ML20080L394
| ML20080L394 | |
| Person / Time | |
|---|---|
| Site: | Harris  |
| Issue date: | 09/26/1983 |
| From: | Matthew Smith CAROLINA POWER & LIGHT CO., METEOROLOGICAL EVALUATION SERVICES, INC. |
| To: | |
| Shared Package | |
| ML20080L376 | List: |
| References | |
| ISSUANCES-OL, NUDOCS 8309300292 | |
| Download: ML20080L394 (19) | |
Text
.,_
=_...
~
1 T
UNITED STATES OF AMERICA p[0 NUCLEAR REGULATORY COMMISSION Before the Atomic Safety and Licensing Board %3 SB)29 Ri z3 r r e 7 r e r 3 t.
In the Matter of
)
_e'
)
CAROLINA POWER & LIGHT COMPANY
) Docket Nos. 50-400 OL AND NORTH CAROLINA EASTERN MUNICIPAL
)
50-401 OL POWER AGENCY
)
)
(Shearon Harris Nuclear Power Plant,
)
Units 1 and 2)
)
AFFIDAVIT-OF MAYNARD E.
SMITH 4
County of Suffolk
)
)
SS:
State of New York
)
,Maynard E. Smith, being duly sworn according to law, deposes and says as follows:
1.
My name is Maynard E.
Smith.
I am President and Principal Consultant for Meteorological Evaluation Services, Inc. ( "MES " ) located in Amityville, New York.
I obtained a Master of Science in meteorology in.1942 and have been engaged in the practice of professional meteorology since that time.
A copy of l
my professional qualifications is attached hereto as Appendix A and incorporated by reference herein.
i l
MES, under my direction, has provided meteorological consulting services to a variety of organizations since 1972.
l l
2.
From 1948 to 1972 I was Leader of the Meteorology 1
j Group at Brookhaven National Laboratory.
Our I
research team specialized in studies of micrometeorology l
and atmospheric diffusion.
In 1956-57 Brookhaven was f
requested by the Atomic Energy Commission to undertake a basic study of the safety of nuclear power plants.
i 1
My colleague, Irving A.
Singer, and I served as the 8309300292 830927 gDRADOCK 05000400 PDR 1
_ _ __g
e b
O D meteorologists on the Brookhaven task group that conducted the study.
Our portion of the work became part of the overall report, and it was also published by the American Industrial Hygiene Assoication under the title " Diffusion and Deposition In Relation to Reactor Safety Problems" (Smith & Singer, 1957).
This paper is the published ver-sion of BNL 3391, referred to by Wells Eddleman'in an August 31, 1983 " Motion for Partial Summary Disposition on Eddleman Contention 80" in the above captioned proceeding.
A copy of our paper is attached hereto as Appendix B.
3.
On page 2 of the Eddleman Motion, the statement is made that:
I Likewise, there is no dispute that rainout can increase nuclide deposition several orders of magnitude (i.e.,
100 to 1,000 times).3 Reference 3 in the above quotation denotes BNL 3391.
M r.,
Eddleman then concludes that 10 C.F.R. Part 50, Appendix I dose calculations for emissions from routine plant opera-tions at the Shearon Harris Nuclear Power Plant will be at least 100 times greater than calculated by Applicants or the Staff because Applicants dispersion model fails to account for rain.
4.
The attempt to utilize BNL 3391 to support the above quotation reflects at the very least, a complete lack of appreciation of the differences between the hypothetical situation being addressed in the paper and the diffusion, depositicia and rainout from either routine or accidental releases from a modern nuclear power plant.
I t
~
l
~
. 5.
The Smith-Singer publication was a summary of our best judgment at the time about the fate of radioactive gases and particles that were hypothesized to become airborne as a result of a major accident that would rupture the con-tainment of a nuclear power facility.
Even though we were unable to specify exactly how such an accident might occur, we assumed a very rapid release of a -large quantity (50% of a
the total fission product inventory) of mixed fission prod-ucts into the atmosphere.
We then considered ths various mechanisms for diffusion and deposition by meteorological processes.
Thus, our paper did not address and is not in any way relevant to either the routine release of radioac-tivity treated in current Appendix I evaluations or even postulated accidental releases from a modern, elaborately contained nuclear plant.
6.
In contrast to the hypothesized major accident at a gas-cooled reactor evaluated in the Smith & Singer paper, where we assumed both a breach of containment and a rupture of the reactor vessel, routine emissions from a modern nuclear plant involve very small amounts of chemically active gases and small particles.
The magnitude of any rainout depends upon having some inventory of gases and particles present in the atmosphere.
If they are not present, no amount of rain can deposit them.
Furthermore, the larger the particles, the more effective would be scavenging by precipitation.
Note that in a modern nuclear plant, even in the event of an accidental release, most particles and gases that could be deposited by rainfall would have been removed beforehand by the containment and filtration sys-tems.
The situations are simply not comparable.
. 7.
In contrasting the 1957 accident calculations with and without rainout, Smith & Singer were discussing a very short-term event, the important part of which would last no longer than an hour.
We sought to describe the combination of diffusion and rainfall rates that would maximize the deposition and dose calculations.
The Appendix I dose calculation for routine emissions from a modern' nuclear power plant would not lend itself to such maximizing as-sumptions.
Appendix I dose calculations take into consid-eration dose from various pathways over the period of a year, not the dose available immediately after an accident.
At any given site, rainfall occurs at varying rates during some percentage of the year.
The geographical distribution
-~of rainf all around a site is usually also variable, tending to occur more or less frequently with various wind direc-tions.
However, there is no "one area" or " hot spot" that potentially would receive a sharply disproportionate share of the radio-active deposition.
An area receiving an unus-ually large share of rain in one hour or even during one season will inevitably receive less in another, resulting in the long-term smoothing that is reflected in Appendix I calculations.
8.
I emphatically disagree with Edleman that BNL 339' supports his assertion that rainout can increase radio-nuclide depo-sition dramatically, except in the very restricted context outlined in that paper.
Furthermore, I am confident that no other document in the reputable scientific literature would in any way justify his assertion.
9.
Finally, I would like to record my concurrence with the approach employed by Applicants for the Shearon Harris Nuclear Power Plant in making atmospheric dispersion and dose calculations that have been chal-lenged because they do not take into account rainout.
In my judgment a rainout adjustment to these calculations would result in a trivial change.
~
~
Maynard E.
Smith Subscribed and sworn to before me this day of Sept, er, 1983.
\\
d'M M
fotM[ Public MARTIN E. LEONARD NI.io.f 5 My Commission Expires y ' % cGia w x: w 4
l
,.,__.. _. _. ~
c APPENDIX'A MAYNARD E.
SMITH Education:
University of Chicago, Additional work on general circula-tion, single-station analysis, 1942 New York University, MS, Meteorology, 1942 Princeton University, BA,' Economics, 1941 Experience:
Meteorological Evaluation
- Services, Inc.,
"Amityville, New
~
York, 1968 to Pre,sent, President, Principal Consultant Mr. Smith has provided consulting assistance t'o a number of industrial and governmental organizations since the early 1950's.
In
- 1968, Smith-Singer Meteorologists, Inc.
was founded and Mr. Smith became President.
The Company's name was changed to Meteorological Evaluation Services, Inc.,
in 1977.
The Company provides advice and assistance in meteorological and air pollution problems, including atmospheric diffusion studies, design and evaluation of stacks and abatement facil-
- ities, processing and analysis of meteorological and, air pollution survey data, evaluation of wind loads on structures and the preparation of environmental reports.
It _ also con-ducts a variety of applied research projects, such as, field evaluations of cooling tower plumes and building downwash, and the analysis of large-scale pollution patterns.
Brookhaven National Laboratory, Upton, New York, 1948-1972, Leacer, Meteorology Group The original objectives of the Meteorology staff centered on understanding and forecasting the dispersion conditions af-fecting the reactor cooling air.
Following solution of these problems in 1952, the activities were redirected toward basi'c research in low-level diffusion and deposition.
Prominent aspects of this program have been detailed investigations of diffusion from elevated sources, studies of the low-level wind speed structure, and the development of specialized meteorological and sampling equipment.
The study of particu-late deposition over grids of samplers in the open terrain and in forests was also important.
Significant by-products of these studies have included in-struments, techniques and procedures for applying the results to practical problems in air pollution, not only in the atomic energy field but for industry in general.
METEOROLOGICAL EVALUATION SERVICES,INC.
w-
~.
o MAYNARD E.
SMITH American Airlines New York, 1945-1948, Supervisor of Mete-orological Staff, New York Region Staff provided terminal and route weather forecasts for oper-ations in the Northeastern United States.
United States Air Force, 1941-1945, Major -
Research in upper atmospheric analysis and forecasting.
Development of new weather service for the 12th Army Group in Europe to pr. ovide meteorological data and forecasts for a variety of ground-force activities.'
~
Professional Organizations and Committees:
American Meteorological Society Air Pollution Control Association American Meteorological Society / Environmental Protection Agency Steering Committee on Atmospheric Diffusion Modeling, 1981-Invited Participant & Panel Chairman, AMS-EPA Workshop on
" Quantifying and Communicating Uncertainty in Regulatory Air Quality Modeling," Woods Hole, MA, Sept., 1982.
Invited Participant & Panel Chairman, EPA Workshop on "On-Site Meteorological Instrumentation Requirements to Charac-terize Diffusion From Point Sources,"
- Raleigh, NC, Jan.,
1980.
Invited Participant & Panel Chairman, EPA Workshop in Rough Terrain Modeling, Raleigh, NC, July, 1979.
Invited. Participant, EPA Conference on Modeling Guideline, Argonne National Laboratory, February, 1977.
Steering Committee, Large Power Plant Effluent Study, Air Pollution Control Organization, Environmental Protection Agency, 1969-1973.
- Chairman, Task
- Group, Dispersion of Airborne Effluents, American Society of Mechanical Engineers, 1966-1968.
METEOROLOGICAL EVALUATION SERVICES,INC.
r APPENDIX B Diffusion anc Deposition in relation to reactor sa etyproblems MAYNARD E. SMITH end IRVING A. SINGER Brookhaven National Laboratory Upton, Long Island, New York AT THE request of the Atomic Energy figure for this probability was Included, Commission, Brookhaven National since the various means of deriving such Laboratory has recently completed an in-figures - (1) past experience, (2) com -
tensive study of the hazards associated posite probabilities of successive or simul-with nuclear power plants. Instead of con-taneous events, and (3) average valu,es of fining the investigation to an individual considered opinions - all failed to produce reactor in some particular location, an at-useful information.
tempt was made to derive conclusions It is surprising that with the variety of applicable to the majority of power re-power reactors, containment schemes, sites, actors and sites now under consideration. 'and population distributions currently un-The study was also unusual in that it was der consideration, anything typical could be a genuine group effort and represents a chosen, but it is believed that the hypotheti-wide variety of viewpoints and disciplines, cal model selected contains quite representa-Officially, nine individuals are listed as the tive features. The reactor was considered to study team, but more than a score contrib-be a 500.000 thermal KW unit, fueled with uted substantially to the investigation, and On utilized in a fuel reprocessing cycle of many others provided helpful advice and 180 days. Such a reactor was calculated to criticism. The full report will shortly be contain total fission product activity in the I
available from the Government Printing amount of 4.1 x IOS curies. Some of the Office, under the title " Theoretical Possi-important constituents are as follows:
bilities and Consequences ~ of Major Acci-Iodine 5.0 x 107 curies
~
dents in Large Nuclear Power Plants."
Noble gases 3.4 x 107 curies The purpose of this paper is to review Strontium 9 1.7 x 107 curies the role of atmospheric diffusion and depo-Strontium" 3.0 x 103 curies sition processes in hazard evaluation, but Ceriumm 8.0 x IOS curies a few comments concerning other parts of Two types of accidents resulting in a the work are both interesting and necessary release to the atmosphere were postulated, as background information. Any reactor The first assumed a breach in the container, hazard analysis has three natural subdivi-allowing volatile fission products and part sions, dealing, respectively, with release, of the strontium inventory to reach the at-distribution, and consequences of the dis-mosphcie. The second was a major accident charge of fission products. Clearly, the accompanied by enough heat to rupture the study group concluded that a major release containment vessel, from which 50% of the
- was possibic, despite the elaborate control entire fission product inventory escaped. It devices and containment vessels included in is not important in terms of this paper to all power reactor designs. Had this not been specify precisely how these events were pre-true, there would have been no need for the sumed to occur. Mechanisms for each could work on diffusion and deposition. It must be visualized, and neither could be ruled out be emphasized, however, that a large scale as impossible. It is significant, however, release of fission products was considered that both were considered improbable, with an extremely remote possibility. No specific the major 50% release significantly less
~ '
Research carried out under the auspices of the United probable than the volatile fission product States Atomic Energy Commission.
release.
I
-.,r c
y-y
- ~= -.
site, shown in Fig.1. The two-mile valley 1.tsr or Sntsor.s width with 200-foot ridges parallel to the symbol Description U nits.
oMMaMn h pdy imaginative, but in no way unrealistic.
h height of cloud m
From a survey of U.S. climatological n
stability parameter dimension-records and available micrometeorological studies, a synthetic microcliraatology was i
mean dosage eu developed and is presented in Tables I and l
sec/m3 II. The first entry in Table I may be ex-Q or Q pollutant emission curies C*
horizontal diffusion pected to surprise individuals unfamiliar parameter m=/2 with microclimato19gy, for the temperature C,
vertical diffusion inversion is still generally thought of as an parameter m*/2 uncommon phenomenon. Actually, the as-is mean wind speed m'see signment of 50ch of all hours to the inver-x distance downwind m sion category is quite in keeping with micro-y distance crosswind m climatological st'.: dies such as those at
- Savannah hd and hoohann? Oak of dep t on uries/m2 Ridge datas would even suggest the possi-P rate of rainout curies >me V'
velocity of deposi-bility of a value greater than 50th for a valley location.
tion em 'see A
proportion of cloud The selection of typical lapse rates and removed per sec.
mean air temperatures needs little explana-by rain 1/see tion or justification. There are wide varia-
'Those not specified are included in the text tions in both over the United States, but the
**W"-
numbers used are realistic. The same may be said of the percentage of hours of rain-From the meteorological point of view, fall and the association with lapse and the task was to translate these releases into inversion cases. The normal and moderate cloud dosage and particulate deposition rainfall rates were selected from unpub-downwind of the reactor site. Thus, the ob-jective was definition of the distribution of C
N l
~
the fission products. Evaluation of the con-sequences of contamination formed yet an-h {j niven other part of the study, and is not consid-q ered in the present paper.
zoo rf i
f RIDGES Q' (
6 I
Site and Climatology N
THE FIRST step was to establish a typical i
site complete with appropriate climato-
}
logical statistics. This, like many other of
. k{
city 1
the generalization problems, proved simpler
)y than had been anticipated, and the very fact f
that the reactor was restricted to electric
)$
){( o power generation helped considerably. Such ip 2p a reactor is usually designed to serve a mu.
f uitts f)4 k
nicipal area, and would logically be placed at
) (i i
a distance from the population center com-mensurate with economical land and trans-
{
,-Possiett atActon mission costs (30-35 miles). A substantial LOCATION source of fresh water is a necessity, so that f
e" it was also possible to specify location near
{
a lake or river. Of the two, the river is be-lieved more probable. These facts permit
$tCT N
FL0 construction of a rough map of the idealized Fig. l.
lished studies at Brookhaven.
Imasiaary reactor location la 4. United Sta++s.
~--n,,-
-,-e--
m e--
, -~-
w----,-
-w+-
x--,~,m n.
lapse cases, respectively, were chosen to Tm 1.
I*
METEonotocICAL PARAMETERS FoR AN the ground over relatively long distances.
IDEALIZED SITE The upper level winds would not be expected mnn.1 stamt*
to show any important diurnal variations, L*n" 1"'y**
and were chosen as 15 m/see for all cases.
(nonn.ur egennany Tables I and II provide all the usual
- 'in*y"'
"fr's"'
meteorological parameters necessary for the rrecener er m
occurrenee trercenti night. al )
morning) computation of the probability of various 1
Lapse rate (*C/100 m)
-1.0
+ 1.0 weather conditions existing at the time of 1
Mean air temperature an accident. While a much larger number
(*F) 60 40 of specific conditions should be investigated Ilours with precipitation to give a complete representation of the (percent) 13 2
probable fate of'the fission products, it is Mean wind speed at ground-believed that the range of injury and dam-age reflected by various combinations of M
in s d at 400-800 these includes about 954 of the possible meters (mJsec) 15.0 15.0 cases.
Total annual precipitation 40 in/yr Most probable rainfall rate
.0.02 in/hr One important feature of the meteoro-Rainfall rate exceeded by 10 logical analysis, as well as much of the other percent of hours..
.. 0.15 ins hr material in the report, is that there was a lack of conscious bias toward pessimistic TABLE II.
values of the parameters. The meteorological WIND DIRECTION Fr.EQUENCIES FoR AN approach, in all its details, was completed
^I (pfeen before any of the consequences of the as-sumptions were known. It should not be cround. Level Aloft (4oo. soo m) inferred from this that the selection of Dir-twn L.p.e Inver. ion L r., and Inver. ion theoretical models and values of the param-eters is considered good. On the contrary, North 7.5 22.5 10 Northwest 10.0 5.0 20 many portions are very uncertain, as will West 5.0 5.0 20 become evident in the development of the Southwest 7.5 5.0 20 various phases.
South 10.0 2.5 10 Southrast 2.5 2.5 5
Initial Cloud Behavior F the most important and most East 2.5 2.5 5
ONE Northeast 5.0 5.0
~^
dimeult portions of the meteorological Totals 50.0 50.0 100 study was specification of the cloud behav-ior immediately after release. This was true The upper level wind distribution in the both because of the difliculty of specifying final column of Table II reflects a predomi-the exact nature of the release from the nant westerly flow very common in middle reacter, and because of scanty knowledge latitudes. The division of the ground-leul concerning the behavior of small clouds and winds into lapse'and inversion hours is in jets. In terms of the latter, the difficulties accordance with the well-known " channel-stem largely from the lack of suitable field ing" effects of valleys, and is especially tests for comparison with existing theoreti-noticeable in the 22.59 frequency assigned cal models. Much recent attention has been to northerly (down valley) inversion winds. devoted to nuclear bomb clouds, but these It is also common knowledge that a relation are too large and too hot for the current usually exists between wind direction and scale of interest. Data of the right scale, rainfall, but this is such a widely varying derived from chemical explosions or combus-feature of climatology that it was decided tion, are usually by-products of other stud-to consider rainfall hours independent of ies and do not include suflicient information wind direction.
for an acceptable analysis.
l The wind speeds of 3.0 and 5.0 m/sec In this study, the equation of Sutton* was selected for the ground-level inversion and selected for the daytime ascent and the Hol-I l
l l
1'
l land modification 4 was applied
- ggy, to nocturnal conditions. These DIrrestox PAnAMETEas UsED IN THE STimY predicted heights of 860 meters in lapse conditions and 400 me-c, c,
u ters during inversions for clouds ygggea Hgry y
,,g of su'ficient heat and pressure to rupture the containment shell.
Typical 0
0.25 0.40 0.40 5.0 It was recognized that release I. apse 800 0.25 0.40 0A0 15.0 of volatile products probably Typical 0
0.55 0.40 0.05 3.0 would not be hot, and also that Inwrsion 400 0.55 OA0 E05 15.0 the major failure could occur slowly enough to prevent any effective as-is substituted for the rate of release (Q),
cent. The study therefore proceeded on the result, x, is given in the. convenient the basis of essenthily three types of dosage unit of curie-sec/m:. This is of release: (1) a hot cloud rising to 860 meters direct use to individual.s evaluating the con-during the daytime, (2) a hot cloud reach-sequences of an accident.
ing 400 meters at night, and (3) a cold The added considerations necessary in a cloud at ground-level regardless of the time radiation study, such as attenuation, decay of day.
and build-uri factors, were all included in The effects of the initial behavior are far-the establishment of specific dosage values reaching in terms of both direct dosage that were used to delineate injury and dam-from the radioactive cloud and deposition age categories.
of particulates in dry weather. In succeeding Selection of appropriate values for the sections this is illustrated by an analysis of diffusion parameters was hampered by the the major release only, The volatile case same lack of appropriate field data that had I
can usually be treated as a simple percentage been a problem in determining the ascent of the cold 50% release, and is of no partic-of the cloud. Either measurements had been ular interest in the analysis of distribution made from a source which was too large, as such.
such as nuclear bomb tests, or low-level sources had not been studied to sufficient "80" distances. Clearly, a cloud ditIusing in the MANY attempts have been made to derive stratosphere is of little use, and the extra-expressions for the diffusion of gases polation of data such as Brookhaven's, which or aerosols in the atmosphere. It is question-is measured only to within 10 kilometers, is able that any can be defended rigorously questionable. Another limitation brought on theoretical grounds, but several could be forth in the analysis of diffusion data was used with acceptable accuracy if appropriate that most studies have been made under values of the diffusion parameters were in-lapse conditions. The nocturnal inversion troduced. The treatment of Sutton* seemed case, which occurs approximately 50% of especially appropriate, since it possesses the the time, has seldom been studied quantita-flexibility necessary for the study, and is tively. Fortunately, it was possible to obtain familiar te all engaged in micrometeorology.
some information concerning the dispersion In particular, the basic equation for a con-over long distances for the inversion case tinuous point source was used as the diffu-from Thomas 30 of the Tennessee Valley sion model:
Authority. These were helicopter measure-ments of the concentrations of SO2 up to E (x,y,o) =
exp. -
iin C'C, x -=
50 miles from a steam plant. They supple-ment local inversion studies made at Brook-
~~
M*- ], h2 havent with Kytoons. Both are continuous C2 x -=
C2 x,
sources rather than isolated clouds, but they z
(1) have helped greatly in the final selection of As had been stated, it is diflicult to speci-the parameters given in Table III.
fy exact nature of the release, but if the S nee the diffusion parameters are criti-total quantity of fission products in curies cal to all aspects of the study, it is impor-333,y g i,,,, a,,i,,a i, i3, Lt.t or symboi -
tant to examine Table III carefully. The O
c..
.,.___-s m.
_. ~, - - _.,.,, _
most noticeable feature is that no allowance has been c'r' i:
o' ',',$$3; ;;yc;5.,y,op,9ll,r,vst cLoa made for variation of the
[
parameters with height, ex-L g
cept in the case of the j
\\
.csc.w m..... *ci mn mean wind speed (u). This a[
gg';5,I, h,,,*:!n..,,,,u,',.',,,,
is based on the fact that i
the diffusion occurs over,
greater distances than any i
\\
"'d''"**
._.,;i"yl,8 l,"J,*'!c'f;.','..a previously studied, and it i,
.eun... n ce..i. n eg i
is doubtful that the varia-g **i-
' 's O' D'l "" " """'
1' tions with height shown
'S I"c*."u'n'".n'".,."..*.*..,'
i in short-distance tests are
\\'.,s e
pertinent. In terms of wind speed, it is well known that
. \\'s, i
',e an important increase usu-
,oc
~
ally exists in approxi-
.+u mately the first 2000 fect above ground, and this is Fig. 2.~
reflected in the values Th. dos.g. dir.c,Ir do-n.ind of th ructor er. sko n, since d8 Position and r.inout.ctu.Ily remove. portion of the particles, shown*
the correrted curv.
.r. included.
h.o serious quarrel is an-ticipated with the use of
'e :
o,no e e. 4s.e amets 0.25 for n in the daytime t
r. ca. ocrw
.6 c.our.
saev e-te <t 6 co o.v e s conditions,since most inves-tigations have shown val-s
\\
i n$< + m 'c a
ues close to this. The same
!!!,0*,7.. I !$
for both C, and C, during
-3 s gryl;,,,ol!@;,i g ~.-
be said of the 0.40 value
,\\
lapse conditions. The use I -
','\\ 3
.'\\
.\\.
\\
of 0.55 during the inver-4 sion case is considerably E
$ j:s
- t'"" *llJ'.!%3'l',1%
8*9 Ni less firm, and the main
\\
\\\\.
justification is that it fits the TVA data rather well.
"-~ O'llEU "*'"*
t.,
ip The choice of 0.05 for the
-- E*;F.'i','" "*'"*
..D, s-
,3.
nocturnal C, again is in rough accordance with the
, y,L,e......... */u.,
meager information avail-m able. It is almost certainly F. " '"
ig. 3.
no larger, and it may be Th. dos.g. dir.ctly do-n irid of the r..ctor er. sho n. Sinc.
as small as 0.02. The use d.poimen and reinout.etually remov.. portion of th. particles, of 0.40 for C, during the the correct.d curves
- r. included.
inversion retlects the belief h
that strong horizontal wind shear is main-night and day is immediately evident, for tained even during very stable conditions. important dosages extend to hundreds of The importance of these parameters is kilometers at night as compared to appro::i-illustrated by the dosage configurations mately 10 during the daytime. This is as-l i
shown in Figs. 2 and 3, in which the solid sociated with the very small vertical diffu-i lines represent the dosages directly down-sion (C, = 0.05) and the very large stability wind of the reactor for a cold release in parameter (n = 0.55) applied to the inver-which the cloud centerline begins and re-sion condition.
mains at ground-level. This approximation Figs. 2 and 3 are indicative only of condi-is obtained by settling y = h, o in equa. tions along cloud center-line. Obviously, the tion (1). The very great differ. ice between width must be defined if the rature of the l
9 0
e
i i
e
\\
also, the difference between lapse and in-
. a di.
l
- [.
.,n.
s.. u u..c w nu n as.u a.i. caa "!
version is evident in terms of distance, but in neither case does the dosage reach the
't-minimum damage level.of 10 c-sec/m,
~"
s Deposition and Rainout I
r,'
.. ! DIFFUSION studies provide a basis for 3
!.L evaluating the direct effects of the cloud l
as it passes over the countryside, but they
%% Y.':$'.'"
T J
give no indication of the particulate residue
"~'" M;'.
y that may be transferred to ground surfaces,
-u vegetation, and buildings. The term "trans-
,i I
ferred" is used initially instead of deposi-
-a tion or rainout to suggest the lack of quan-Fig. 4 titative knowledge concerning the processes isolines of constant dosage (C-sec/m3) are that are actually involved. It is known that shown in a horizontal plot. No correction for small particles may be influenced by many deposition or reinout is included.
forces other than simple gravitational set-tling. Ranz and Johnstone5 have shown that
.i.
.c.. [.. n..e..$,.:
n.u. n.i e, e.a.i. caa Impaction, electrostatic and thermal forces I
-u
{
may be equally important for lg particles u..-
u
.e" This is also shown by the recent experiments l
in which deposition of uranium fume was
,I,.
- - - - ~ '
I measured. Similarly, a simple treatment f
!'~
.u.'*..
' 'I probably does not describe scavenging by
- rain where the hygroscopic nature of the
- eg Q.
- [L particles may be as important as their size
}
- ,
- '.*::" N.,tt ~
and shape. Forms of precipitation other
.p
- u l
'I TABI.E IV.
'=
SUGGESTED LI5!!TS
'=
=
aa n
=
=
RADIATION DOSAGE AND DEPOSITION Fig. 5.
Similar to Fig. 4. except for diWerent distance F
seafe.
7,p Category C-sec/m3 hazard is to be evaluated. This was accom.
A Lethal exposure
>400 B
Illness likely 400 - 90 plished by computing dosage isolines. Those shown in Figs. 4 and 5 represent significant I"IT #*IF'm"ay*
- "l'
[ nc r,enu P
dosage limits as specified by the health e
90 - 10 physicists and biologists of the Brookhaven D
No injury or expense
< 10 team (Table IV-). The important point to Range C/m2 note is the tremendous difference in area I
Urgent evacuation resulting from changes in diffusion param-(within 12 hr)
T eters. The 10 c-sec/m3 isopleth, for example necessary
>0.2 encloses an ellipse roughly 600 x 8 km dur.
II Evacuation necessary
> 10-2 j
ing inversions as compared to one 13 x 2 III Severe restrictions on km during the day, land use, possible
, The other interesting feature of the diffu-t on restric$on on slon analysis appears when the hot cloud outdoor work 10-2-103 sources are compared to those at ground-IV Probable destruction level. Figs. 6 and 7 are plots derived from of standing crops.
precisely the same assumptions as those restrictions on agri-i used to obtain Figs. 2 nd 3, except for the culture for first year 10 10-4 t
cloud heights and the wind speeds. Here V
No expense likely
< 10-4
)
l l
i i
f i
_,.,-.....w.-
~
l ic' ic a
j DIFFUSION oF FIS$ ION PRODUCTS DIFFU$r0N OF FIS$0N PRODUCTS HOT CLOUD ALorf MOT cloud ALOFT TYPICAL DAYTiwE CONDITION $
TYPICAL NOCTURNAL CONDITONS d
m d
e pertYs
~
UttTs og gas sol e f a to' Cue'Es
- fleas I0l e Is@'
stits j
[g og
...i n.s
.....i n.s
.gg
=:=.e:.n=-
= = e:.,.:.= -
d i
r m* -
"s
)
N I
i O
L i
i
.l ge r g.
I e
s N
\\
t ri,es i
i
-e oisva no... =o to
.oo oista=ct oo..imanito.ctres
.o. tes
. m,i,
.g
..g i
l Fig. 6.
Similar to Fig. 2. except that dosage does not Fig. 7.
occur continuously at the ground. Curves ad.
Identical to Fig. 3. except for meteorological justed for reinout and deposition have not been conditions, inclu3ed since the maximum dosages are so low.
Paint pigments, such as zine or cad-tha'n rain present even more complicated mium sulphide, have been used largely as problems.
tracers rather than for quantitative anal-Unfortunately, scientific knowledge of the yses.
right type for this study is even more inade-In any case, it is apparent that little quate than that applying to diffusion. The justification exists for the use of analyses most probable reasons are that complete more sophisticated than simple settling ap-field experiments in deposition and rain-out proximations until more research has been of small particles are extremely difficult to completed. Accordingly, the straight-for-conduct, and there has been little need until ward approach to dry deposition and rainout j
the present. All the limitations confronting prepared by Chamberlain was used without 2
meteorologists mentioned in the section on alteration in the study. In dry weather it is b
- diffusion can be repeated here, together assumed that the small particles are brought with a host of others. Theoretical work and close to the ground by turbulent diffusion, laboratory studies have dealt with idealized and that deposition occurs from the lower spherical particles under conditions unlike portion only. This is stated in equation (2),
those in the atmosphere. Field experience which is simply the basic diffusion equation with bomb tests has covered a satisfactory multiplied by a velocity of deposition (V,)
range of particles, but the most interesting and corrected for the removal of particles sizes froir, the point of view of reactor haz-by the first exponential term.
ards are largely confined to the strato-A serious question arose in regard to the sphere.
probable particle size range of fission prod-
-n e
-n
m.
1
^
~
.I I
l 2Q, V, 4V,x*/2 TABLE V.
6=_
exp exp PARTICLE S!zt DisintscTroNS' nur /20,.
t uwC, C,x2-n Perecet Percent Si,ze by number by weight y2 (h2)
(2) Group having 1.Op j
0.5 95 40 C2 x2-a C2 x2.n mass median diameter e 1.5 5
60 r
=
ucts released from the reactor. it seems 1.5 15 1
i most probable that a release would occur as a result of or in combination with combus. Group having 7.0 3.5 60 16 mass median diameter 7.0 23 45 t
tion, with the particles having the general 15.0 2
38 characteristics of a fume. The size dis.
tribution fitting this description would cer-
- Typical fume and dust distributions are tainly be very small. However, one cannot shown.
rule out an accident of different nature giv-ing a much larger particle size distribution.- much the same relationship with meteoro-Table V shows the two distributions selected logical conditions, in that the important for consideration in the study. Figs. 8 and 9 deposition values extend much ' further represent the deposition for these two par-downwind durir.; inversions. The small par-ticle size groupings during day and night ticulate group deposits rr. ore slowly, and conditions from ground-level clouds. The results in contamination smaller by a factor figures are essentially duplicates of Figs. 2 of about 100 at any distance of importance and 3, except that deposition rather than during the daytime. The depletion of the j
direct cloud dosage is plotted. They show DEPOSITION OF FiS$ ION PRODUCTS i
,o "
4 GROUND-LEVEL CLOUD OEPCSITioN OF FISSION PRODUCTS l
TYPICAL NOCTURNAL CONDITIONS '
GROUND-LEVEL CLOUD 1
TYPICAL DAVTiUE CONoit?ONS wts 9
"?,'.'s, " ' 'f '** o n"
. cites g,. g ao,
- a
=om.,-
"YI ep = ass utoia=
't..t.e v c =
I o
o a cran q
[#
r 4
I b 5;
l o".*!St"g'S"
unas ef i<, ;wie >r:te" gg raaa-treas y d,~
Wo' ',!A.', * **
\\Y'o.a.,.'!.'.
't "
,e J
as r.
"lEmsae= roa mort oc*osaen roa votarat Ev Utvi I*26rI'v"iIE vYtu'IsY,*Eaa N'o'Ed=IE,cs 'w'v c a-etviu vau a 4,,
,ooo 4
ovsta=CE ocomm840.R.LOWEftes
(
o'sTa E.0**w s40. R.L YE#s
-l,
o
~ia i
Fig. 9.
Fig. 8.
Dry weather deposition is shown for the two Counterpart of Fig. 7. Note that 1.0u particles particle size distributions shown in Table 111.
become most important beyond 300 km.
e-,
.---m e
.m.
cloud of larger particles becomes noticeable at great distances during inversions, how-
?\\. \\- - - i
"'""o',Ttho"r?"* E ever, and beyond 300 km the I cloud gives
'c DaE co*Dons -
greater deposition than the 78 simply be-
'\\,
plgEIk=',yri _
- aa;A*,,$g
- g;;;p,g,,,'=q cause there is much more material remain-N aa m<=
ing in the air.
[
'-\\
Plots of the deposition curves associated with the hot cloud cases are not included,
\\
for they add nothing to the discussion. De-N position from the 7p group exceeds that
'N Y
'\\ g 3
from the 1p cloud at all distances and levels I
s of interest, and in neither lapse nor inver-
)
}
g
\\,
]
N heavy deposition.
1
,h',
sion cases are very large areas afTected by Rainout has been treated in a similar g.
N,\\
_i manner using the same two particle distrib-
[
_. ' f a,5,5,,,"ga 4
y h
"8 g".,y" ',a "y",#
'\\,
utions and settling rates, and considering all precipitation as rain. Chamberlain's2 s -
formula describing the process is as follows:
i E aE P=
exp -
E 5"J'II EE#
\\ 'i-Q, cxp (-A/u) y2
'o
)
L2/2 C x2 n/2 C2 x2 n (3) r y
This equation ditTers sharply from (2) in that all terms associated with height (h) or ie,
> ~;
" i, g
vertical diffuson (C) disappear. This simp-o $=ct,oo*"*a ~T5a5
'co ly means that the relatively rapid fall of the
,,, a.
raindrops brings any collected particle to p;g, 3 0, the ground almost instantly, no matter what cur.. for th.
t.. p.r+;ct, i::. d;stributions its initial location within the cloud. The im-and two retnf.H rates are sho-n.
plication of this is very important. No advantage is gained from the rise of a hot It is necessary to compare carefully the cloud, so that all releases are effectively plots for the various meteorological condi-treated as cold except for the important tions, particle sizes and rainfall rates to see variation of wind speed with height.
the possible ground contamination patterns A second important feature becomes ap.
(Figs.10-13), IIowever, it is almost im-parent only when the data for various cases mediately apparent that the lutfal height are plotted. This is the very marked attenu. of the cloud offers no advant age. In fact, ation of the total content of the cloud by the comparison of the hot and cold eses during rainout process, given mathematically by nocturnal conditions reveals t!.e the ele-vated cloud results in greater wosition A
the term exp ( 7). A table of A values except very close to the source. Thu inay be as a function of particle size and rainfall attributed primarily to the greater wind rates is shown in Table VI.
speed aloft, which carries the cloud further before the same percentage of depletion is achieved.
TAB 12 VI, The rainout process is such a complex RAINOUT or SMALL PARTICIIS relation of rainfall rates, diffusion parame-p cloud nemoved per second n>
ters, wind speeds and particle sizes that it p,,,
Diameter Gl Rainfall o.02 in/hr Rainfal10.15 tn/hr is hard to ilt any very simple rules. In most 0.5 1.0 x 104 2.0 x 104 cases, however, small particulates may have 1.5 1.5 x 104 3.0 x 104 much more serious consequences in terms 3.5 6.0 x 104 3.0 x 10-4 of the size of the contaminated area. Re-7.0 1.5 x 10 4 7.0 x 10-4 strictions on agriculture (104 to 10-4 c/m2),
15.0 2.0 x 10-4 1.0 x 10-s for example, would extend further down-l l
io
.. \\
Mi & FM WT5 }
'o
,,,,n, g
.y3 E
g\\
RA%Out oF FiSSiCN PRODUCT 5 :
g
\\ caotno ttvtt ctovo
(',
g\\ tvPiC A CAvitut,, CONDITIONS g
g 3vy3.Livtt CLouo
\\tTPICAL NOCTU5that N
coNomoNs N,
o.,,
.\\
4 ;rg,.,sa,io.
Teyes s
., \\
.....s c,..
i
\\\\
o'" E.'.I'I.. """'
k
]
,.h.. \\.
T 2
\\,
\\
f
'\\
3.\\.
\\ '.
,.o' r i
.l
,g.
N
~
unns.c. cun ts
\\
\\,
e
..ra.sr esi.
}
i s
i s
. o r.a.. tr==
c,.o ss
_o,,.,...,.,,<a.s
\\
,g
\\
,g
- ,c.ie..- -- e...ss..c o..=
\\\\
l o.o 3
'. o sm.i.oir,i.'i.olue
's' i
\\
i i
\\
e,s.u,. c.a.s ctouc
,o o
v
\\
g
~
4
,o "
{
- e ns.,.
o,..
\\.
\\.
\\
i
. -.. o... ss. c o...
g c on m,- c.a.s..
o or m,
ca.i.
u
.u
\\
'\\.
n.
\\ ~
v..ss.c o.*
..ss.c o...
- ~:::,' 2,,
\\
- o....s.s.,,c.a. m
- i. '!. ~
- ',* ".'
- :3
\\
\\.
r,
\\
s
\\
to,r com m o
\\ \\
i
\\
\\.
\\.
\\. 3 t,.
\\
o..,.ss. c oi.=
\\
i e ro g
,u,. c.a..s no,..+o.v
. a.a. <s, voc.r a....\\\\.
g, m e,..
i.
- j a
em o.
i e.
r o.:
- no c.. w ro. vou,o.*to
\\, g t-g
.tts.se c.ni et oe?
g g
~(
" "[, O ' " [,i,,,,
\\,1,\\
e.s, o u
'6.,.
=2 oe..mo...to.e r,r es
,c
,x
,c oo oista=ct oc.=.+o..itc. cts.s
~,y
..as
. Lis Fig.12.
Fig. I 1.
Note that in reinout. the rise of the cloud does S;milar to Fig.10, but much more rapid deple-not prevent important deposition on the ground.
tion removes most of the material dose to the source.
itself is most improbable 1.as been stressed.
but it is also interesting to review the prob-wind with small particles in each of the four abilities of the various meteorological occur-cases shown.
rences in the event of such a failure. This All other factors remaining constant, an is a relatively simple matter of combining increase in the rainfall rate tends to aug-the synthetic climatology and the site map ment rainout close to the source and decrease given Earlier.
it at greater distances. This effect is much For the conditions postulated, no matter more prominent with light winds than with what the location or the nature of the acci-strong ones.
dent, it is always probable that there will be Comparison of these curves with dry no rain and that the cloud will not move deposition cases shows that rainout has toward the city, with a slight preference more serious consequences. This is often for an inversion condition over a lapse. Only true of the hot clouds which would produce in the case of the northern location is there only minor deposition in dry weather, but a relatively large change of movement of could affect tremendously large areas with the cold, ground-level cloud toward the city rain fall.
during an inversion, and this reflects the channeling of the wind flow at night. Rain Probabilities occurring at the time of an accident would
]N THE foregoing, the major factors affect-usually be associated with lapse conditions ing the dispersion and deposition of fission and the movement would tend to be away products have been discussed as individual from the city. The combination of rain and subjects. The fact that the major release inversions is relatively uncommon.
)
1 h
which important concentrations and con-
, g..
" c N S E O Y *** l tamination may extend. It also has a most 5
'g \\
"CTgs j
significant effect on the magnitude of the dosage and deposition values, but the in-
.N crease in distance with increasing stability
'N
\\
is virtually always present.
\\
A second generalization may be made e
4 N
concerning normal rainfall associated with N,
\\
clouds composed of small particles. The typ-
\\
'- N\\
leal rainfall, which is relatively light, great-ia '
g'\\\\
ly increases the deposition close to the g
source, and usually extends the isolines of
. ns 5
'y,,5g' 4'C', 7 5 significant contamination to much larger
""I'*
gg;;.
N distances than would be the case in dry
[
' Wo"I'!'fi#.!/,' "*
\\g -
weather.
3
/
\\
Not to be neglected is the ascent of a hot g,og g.:
S cloud. In dry weather, an initial rise of only a few hundred meters may reduce.contami-i
- ,0,",',T,5,,
k ;*1 nation downwind by several orders of mag-
,,,,,,,,m,,,
nitude. This rule cannot be applied during
'k _
precipitation, and may in fact be reversed
...ss.co,o
'~ S','.",7'3*d',,
because of the more rapid movement of the 5
-. ;;"'ll,"{fl,"
3 ga debris.
ooa a a"aa
- - - ;?2',."!!,'l p
Finally, careful site selection in relation to meteorology and topography can reduce r
,,,,,,,J the probability of injury and damage to an
,,.i
, ma o,su?cc ecm.o.n'! tores important degree. This paper has used the distances and areas of contamination as yardsticks, and the evaluation of the im-
""5 Fig.13.
portance of meteorology is based upon them.
Similar to Fig. 4.
Ilowever, consideration of population dis-tributions, land values and land use, as has Summary and Conclusions been done in the full study, in no way dimin-TI{E foregoing is a resums of portions of ishes the importance of meteorology in site ~
the recent Brookhaven study of the haz-planning.
ards associated with power reactors. Atten-tion has been centered on the distribution Acknowledoments of fission products that might accompany Tite AUTitons have already stressed the a major reactor failure, and therefore is fact that much of the material represents largely concerned with meteorological the work of other team members involved in this study, but it is felt that Kenneth W.
processes.
Despite our inability to achieve high ac-Downes, the Project Director, deserves curacy in many of these estimates, it is particular credit for his help in keeping all perfectly clear *that the possible range of of the efforts carefully oriented toward the direct cloud dosages and ground contamina-central objectives. In terms of the meteoro-tion by fission products is very large logical study itself, there were many im-e indeed. Changes in meteorological condi-portant contributors. Foremost are F. E.
tions alone could mean the difference be-Bartlett, R. M. Brown, and S. Rottenberg tween a relatively minor event and a major of the Brookhaven Meteorology Group, who catastrophe. It is not easy to generalize have devoted countless hours to the tedious about the problem, because of the many process of completing and checking the cal-factors involved, but a few obvious features culations. In the planning and development 1
of the study, P. H. Lowry, of Johns Hop-do stand out.
Most apparent is the major influence of kins, and L. Machta and D. Pack, of the atmospheric stability on the distance to U.G. Weather Bureau, have been most help-
ful. Many individuals have assisted in the 2.
Fux. L. L et at: Savannah River Plar.t Stack Gas Disrersion and Microclimate Survey. E.1. du 2,ont de procurement of data and background infor-Nemours. DP te. is.u.
mation, but particular credit is due F. W.
3.
HouA ND. 8. U A MMrologien! Surver of the Thomu and F. E. Gartrell, of the Tennes- 0*(Ridge ^
S her Bur u and U. S. Atomic Enem*
see Valley Authority, who provided very Comminion: MeteoroNy and Atomic Energy. U.
S.
Valuable new data on Inversion Concenha.
Government Printinr 06.ce.1955.
lions as Joon as thef.- field tests were com*
5.
RANz.
W.
E.,
sai JoH NeToNE.
H. F. : Some Aspects of the Physical Behavior of Aerosol Particles in pleted. Dr. A. C. Chamberlain, Atomic the Atmosphere. Proc. rad Nar*L Air Pollariein Sympo.
Enetty Research Establishment, made an
'i"7sEII.'I.A.,and SI 3mrtu. M. F.
Reistion of outstanding contribution in summarizing Gustinen to other. Meteoroionical Parameters. J. Afe-recent British experience in this field, and w=t.10.121 atS50.
in his pertinent s'aggestions. J. Z.
7.
Surm. M. t Srwen. t. A. BAnTHT. I. E. and of the U.S. Atomic Energy Comm. Holland' MAnCra. L.: Yariation of Efftv nt Concertrations During ission, has Temperature inversions. inst (n pres.p.
reviewed each step of the development with 8.
S tyroN, O. G.: Note on "Entrainment and the a thorough critical appraisal inv'aluable to
- ["a",'" C*o','[30 ^,E"'2;,Y,"3Y," ' "*
- C the tuthors.
j
- e. scrroN. o. G.: The Problem of Difrusion in the References Lower Atmoephere. Quart. J. Roo. kleteoroL dee Ts.
{
j esto m.
1.
CHa mstm.Arx. A. C.: Aspects of Travel and De.
- 10. Thomas. F. W., Cantata. F. E and CAaraNTrn, position of Aeropol and Vapour Clouds. AERE HP. R S. B.: Internal Publications and Pusonal Corumunica-1261. 1955.
tions. TVA. November and December,1956.
.s, l
l s
t 1
\\
J I
l Reprinted from AMERICAN INDUSTRIAL HYGIENE ASSOCIATION QUARTERLY, 18:4, 319-330, December,<195't.
(Copyright,1957, American Industrial Hy.tiene Asrociation)
^
l
\\
3' s
2 4
I
..-~
__g
_ _ _ _. _ _ - -. _ - _ _ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - " - -