ML20024E916

Affidavit of Bd Mcfeaters Supporting Util Summary Disposition Motion on Eddleman Contention 80 Re Atmospheric Dispersion Model.Models Assure Actual Concentrations of Radionuclides Should Never Exceed Estimates from Models

ML20024E916
Person / Time
Site:	Harris
Issue date:	08/31/1983
From:	Mcfeaters B CAROLINA POWER & LIGHT CO.
To:
Shared Package
ML20024E900	List: ML20024E899 ML20024E903 ML20024E906 ML20024E910 ML20024E915 ML20024E916 ML20024E917 ML20024E919
References
ISSUANCES-OL, NUDOCS 8309070206
Download: ML20024E916 (41)

Text

--

-d UNITED STATES'OF AMERICA NUCLEAR REGULATORY COIMISSION

~BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of .) ,

)

CAROLINA POWER & LIGHT COMPANY )

AND NORTH CAROLINA EASTERN )' Docket Nos. 50-400-OL ,

MUNICIPAL POWER AGENCY ) 50-401-OL

)

(Shearon Harris Nuclear Power ')

Plant, Units 1 and 2) )

AFFIDAVIT OF. BRIAN D. MCFEATERS IN SUPPORT OF APPLICANTS' MOTION FOR

SUMMARY

DISPOSITION OF INTERVENOR WELLS EDDLEMAN'S CONTENTION 80 County of Wake )

SS:

State of North Carolina )

BRIAN D. MCFEATERS, being duly sworn, deposes and says as follows:

1. I am a Project Scientist - Meteorological Supervisor employed by Applicant Carolina Power & Light Company. My business address is 7C3 Center Plaza Building, 411 Fayetteville Street Mall, Raleigh, NC 27602. A summary of my professional qualifications and experience is attached hereto as Exhibit "A". I have personal knowledge of the matters stated herein and believe them to be true and correct. I make this Affidavit in support of Applicants' Motion for Summary Disposition of Intervenor Wells Eddleman's Contention 80 in

. this proceeding.

2. The function of the Physical Sciences sub-unit, meteorological operations at Carolina Power & Light Company is to provide the company with professional expertise in all matters associated with meteorological data collection, analysis and assessment as they affect Carolina Power & Light Company. This function includes operation of onsite meteorological monitoring stations at nuclear plant sites, emergency preparedness support through l

operational synoptic forecasting assistance, atmospheric analysis of diffusion j and transport for potential accident and routine nuclear plant operation and l professional consultation services as related to meteorological concerns affecting the company. As a member of the Operational Training & Technical Services Department, I have direct supervisory responsibility for all meteorological monitoring and assessment activities and have prepared the l diffusion studies and assessments for the Shearon Harris Nuclear Power Plant.

8309070206 830901 1 PDR ADOCK 05000400 l 0 PDR

I 3 I have prepared a technical paper attached hereto as Exhibit "B",

the statements contained therein are incorporated into this Affidavit as if set forth in full herein. The purpose of Exhibit B is to demonstrate that there is no factual basis for Eddleman Contention 80, because Applicants' mixing and dispersion models employ state-of-the-art techniques and utilize conservative assumptions that assure that actual concentrations of radionuclides should never exceed the estimates obtained from use of the models.

. Q Wt t9 Brian D. McFeaters soitistri,,

Subscribedpndsworntobeforeme this .3/d day of August, 1983 O.M

... ....4.,

& --NotaryA*Public _ f ,>_w . fj'goTARy

• A O g t, t C f My commission expires W M I ,/99[. .  %,, .....,..' ,

[

' EXHIBIT A'

~

BRIAN D. MC FEATERS-PROJECT SCIENTIST - METEOROLOGICAL SUPERVISOR OPERATIONAL TRAINING & TECHNICAL. SERVICES DEPARTMENT CAROLINA POWER & LIGHT COMPANY EDUCATION: The Pennsylvania State University - B. S.

Meteorology -; June 1972 Member American Meteorological Society Member American Nuclear Society Member Eastern North Carolina Nuclear Society PROFESSIONAL EXPERIENCE:

Present to Project Scientist - Meteorological Supervisor -

August 1981 Operational Training & Technical Services Department. Presently supervising the Physical Sciences sub-unit which is responsible for all meteorological and seismological concerns of the company. The sub-unit supports nuclear power plants with operation of onsite meteorological monitoring stations and diffusion analysis, with emergency preparedness support through operational synoptic forecasting, with analysis and assessments of atmospheric transport and dispersion and through general professional consultations on all meteorological and seismological activities as they relate to the company.

August 1981 to Senior Scientist - Meteorologist - Technical Services September 1976 Department, Licensing & Permits Section. Responsible for the meteorological program and operation of the meteorological monitoring stations, assuring that all regulatory requirements had been fulfilled.

Additionally responsible for the preparation of the FSAR and ER sections pertaining to meteorology and atmospheric dispersion.

September 1975 to scientist - Meteorologist - Westinghouse Electric j May 1973 Corporation, Environmental Services Division, i Pittsburgh, PA. As a staff scientist, responsible for i conducting and assisting in the preparation of i environmental impact statements for both fossil and nuclear power plants. Conducted and wrote meteorological analysis for both the FSAR and ER at the Clinch River Breeder Reactor as well as other nuclear power and fuel fabrication facilities.

Developed computer models to assess dispersion from fossil, nuclear and cooling tower facilities.

v--. - , - ~ .,r,.,.,.,-,,,. . , - , . . . . - . , - . _ , ,, y .r.y_. ,,..._y, _--...,..,,_____,y,,_ ...,n., , , _ . , . , _ _ , ,. _,,_ _.__. _

May 1973 to Forecast Meteorologist - DeNardo & McFarland' Weather January 1976 Services, Inc., Pittsburgh, PA. A staff forecaster responsible for the preparation of public weather forecasts for radio and television and for the briefing of private corporate clients. Assisted in the environmental assessment of fossil facility impact upon local air quality regulations. Performed field measurements of meteorological and air quality parameters using state-of-the-art instrumentation.

l l

l l

l I

l l

1 ,_, -- . _ . _ , . _ . . . _ . _ _ _ _ _ _ . . . . _ _ . _ _ _ _ , _ _ _ __ . _ , . _ _ , . _ _ _ . _ . _ _ . . , _ . _ _ _ _ _ _ .

EXHIB t B

. I. INTRODUCTION The purpose of this paper is to substantiate.and demon-strate that the atmospheric dispersion model used by Applicants to estimate the mixing and dispersal of radioactive effluents which could be released from the Shearon Harris Nuclear Power Plant (SENPP) into the air under accidental and routine conditions:

1) represents state-of-the-art atmospheric dispersion modeling for nuclear applica-tions;

2) is consistent with the various NRC Regula-tory Guides cited below;

3) is adequate, reasonable, and useful to estimate the relative concentrations of radionuclides, and subsequently, the radia-4 tion and inhalation doses which could be observed at SHNPP;

4) accounts for site specific atmospheric factors;

5) contains many conservative factors that significantly reduce the transport, diffu-sion and depletion of the plume as compared to realistic and a;tual conditions, thus

. _ . , .~ , - _ ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ , .

assuming very small and' incomplete plume dispersion and mixing. (The conservative

-assumptions embedded in the dispersion mod-eling methodology result in estimated high relative concentrations and doses. These relative concentrations and doses are much higher than occurrences during realistic and actual conditions, accounting for the SHNPP' site specific-adverse meteorological conditions); and

6) provides very conservative estimates of ground level relative concentration and ra-diation and inhalation doses due to the ex-clusion of the rainout considerations from .

an atmospheric dispersion model.

II. GENERAL NETHODOLOGY -

Section 100.11 of 10 C.F.R., entitled " Determination of 1

Exclusion Area, Low Population Zone and Population Center Dis-tance" establishes numerical limits for the individual exposure to a total radiation dose due to an accidental release of ra-dioactive effluents. 1/

Appendix I of 10 C.F.R. Part 50 entitled " Numerical Guides for Design Objectives and Limiting Conditions for Operation to Meet the Criterion 'As Low as is Reasonably Achievable' for Ra-dioactive Materials in Light-Water-Cooled Nuclear Power Reactor i

l l

Effluents" establishes numerical limits for individual exposure to a total radiation dose under.coutine releases of radioactive effluents. 2/

Compliance with the numerical limits for_ total radiation dose is accomplished through the implementation of:

.1) the necessary design objectives for equip-ment to control the release of radioactive material in effluents from nuclear plants (10 C.F.R. Part 50.34a) 3/; and

2) management tools that enable both the NRC and the Applicants to estimate the total exposure assuming a particular release rate and the environmental conditions based on on-site observations at SHNPP.

One of the key management tools most commonly used to relate total exposure to the release rate and environmental conditions (often referred to as establishing source-reception relationship) are mathematical dispersion models. These models provide an estimate of the dispersion of the released effluent and of its spatial concentration in the medium under considera-tion (air, water, soil).

Mathematical dispersion models have been in use in nuclear applications for approximately four decades. In general these models account for the transport, diffusion, and depletion mechanisms involved in dispersion of radioactive materials.

The depletion mechanisms generally accounted for in air are:

dry and wet deposition and radioactive decay.

n '

. _o -

y Thus, the objective of using dispersion models nfor -atmo -

spheric releases of.radioactiveLeffluents, is:

to provide conservative estimates of rela - ,

tive concentrations under various meteoro '

logical conditions for routine and- _ _ . ---

xw .

accidental release conditions, with .

r

-u emphasis on accidental release coupled with 's - ,

worst case meteorological conditions. _.

r

. -- c'

'\

a This objective is further delineated in several of the NRC Regulatory Guides that provide the methodology for atmospheric dispersion and dose assessment.

These include:

x

1) Regulatory Guide 1.4, " Assumptions Used for Evaluating the Potential Radiological Consequences of a Loss of Coolant Accident-for Pressurized Water Reactors" 4/; -

2) Regulatory Guide 1.24, " Assumptions Used for Evaluating the Potential Radiological '

i Consequences of a Pressurized Water Reactor I

Radioactive Gas Storage Tank Failure" 5/;

! 3) Regulatory Guide 1.25, " Assumptions Used for Evaluating the Potential Radiological Consequences of a Fuel Handling Accident in t

l the Fuel Handling and Storage Facility for l Boiling and Pressurized Water Reactors" 6/;

1 l

,+ c -y, 3.-

n: ._ q '2 2:

n-

" ' ^

'( .

q .. -} r .p /

1, -* '

..I

.l,y2NJ f

4y; a ' i 4): . Regulator 9~Guida'l.77,," Assumptions Used, x; m- .

dfor Evaluating aLControl Rod Ejection

^

Y

,, ..f Acci' dent' for Pres,surized Water Reac g 2 + o. 3 3 l tors" 7/e

^

i: f  !-

9. .

~.c -

5)-T' Regulatory Guide 1.109, " Calculation of Annual. Doses to Man from Routine: Releases.

- \

of Reactor Effluents for She Purpose'of' l ,' _

Evaluating Compliance with 10 C.F.R. Part'

- +

~

f/

, 50, Appendix I" 8/; 3-v,, -

~

6), Regulatory Guide 15111, " Methods for J Ectimating Atmospheric Transport and Dis-

,persion of Gaseous Effluents in Routine Releases from Light-Water-Cooled Reac-tors" 9f;

, : 7) Regulatory Guide 1.145, " Atmospheric Di's-

' persion Models for Potential Accident

- Consequence Assessments at Nuclear Power Plants" ,10/.

III. ATMOSPHERIC DISPERSION MODEL USED IN ROUTINE AND ACCIDENTAL RELEASE ASSESSMENT.

A. Rationale for Use of Gaussian Model The Gaussian plume model is the basic, most widely and most commonly used atmospheric dispersion model 11-16/.

i The Gaussian name has been adopted to emphasize that the plume dispersion in the vertical and horizontal (lateral) directions follows the Gaussian distribution function. The model

i formulation is therefore based on the Gaussian equation. Key factors that make the Gaussian dispersion model so versatile in nuclear application.are:

1) It approximates,the dispersion of windborne pollutants in the atmosphere under various meteorological conditions;

2) It represents a state-of-the-art dispersion-modeling technique and methodology, used in air pollution metecrology both for nuclear and non-nuclear applications;

~

3) The formulation is basic and easily imple-

~

mented by the user. It is based on fundamental principles of atmospheric physics and represents a solution to the Fickian diffusion equation with constant diffusivity and wind speed. The formula-tion is also consistent with the random nature of turbulence;

4) It is easy to use in practical applications relating source to receptors under various meteorological conditions, thus making the model's utility quite high;

5) It produces results that agree well with

'N other types of atmospheric dispersion N models 17-18/ including the most detailed and complex ones available. Justification i

s lm n

. 'c

<.h for'the extra effort and_the potential complica'tions associated with the use-of a very complex model is quite; questionable due to the small advantage,;if any, that a more complex model.has over the Gaussian-dispersion modeli

6) It tends to. provide conservative estimates of plume dispersion from low-level or

. ground-leval releases under certain limiting meteorological conditions condu-cive to high radioactive concentrations, such as stable atmospheric conditions with very light winds. ,

Because of these key factors, the NRC considers the Gaussian dispersion model to be the most prevalent tool in nu-clear applications for estimating relative concentrations of radionuclides and radiation and inhalation doses due to accidental and routine releases. This Gaussian model is an in-tegral part of the dispersion calculation methodology outlined in many NRC. Regulatory Guides such as: Regulatory Guides 1.4, 1.24, 1.25, 1.77, 1.09, 1.111, and 1.145 4-10/. In addition, the NRC has. developed computer codes that are based on the Gaussian dispersion model, for use by its staff in meteorologi-cal evaluation of routine 19/ and accidental releases 20/ from commercial nuclear plants.

l l

l .

p,., ,

,n .,-, . . -.,-. . . . - . , ,,n.-,.,n.-- . ~ , , v.a., , e , ...-

Based on the above information and consistent with the recommended NRC approach to dispersion modeling,' Applicants have used the Gaussian dispersion model in estimating the rela--

tive concentrations of radionuclides'under routine and accidental releases. The exact methodology is detailed in NRC Regulatory Guide 1.'145 10/ for .ccidental releases and in Regu-latory Guide 1.111 9/ for routine releases.

B. Assumotions and Accuracy of the Gausian Disp,ercion Model for Nuclear Applications.

The Gnussian dispersiori model is based on several as-sunptions i. hat should be realizec by the users of the model.

It is imperative to understand these assumptions in order to put in perspective the applicability, capability, and perfor-mance of the model for a specific application. Once these as-sumptions are fully understood, the model may be used in specific applications. The resulting calculated relative con-centrations, and subsequently the calculated radiation and inhalation doses, then can be interpreted in order to determine whether a particular plant meets the numerical guidance set up in 10 C.F.R. Part 100.11 1/ and Appendix I to 10 C.F.R. Part 50 2/.

The Gaussian dispersion model incorporates assump-tions associated with both its formulation and the augmentation and refinement of its capabilities and performance. Assump-tions associated with the model's formulation are-

r

1) Plume dispersion follows the. Gaussian dis-tribution in the horizontal and vertical' directions. The plume dispersion-in.these directions is measured by the horizontal and vertical dispersion parameters-(often called coefficients). The dispersion parameters vary with the distance downwind from the source-and atmospheric stability 13,16,21/. For a fixed atmospheric stabil-ity the dispersion' parameters increase with downwind' distance. For a fixed distanca downwind the horizontal and vertical dis-persion parameters attain the largest values under extremely unstable conditions and the smallest values under extremely stable atmospheric conditions. For a ground level release the relative concen-tration is inversely proportional to the product of the horizontal and vertical dis-persion parameters. For example, if the plume spreading in the horizontal and

. vertical directions decreases by a factor of two and three, respectively, the rela-tive concentration of radionuclides in-creases by a factor of six;

_9_

L

. o 2)- The relative concentration is inversely.

proportional to the-wind speed for a given

~

release height. Thus, if the wind speed is reduced by a factor of two, the relative concentration increases by a factor of two;

3) The most1widely used Gaussian dispersion model assumes a steady state continuous point source release. The formultaion in-volved does not consider explicitly the temporal (time) variation of the plume dis-persion. Further, it assumos inutantaneous ,

travel of the plume from the source to a receptor located at a certain distance downwind from the source'regardless of the

-distance between the source, the receptor, and the plume travel time. This means that the model assumes that the. plume spreads instantaneously. [If the plume actually travels at the speed of the wind, it is possible to estimate the plume travel time from the release point to a receptor. For example, with a 1.0 mph wind speed it takes the front part of the plume two hours to reach a receptor located 2.0 miles downwind from the release point, assuming that the wind will persist in the same direction for this time duration];

a a

4) The Gausian dispersion model is a~ straight line model. Since it is not time dependent and assumes instantaneous spread of the -

plume, it does not account for variations in wind directions with time. The wind variations with time are smoothed and aver-aged to' yield an average wind direction;

5) For routine releases the model incorporates the sector average concept that is based on the assumption that_the cencentration is unifornly distributed in the lateral direc-tion within a wind sector of 22.5 degrees.

6) The model assumes homogeneous wind field; i.e. wind is invariant with space coordi-nates (x, y, z).

Assumptions associated with the augmentation and re-finement of the Gaussian dispersion model are:

1) Building wake considerations are in-cluded in the Gaussian dispersion model to simulate the effect of the wake on low-level releases, such as vent, roof, or ground level releases near large obstructions 4,9,10,22-26/.

A building wake adjustment term (or factor) has been included in the model to simulate an increase in the mixing O 8 of the plume'in the vicinity of.

obstructions due to the wake effect.

The adjustment term is a function of obstruction height and width. This methodology is the most commonly used in nuclear applications;

2) The values of the horizontal and vertical dispersion parameters used in the caussian dispersion model under stable atmospheric conditions are small be ause they simule.te the small spread and growth of the plur.e as a

, function of distance downwind under these atmospheric conditions. After a decade of elaborate and extensive field tests, researchers have conclud-ed that the dispersion parameters commonly used 13,21/ to estimate the relative concentration of radionuclides released from low-level or ground-level releases under stable atmospheric conditions and light winds, are consistently smaller than ones derived from field tests 27-33/.

The results of field tests show that the relative concentrations estimated

by the Gaussian dispersion model are one to two orders of magnitude higher than the measured values depending on the site specific characteristics.

31-32/ The reason for this large dis-crepancy is the actual measured large horizontal meander and the large vertical sptead of a plume released near the ground, that are not accounted for by the commonly used dispersion parameters. The increase in plume meander and vertical spread are attributed in part to surface roughness and building obstruction effects that are not included in the commonly used dispersion parametera;

3) The current Gaussian dispersion model utilized in nuclear applications for accidental releases include a wind me-ander adjustment factor in the hori-zontal dispersion parameter that par-tially compensates for the meander effect. 10,20/ No adjustment factor is included for the vertical disper-sion parameter, that compensates for the observed increase in the vertical

O spread of_the plume under stable:

conditions in the. wake of buildings;

4) The Gaussian dispersion model applied-in nuclear applications allows for plume depletion terms that account for dry and wet deposition on a case by case hacis 34-41/. Plumo depletion dua to dry deposition is attributed to gravitational rettling of particles and absorption of gases and-aeroscis onto grc nd surfaces. Plume depletion due to wet disposition, often referred tc as precipitation scavenging, consists of two removal mechanisms:

a) rainout which is the wet deposi-tion component due to precipita-tion in clouds, and b) washout, which is the wet deposi-tion component due to precipita-tion that takes place below clouds.

5) The influence of the wet deposition .

term that simulates the rainout and washout effects has a small effect on calculated relative concentration within about 30 miles (50 km) around the release source. Consequently, the effect of wet deposition on radiation and inhalation' doses is small espe-cially for long-term / annual average doses 34/. Therefore, the more con-servative approach for estimating ra-diar. ion and inhalation doses ignores the influence of deposition on the relativo concentration. Thi.c approach is conni. stent with NRC Begulatory Gnide 1.4 4/ that calls for no correc-tion for the depletion of.the effluent plume of radioactive iodine due to deposition under accidental release conditions. In the event of routine releases, dry deposition is accounted in the Gaussian dispersion model 9/.

Wet-deposition, however, is accounted for on a case by case basis, depending on the release height (elevated) and

the nature of the precipitation, for l

l the area under consideration. If l

releases will be elevated and the area has a distinct rainy season corre-l sponding to the grazing season, wet l

deposition is taken into account 9/.

l l

i

- This approach again. allows for a conservative radiation and inhalation doses assessment.

Considering the various assumptions incorporated into the Gaussian dispersion model for nuclear applications, it is apparent that its accuracy is not perfect. Accuracy is a measure of how well model results compare with actual field data. .In general the estimated accuracy of th; Gaussian dis-persion model is 42-44/:

.1) Within about a factor of two for real world applications; with meteorological parameters reason 1bly well known and steady and without exceptional circumstances 43/;

2) Larger than a factor of two under ex-

=ceptional circumstances such as build-ing wakes, varied surfaces and ex-tremely stable conditions.

The estimated accuracy of the Gaussian dispersion model is supported by field studies, in particular, in cases of ground or near ground level releases with building wake effects and stable atmospheric conditions. In such situations, the relative concentration calculated from the Gaussian dispersion model consistently exceeds the actual measured relative concen-tration by more than a factor of two and often by more than an order of magnitude 31-33/.

IV. CONSERVATISM IN THE-SHNPP DISPERSION MODEL In performing dispersion calculations for postulated accidental and routine releases of radionuclides from SHNPP, Applicants have utilized the Gaussian dispersion model for nu-clear applications. The calculated methodology implemented.by Applicants is consistent with NRC Regulatory Guides 1.4 4/ and 1.145 10/ for accidental release and NRC Regulatory Guide 1.111 9/ for a routine release of radionuclides.

Based on the discussion presented in Sections II and III, it is evident that the Gauscian disperr, ion model applied to at-nospheric releases of radioactive effluents from the SENPP:

1) represents state-of-the-art atmospheric dispersion modeling used for nuclear appli-cations;

2) is consistent with the various NRC Regula-tory Guides applicable to dispersion calcu-lations 4-10/;

3) is adequate, reasonable, and useful to estimate the relative concentrations of radionuclides and subsequently the radia-tion and inhalation doses specific to SHNPP.

The dispersion modeling conducted by Applicants for SHNPP accounted for site specific meteorology obtained from the on-site meteorological station. In addition, the containment building cross sectional area was accounted for in the building wake term o'f the Gaussian dispersion-model. Thus, Applicants' model accounts fully-for site specific-atmospheric factors.

It is the purpose of.this section to demonstrate that the Gaussian dispersion model for SHNPP, used for accidental release calculations, is' conservative because of several as-sumptions included in the calculational methodology.

The assumptions that make the SHNPP Gaussian dispersion model conservative are:

1) The release height in assumed to be at ground level and this is a worst case as-sumption. In rea;.ity the release height for the SHNPP could be more than 150 ft.

above the ground even under accidental release conditions: for example, the con-tainment building is about 182 ft. high.

Moreover,-no consideration is given to the possible rise of the plume due to the release of the energy stored in it. The ground level release assumption coupled with the Gaussian dispersion model implies that the horizontal cross section of the plume containing the plume centerline is at ground level. Since the relative concen-tration is the highest at the plume centerline for any selected distance downwind from the source and for any i -

18-l t

p- y- w-- e -

pw. -,

atmospheric conditions it is evident that the ground level release assumption adds  ;

conservatism (overestimation) to the calcu-lational methodology of the' relative con-centrations;

2) The atmospheric conditions assumed in the accidental release calculations for SHNPP (up to eight hours) represent adverse mete-orological conditions condacive to ex-tremely high calculated relative concentra-tions of radienuclides. The atmospheric conditions are: Atmospheric stability G -

and wind speed of 0.75 mph (0.33 m/sec) which represents the wind instrument lowest detection threshold. As stated in the SHNPP FSAR, 52.3, Table 2.3.6-10, this com-bination of meteorological conditions occurred only 4.95% on an annual basis in-dependent of wind direction and on a wind dependent basis only .549% on an annual l

basis when the wind was blowing from the

north (maximum occurrence sector). The at-i mospheric stabilities commonly used in air pollution meteorology consist of six sta-bility classes A through F, designating A i

l as extremely unstable, B as moderately i

l

unstable, C as slightly unstable, D as neutral, E as slightly stable,' and F as moderately stable. These six atmospheric stability classes have been in use since Pasquill 13/ set up the stability classifi-cation scheme. Gifford 45/ shows in his ,

review of the different diffusion typing schemes that stete-of-the-art atmospheric stability schemes are hasad on six or less classes. The same six stability classes also are included in 12C Regulatory Guide 1.4 4/. In the ear 1*j seventies the NRC in-troduced a seventh stability clasts anr2 labeled it G. This stability class represents extremely stable conditions with the horizontal dispersion parameter being 2/3 of that under the F stability and the vertical dispersion parameter being 3/5 of that under F stability 10/, for any dis-tance downwind from the source. This means that under the same wind speed conditions the relative concentration for G stability is 2.5 times higher than the one for F sta-bility. Moreover the shift from the small wind speed of 1.0 m/sec used in NRC Regula-tory Guide 1.4 4/ to 0.335 m/sec used with

the G stability makes the relative concentration for F stability exceed the one for G stability by a factor of about 7.5-( 2.5 x 1/0.335). Thus it is quite clear that the G stability and light wind speed assumption coupled with the ground level release assumption further increase the conservatism of the Gaussian dispersion model for SHN?P. These conditions assume an extremely restricted dispersion of the plume both in the horizontal and vertical directions;

3) As discussed in Section IV(2) above, conditions with G or F stai>ility the light wind speed coupled with a wind direction persisting for an extended period of time are unlikely 3/. The Gaussian dispersion model for SENPP assumes the existence of wind persistence under G stability and 0.75 mph wind speed for accidental release. If the wind persists in one direction, it will take the plume about 1.75 hours8.680556e-4 days <br />0.0208 hours <br />1.240079e-4 weeks <br />2.85375e-5 months <br /> to first reach the SENPP site boundary (about 7000 ft. from the plant). Since it is unlikely for the wind to persist in one wind direc-tion under the assumed atmospheric

-stability 'and wind conditions, ' the . wind persistence assumption coupled.with

G:sta-

-bility and lignt winds further increases the model's conservatism. The' conservatism is becausesthe plume is restricted to a single direction while it undergoes very limited spread;

4) The classification of atmospheric stability t

for SHNFP is based on the A,T/az scheme _ ,

(often called delta-T) 46/. This scheme provides a mechanism for estimating atmo-spheric stability based on the measured changes in ambient temperature with height.

The delta-T scheme is considered reasonable for estimating atmospheric stability in the vertical and subsequently for determining the appropriate vertical dispersion parameter. It is well known that the use of the delta-T scheme for estimating the dispersion of a plume both in the horizon-tal and vertical directions is deficient under certain meteorological situations such as stable atmospheric conditions and light winds. This is because the mecha-nisms that govern the plume dispersion in the horizontal are not necessarily coupled y,- . ~ , . , .,,.c ., , - - , . - - . . , e-- .c..,.y ,-.,r- _ _ _ _ . - - - ._, , -,- , - - ~ , , . .,, . . , , _ - .-,-w-._n7 -,,-,e

to the ones that govern plume dispersion.in the vertical. Results of field test studies conducted under stable' atmospheric conditions.and light winds reveal that the delta-T stability classification scheme provides overly conservative estimates of the relative concentrations of radio-nuclides 27-33/. This overconservatism is due to the failure of.the delta-T scheme to account for the large horizontal meander of the plume under such conditicas. Results of field tests show that for topography similar to the SHNPP site the calculated relative concentration using the Gaussian dispersion model exceeds the measured rela-tive concentrations by factors that range between 20 to 40 under stable atmospheric conditions and light winds 31/. Incorpora-tion of the wind meander adjustment factor 10/ discussed in Section IIIB closes the gap between the calculated and measured relative concentration. Still, on the av-erage, the calculated relative concentra-tions are overpredicted by about a factor l

of two 33/. Thus the use of the delta-T atmospheric stability scheme coupled with 1

..s- ..

the wind ~ meander factor provided'by NRC Regulatory Guide 1.145 10/ adds additional conservatism to.the Gaussian dispersion model for SHNPP;

5) The Gaussian dispersion model for SENPP in-corporates building wake considerations in accordance with the discussion in Section IIIB and consistent with the methodology outlined in NRC~ Regulatory Guides 1.11 9/

and 1.145 10/. The building wake factor depends on the dimenrions of the building i

under consideration. In the case of SHNPP the smallest cross sectional area of the reactor containment building is used. The building wake factor contains a considera-tion of the fraction of the cross sectional area over which the plume is dispersed by the wake. This consideration is measured by the inclusion of the parameter c 22/,

1/2 4 c { 2.

In nuclear applications the parameter e is assigned the value 1/2 on grounds of conservatism 23/.

In addition, field studies 26,30,32/ show that the existing approach to building wake effect, as incorporated in the Gaussian dispersion model, tends to underestimate the effect of the wake on the spread of the

~

plume in the vicinity of buildings at nu-clear power. plant sites. This causes the Gaussian dispersion model'to overpredict the relative concentrations. Van der Hoven 32/ states that the present building wake factor accounted.for in the Gaussian dis-persion model used for accidental release conditions 4,10/, which is the model used for SENPP, is hardly cufficient to allow

~

for the factor of ten or more which represents the ratio between the measured

.to calculated vertical dispersien parameters under stable atmospheric conditions and lightwinds.

These points clearly demonstrate that the Gaussian dispersion model for SHNPP contains another conservative element based on assumptions embodied in the building wake factor incorporated in the model.

This conservatism in the building wake factor should result in an over-prediction of the relative concentrations for SHNPP as compared to what actually will be measured in reality;

, s .:

6) The Gaussian dispersion-model for.SHNPP does not-include a wet deposition term that accounts for the rainout and,(or washout of portions of-the plume underLaccidental and routine releases. As explained in Section IIIB the exclusion of the wet deposition term from the calculations:

is consistent with the NRC approach outlined in Regulatory Guide and 1.4 4/.

provides more conservative estimates of ground level relativo concentra-tions and subsequently of radiation and inhalation doses.

The inclusion of the wet deposition term in the Gaussian dispersion model for SHNPP would have a small effect on the relative concentrations under normal precipitation events (about 0.10 inch /hr) and is well within the realm of accuracy of the Gaussian dispersion model. Plume depletion due to wet deposition under severe thunder-storm with a rainfall of about 1.0 inch /hr is estimated to be about 50 to 70 percent.

This corresponds to a reduction in the rel-ative concentration without the wet

. _ _ . _ . _ _ _ - _._ . _ . - _ . ._-- _._ _ _ _ ~ . _ - . _ __.. . - , , .

deposition term,- by a factor cf about two to three. The net result is a corre-sponding reduction in the radiation and-inhalation doses.

Applying mass conservation principles to the radioactive plume-it is logical to assume that the amount of radioactive material depleted from the plume in the air will be deposited onto soil, other sur-faces, vegetation, and water bodies. Thus, the portion of radioactive material that has been depleted from the plume and deposited onto various surfaces will be accounted for in the ingestion pathways.

The exclusion of the wet deposition term in the Gaussian dispersion model for SHNPP-is consistent with other conservative elements built into the model, as applied to air-borne atmospheric releases under accidental conditions, since radiation and inhalation doses are of more concern than ingestion dose and since radioactive material that has been deposited on surfaces can be contained better than radioactive material in air, (food chain can be better controlled).

4

.This-discussion. clearly shows that the.

exclusion'of the wet-deposition term from the Gaussian. dispersion model for SHNPP.

adds more conservatism to the'model.

4 e

N 4

d e

O

REFERENCES

1. 10 C.F.R. Part 100.11, " Determination of Exclusion Area, Low Population Zone and Popul~ation Center Distance."

2. .10 C.F.R. Part 50, Appendix I, " Numerical Guides for Design Objectives and Limiting Conditions for Operation to Meet.the Criterion 'As Low as is Reasonably Achievable' for Radioactive Materials in Light-Water-Cooled Nuclear Power Reactor Effluents."

3. 10 C.F.R. Part 50.34a.

4. U.S. Atomic Energy Commission Regulatory Guide 1.4, "Ascumptions Used for Evaluating the Potential Radiological Consequences of a Loss of Coolant Accident for Pressurized Water Reactors," Revision 2, June 1974.

5. U.S. Atomic Energy Commission, Regulatory Guide 1.24, l " Assumptions Used for Evaluating the Potential Radiological Consequences of a Pressurized Water Reactor Radioactive Gas Storage Tank Failure,"

March 1972.

l l

~ _ _ _ . . . . _ _ , _ _ . . . _ _ _ , _ _ - . .__

6. U.S. Atomic Energy Commiss' ion, Regulatory Guide 1.25,

" Assumptions Used for Evaluating the Potential Radiological Consequences of a Fuel Handling Accident in the Fuel Handling and Storage Facility for Boiling and Pressurized Water Reactors,"

March 1972.

4

7. U.S.' Atomic Energy Commission, Regulatory Guide 1.77,

" Assumptions Used for Evaluating a Control Rod Ejection Accident for Pressurized Water Reactors,"

May 1974.

8. U.S. Nuclear Regulatory Commission, Regulatory Guide 1.109,

" Calculation of Annual Doses to Man From Routine Releases of Reactor Effluents for the Purpose of Evaluating Compliance with 10 C.F.R. Part 50, Appendix I, October 1977.

9. U.S. Nuclear Regulatory Commission, Regulatory Guide 1.111,

" Methods for Estimating Atmospheric Transport and Dispersion of Gaseous Effluents in Routine Releases from Light-Water-Cooled Reactors," July 1977.

10. U.S. Nuclear Regulatory Commission, Regulatory Guide 1.145,

" Atmospheric Dispersion Models for Potential Accident Consequence Assessments at Nuclear Power Plant,"

February 1983.

. . w.

11. Sutton,.0.-G., "A Theory of' Eddy Diffusionfin the: Atmosphere,"

Proceeding R. Soc (London) Ser A, pp. 135-143, 1932.

~

12. Sutton, O. G., "Micrometeorlogy,".McGraw-Hill Book Company, New York, 1953.

13 '. Pasquill, F., "The Estimation of Dispersion of Windbourne Material," Meteorological Mag., Volume 90, pp. 33-49, 1961.

14. Pasquill, F., " Atmospheric Diffusion," 2nd Edition, John.

Wiley and Sons, New York 1974.

15. Gifford, F. A., "Use of Routine Meteorological Observations fbr-Estimating Atmospheric Dispersion, Nuclear Safety,"

Volume 2, No. 4, pp. 47-57, 1961.

16. Slade, D. H. Editor, " Meteorology and Atomic Energy, U.S.

Atomic Energy Commission," TID-24868, July 1968.

17. Lewellen, W. S., R. I. Sykes and D. Oliver, "The Evaluation of MATHEW/ADPIC as a Real-Time Dispersion Model," NUREG/CR-2199, ARAP (Aeromated Research Associates of Primcelon, Inc.) Report No. 442 RB, March 1982.

i 18. Hanna, S. R., G. A. Briggs and R. P. Hosker, Jr., " Handbook l

on Atmospheric Diffusion," DE 81009809 (DOE / TIC-22800), 1982.

l t <

l

.s.

l

~

~

' Ti_<c w w

.19. Sagendorf, J. F., J. .T. Goll'and W. .F.'Sandusky, " User < j i

Guide.for.XOQDOQ: Evaluation Routine Effluent s Releases at Commercial Nuclear Power Statu's,"

.. /

NUREG-CR-2919,'1982. "

s

20. Bander, T. J., " PAVAN: An Atmospheric. Dispersion Program for. Evaluating Design Basis Accidental Releases for Radioactive Materials from Nuclear Power Stations,"

NUREG/CR-2858, November 1982.

21. Hilsmeier,'W. F. and F. A. Gifford, Jr._, " Graphs for Estimating Atmospheric Dispersion, U.S. Atomic Energy Commission Division of Technical Information,"

ORO-545, July 1982.

22. Gifford, F. A., " Atmospheric Dispersion Calculations Using the Generalized Gaussian Plume Model, Nuclear Safety,"

Volume 2, pp. 56-59, 1960. ,.

23. Gifford, F. A., " Atmospheric Dispersion Models for i Environmental Pollution Applications," Air Resources

( Atmospheric Turbulence and Diffusion Laboratory NOAA, Oak Ridge, Tennessee, ATDL Contribution File No.

l l 7517, June 1975 (presented at the'AMS Workshop 1on' Meteorology and Environmental Assessment, Boston, i MA, October 1975).

l

vt ,

, ,,:'s l.

f .,

dj -

24. . Hosker, P. R., Jr.-, "MetN$ds'for Estimating Wake Flow from

..a-Effluent Dispersion Vear Simple Block Like Buildings,"

NOAA Air Resou'rces-Laboratories, NOAA. Technical Memorandum ERL ARL-108, Section 3.0, May 1981.

25. Halitsky, J.,~" Gas D.iffusion^near Buildings in Meteorology eind Atomic Energy," TID 25314, pp. 221-225, 1968.

~

26. Halitsky, ' J.', " Wake.and Dispersion Models for the EBR-II Building Complex, Atmospheric Environment," Volume II, pp. 577-596, 1977.

27. Pickard, Lowe, and Associates, Inc., "The Research Corporation of New England and General Public Utilities Service Corporation, Atmospheric a Diffusion Experiments with SF 6 Tracer Gas at Three Mile Island Nuclear Station, Under Low Wind Speed Inversion Conditions,"

GPU Report,- January 1972.

28. Sagendorf, J. F. and C. R. Dickson, " Diffusion Under Low Windspeed Inversion Conditions," NOAA Air Resources I '

Laboratory, NOAA Technical Memorandum ERL ARL-52, December 1974.

s

~)

+ , ,

. ~ ' *

..r.,

s ,

4~_ - -

+

~

11, ,

- M: ,, ,

g' ,,

29. ~ Wilson,'R. B'.,7.G.'E.-Start,.C( R..Dickson and N.:R. Ricks,.

y> -j '

." Diffusion Under Low Windspeed Conditions Near Oak f~f b  ?"

N Ridge, Tennessee," NOAA Air

• Resource Labor,atories, NOAA .

1 . Technical Memorandum ERL ARL-61, August.'1976'. .

, 30. Start, G.~E., N. F. Hukari, J. F. Sagendorf,'J. H. Cale and - -

C. R. Dickson, "ECOR Building Wake Effects on Atmospheric' Diffusion, Air Resources Laboratories," NOAA

\ . Technical Memorandum ERL ARL-91, November 1980.

q '

31. . Van ~ der Hov'en, I .",' "A Survey of Field.Mbasurements.of Atmospheric Diffusion Under-Low-Wind Speed' Inversion J

Conditions," Nuclear' Safety, Volume 17, No. 2, pp. 223-230, 1976.

m ,

• - 32. Van der Hoven, I., "A Comparison of Measured Versus

~ l ._ - - Model-Predicted Effluent Diffusion for Ground

\'t

( Releases," NOAA Air Resources Laboratories, NOAA i\

e- Technical Memorandum ERL.ARL-105, August 1981.

qg .y qe O .) ks _

3,3 '.' V,an der Hoven, I., " Measured Versus Emperical Techniques

.) -

\

to Determine the Plume Sigma -y for Ground Sources,"

s 3 NOAA Air Resources Laboratory, NOAA Technical i

Memorandum ERL ARL-120, .May 1983.

\

34. Chamberlain, A. C., " Aspects of Travel and Disposition of Aerosol and Vapa Clouds," United Kingdom Atomic Energy i.J' l l '

Authority, AERE-HP/R 1261, 1961.

l

. ]'

, ,1 -

l a D4 q s 3 kAI , } k 3 \f';

s

. . m i se ^s -

a

35. Slinn, G. N., "Some Approximations:for the Wet 2and Dry?- -

Disposition-of Particles ~and Gases 4 from the Atmosphere, Water,. Air, and Soil Pollution,"' Nuclear Safety, Volume V, pp. 513-543,(1977.

536. Slinn, G.-N., "Parameterizations for Resuspension and for-Wet and Dry Disposition of- ,

Particles and Gases for Use in Radiation Dose Calculations," Nuclear Safety, Volume 19,-No.~2,.

pp. 205-219, 1978.

37. Heffter, J. L., " Air Resources Laboratories Atmospheric Transport and Dispersion Model (ARL-ATAD)," Air Resources Laboratories, NOAA Technical Memorandum ERL ARL-81, February-1980.

38. Ritchie, L. T., W. D. Brown and J. Robert Wayland,

" Effects of Rainstorms and Runoff on Consequences of Atmospheric Releases from Nuclear Reactor Accidents,"

Nuclear Safety, Volume 19, No. 2, pp. 220-238, 1978.

39. -Ritchie, L. T., W. D. Brown and J. Robert Wayland, " Impact of Rainstorm and Runoff Modeling on Predicted Consequences of Atmospheric Releases from Nuclear Reactor Accidents."

i

=~

r  ? 3 - -  :

. s. +

u

40. - Ho ske r', R. P..,1Jr , " Practical Applicationiof Air.

Pollution Disposition Model - Current Status,. Data

' Requirements'and Research Needs," Air Resources Atmospheric Turbulence'and Diffusion Laboratory, NOAA,. Oak Ridge, Tennessee, ATDL Contribution 80/8, 1980.

41. Brenk, H. D. and.K. J.-Vogt, "The Calculation of Wet Disposition from Radioactive Plumes," Nuclear Safety, Volume 22, No. 3, pp. 362-371, 1981.

42. Committee on Atmospheric Turbulence and Diffusion, " Accuracy of Dispersion Models," Bulletin American Meteorological Society, Volume 59, No. 8, pp. 1025-1026, 1978.

43. "EPRI Mathematical Model For Atmospheric Pollutants,"

Appendix D, Available Air Quality Model, Electric Power & Research Institute, EA-1131, pp. 5 5-33.

44. Buckner, M. R., Compiler, " Proceedings of the First SRL Model Validation Workshop (November 19-21, 1980)

Hilton Head, South Carolina," DP-1597, UC-11, October 1981.

45. Gifford, F. A., " Turbulent Diffusion - Typing Schemes: A Review," Nuclear Saftey, Volume 17, pp. 68-86, 1976.

', T. . .

46. U.S. Nuclear Regulatory Commission,. Regulatory

Guide:1.23. .

Meteorological Programs in Support.of Nuclear-Power Plants."

I 1

l l

l i

r.

i

. - - . _ . _ . . _. _ . _ _ _ _ _ _ _ _ _ _ . _ _ ..__ _._. _ ._ _._.._ _ _ _