _

(4)' dose-response relationships for early fatality with three types of medical treatment The first of these parameters was treated by system analysis and standard

- methods of combining uncertainties.

The other three were treated by a sensi-tivity study using the CRAC2 code to provide a large number of conditional CCDFs for the 11 different sets of release fractions (source terms).

These CCDFs were used to define the upper and lower conditional CCDFs for the source terms.

The upper and lower CCDFs were combined probabilistically with the un-certainity distribution on source term probabilities in order to generate the uncertainty bands on'the overall CCDFs.

The variations used in source term parametrization.were mostly subjective.

For offsite emergency response the evacuation speed was varied from~2.5 to 10 mph (4 to 16 km an hour), while the delay time before evacuation ranged from 1 to 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />. For severe earthquake conditions, no variation in the parameters of offsite emergency response was made.

For the 10- to 25-mile (16-to 40-km) region, sheltering in basements for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> followed by rapid relocation was used; the 25 mile (40-km) distance was also extended to 50 miles (80 km).

Con-sidering that the state of the art of uncertainty assessment in consequence analysis is not well developed, this method of uncertainty analysis in the ER-OL is acceptable to the staff. However, the lack of any variation in the pessimis-tic direction in offsite emergency response parameters for the severe earthquake-conditions and too many variations in the optimistic direction for nonsevere earthquake conditions, and the lack of variation in the source terms to encom-pass some of the high values of the release fractions as used in the staff anal-ysis, lead the staff to disagree with the upper estimates of the overall CC0Fs in the ER-OL.

By letter, dated March 13, 1984, PECo states that errors had been discovered in the ER-OL consequence analysis.

PECo has further stated that these errors, when corrected, will not significantly alter the ER-OL conclusions.

The staff also performed a limited sensitivity analysis. With respect to varia-tion of probability of earthquake-induced release categories, the staff con-cluded that the staff's point estimates of risks could be exceeded by factors of up to 6, but could also be lower by factors up to 3.

With respect to param-eters of offsite emergency response the overall risks could be increased by up to 20%. With respect to medical treatment, the risk of early fatality could have a spread within factors of 2 to 3.

The staff has not-performed a sensi-tivity study with respect to probabilities of release categories initiated by causes other than severe earthquakes, source terms, and other elements that

^

contribute to uncertainties.

Based upon the insight gained from review of similar PRAs for Indian Point and Zion, it is the judgment of the staff that the staff's Limerick risk estimates could be too low by a factor of about 40 or too high by a factor of about 400.

~

i Limerick FES N-5 L

4

_