United States       Washington        EPA 520/1-78-001A
           Environmental Protection    DC 20460
           Agency
&EPA      Protective Action Evaluation
           Part I
           The Effectiveness of
           Sheltering as a
           Protective Action Against
           Nuclear Accidents Involving
           Gaseous Releases


-------
                         LEGAL NOTICE
     This report was prepared as an account of work sponsored by
the Environmental Protection Agency of the United States Govern-
ment under Contract No. 68-01-3223.  Neither the United States
nor the United States Environmental Protection Agency makes any
warranty, express or implied, or assumes any legal liability or
responsibility for the accuracy, completeness or usefulness of
any information, apparatus, product or process disclosed, or
represents that its use would not infringe privately owned
rights.

-------
    PROTECTIVE ACTION EVALUATION
               PART I
 THE EFFECTIVENESS OF SHELTERING AS A
  PROTECTIVE ACTION AGAINST NUCLEAR
 ACCIDENTS INVOLVING GASEOUS RELEASES
              APRIL  1978
            George  H. Anno
           Michael  A. Dore
            Prepared for
U.S. Environmental Protection Agency
    Office of Radiation Programs
       Washington, D.C.  20460

-------
                              Ill
                           FOREWORD
     The Office of Radiation Programs carries out a national
program designed to evaluate the exposure of man to ionizing
and nonionizing radiation, and to promote the development of
controls necessary to protect the public health and safety
and assure environmental quality.

     Office of Radiation Programs technical reports allow
comprehensive and rapid publishing of the results of intra-
mural and contract projects.  The reports are distributed to
groups who have known interests in this type of information
such as the Nuclear Regulatory Commission, the Department of
Energy, and State radiation control agencies.  These reports
are also provided to the National Technical Information Service
in order that they may be readily available to the scientific
community and to the public.

     Comments on this report, as well as any new information,
would be welcomed; they may be sent to the Director, Environ-
mental Analysis Division  (AW-461), Office of Radiation Programs,
U.S. Environmental Protection Agency, Washington, D.C.  20460.
                               W. D. Rowe, Ph.D.
                         Deputy Assistant Administrator
                             for Radiation Programs

-------
                               iv






                            PREFACE






      The material contained in this report was sponsored by the




U.S. Environmental Protection Agency under the technical guidance




of Mr. J. Logsdon of the Office of Radiation Programs, Environ-




mental Analysis Division.  Based on a study to assess the




application and utility of sheltering and evacuation as specific




protective measures in the event of accidental releases of gaseous




radioactive material from nuclear power plants, this report is




the first of tvio that deal specifically with the effectiveness of




public shelter structures.






      The second report evaluates both sheltering and evacuation




protection measures from the standpoint of providing technical




guidance in formulating emergency planning procedures.






      The purpose of this contract report is to provide a technical




basis for EPA  to develop guidance with regard  to actions to protect




the public from accidental airborne releases of radioactive material




from nuclear power  facilities.  The information in  this report should




not be construed as guidance from EPA to State and  local officials




in  development of their  radiological emergency response plans.   Such




guidance will  be published  in  the "Manual of Protective Action Guides




and Protective Actions  for Nuclear  Incidents," currently under de-




velopment by the EPA Office  of Radiation Programs,   The Environmental




Protection Agency  is making  this report available as a source of




technical information.

-------
                         TABLE OF CONTENTS

PREFACE	ill
LIST OF FIGURES	vii
LIST OF TABLES	   ix
I.   INTRODUCTION 	   1
II.  ANALYSIS	3
       RADIONUCLIDE SOURCES 	   3
       SHELTER STRUCTURE MODEL 	    5
       FALLOUT GAMMA-SOURCE ATTENUATION 	  19
       TIME-FRAME MODEL	28
       DOSE REDUCTION FACTOR	32
       DOSE COMPONENTS—UNSHELTERED	35
       DOSE COMPONENTS—SHELTERED	37
       SHELTERING AND EVACUATION 	   43
III. RESULTS	51
IV.  CONCLUSIONS AND RECOMMENDATIONS 	   85
Appendix A:  Fallout Gamma Source 	  92
Appendix B:  Dose Reduction Factor 	   94
REFERENCES	103

-------
                                 vi
                           TABLE OF CONTENTS






Fig.  1—Attenuation Comparisons—Infinite Water Medium 	  14



      2—Shelter-Structure, Cloud Gamma-Attenuation Geometry ... 15



      3—Attenuation for Structure Walls and Roof—Cloud Source    17



      4—Finite Cloud, Gamma Dose-Correction Factors Versus

         Gamma Energy	20



      5—Finite Cloud, Gamma Dose-Correction Factor Versus

         Effective Shelter Radius 	  21



      6—Gamma Attenuation for Structures—Fallout Source ....  22



      7—Finite Plane Source, Geometry-Correction Factor for

         1 MeV Gammas	27



      8~Sheltering-Model Time-Frame	29



      9—Sheltering and Evacuation 	 45



     10—Air Exchange and Infiltration Rates in Closed Passenger

         Compartment When Air Conditioning is Set at a Maximum .  . 47



     11—WB DRF Versus T,, (T -0,T =1)	53
                        £•    J.    cl


     12—Thyroid DRF Versus T , (T -0,T -1)	56
                             ^    JL    &


     13—WB DRF Versus T  , (T -0,T -0), L - 1	57
                        3.    X    £


     14—WB DRF Versus T  , Case A, (T -0,T -0),  (T-0.25.T -0.5),

         L - 1	a	 58



     15—WB DRF Versus T  , Case B, (T -O.T -0),  (T -0.25,T -0.5),

         L - 1	a	!. .2. ../.  ...*.... 59



     16—WB DRF Versus T  , Case C, (T.-O.T -0),  (T -0.25,T,-0.5),
         *   *          <1            1    fc       X       fc
         L - 1	60



     17—WB DRF Versus L,  (T -0,T -0,T -1)	62
                            X    t,    SL


     18—Thyroid DRF Versus L, (T »0,T -0,T -1)	63
                                 X    £    SL


     19—WB DRF Versus L, Case A, (T -0,T -0), (T-0.25,T,-0.5),


         T  - 1	• • • • - •  •	64

-------
                                  vii
Fig. 20— WB DRF Versus L, Case B, (T -O.T.-O),  (T.-0.25,T9-0.5) ,
         T  - 1 .......... ../... .. . . . ..... 65
          a

     21— WB DRF Versus L, Case C, (T,-0,T,-0),  (T.-O^S.T.-O.S) ,
         Ta - 1 ...... ............... ..... 66

     22— Thyroid DRF Versus L, (T -0,T -0),  (T.-0.25.T--0.5) ,
         T  - 1 ......... .. 7 ....... 7 ...... 67

     23— WB DRF Versus T,, Case B, LS, T  -  1  ........... 69
                        ^               £1

     24— WB DRF Versus TZ, Case B, SS, T  -  1  ........... 70

     25— Thyroid DRF Versus T,,, Case B, T  - 1  ..........  71

     26— WB DRF Versus T,, Case B, SS (A-0.4,0.6,0.9,L-0.125,
         1.0,2.0) . . . f ..................... 72

     27— WB DRF Versus T  , Case B, LS (A-0.05,0.1,0.2,L-0.125,
         1.0,2.0) . . . f ..................... 73

     28— WB DRF Versus T  , Case B, (T -1,T,-0), SS  (L-0.5,1.0,
         1.5), LS (L-2) I ...... a. . f  ............ 75

     29— Thyroid DRF Versus T. , Case B, (T -1,T_-0), L -  0.5,
         i n  i <»  9 n       J-            &     i                    -,
         .i..w, A»J, ^..u«  • • . •  • . . . . . .  .  .  .  . .  .  .  .  .  .  /o
     30— WB and Thyroid DRF Versus TI§ Case B,  (Ta«l,T2-0,L-0.125)   77

     31— WB and Thyroid DRF Versus T. , Case B,  (T -1,T0-0.25,L-1.0)  78
                                    X            a     £

     32 — WB DRF Versus Iodine Ingress Fraction, Case  B,
                                                                     79
     33 — Sheltering with Evacuation, WB, SS — Transit  Time Versus
         Shelter Time (T -0.5)  ..................   81
                        cl

     34 — Sheltering with Evacuation, WB, LS— Transit  Time Versus
         Shelter Time (T «=0.5)  ..................   82
                        d

     35 — Sheltering with Evacuation, Thyroid — Transit Time Versus
         Shelter Time (T -0.5)  ..................   83

-------
                                 viii
                            LIST OF TABLES


Table 1.  Radionuclide Source Data	4

      2.  Air Changes Taking Place Under Average Conditions in
          Residences, Exclusive of Air Provided for Ventilation ... 8

      3.  Representative Cloud-Gamma Attenuation Factors 	 18

      4.  Representative Reduction Factors for Surface Source ...  .24

      5.  Dose Components	34

      6.  Fixed Parameter Summary 	   52

-------
                          I.  INTRODUCTION

     In the event of an airborne release of radioactive material from a
nuclear power plant accident, sheltering of individuals is an important
consideration in emergency protective action planning as it may be 1)
an effective means of significantly reducing radiation dosages; 2) the
only practical option in view of possible time and logistic constraints.
Moreover, most people in urban areas, for example, spend 75 percent of
their time indoors.
     This report describes an analysis to estimate the effectiveness or
benefit that might be derived from sheltering following a release of
gaseous fission products from an operating nuclear power station.  The
objective of this effort is the development of sheltering effectiveness
information that could provide 'general guidance to those responsible for
formulating required emergency plans for nuclear power plant siting.
Accordingly, the approach taken here does not lend itself to the specific
evaluation of shelter structures involving detailed descriptions and
operational scenarios; but rather focuses more broadly on vhat are deemed
to be the essential parameters and their variations, and the general
characteristics of small and large categories of shelter structures
available to the public.  Shelter effectiveness as referred to in this
report is the ratio of the dose that may be incurred with sheltering
conditions to that without sheltering in the open, specifically defined
as the dose reduction factor (DRF).  DRF estimates for different con-
ditions of source release, shelter structure assumptions, and operational
time parameters are made for both whole-body and thyroid doses separately,
based on a single-compartment structural model of the time-varying out-
side and inside gaseous radionuclide sources of krypton, xenon, and iodine,
     Design basis accident (DBA) assumptions are made for the gaseous
radionuclide release to define the proportion of rare gases and radio-
iodines.  The magnitude of the release and dose estimates are based on
radionuclide data from The Reactor Safety Study (WASH-1400) [1].  How-
ever, inasmuch as the DRF, as defined above, is the key index used to
characterize the effectiveness of sheltering, it is not sensitive to the

-------
absolute source release magnitude insofar as an approximate proportion-
ality is maintained among the individual radionuclide sources.  Source
release time and duration assumptions are related to release categories
given in Ref. 1 as PWR 1, PWR 3, and PWR 4, for which release times
range from 1.5 to 2.5 hr and the release duration ranges from 0.5 to 3 hr.
     The basic shelter model characteristics considered are gamma ray
attenuation, source geometry, gaseous fission-product ingress, and air
change rate.  Numerical values used for DRF calculations are based on
a literature review and some assumptions that are made where data are
sparse or lacking.
     Temporal parameters considered are source release time and duration,
cloud travel time, and time spent in the shelter structure.  These para-
meters are used to illustrate the sensitivity of sheltering effectiveness
to variations in parameter values.  Also, the analysis of shelter effective-
ness is based on a time-frame model, which can be conveniently related
to other operational times important for emergency planning (e.g., in-
formation time-delay and reaction time) required to accomplish the pro-
tective action—in this case, sheltering.  In addition to developing
shelter-effectiveness estimates parametrically, the advantage of exiting
and evacuating the vicinity of the shelter area after some initial time
in the shelter is analyzed from the standpoint of the DRF and temporal
considerations.

-------
                            II.  ANALYSIS

RADIQNUCLIDE SOURCES
     Table 1 gives the radionuclides and associated data used in this
study to simulate a fission-product release of the rare gases (Xe and Kr)
and the significant radioiodines.  The Xe, Kr, and I radionuclide sources
and parameters shown are essentially the same as those given in WASH-
1400, Appendix VI [1], with the exception of Xe-133m and Xe-135m, which
have been added for completeness only, since they would not affect results
significantly.
     Fission-product source inventory data based on ORIGEN Code calcula-
tions [2] were used to estimate the Xe-133m and Xe-135m sources listed
in Table 1, based on a 550-day irradiation period (same as Ref. 1).
Since the decay half-life of Xe-135m decaying to Xe-135 is short  (15.6
min) considering the times of interest (hours) in this study, the esti-
mated shutdown zero-time Xe-135m inventory was added to the Xe-135
source on a mass basis and converted to Xe-135 on an activity basis,
which increases to 0.27 Ci instead of 0.26 Ci given for Xe-135 in Ref. 1.
The metastable decay half-life for Xe-133 is by comparison appreciable,
and no similar adjustment for the Xe-133 source inventory was made.
     The average decay gamma energies listed in Table 1 for the metastable
Xe radionuclides were taken from Ref. 3  (pp. 32-33); the whole-body  (WB)
cloud gamma-dose factors, from Appendix D, Ref. 1.  These dose factors
for the ground-y (surface deposition source) do not take ground roughness
into effect (such as a factor of 2).  The estimated effectiveness values
in terms of a dose-reduction factor would not be affected significantly
whether or not the ground roughness adjustment were included.
     An estimate of the average gamma decay energy was made for the
source nuclides to serve as a guide in 1) estimating gamma ray attenua-
tion factors for shelter structures and  also in 2) making estimated
adjustments for finite source geometries of cloud-source volume and
contaminated  floor-surface spaces inside the structure, since the dose
factors for cloud-y and ground-Y apply to infinite source geometries

-------
                                             Table 1




                                      RADIONUCLIDE SOURCE DATA
Nuclide
Kr-85
Kr-85m
Kr-87
Kr-88
Xe-133
Xe-133m
Xe-135
Xe-135m
1-131
1-132
1-133
1-134
1-135
Half-Life
(hr)
93,600
4.32
1.27
2.78
127
55.2
9.12
0.27
193
2.4
21
0.864
6.72
Source
(Curies * 108)
(Q)
0.006
0.26
0.52
0.76
1.7
0.04a
0.27a
0.27s
0.85
1.2
1.7
2.0
1.5
Average Gamma
Energy (MeV)
(E)
0.0
0.16
0.82
2.21
0.08
0.23b
0.26
0.52b
0.39
2.3
0.63
2.4
1.45
Dose Factors
Cloud-y
(r em/ sec)
(Ci/rn3)
0.0
0.036
0.36
0.42
0.007
0.0075C
0.06
0.0972C
0.09
0.55
0.12
0.6
0.42
Ground-Y
(rem/hr)
(Ci/m2)
0.0














2.8
17
3.7
16
12
WBD
(50-yr)
(rem/Ci)
0.0






i







2,600
130
570
40
290
Thyroid
(50-yr)
(rem/Ci)
0.0






1







1.47xl06
5.3xl04
3.96xl05
2.5xl04
1.23xl05
 Based on Refs.  1 and 2.
 Ref.  3.
-Ref.  4.

-------
(discussed below, p. 13ff.).  The average gamma decay energy was estimated
to be ^1.2 MeV, based on the following simple weighting relationship:
where Q. and E, are the radionuclide source activities and gamma energies,
       J      J
respectively, listed in Table 1, and y (E.) is the gamma-ray linear energy
absorption coefficient as a function of energy for air given in Ref. 5.
The estimate of the average gamma ray energy was based on a summation over
all the radionuclides shown in Table 1, with the exception of Xe-135m—
again because of its short half-life for the times of interest in this
study.  The gaseous radionuclide data in Table 1 are used to estimate
shelter effectiveness in dose reduction by summing each nuclide contribu-
tion (assuming single radionuclide decay) to obtain the unprotected (out-
side shelter) and protected (inside shelter) dose.  Design basis assumptions
(DBA) are made for the source release—100 percent of the noble gases and
25 percent of the radioiodines available for release.
SHELTER STRUCTURE MODEL
     A simplified approach rather than a detailed investigation was
adopted to account for those factors that might contribute to the bene-
fits of seeking structural shelter in the event of a gaseous, radioactive
fission-product release from a nuclear power facility accident.  The
reasons for taking this approach are as follows:
     1.   Explicit consideration of all types of possible structures
          that may be available for shelter—single-family dwellings,
          apartment buildings, office buildings, subways, tunnels,
          factories, and vehicles, etc.—would require an analysis.
          beyond the scope of this effort because of the large varia-
          bility in the parameters that determine effectiveness.

-------
     2.    In many  instances,  reliable parameter data are not available;
           e.g.,  the actual  chemical/physical  form of the "gaseous"
           constituents  at some distant point  from the source release
           and  the  ingress of, in particular,  radioiodine into shelter
           structures.
     3.    The  main purpose  of this study is best served by providing
           overall  technical guidance information as to the effectiveness
           of using a shelter structure based  on some assumed conditions
           for  shielding and ventilation rates without specifically focusing
           on detailed physical description and analysis of shelter structures*
     After a review of  the  literature dealing with the key parameters of
this study's simplified model (in keeping with the above reasons),
calculations were performed using parameters  selected to simulate
what this  study classifies  as "small" and "large" structures (SS
and LS) to illustrate the relative effectiveness of typical single-
family dwellings and of larger structures such as office buildings,
auditoriums, apartment complexes, etc.  In developing the shelter model,
consideration was given to  account for the following possible avenues of
exposure to shelter inhabitants:
     o     External WB dose  from airborne radioactive material both
           outside  (shielded) and inside (unshielded) the shelter
           structure.
     o     Inhalation WB and thyroid dose from airborne radioactive
          material inside the shelter structure.
     o    External WB dose  from radioactive fallout material deposited
          both outside  (shielded) and inside  (unshielded) the shelter
          structure.
     In this study, beta skin dose was not considered, as it is assumed
to be of secondary importance as compared with WB and thyroid dose con-
siderations.  The external WB doses (cloud-y and fallout-Y) are based
solely on radionuclide-decay gamma radiation  in which both shelter-
structure attenuation and finite source geometry factors are included
in the model as discussed below.

-------
     The entry of outside airborne radioactive cloud material is assumed
to be dependent on the shelter-structure ventilation rate (forced,  natural,
or both) assuming constant homogeneous mixing based on simple one-compart-
ment outside/inside air exchange.  This type of stirred-tank. mixing and
ventilation model has been applied in studies of the relationship between
indoor/outdoor pollutants (e.g., NO, NO ,  CO, and 0_) and has predicted
concentration versus time profiles that are similar to those measured 16].
The radioiodine fallout deposition inside the shelter is then also  assumed
to be dependent on the ventilation rate as well as the fallout deposition
velocity; these aspects are also discussed below.
Shelter Structure Ventilation
     A review of literature on ventilation rates of homes and buildings
indicates a wide variety of air change estimates ranging anywhere from
^0.1 to 6 per hour for single-family dwellings to *v
-------
 Health Division of  the Los Angeles City Building and Safety Department
 [8]  indicated  internal air turnover  time of from 5 to 10 min depending
 on  occupancy requirements, with  a representative value of about 7 min—
 with 15  percent outside air makeup as a comfort-level requirement—
 which corresponds to 0.9  and  1.8 air changes per hour and ^1.3 air changes
 per  hour.  Considering the above data, the rates for single-family
 dwellings and  large structures are generally comparable, assuming internal
 forced-air systems.
      In  the absence of forced-air ventilation systems, home and building
 air  change rates would be expected to vary much more widely—as indicated
 by  the published data examined by Handley and Barton [7].  This conclusion
 is  also  supported by observations of Yocom, et al.  [9] who note that
 particulate pollutant levels  are lower in public buildings than in homes.
 The  AS11RAE Handbook of Fundamentals  [10] points to the lack of published
 data on  air change rates  for  different buildings, exclusive of air pro-
 vided for ventilation, when utilizing the air change method for estima-
 ting infiltration  requires experience and judgment.   Table 2 gives ASHRAE
 Handbook values that may be used with reasonable precision in making
 infiltration estimates for residences with different room conditions.
                               Table 2
      AIR CHANGES TAKING PLACE UNDER AVERAGE CONDITIONS IN RESIDENCES,
                EXCLUSIVE OF AIR PROVIDED FOR VENTILATION

           ...,,-,       _  ,, ,.                 Number of Air Changes
           Kind of Room or Building               Taking place per ^a

 Rooms with no windows or exterior doors                  1/2
 Rooms with windows or exterior doors on one side       1
 Rooms with windows or exterior doors on two sides      1 1/2
 Rooms with windows or exterior doors on three sides    2
 Entrance halls                                         2
 o
 For  rooms with weathers tripped windows or with storm sash, use two-thirds
 these values.
      The other is the "crack method" based on measured leakage character-
istics of the building components and selected pressure differences.

-------
     Another approach in making air change estimates due to natural
ventilation for houses is given by Coblentz and Achenbach [11],  who suggest
the following empirical relationship in which the air change rate is
proportional to the outside wind speed and inside/outside temperature
differential (i.e., without inside forced ventilation):

                I (changes/hr) = A + BW + CAT

where
      A = air change rate for W - 0, AT - 0 (0.12 to 0.18),
      B - 0.013,
      C « 0.005,
      W » wind speed, mph,
     AT B T         T         ° T*1
           inside    outside'
Assuming the upper limit of A = 0.18 and AT = 20°F  gives air change rates
of about 0.35 per hour for a 5-mph wind speed and about 0.5 for 15- to
20-mph wind speeds, which appears to be somewhat on the low side compared
with other data reviewed.  This difference, however, may be due to new,
well-built houses that made up part of Coblentz and Achenbach's field
samples.  In contrast, measured air change rates given by  Megaw  [12] for
a hut structure that were made in conjunction with  radioiodine penetration
experiments were substantially higher, ranging  anywhere  from about  2 per
hour to 8 per hour  (the  latter, however,  for open windows).  An examina-
tion of Megaw1s data reveals an indication of air-change-rate proportionality
with outside wind speed  that, roughly, was about 0.5  (changes/hr)  per
 (mi/hr).  This  figure  corresponds  to  only an "eyeball"  estimate  from
Megaw's data, which are  complicated by variations in wind  direction.
Such variations would  give rise to  different pressure  differential  dis-
tributions due  to asymmetric  flow  patterns, which would  affect  the internal
air change rate.
     Based on the above  review of  air change rates  that  might be expected
for single-family dwellings  (small structures)  and  various building
structures that could  be used as  temporary public shelters, values  of
from 0.125 to  3 air changes  per hour  were assumed in  performing shelter-
structure effectiveness  calculations.   It was  felt  that  ^0.125  changes

-------
per hour might represent relatively "tight" structures (either large or
small) and that ^3 air changes per hour might represent a practical
upper limit of structural ventilation.  Of course, as indicated, much
larger values of 6 to 9 air changes per hour have been measured; but it
was felt that these values would represent extreme cases (e.g., open
windows or portals), which do not represent practical cases if good
planning is assumed.
Gaseous Fission-Product Ingress
     The extent to which radioiodine will penetrate a structural shield-
ing facility is dependent on the gross tightness of the structure, the
ventilation rate, filtration, and the chemical and physical properties
of the released material and the interacting species.  Many of these
facets of a gaseous fission-product release from a nuclear accident
are currently unknown, particularly for radioiodine, which leads to
difficulty in accurately predicting the ingress of gaseous radioactive
material into shelter structures.  For the rare gases (Xe and Kr), most
are willing to accept virtually no effective "structural filtering,"
because of their inertness and stability as gaseous forms.  Accordingly,
in this study no effective filtering action has been included in esti-
mating their internal structure concentrations.
     For the halogens, which are here assumed to be all radioiodines,
the case is more complicated and suffers from scarcity of experimental
work on indoor/outdoor pollutant-level relationships dealing with the
ingress of radioiodine into various potential sheltering structures.
The radioiodines are of course particularly important sources due to
their large contribution to the WB dose, as well as being totally
responsible for the thyroid dose.
     Three known chemical forms of radioiodine present as airborne
gaseous species in power-station areas during and after handling defec-
tive fuel elements are elemental iodine (I2), hypoidous acid (HOI), and
organic iodides (CH.I).  The ratio of the three species would depend
on the conditions under which an accidental release might take place.
Elemental iodine is thought to be the primary form released from
uranium-oxide fuel.  It hydrolyzes rapidly in water, generating HOI, or

-------
                                 11
it forms organic iodides through a slower reaction with organic  com-
pounds, with relative stability in air increasing in the following
order [13]:
                        I2 < HOI < CH.I
     The actual chemical/physical form of the radioiodine that would be
present at some off-site point is yet another question; however, pro-
bably very little protection would be offered by a structure against
the ingress of HOI and CH_I, both unreactive gaseous forms like Xe
and Kr [14].  The Reaator Safety Study [I] did consider other possible
forms of radioiodine that could be released (e.g., HI, Csl, and Zrl),
but concluded that these forms would not be major species as they had
not been verified experimentally.  In its dose calculations, Ref. 1
assumes, primarily, elemental iodine; and, to a much lesser extent,
organic iodide (approximately a factor of 100 less).  However, assuming
elemental iodine release to the atmosphere, controlled field release
tests  [15] involving elemental iodine (I2-131) indicated a rapid trans-
formation in apparent particle size from the source—in that the field
                                                                     e
sampling results for the released gaseous product (effective HMD * 2A)
revealed a much broader spectrum of sizes, closely resembling  the normal
distribution by size, of particles in the atmosphere  (with an  HMD «  0.4
microns).  The above would suggest some effectiveness  of shelter structures
in reducing radioiodine ingress released in the elemental  form,  depend-
ing, of course, on overall integrity, ventilation, filtering,  etc.
     Estimates of radioiodine ingress into structures  for  this study are
primarily based on the observations and work of Megaw  [12], which repre-
sents  essentially the only source of published information applicable
to this study; other related, but not applicable, work [16] has  been
sponsored by the Office of Civil Defense  (Defense Civil Preparedness
Agency),  Megaw's work originated from the accidental  Wind scale
incident in which it was estimated that dose rates inside  build-
ings may be from about 14 to 25 percent of those outside.  Subsequent
experimental measurements were made by Megaw involving radioiodine

-------
                                   12
 releases and a reasonably tight wooden hut; and he concluded that the
 time integral of the inside concentration (dose) may be from 20 to 80
 percent of that outside, depending on wind velocity and direction.
      An examination of Megaw's published data [12] does not suggest any
 correlation of the inside-to-outside dose ratio with either outside wind
 velocity or ventilation rate,  probably because of the varying conditions
 under which measurements were  made; e.g., measurements were made for
 unique sets of wind direction  and velocity.   A simple statistical
 analysis of the data indicates a protection factor (ratio of inside to
 outside dose)  of 0.51 ± 0.12 (pooling the data from two experiments
 described).   From Megaw's work,  however,  it is not possible to identify
 precisely the  extent of the radioiodine filtering action or resistance
 to  ingress for use in a simple mixing model such as is assumed for this
 study,  even for the test structure used in the experiment,  because of
 the absence of experimental information regarding source release time
 and intensity  distribution and the absence of  any correlation of the
 inside-to-outside dose ratio with ventilation  rate.   The dose reduction
 factors  given  by Megaw are therefore  effective values that  would include
 any filtering  or ingress action  of the shelter structure used in the
 experiment plus the specific test conditions and  parameters.   However,
 to  take  into account what is felt to  amount  to some gross filtering
 action  for radioiodine—whether  assumed  to be  due to trapping or
 deposition in  small cracks or  openings—the above-mentioned value  of
 0.51  has  been  tacitly assumed  in approximating the explicit filtering
 action for shelter  structures.
 Radioiodine Deposition
      Shelter effectiveness estimates  in  this study take  into  account
 external  WB dosages  from outside and  inside radioiodine  source deposition,
 using estimates  for  the  deposition  velocity, V .   Values  ranging from
                                               O
 0.1 to 1  cm/sec were  obtained  from  controlled  environmental  radio-
 Iodine tests made at  the  National Reactor  Testing  Station in  Idaho  [15].
 For outside radioiodine deposition velocity, the Heaator Safety Study [1]
used a value of 0.5 cm/sec, which is also assumed  in this study for
surfaces outside a shelter structure.

-------
                                  13
Inside a shelter, a value of 0.025 cm/sec was assumed for the radio-
iodine deposition velocity for the floor surface, based on Megaw's
work in which he estimated the inside deposition velocity to be only
about 5 percent of the outside deposition velocity [12].
Cloud-Gamma Attenuation
     The attenuation of cloud-gamma radiation that might be afforded by
building structures has been estimated by Burson and Profio [17]; the
results of their analyses served as a guide for estimating the cloud-
gamma attenuation factors used in this study.  The source basis for the
attenuation calculations that were performed applies to the PWR Category
2 accident [17]  ten miles from the plant, under average dry meteorological
conditions.  Figure 1 shows comparison of mass-path attenuation for
different energies based on dose buildup and exponential attenuation in
water.  Since,  for the source energies of interest, most of the attenua-
tion will be due to Compton scattering—where Z/A remains relatively
constant at about 0.5—water data (mass-path) are suitable  for applica-
tion to the usual structural materials such as wood, concrete, brick,
and even steel  [17].  As shown in Fig. 1, for mass-path  thicknesses
                            2
of interest up  to ^45 gm/cm , attenuation values—particularly,
for reactor-accident spectra  (ground  and cloud source)  and  Co-60—are
all quite close.  Moreover, slight variations in the spectrum are not
considered significant, since the higher-energy  gamma  rays  would
be the most penetrating and any differences  in attenuation  would  not
amount to any major source  of uncertainty considering  the other  assumptions
made in this study.
     Burson and Profio's attenuation  factors  [17] used in  this study are
based on calculations assuming a  simple  hemispherical  shell model (Fig.
2).  Estimates  were made of the gamma attenuation (with the dose  point
at the origin)  for the portion of radioactive cloud  material  outside the
shelter, based  on numerical evaluation of the attenuation,  A(x),  given by a
relationship of the form

-------
    1.0
    0.8
    0.6
s

§
    0.4
    0.2
   0.1
                                14
               reactor  f ground  source
               accident (cloud source
                                          \
                           1.12 hr fission
                         \ 6     products
                             'Co
                     \
                                              \
                                                \
                                                  137
                                                     Cs
               10
20      30
40
50      60
70
                        Mass thickness,  x  (gm/cm  )
        Fig.  I—Attenuation  comparisons—infinite water medium

-------
                                  15
          dr.x'
                                 (out)
air
    Fig. 2—Shelter-structure, cloud gamma-attenuation  geometry

-------
                                  16
                                             •   °  V'a
                                                        dr
                   gamma
           A(x)  . energies.
                                R
                                          "V
                                  B(y r)  e  a  dr
                                     £L
                   gamma
                  energies
where  p   and  p   are  the  energy-dependent  gamma-ray absorption  coefficients
              w
 for  air  and water,  respectively;  and  K  and  S  are  the energy-dependent dose-
 conversion and volume-source  terms, respectively;  and B  is  the dose buildup
 factor.   Figure  3,  a plot of  the  gamma  ray  attenuation for  accident spectra •
 for  a =  3 m and  pR  = 3,  is applicable for estimating gamma  ray attenuation
 for  structures of a wide variety  of enclosure sizes  (effective radii, a),
 since A(x) is relatively insensitive  to a,  because of the low density
                  -3      3
 of air (1.293 x  10   gm/cm )  as compared to structural material.  Also,
 very little dose contribution would be  expected from cloud  sources
 beyond about three  mean-free  paths in air;  therefore, yR »  3 is a
 reasonable approximation for  an infinite cloud source with  regard to
 the  gamma-radiation transport considerations in this study.
     The  estimates  of gamma attenuation for an outside radioactive cloud
 source that can  be  made  from  Fig. 3 depend  on the  structural assumptions.
 For  example, for a  wooden frame house with  roof and ceiling consisting
 of 1/4-in. wood  or  asphalt shingles,  3/4 in. of wood sheathing and
 rafters,  and 1/2 in. of  gypsum board, the mass thickness would be

       1.5 (in.) x  2.54  (cm/in.)  x 0.84 (gm/cm3) « 3.2 gm/cm2

where A(x) =0.9.   Better protection  would  be afforded by a small house
with a wooden roof  and masonry walls.   For  example, assuming half the
2rr solid  angle (Fig. 2)  to be subtended by  the walls and the other half
by the roof,  the overall attenuation  factor would  be

-------
                                17
5
o
o
•r-
*J
fl
    o.i
                         Mass  thickness, x  (gm/cm  )
    Fig.  3--Attenuation  for structure  walls  and roof—cloud source

-------
                                  L8
                 0.5 x (0.9) + 0.5(0.38) - 0.64

 where the attenuation for the walls (0.38) is based on a wall-mass thick-
 ness of

          4 (in.) x 2.54 (cm/in.)  x 2.7 (gm/cm3) * 28 gm/cm2    ,

(assuming 8-in.  concrete bricks with a 50-percent void volume).
      Attenuation of cloud-gamma radiation for large structures  such as
 office buildings and multistory structures could be significantly more
 than for simple structures such as single-family dwellings.  Attenuation ol
 8-in.-thick solid concrete, either exterior walls or interior walls
 (e.g.,  fire-resistant stairwells)  may be equivalent to mass thickness
                         2
 of  around 45 to 50 gm/cm ,  corresponding to attenuation factors of 0.2
 to  0.17 (Fig.  3).   Table 3 summarizes representative cloud-gamma  attenua-
 tion factors for the types of structures noted.

                                Table 3
           REPRESENTATIVE CLOUD-GAMMA ATTENUATION FACTORS
            Structure
      Wood  frame house, no basement
      Masonry house,  no basement
      Basement  of wood house
Attenuation Factor
     0.9
     0.6
     0.6
      Basement  of  masonry  house                      0.4
      Large  office or  industrial  building            0.2 or less

      The  above values do  not  suggest  any  substantial  protection  from
external  cloud-gamma  radiation afforded by  lightly  constructed,  frame
single-family  dwellings.   In  this  study,  however, estimates of shelter-
ing effectiveness were made assuming  somewhat more  substantial gamma-
attenuation protection, A(x)  - 0.4 to 0.9 for small structures.   For large

-------
                                  19

structures, shelter effectiveness estimates were made for A(x) - O.OS
to 0.2; gamma attenuation could be even greater, amounting to values
much less than 0.05 for veil-protected areas within large multistory
structures.
     The estimated external WB gamma-dose contribution from airborne
gaseous radioactive material that enters a shelter structure is based
on a finite cloud-source geometry correction factor, since infinite
cloud-dose conversion factors (Table 1) are used in estimating shelter
effectiveness.  The source geometry correction factor is defined as

                      G(E,R) - D(E,R)/D(E,-)

where D(E,R) and D(E,«) are gamma doses at the origin of a hemispherical
cloud source for finite and infinite radii, respectively.  Values for
G(E,R) based on point-kernel integration over a hemispherical source
volume in air, assuming Berger's expression for a dose buildup  factor,
are given in Ref. 18 for various energies and source radii; values of
G(E,R) are plotted in Fig. 4.  Figure 5 gives finite cloud-geometry
correction factors for a couple of gamma energies of interest in this
study, where very little difference is seen between 1 and 1.2S  MeV gammas.
     Simulation of small and large shelter structures in  this study
assumes effective hemispherical radii of 3.4 and 10.3 m to  represent
                                                    7
shelter enclosures of approximately 400 and  3600 ft  of  floor area,
respectively.  From Fig. 5, estimated small- and large-shelter-structure-
geometry correction factors are 0.01 and 0.034, respectively.   These
values are assumed in estimating  the effectiveness  of shelter structures.
FALLOUT GAMMA-SOURCE ATTENUATION
     Considerable analytical and  experimental work  has been done  to
determine  the protection against  fallout-source gamma radiation afforded
by various  types of building structures, primarily  for civil defense
applications.  Burson and  Profio  [17]  reviewed  much of this work  for
application  to nuclear  power plant  accidents, and performed additional
calculations using the  method  given in Ref.  19  to estimate  attenuation
factors for  some simple rectangular structures  (Fig. 6).  Experimental
results  [20-24] generally  indicate  protection factors  (PF), often

-------
                                   20
      u
      o
      »J
      o
      ,1)
      c
      o
      
-------
                             21
a:

10
 o
 *-»
 o
 rO
4J
O
 O
   0.001
                           Radius, R  (meters)
Fig.  5--Finite cloud, gamma dose-correction  factors  versus effective
        shelter radius

-------
                                22
     1.0
«'
 I
•••
••
 -

.
     0.1
   0.01
                                               2
                                 Floor space  (ft )  -
                                     10 x 10
                                     20 x 20

                                     30 x 30


                                     40 x 90
                       large
                    structures
             small
             struc-
             tures
       0
10
20
30
40
50
60
70
                        Mass thickness, x (gm/cm )
       Fig. 6—Garana attenuation for structures—fallout source

-------
                                 23
referred to as the reciprocal-of-attenuation factor, from 2 to 5 for a
wood frame home (without basement) and from 3 to 10 for block and brick
homes.  Most attenuation-factor estimates for fallout gamma sources in-
clude the effect of ground roughness, which can vary accordingly as
tabulated below by the Defense Civil Preparedness Agency [19].
          Ground Roughness Condition
          Smooth plane (hypothetical)
          Paved areas
          Lawns
          Gravelled areas
          Ordinary plowed field
          Deeply plowed field
Reduction Factor
  1.00
  1.00 to 0.85
  0.85 to 0.75
  0.75 to 0.65
  0.65 to 0.55
  0.55 to 0.47
     Many other aspects affect protection against fallout sources, in-
cluding structural materials, wall-exposure areas (taking into con-
sideration basements and multilevel dwellings), topographical varia-
tions (hillside or flat ground level), mutual shielding offered by
nearby buildings and structures, and the internal location within a
shelter structure.  For example, protection factors for basements may
be from 10 to 50; and material shielding of nearby buildings may offer
protection factors of from about 1.7 to 2.5 [25].  Complex structures
such as multistory office and apartment buildings offer protection
factors of 20 or more (away from doors or windows); this factor is
supported by experimental measurements [26-28].  Table 4 summarizes
recommended attenuation or reduction factors for some representative
shelter structures and also vehicles [17].
     The reduction values in Table 4 are relative to 1 meter above a
hypothetical, uniform infinite plane of homogeneous source con-
centrations.  The values given are only representative and not to be
taken as exact; and as indicated above, different values will result
because of wide variations in constructional details and topography.
     Estimates of the external WB dose from radioiodine fallout inside
a shelter structure are based on a dose detector point 1 meter above

-------
                                 24
                                Table 4
          REPRESENTATIVE REDUCTION FACTORS FOR SURFACE SOURCE
     Structure and/or Location
Reduction Factors
1m above a hypothetical, infinite, smooth plane
1m above ordinary ground
1m above center of 50-ft roadway half contaminated
Cars, pickups, buses, and trucks on 50-ft road:
  Road fully contaminated
  Road fully decontaminated
Trains
1- and 2-story wood frame homes (no basement)
1- and 2-story block or brick homes-(no basement)
Home basement—1 or 2 walls fully exposed:
  1 story, less than 2 ft of basement walls exposed
  2 story, less than 2 ft of basement walls exposed
                                           2
3- or 4-story structures,  5000 to  10,000 ft  per floor:
  First and second floors
  Basement
                                 2
Multistory structures, >10,000 ft  per  floor:
  Upper floors
  Basement
     1.00
     0.70
     0.55

     0.5
     0.25
     0.4
     0.4
     0.2a
     O.la
     0.05a
     0.02a

     0.053
     0.01a

     0.013
     0.005£
 Away from doors and windows.

-------
                                  25
a circular area for small and large shelter structures in which infinite-
plane dose-conversion factors were used  (see Table 1, p. 4).  Therefore,
a finite-plane geometry correction factor was applied in calculating
dosages, defined as

                          G' (R) - D(R)/D(~)    ,

where D(R) and D(«) are the finite plane  (radius, R) and infinite-plane
doses for d = In above the surface.  G1 (R) may  also be expressed as
where D(R,») is the plane-source dose for source radial dimensions from
R to », and D(») = D(0, ).  The dose D(R) for a flat plane source is
given by
                                 CO
                               f
ncr»> _ k o     /    B(pr) e    dr
DQi-0 - Y sa
                              >/R2+d2

where
         Sa = source strength per unit area,
          k = dose-conversion constant,
          R = distance from source plane  (1 m),
            = gamma-ray absorption coefficient in air,
      li(ur) = 1 + Cure yr  (Berger buildup factor).
Integrating the above over the appropriate source-plane  upper  limits
(see Appendix A) yields
and

-------
                                  26
                           kS
                    D(-)  -
(yd)
                                                 -(l-D)yd
where  E.. (x)  is  the  first-order exponential integral function,  and C and
D  are  the  Berger buildup factor coefficients  for  air  given  in  Ref.  29.
Assuming 1.293  x 10~   gin/cm  for air,  calculations of G'(R) were made
for  0.5, 1.0, and 2.0  MeV gamma rays  for various  values of  R using the
following  data  from Ref.  29.
Energy (MeV)
0.5
1.0
2.0
C
1.6001
1.1571
0.8363
D
1.0094
0.05749
0.0243
M/cm \
P\gm 1
0.088
0.063
0.046
Some results are  given below  for  R =  10  and  30m:

Energy (MeV)

0.5
1.0
2.0
G'(R)


10m
0.413
0.414
0.419

30m
0.620
0.624
0.622
     As indicated above, very little variation  exists  from 0.5  to  2.0
MeV; the 1-MeV values plotted in Fig.  7 are assumed  to be  representa-
tive for this study.  Again, assuming  3.4 and 10.3 m as  effective
radii applicable for small and large shelter  structures, yields finite-
source geometry correction factors of  0.28 and  0.43, respectively,
which are used in the shelter model calculations.

-------
     20      40
60      80      100      120
     Radius, R  (meters)
Fig. 7—Finite plane source, geometry-correction factor for 1 MeV gammas

-------
                                  28
 TIME-FRAME MODEL
     The question of  shelter  protection  effectiveness  from airborne
 radioactive material  accidentally  released  from  a nuclear power plant
 is dependent  upon the time  required  for  individuals  to gain entry into
 a protective  structure;  and the  length of time they  remain, as compared
 with the time of cloud arrival and passage.  The required entry time
 assumes that  individuals are  transferred from either unprotected or
 protected locations to another location  affording maximum protection,
 considering logistic  constraints,  etc.   On  the other hand, individuals
 could also be located in houses  and  buildings already  providing adequate
 shelter so that effectiveness would  not  depend on access time.
     Figure 8 shows the  time-frame model assumed in  estimating the
 effectiveness of sheltering,  as  well as  other times  of interest (to
 put them in perspective).   Measured  from initiation  of a possible
 incident, (T +T ) is  the estimated time-of-arrival of  the assumed lead
            K  a
 portion of a  radioactive cloud.  The time from source  release, T_,
 measured from incident initiation, may vary from about 1.5 to 9 hr for
 the more severe accident categories  [1]; although in one instance
 (PWR 4 Category), a value of 28  hr was indicated.  Source release times
 of from 1.5 to 3 hr were considered  to be of more interest in this
 study,  since  protective  evacuation action might  very well be more
 appropriate,  considering the greater time that would be available.
     Cloud arrival time,  T  , would depend completely on the location
                          <1
 of a shelter  from the point of release and  the prevailing meteorological
 conditions (primarily, wind speed  and direction) during cloud travel
 time.  Assuming a given  sustained  average wind speed (and direction),
 x/u is an estimate of T  , where  x  is the distance from the release and
                       Si
 u the average wind speed.   For example, for an estimate of the average
 low-population zone distance of  around 3.4 mi based  on siting data
 given for 76 nuclear  power plant sites [30], cloud arrival time would
be approximately 1-1/2 hr to 20 min  for wind speeds  of  from 2 to 11 mph,
 respectively.   The effective time  for sheltering from  incident initiation
 is shown in Pip. 8 as (TD+TT), where TQ is the delay time for the initiat-
 ing event to the sheltering order, and T_ is the actual time spent in taking

-------
                        TR + V
                       •TD + V
                                        (or Ts)
                                            Time in shelter
                                                                                                                  K>
                                                                                                                  VO
 Incident
initiation
TR = source release time


T, = cloud arrival time
 a

T  = cloud passage time

T  = cloud-source release duration
     (forTe>Ts, Te=Ts)


T, = shelter entrance-delay time
     after cloud arrival
T2 = shelter time after cloud passage


T  = evacuation time after leaving
 v   shelter

Tn = delay time from initiating event
     to sheltering alert

TT = actual  time spent taking shelter
                                         Fig. 8--Sheltering-model  time-frame

-------
                                  30
 shelter (assuming individuals are not already in a suitable shelter).
 Delay time estimates (T ) have been discussed by the EPA [31]  with
 regard to evacuation that may be somewhat applicable to sheltering,
 since the time components of T_ are similar or may in fact  be  one and
 the same In terms of a local decision process.   As assumed  here,  T_
 represents the total delay time from initiation of an event to onset
 of physical movement to a shelter.   For evacuation,  the EPA estimates
 this delay time as being from 0.9 to 4.5 hr [31].   Also,  for evacuation,
 the EPA estimate for T_ is from 0.2 to 1.5 hr,  which may be excessive
 for sheltering on the high end.   That is,  reasonable sheltering times
 may be anywhere from a few minutes  to half an hour.
      Allowance is made in the time-frame model  for a shelter-entrance
 delay time measured from time of cloud arrival,  which would be dependent
 upon T_,  T , T_,  and T_.   The shorter T,  is,  the better is  the shelter-
       K   a   u       i                 i
 ing effectiveness with the maximum  advantage  for T.  equal to zero,
 (TD+TT)_<  (T-+T ).   Normally,  T-  would be  expected to be  either zero
 or small  except for relatively high sustained wind speeds and/or for loca-
 tions relatively  close to a release.
      The  cloud passage time,  T ,  would depend on source release duration
 (T )  and wind  persistence time (direction  and speed).   T  may  range
   s                                                    s
 from 0.5  to 4  hr,  depending on the  accidental release events [1]  that
 would be  of interest for  seeking  shelter.   Estimates  of wind persistence
 time should be based on particular  site meteorology.   In  terms of pro-
 tective action by the  public (i.e.,  taking  shelter or evacuating), the
 wind persistence  time  estimates made  at the time of and during postincident
 phases of  an accident  are among  the most important parameters  affecting
 the effectiveness  of the  protective action.   Ideally,  the most useful
 type  of information  on persistence, when making  protective  action
 decisions,  would  be  an estimate of  the  mean or expected wind-direction
 persistence time—given a particular  time of  the day  and  that  a particular
 direction has been maintained  up  to that point.  Such  predictive ability
would have  to be  formulated  from  a  detailed statistical analysis of
site meteorological  data  of  record  requiring  frequent  observations
 (perhaps every  15 min) over an adequate period of  time.  A means of

-------
                                 3)
computing source-cloud trajectory based on real-time analysis of site
and regional meterological data is described as a feature of the ARAC
program currently being developed at the Lawrence Livermore Laboratory
[32].  This kind of capability would obviously be very useful in planning
emergency public actions such as sheltering.
     In this study, the persistence time is simply related to the cloud
exposure time designated in Fig. 8 as T ; such that, if T is an estimate
                                       G
of the persistence time, then T  - T  for T  < T, otherwise, T  - T.
                               68      S """"                G
     The time-frame model for sheltering also considers the time that
individuals may have remained in a shelter after passage of the radio-
active cloud.  For example, although exiting a shelter may afford
more protection and thus avoid exposure to accumulated internal con-
tamination, precise exiting with regard to cloud passage may not be
practical, and the overall time spent in a shelter could be as indica-
ted by the shaded portion of Fig. 8.  This shaded portion, then, designates
an "internal receptor" with respect to radioactive gaseous fission-
product sources.  That is, during T., after cloud arrival, unprotected
individuals may be exposed to airborne radioactive material by means of
direct WB gamma radiation from both airborne and ground-source fallout
material and from radioactive material entering the body via inhalation.
During the interval inside the shelter, (T_+T +T ) - (T_+T_), the inter-
                                          K  a  e      D  T
nal receptor is exposed to WB radiation from airborne and surface-source
(fallout) material, both inside and outside the structure, and internal
radioactive material entering the body via inhalation.  During T.,
after cloud passage, the internal receptor is assumed to undergo the same
type of exposure with the exception of that due to airborne  gaseous
fission products outside the structure.
     Finally, after TZ, the time-frame model makes allowance for the
time that may be required for leaving the area (where an individual
may be exposed to outside fallout in transit either on foot  or by
vehicle).  If vehicles are used for transport, simulation can include
the effect of shielding attenuation of the fallout-source gamma radia-
tion.  In terms of the time-frame model, the shielding effectiveness
is defined as the ratio of the dose received under unprotected and

-------
                                  32
 protected conditions to that received under unprotected conditions over
 the interval (Tg+T2+Tv) due to the exposure modes  mentioned above.
 The time-frame model is thus formulated to Indicate the effects  of the
 time parameters on shelter effectiveness.   The effectiveness estimates
 in  this  study are mainly concerned with times  commencing at cloud
 arrival,  (TR+Ta),  in which simple  radioactive  decay by  each source
 isotope  is  considered over (TD+T ).   Note  also that the time-frame
                              K  a
 model assumes an abrupt boundary at both the leading and trailing  edges
 of  the radioactive cloud material.   Of  course, in  reality,  this  is not
 true,  as  it  is  well-known that turbulent diffusion in the atmosphere
 gives  rise,  on  the average,  to continuously changing airborne source
 boundaries—whose  dimensional  scales, however,  are such that the above
 model  would  be  a reasonable  approximation,  considering  the  source
 release intervals  of  interest  (excluding an instantaneous puff).
 DOSE REDUCTION  FACTOR
     The  estimated measure of  effectiveness afforded  by  a shelter
 structure—based on  the models  and assumptions  discussed  above—is
 referred  to  here as  the dose reduction  factor  (DRF).  This  value is given
 by  the ratio  of  the dosage received during  shelter  protection to that
which would  be  received in the  open.  DRF values are  estimated for both
 thyroid and WB  exposures.  The  DRF for WB gamma dose  is given as
                 (DRF)
                      Y   EC  + 1C  + FD
                            o     o     o
where
     EC  «* External gamma airborne source dose, sheltered,
     EC  » External gamma airborne source dose, unsheltered,
     1C  • Inhalation airborne source dose, sheltered,
     1C  » Inhalation airborne source dose, unsheltered,
     FD  = External gamma surface source dose, sheltered,
     FD  » External gamma surface source d
The DRF for thyroid gland dose is given as
FD  » External gamma surface source dose, unsheltered.

-------
                                  33
                                   TV
                    (DRF)Thyroid " fT    '

where
     TC  - Thyroid inhalation dose, sheltered,
     TC  - Thyroid inhalation dose, unsheltered.
     Table 5 summarizes the dose components given above, relating the
source, receptor, and time-frame conditions that were considered in
performing DRF calculations.  For example, EC and FD values include
estimates of external gamma WB dose for sources both inside and outside
a shelter structure for the exposure times (defined in Fig. 8, p. 29)
indicated.  Shelter dose components (EC, FD, 1C, and TC) also include
a portion of unsheltered dose contributions accumulated over the ex-
posure period, T-, to simulate the effects of shelter-access delay times
that assume no protection during that interval.  The remainder of this
section describes the development of these dose components used in the
calculation of the DRF values.
     Doses downwind from an accidental release of airborne gaseous
fission products are dependent upon the concentration of the airborne
radioactive material that can be expressed as follows for continuous
source release conditions:

                 X(r,t) =  (X(r)/Q) Q(t)           (Ci/m3)

                   3                                              3
where x(r)/Q  (sec/m ) is  the  ratio of the  concentration x(O  (Ci/m ),
                                                             •
at a distance r  from the  release to the source  release  rate  Q  (Ci/sec);
     •
and Q(t-x/u)  (Ci/sec) is  the  time-dependent  source-release  rate  function.
In general, the  dose at r is  given by the  integral  of the concentration
over the period  of exposure,  T ,
                              e
                  D(r) -IK.. x(r,t) dt        rem

-------
                                  34
                               Table  5


                            DOSE  COMPONENTS
Dose
Component


EC, FD


1C, 1C

EC , FD
o' o

1C , TE
0* 0
Source
In



X


X





Out
X

X

X




X


Receptor
In


X
X


X





Out
X



X




X


Exposure Times
T,, T 3
1 V
(1,-Ij) + T/
CT.-V + T2
T.
1
(Te-Tl) + T2


(T +T +T )
e 2 v

u
  TV and T2 are post-outside airborne cloud times and apply to the

fallout dose (FD) only.

-------
                                  35
where K_ is a dose conversion factor.  For this study, calculations
were performed assuming x(r)/Q unity (i.e., a unit dilution factor),
since it is a common multiplier for all dose components and therefore
does not affect DRF values.  Accordingly, the dose component estimates
described here are based on integrations of the time-dependent sources
both inside and outside the shelter structure.  The release rate at
the source is assumed to be constant with a correction for simple radio-
active decay over a release period, T ,
                                     8
                    Q(t)
                                 -xt
(Ci/sec)
where QQ is  the initial  radionuclide  activity  inventory  in  the  re-
actor at the time of an  accident  (Table 1,  p.  A),  f   is  the radio-
nuclide release fraction  (DBA  assumptions), and X  is  the radioactive
decay constant.  For ease of illustration,  the development  of the
following dose components does not  use  subscripts  designating each
radionuclide source, and  it should  be understood that summations over
radionuclide sources are  performed  in making computations.
     The calculations are obtained  from differential  rate equations  and
integration  over the time-dependent sources.   Derivation of these dose
component relationships  are detailed  where necessary  in  Appendix B.
 DOSE COMPONENTS—UNSHELTERED
      Whole-body  cloud  and thyroid dose components assuming no shelter
 protection  are  given as
                                          Q(t) dt
"EC 1
o
1C
o
TC
** nj




"K,
1
K ,-B
2
v • R
.K3 J

-X(TR+Ta)
e



-------
                                   36
"Kl "
_K 'B_
frQo e R *
T A
s
      -AT

  (1-e    )
                                                       ren
 where K-, K-, and K. are the dose conversion factors given for WB cloud


 gamma, WB inhalation, and the thyroid inhalation dose, respectively, and

                                               -4   3
 B is the breathing rate assumed to be 3.4 x 10   (m /sec).  In the above,


 the source release duration, T , is assumed to be the downwind exposure
                               S

 time, T .
        e

      The local fallout deposition rate outside the shelter is assumed

                         2

 to be V v(r,t) (Ci/sec-m ),  and the depletion rate to be due to only radio-
        O

 active decay.  Expressing the airborne concentration as x(r,t)

 where x  includes the exp [- x(TR+T )I term, the outside ground-fallout


 deposition,  F(t)  (Ci/ra ), is obtained from the following equation:
                                        - AF(t)
 Integrating,
                      F(t)
                          out
                   V x t
                    go
                                       -At
       (Ci/m2)
      The  fallout dose  component  is  given  by  integration  of  F(t)     over the


 time  of cloud  passage,  T  ,  plus  the contribution  from the fallout  source


 after cloud passage  integrated over the reference time,  (T2+T ).
   FD  = K.
     o    4
/'
                                    0
                                    2  v
 -At ,
e    dt
                                              -AT
                                                                    rem
where K, is the ground-source gamma-dose conversion  factor.

-------
DOSE COMPONENTS—SHELTERED
Airborne Source—Inside
     The source intake rate for a shelter structure is assumed to be
eLxo e   , where e is the ingress fraction (discussed above); and
L (time)   is the air change rate that is (f/v), where f is the volu-
metric air-inflow rate and v is the enclosure volume.  The internal
concentration C(t) is assumed to be reduced by air outflow, radioactive
decay, and internal radioiodine surface deposition at the rates given
by (L+X) C(t) and (V'/i.) C(t).   For internal surface deposition, V is
                    8                                             g
the deposition velocity inside the shelter (discussed above) and I is
the mean fall distance for iodine fallout material in the shelter
enclosure, assumed to be one-half the average floor-to-ceiling distance
or about 1.5 meters.  The significance of this coefficient as compared
with (L+A) per hour can be seen from
              _ 0.00025(m/sec) x 3600 (sec/hr)   . , .  -1
                            1.5 (m)            " °-6 hr

where (as indicated above) a range of 0.125 to 3 hr   for L was chosen
for the DRF calculations, and  A can range from about 0.0036 to 0.8 hr"1
for the radioiodines.  Since V - 0 for the rare gases, the (V'/A)
                              O                               £
coefficient is zero for determination of the internal noble-gas con-
centrations.  Based on the above, the internal structure concentrations
are determined from the following differential equation:
where
             K •= L + X + Kf   ,   and   K, - V'/£
                          1              f    8
Integrating,

-------
                                  38
 The dose accumulated in the shelter structure over the time interval
 (T1.Te)  is
      D -  G  K^  B  /    C(t)  dt

                 T,
             (L+Kf)
                           -XT,     -XT
                                -  e
  , / -KT.    -KT
-Me   ^e   €
                                                                 rem
     Specific dose components are  given by  the  above  equation,  depending
on the values of the dose conversion  factor,  K  ,  the  breathing  rate, B,

and the finite cloud-geometry correction  factor,  G, as  listed below:

EC1
IC1
TC1
KD
Kl
K2
K3
B
1
3.4 x l(f4
3.4 x 10~4
G
<1
1
1
After the cloud has passed the vicinity of the structure, the internal
concentration is
                 C'(t) - C(T ) e
                            e
                                -Kt
and the dose accumulated in the shelter due to the internal airborne
source is
                         r2
              D = G KJJ B /   C'(t) dt
                         0
                  G 1C  B             -KT,
                  —j^— C(Te) (1 - e   *)     rem

-------
                                 39
where
ex L
-XT
                                         -KT
                                      -e
The dose components EC-, 1C., and 1C. are obtained in the manner given
above for EC., 1C.., and TC...  Dosages due to airborne gaseous fission-
produced sources inside the structure are given by (ECj+EC-) +  (IC.+IC.)
for the WB and (TC+TC   for the thyroid.
Airborne Source — Outside
     During the time interval T., it is assumed that individuals are
unprotected and dose components are similar to those given for unpro-
tected exposures over the exposure interval to the cloud, T  , i.e.,
fEC'l
0
ic;
TC'
u 0J
«
V
YB
-K3-B-
-X(TR+Ta)
f Q e
rxo
V
                                                -XT
                                           (1 - e
                         rem
 The  attenuated WB  gamma-ray exposure  in the shelter structure  from the
 outside airborne cloud  source over the internal (T.,T )  is given by
                               j.

             j_ - A(l-G)  KjX0 f
                             T,
      dt
                                      -XT
                                   - e
                                                 rem
 where A is the cloud gamma-ray attenuation, (1-G) is the source-geometry
 correction factor for the outside cloud, and x —the reference concentra-
                 •                             o
 tion per unit x/Q—is

-------
                                  40
                               -X(TR«)

                              e
                          T
                           e
 Surface Source—Inside


      The differential equation for the surface deposition rate for radio-


 iodine in the shelter structure is
                    ^-V'C — —1
FD1 "
;Gie*0LK4 > i
K' JA2
r -ATI
(ATj+l) e
-AT \ . / -Kl
e 1 , ! L
-IT
(AT +1) e e
\ -KT \ f
1 e\\
                                                            retn
After cloud passage, the WB dose from internal radioiodine surface de-


position  that accumulated during cloud passage is given by

           F(Te)in  I   e"   dt
                   0

                                                               \1
                                                               ;J
VeG>CXQLK4 I  Te G   "    1  /-U    -KtW,    -AT2
  —_	 |  	.	___ j e    _ e   J (1 - e     /|  rera
where F(T ),  is the internal fallout level at T .
         e in                                   e

-------
                                 41
     After cloud passage, internal radioiodine fallout deposition con-
tinues to take place owing to the residual airborne source inside the
shelter structure.  The rate of internal fallout deposition is

                                        - XF(t)
Integrating the above yields the fallout deposition from the post-cloud
passage internal-airborne source in the shielding structure, which is
in turn integrated to yield the WB dose written as (see Appendix B)
FD.
V'G'ex I
 g   AQ
   K1
                  -AT
 -KT
                       - e
""
                                        - •
                  ) - * fr -
                                                                     rem
The WB external gamma dose from internal radioiodine fallout in the shelter
structure is given by (FD +FD_+FD_).
Surface Source—Outside
     During the time interval TI, the accumulated WB dose while seeking
shelter (unprotected) from outside ground fallout deposition is
                FD1 = K.
                  o    A
 T.
/
F(t)  ,. dt
    out
                      V Y K.
                       g*o A
 r
 ll -
                       ii
                       H
                                                       rem
     During the time interval (T1,Te), the WB gamma dose inside the
shelter structure from outside ground-fallout deposition is

-------
                                  42
                     Xl
                        ( i   r         -XTI            -XT 1 )
              - A> W«V21 <*Ti+1)  e    -  (xvl} e   ej (     rem

where A* is the shelter-structure  attenuation of gamma rays from the
ground-fallout source.
     The WB gamma dose  accumulated inside from residual outside ground
fallout deposition is
                               T2
          FDl - A'K.  F(T )   k  f   e"Xt dt
            2      4     e outJ
                              0
                             -XT
                         T   e  e       -XT
     After !», the time interval  assumed during which people may continue to
be in the shelter structure after outside cloud passage, the computational
model assumes that individuals  leave the vicinity of the shelter over a
time interval, T , either  unprotected  (e.g., on foot) or protected from
residual ground-fallout source  gamma radiation while leaving the con-
taminated area in a vehicle with  i
Accordingly,  the WB gamma  dose  is
taminated area  in a vehicle with a shielding attenuation factor  of A1.
                               T+Tv
                                //  v
                                      -At  .
                                     e    dt
                AM,   *  Te e"^ F -XT2    -*
-------
                                 43
     The external WB dose from shielded gamma radiation emanating from
ground-fallout deposition outside the shelter structure (with the
exception FD', where the receptor is assumed outside) is FD'  + FD,1  +
            o                                              o      1
FD^ + FD^.  Note that for FD^, FD^, and FD' (receptor inside) a geometry
factor—i.e., (1-G1)—is not assumed, which is consistent with the attenua-
tion values, A1, for ground-source fallout deposition.  That  is, the
fallout source on the roof of a simple structure would approximate the
ground-fallout source deposition in terms of source geometrical effects
for the reference dose point 1 meter above an idealized ground-plane
source, (see sketch below).
     Shelter
     structure
Fallout with no
structure
SHELTERING AND EVACUATION
     An investigation was made to determine the utility of the combined
protective action of sheltering and evacuation.  That is, both from the
standpoint of time constraints and th.e DRF, the combined .protective-
action measures may offer an advantage over sheltering only.  For example,
for individuals located relatively close to the point of the accidental
release in terms of either distance or cloud-arrival time, sheltering
may be the only option.  Furthermore, if the duration of the source
release were to continue longer than expected because of, e.g., wind
persistence or miscalculation, exit and evacuation from the shelter
structure may be advantageous in terms of dose savings as opposed to
remaining inside over the whole cloud-passage time.  The important
considerations in addressing this question are exit-time from the

-------
                                   44
 shelter structure, Tg,  evacuation transport time,  T,_,  and  cloud  exposure
 time,  T ;  and  protection characteristics  of the  shelter  structure  (see
 above) and any evacuation vehicle(s)  that may be used  from transporting
 people out of  a radiocontaminated area.
     The analysis of  the above  situation  is based  on a simple model
 (Fig. 9) assuming ideal sheltering conditions where both shelter-entrance
 delay time and residence time after cloud passage  are  zero (T. - 0,
 T_ - 0; see Fig. 8).  Figure 9  is a plot  of accumulated  dose as  a  function
 of shelter time  up to the cloud passage time, T  ,  where  the dose is D_.
                                               e                    o
 Values D..   and  D. are  to suggest possible  accumulated dosages for shelter
 exit at TS  and evacuation time  T_.  During  the interval  T  , the  model
 assumes that individuals  are exposed  to airborne and ground-fallout
 source radioactive material while in  an evacuation vehicle that  offers
 some degree of protection discussed below.   The  decision is' simplified
 to making  a comparison  of the estimated dose values.   That is, for
 D- < DC, exit  from the  shelter  structure  and evacuation  would be a serious
 consideration; whereas,  for D0  >  D_,  it would be more  advantageous to
                             L.     &
 remain inside  from the  standpoint  of  dose savings.
     An equivalent means  of deciding whether to  effect shelter exit and
 evacuation is  based on  DRF comparison.  The  actual numerical approach
 taken in this  analysis  is based on the question  of for what values of
 Tg, TT, and T£ is the relationship  (DRF)S/E  <_ (DRF)g satisfied,  where
                         D(T ).
                (DRF)S =  (Te)in      (shelter only)
                            e out
and
              . D(Vin - D(TS'Te>in * D(TS'TT>ev     (shelter and
          S/E               D^Vout                   evacuation)
where

-------
     D,

8
                                                                                                     Ln
                                                         I  	     1
                                    Shelter time
                    Fig. 9--She1tering and evacuation

-------
                                  46
           D(T )  f   " dose outside shelter structure,
               ^i     " dose inside shelter structure,
           D(T_,T ).  - dose inside shelter structure over interval
              o  e in
                        
-------
60r      1.0
50
          0.8
40
          0.6
30
          0.4<
20
          0.2
10
                   no A/C me tor
                    Tur
ning
                      10        20
                  30        40        50        60

                       Vehicle speed (mph)
70       80
90
                  Fig. 10--Air exchange and infiltration rates in closed passenger compartment
                           when air conditioning is set at a maximum

-------
 outside  air.   At  higher speeds  and with the same settings,  Q/V (air  change
 rate) would be expected to  approach the values  obtained with  the  air-
 conditioning  unit in  operation.   This  expectation is  based  on the
 assumption that general leakage  rather than the fan is  the  dominant
 factor determining f/v at these  speeds.   A value of 0.5 min   (30 hr  )
 was  chosen for this analysis, which corresponds to ^35  mph  when general
 leakage  is the dominant factor  for f/v.   Penetration  of gaseous fission
 products  into  evacuation vehicles  was  assumed to be 100 percent for  the rare
 gases and 80 percent  for radioiodines, which corresponds to the upper  limit
 of the estimates  of Megaw [12] based on  simple  shelter  structure  experiments.
     For  sheltering,  the DRF and  dose  component relationships are as given
 above (p. 28ff.);  where, for D(Tg,T )±   above,  Tg - 1^  and  TZ - 0.   For
 evacuation, the vehicle is  assumed to  be analogous to a shelter structure,
 and  dose  estimates for  D(TS,T_)    for  the exposure-evacuation time,  TT,
 are  based on the  same dose  components  considered for  shelters—with  the
 exception of radioiodine deposition inside the  vehicle, which is  assumed
 to be insignificant and, moreover,  cannot be modeled  accurately without
 some experimental  verification.
     External  WB  cloud-dose accumulation from gamma-ray penetration  of
 the  evacuation vehicle  is
                                 T +T
                                /S  T  -At
          EC_ - A.(l-G..)KlXo J      e    dt
out    v    v
                   T
                                S
                                -ATS 1
                ' V^Wo e     I (l - e    * j     rera

where A  is the vehicle attenuation for cloud-source gamma radiation,
G  the finite cloud-source geometry factor, and  K., the dose conversion
factor.
     Inside the evacuation vehicle, the rate of  concentration change
is

-------
                     dC(t)      T   -Xt   „ „, v

                     "It     exoLve    -KvC(t)
where Ky - L + A.




     Integrating the above for C(0) - 0 gives
C(t) •= eX0(e~Xt - e
                                  -K t
                                              (Ci/m3)
for the concentration in the vehicle.  The dose accumulated in the vehicle


over the period T  is given as
           in
                          T +T
                           S  T
                                C(t) dt
                G ex BK_ e
                 v Ao  D
                         -XT.
                         -XT
                                                      rem
     Specific dose components are obtained based  on  the  values  for  the


constants as given below:

EC,
in
1C.
in
TC.
in
S
Kl
K2
K3
B
1
3.4 x 10~
3. A x 10~4
K
v
<1
1
1
                                                                    -Xt
     The  ground-fallout deposition  given  above  is  F(t)     =  V  Y  t  e
                                                      out    go

The external WB dose  from  ground-fallout  source gamma-ray  penetration


of the vehicle during evacuation  is

-------
                                  50
 /•
J
Ts
                             F(t)out dt
                A'V x K,  r         -XT-                -X(T_+T,_)-i
              -  VR2° 4  I (XTS+1)  e   S - (XTS+ TT+1) e    S  T  I    rem   .

      The dose accumulated during  evacuation,  D(T ,T )  ,  corresponds to
 either the  WB or thyroid.   For the WB, the dose is (EC., +EC  >IC.  +FD   ):
                    '                                   in    out   in   out '
 for  the thyroid, TC.  .   Note that internal exposures for  EC.   and 1C.   are
 assumed to  accrue only when the evacuation vehicle is in  the vicinity
 of the airborne radioactive gaseous material  over the period T_,  which is
 a very good approximation considering the large values of L .   That is,
 in actuality,  once the vehicle leaves the vicinity of the airborne  radio-
 active material, x  in the differential equation above is zero and  the
 internal vehicle concentration drops very rapidly within  a few minutes—
 which  would not give  rise  to any  significant  dosage as compared with con-
 ditions when the vehicle is  assumed to be in  the vicinity of  the  airborne
 radioactive material,  provided of course T_ is  more than  a few udnutes.
 That is,  the equilibrium concentration that would be reached  in the
 vehicle within a few  minutes is given as
                       -Xt        -Xt
                      e      £*0 e           -At
                                  Lv
which would be approximately that outside the vehicle.  Then,  after  leaving
the vicinity of airborne contamination,  the concentration of gaseous radio-
active material in the vehicle falls off as exp[-(L +X)t], where L   is
M).5 min"1 (30 hr"1).

-------
                                  51
                           III.   RESULTS

     Estimates of shelter effectiveness have been made using the DRF
calculational model and assumptions discussed in Sec.  II.   It is, of
course, impossible within the scope of this effort to  develop informa-
tion comprehensive enough to anticipate what might be expected for all
practical situations.  Accordingly, assumptions are made regarding
input parameters and ranges of variables in order 1) to demonstrate the
degree of shelter effectiveness in a general sense, and 2) to indicate
sensitivity variations for some specific situations.
     Input values used in the sheltering calculations fall under two  .
categories.  First, a set of fixed parameters were selected  (summarized
in Table 6).   These values are in part based on study ground rules
(DBA gaseous-release assumptions), review and analysis of existing data,
and an attempt to develop representative information that can also be
related to the Reactor Safety 5tu<% [1].  The notion of "fixed para-
meters" obviously applies only to this particular analysis;  in reality,
there may be appreciable variations in shelter characteristics and,
for example, iodine deposition velocity.  The second input category
consists of the  temporal and ventilation rate variables selected  to
indicate the sensitivity and degree of sheltering effectiveness  over
their range of values.  In some cases, extrapolation can be  made  (with
care) to estimate shelter effectiveness beyond the  specific  range limits
used in making the calculations.
     Shelter effectiveness results are given in terms of the DRF  in
Figs. 21 through 31.  Enough data are given in Figs. 11 through  31 to
enable a fair amount of cross-plot extrapolation.  All the time variables
have units of hours and are identified in Fig. 8  (p. 29).  The ventila-
tion rate, L, is in units of hr   ; and SS and LS designate  the small  and
large shelter structure categories, respectively.  The plotted results
in Figs. 11 through 31 are discussed below.
     Figure  11 gives WB  DRF  as  a  function  of  time (T  )  in the shelter
structure  after  passage of  the  airborne  cloud  source,  assuming no delay
(T-0) in  shelter access after  initial  cloud arrival and  1-hr cloud

-------
                                   52
                                 Table 6


                           FIXED  PARAMETER SUMMARY
SOURCE TIMES
Release Time
Case TR' Hr
A 1.5
B 2.0
C 2.5
GASEOUS FISSION PRODUCTS
Release
Fraction
Kr and Xe 1.0
I 0.25
SHELTERS
Small
Structures (SS)
Cloud gamma attenuation, A 0.6 (0.4, 0.9)a
Fallout gamma attenuation, A* 0.2
Finite cloud factor, G 0.01
Finite fallout factor, G1 0.28
Release Duration
V Hr
0.5
1.0
3.0
Ingress
Fraction
1.0
0.51
Large
Structures (LS)
0.1 (0.05, 0.2)a
0.01
0.034
0.43
DEPOSITION
          V (outside) - 0.005 m/sec
          V(inside)
           g
0.00025 m/sec
 variations

-------
    0.5
    0.4
    0.3
OL

O
                               53
                             SS    L = 0.125
                                                        A


                                                        B
    0.2
    0.1
          A & B
                                   L = 1.0
                             LS    L = 0.125
                -L	1	1	1	L
0       0.2       0.4
                                   0.6       0.8


                                   (hours)
1.0       1.2
                  Fig. 11—WB DRF versus T?, (T.-O.T =1)
                                          £    i    
-------
                                   54
 travel  time  (T -1)  from the  point  of  source  release.   Significantly,
              £L
 more protection  is  afforded  by  the large  shelter  structure  (LS)  than
 the small one (SS).  The effect of ventilation rate is also more im-
 portant for  the  LS  than the  SS,  primarily because of  the difference in  -
 cloud-gamma  attenuation.  That  is,  a  larger  portion of the dose  in the
 SS is due to gamma  ray  penetration of the shelter from the outside airborne
 cloud source than in the LS; that  portion of the  dose does not depend
 on the  air change rate.   A value of L * 1 hr  may be somewhat repre-
 sentative, whereas  0.125 hr~  represents  a fairly low value associated
 with a  relatively tight structure  with very  little or no forced  air
 circulation.  The relative positions  of the  A, B, and C categories of
 release duration are also determined  by the  combination of cloud-gamma
 attenuation  and  ventilation  rate.   For the SS, the relatively larger dose
 component from outside  cloud-gamma ray penetration is sufficient to
 offset  the dose  component from  internal airborne  radioactive material.
 For example, for T2 = 0 and L =  1,  the relative positions of A,  B, and
 C are the same for both SS and  LS;  but the spread is  larger for  the LS
 than the SS, indicating the  effect  of a relatively larger number of air
 changes with respect to release  duration  (0.5, 1, and 3 for A, B, and
 C, respectively) for the LS as  compared with the  SS.   The crossover point
 at T. = 0.2  for  the SS  is due to the  increasing importance of the out-
 side ground-fallout dose component, assumed  to be reduced at a rate
 dependent upon only radioactive  decay, as compared with the dose from
 inside  airborne  radioactive material  assumed to be reduced at a  rate
 dependent upon radioactive decay,  ventilation, and internal fallout
 deposition of the radioiodines.  For  low  air-change rates (L=0.125 hr   ),
 the DRF for  the  SS is determined largely  from external sources,  where
the A,  B, and C  curve positions primarily reflect the  differences in
 radioactive  source decay.  For  the LS, the dose components from  outside
sources  are  relatively  less  important than those  for  the SS; and a
clear separation of the A, B, and C release-duration  categories  is
not seen when both inside and outside dose components  are relatively
more comparable.

-------
                                 55
     Figure 12 gives DRF plots for the thyroid for the same conditions
assumed for Fig. 11, which apply to both SS and LS.  In general, the
DRF values indicate somewhat more protection for the thyroid than for
the WB, and are more sensitive to T_; particularly for L » 1 hr~\
since there are no competing outside-source dose components.  The
relative positions for A, B, and C are due to the different number of
air changes associated with each source release duration.  Since the
DRF values in Fig. 12 correspond to a radioiodine ingress fraction of
0.51, they scale accordingly.
     Calculated results of the WB DRF sensitivity with cloud-source
arrival time are given in Figs. 13 through 16.  The DRF variations for
the thyroid, not plotted here, are insignificant as a function of cloud
arrival time, T .  In Fig. 16, the DRF decrease with T  for the SS is
               £1                                      Si
due to the relatively decreasing importance of the WB-dose component
from the outside airborne cloud source.  That is, the model includes
only simple radioactive decay and predicts that the relative contribu-
tion of the noble-gas sources to the WB dose decreases more with time
than does that of the radioiodines.  In reality, that decrease with T
                                                                     cl
may not be quite as prevelant, particularly for times longer than a
few hours.  The countereffeet, however, is indicated for the LS, since
the significance of the external gamma WB-dose component from the
outside airborne source is masked by the greater gamma-shielding
attenuation assigned to the LS as compared with  the SS.  The relative
positions of the A, B, and C curves are, as indicated above, due to
the increasing number of air changes, respectively, during  cloud
passage.
     Figure 14 shows WB DRF as a function of T   for case A, assuming
                                              Si
late shelter access  (T^O.25) coupled with extended residence time
(T2«0.5) after passage of the airborne cloud.  Ideal shelter timing
(T.^0, T2-0) is also shown for comparison to indicate the significant
loss of protection  for non-ideal shelter-access  timing.  Figure 14  also
shows  loss  of  inherent LS protection  advantage  (due to shielding),  as
compared with  the SS, because  of shelter-timing  considerations.

-------
                               56
DC

Q
                                   0.6       0.8


                                   (hours)
            Fig. 12—Thyroid DRF versus T_, (T,»0,T -1)
                                         L    J.    a

-------
                           57
0.5
0.4
0.3
0.2
             LS
0.1
                                              B





                                              A
      J	1	,	L
                      -I	L
                           Ta  (hours)
                             •
                                                       8
Fig. 13--WB DRF versus Ta,
                                               O),  L = 1

-------
   1.0
   0.8
   0.6
.
   0.4
   0.2
    0
             I      i
               J	L
                                          I       I
                            T, = 0.25  , T9 * 0.5
                                /
                                1 s 0
                                  - 0 ,  T0  =  0
                              T   (hours)
                               a
                                                           ss   -
                                                           LS
                                         J	I	I	L
                                                          8
Fig.  14--WB DRF versus Ta> case A, (Tj-O.Tg-O),
                                                          .S) , L - 1

-------
     0.6
     0.5
     0.4
 g  0.3
     0.2
     0.1
        0
                 J	L
                                59
                               T, « 0.25 , T9=  0.5
                                           ^ * 0
                                      0 ,  T2 =  0
                                T, (hours)
                                 a
                                                            LS

Fig.  15-WB DRF  versus Ta§ case B, fT^O.T^O),  (Tj.0.25,1 -0.5).  L

-------
     0.6
    0.5
     0.4
1  0-3
    0.2
    0.1
                              60
                                 I   .   I
                                                            ss
                                                            ss
                                                   J	I
                    23456789


                               T  (hours)
                                a
Fig.  16—WB DRF  versus Ta> case C, (T^O.T^O), (T^O.ZS.T^O.S),  L = 1

-------
                                 61
Figures 15 and 16 are similar plots for cases B and C, respectively,
where timing (T][ and TZ) is relatively less important, because of
longer exposure to the cloud source.
     Calculated DRF results in Figs. 17 through 22 show the effects of
shelter-structure ventilation rate.  Figure 17 gives the WB DRF as a
function of air change rate, L hr"1, for ideal shelter timing (T-0,
T2=0).  For L less about one air change per hour, the SS DRF is based
on a relatively larger external gamma-dose contribution from the out-
side airborne cloud source; for L greater than about one air change
per hour, the DRF is based on a relatively larger dose component from
internal -airborne radioactive material.  The LS DRF, on the other hand,
is based primarily on the relatively larger WB-dose component from
internal airborne gaseous radioactive sources for all values of air
change rate, L.
     Figure 18 gives the thyroid DRF dependence on the ventilation  rate
for ideal shelter timing.  Compared with the WB DRF values in Fig.  17,
the thyroid DRF functional dependence on L is much more pronounced,
since the thyroid dose is based solely on internal airborne radioiodine.
The advantage of low air-change rates less than about one-half per  hour
is quite apparent for protection against inhalation doses.
     Figures 19 through 22 show the comparative effects of ideal and non-
ideal shelter timing.  Figure 19 gives WB DRF values as a function  of L
for case A, showing considerable overlap of SS and LS protection for
ideal (Tj-0, T2=0) and nonideal (T^O.25, T2=0.5) shelter timing,.
respectively.  Figure 20 gives WB DRF for case B as a function of L when
the SS and LS overlap is much less than in case A  (Fig. 19), and Fig. 21
shows no SS and LS overlap for the longer source release duration category
(C) when the inherent LS protection advantage over the SS is maintained,
even for nonideal shelter timing.  Figure 22 gives plots of the thyroid
DRF, comparing the effects of nonideal and ideal shelter timing.  The
inversion in the order of the A, B, and C release-duration categories for
nonideal shelter timing is primarily due to the fraction of time an
individual is assumed to remain unprotected during cloud passage for
shelter-access delay time, T^ i.e., the fraction  of  time an  individual
is unprotected is larger for A than for C.

-------
                           62
0.5
0.4
0.3
0.2
0.1
           LS
: :
; 1 1
i
123
i
•
i
4
                          L (hours'1)
         Fig.  17--WB DRF versus L,

-------
    0.5
                             63
    0.4
OS
a
    0.3
    0.2
    0.1
                                L (hours'1)
              Fig.  18—Thyroid DRF versus  L,

-------
                                64
     1.0
                     I I I  I
I I I I
     0.8
     0.6
    0.4
    0.2
                                              111,1111
 T  .0.25 ,V
                                           0.5
                             L (hours'1)
Fig.  19--WB DRF  versus L,  case A, (T^O.T^O),  (T^O.ZS.T^O.SJ, T  - 1

-------
    1.0
                               65
    0.8
     0.6
     0.4
     0.2
                               = 0
                                        = 0
                                        = 0
LS
                                L (hours'1)
Fig.  20--WB DRF versus  L,  case B,  (T^O.T^O),  (T^O.25,1^0.5),  Ta  = 1

-------
                                   66
    0.6
    0.5
    0.4
u,   r\ i
Q:   0.3
o
    0.2
    0.1
                              L (hours'1)
  Fig. 21—WB DRF versus L,  case C,  (T^O.T^O),  (T^O.ZS.T^O.S)
= 1

-------
                           67
  1.0
                    i i i
                         i i i
        T, « 0.25 ; T9- 0.5
  0.8
  0.6
  0.4
  0.2
                            L (hours'  )
Fig.  22-Thyroid DRF versus L, (T^O.T^O), (T^O.25,1^0.5), Ta  =  1

-------
                                   68
      Figures  23  through  25  show WB  and  thyroid  DRF  values  in  terms  of
 T^  and  L  in parametric perspective.  WB DRF  values  are  shown  in  Fig. 23
 for case  B and for  LS.   DRF values  for  TI  -  1  (also the cloud passage
 time for  case B)  can  exceed unity for the  larger values of L;  this
 situation corresponds to entering a shelter  after cloud passage, which
 gives rise to additional dose  from  lingering internal contamination
 for T_  >  0, although  the relative loss  in  shelter protection  as  a
 function  of T is insignificant  for  all values  of TI and L.   Also,
 calculations  performed during  this study indicate virtually no protective
 advantage in  seeking  shelter after unprotected  exposure to a passing
 radioactive cloud.
     Figure 24 is essentially  the same  parametric perspective plot  as
 Fig. 23 for the SS.   Figure  25 gives corresponding  plots of the  DRF
 for  thyroid,  indicating  considerable overlap for certain parameter
 combinations  of shelter-access delay time, T.,  and  air  change rate, L.
     Figures  26 and 27 indicate  the effects  of  shelter-structure attenua-
 tion of gamma radiation  from the outside airborne radioactive cloud for
 the  SS and LS, respectively.  Figures 26 and 27 contain plots of the
WB DRF for independent variations of the cloud  gamma attenuation, A.
 In practice, such variations would not  necessarily be independent of the
shelter-structure attenuation of ground-fallout gamma radiation, as
normally some correlation would be expected.   The results  in Fig. 26
 indicate, however,  that  a factor-of-two increase in gamma  attenuation
 results in about an 80-percent increase in shelter  protection for the
 SS, whereas a factor-of-two  reduction in the air change rate  (2  to  1
 air  changes per hour) results  in about  only  an  8-percent increase in
 shelter protection.   In  the LS,  the effect of cloud gamma  attenuation
is not as significant.   Results  in Fig.  27 indicate that a factor-of-
two increase in cloud gamma attenuation results in a 50-percent  increase
in shelter protection, whereas a factor-of-two  reduction in air  change
rate (2 to 1 air changes per hour) gives rise to a 20-percent increase
in shelter protection.

-------
                                 69
    1.2
    1.0
                                                         • L
                                                              •1.0
                                                        0.125J
    0.8
                                                   _3.0
                                                   -1.0
                                                   -0.5

                                                    0.125
                                                              0.75
ce.
o
0.6
                        -3.0

                        •1.0

                        •0.5
                                                        0.125
                                                              0.50
    0.4
                                                              0.25
    0.2
          j	L
                                                        0.125
                0.2
                      0.4
   0.6


T2 (hours)
                                              0.8
1.0
1.2
             Fig.  23—WB DRF versus TQ, case B, LS, T  -  1
                                     c.               a

-------
                             70
    1.2
    1.0
    0.8
g   0.6
    0.4
    0.2

            J	L
                0.2
                                     i     i
                           i     i     i
0.4       0.6       0.8

      T   (hours)
                                                        •
                               3.0
                               1.0
                              ;0.5
                              0.125
                                                         3.0
                                                       = 1.0
                                                       -0.5
                                                        0.125
       1.0
                                    0.75
                                                   J	I	L
1.0
1.2
            Fig.  24--WB DRF versus T9, case B, SS, T  = 1
                                    c               a

-------
                             71
                                                 3.0
                                                 1.0
                                                 0.5
                                                 3.0   0.25
                                                0.125  0.5
                                                 1.0   0.25
0.02
                   0-4     0.6     0.8      1.0      1.2     1.4
           Fig.  25—Thyroid DRF versus T9,  case B, T  = 1
                                        f-           u

-------
                                 72
     1.0
     0.8
     0.6
     0.4
     0.2
                              IT     (I
                                                  I     I
                                J	L
               J	L
                              2.0  0.9
                              1.0  0.9

                             0.125  0.9
                              2.0  0.6
                              1.0  0.6
                              2.0   0.4
                             0.125  0.6
                              1.0   0.4
                                                       0.125  0.4
                 0.2
0.4
    0.6
T2 (hours)
0.8
1.0
1.2
Fig.  26--WB DRF versus T2, case B, SS  (A=0.4,0.6,0.9,1=0.125,1.0,2.0)

-------
                                  73
   0.30
   0.20-
cc
a
   0.10-
                                                       1.0  0.2



                                                       2.0  0.]
                                                       2.0  0.05


                                                       L.O  0.;
 1.0 0.05


0.125 0.2
 Fig. 27--WB DRF versus T2> case B, LS (A=0.05,0.1,0.2,1=0.125,1.0,2.0)

-------
                                  74
      Figures 28 and 29 are plots of shelter DRF as a function of shelter-
 delay access time, T.., for the WB and thyroid, respectively.  Figure 28
 plots are given for a range of 0.5 to 1.5 air changes per hour, which
 is recommended by Handley and Barton [7] as applicable to single-family
 dwellings.  The value of two air changes per hour for the LS is also
 consistent with their recommendations for larger structures and apart-
 ment buildings.  For ideal shelter timing, T. - 0, and !„ •> 0, the LS
 provides about twice the protection as the SS, which decreases with
 increasing access delay time.  Figure 29 gives thyroid DRF plots for
 the same conditions as Fig. 28; since the thyroid dose is dependent
 only on the ventilation rate, somewhat less protection is afforded by
 the LS than by the SS,  because of the difference in ventilation rate.
      Figures 30 and 31 illustrate the difference in shelter protection
 afforded for the WB and thyroid under ideal and less-than-ideal con-
 ditions.  In Fig.  30,  DRF values are  given as a function  of T.  for a
 low air-change rate (0.125  hr"1) and  TZ  - 0.   For T  =  0,  the SS pro-
 vides a factor of about 2.8 for WB-dose  protection;  whereas the LS
 provides a factor of about  12.5—a relative protective  advantage of
 4.5—for the LS over the SS.   In Fig.  31,  for less-than-ideal shelter-
 ing conditions (T2=0.25 and L=l hr"1), the SS provides  a  factor of
 about 2.2 protection for the WB dose; whereas the LS provides a WB
 protective factor of ^6.7—a relative protective advantage of about
 3—for the LS over the  SS.   The change in the protection  for the
 thyroid dose between ideal  and  less-than-ideal sheltering  conditions
 is  about eight-fold;  the  shelter protection factor of the  thyroid dose
 is  about 40  for ideal conditions and  about  5  for  less-than-ideal
 conditions.
      Figure  32  shows the estimated effect of  the iodine ingress  fraction
 on  the WB DRF.  The  rise in WB DRF is linear with  the iodine  ingress
 fraction, with  the slope primarily dependent on the ventilation  air-
 change rate.  The increase is most apparent for the LS, for  repre-
 sentative air change rates,  and least apparent for the SS, for low air-
change rates.  Again, this difference is due  to the relative  contribution of
the dose components from radioactive sources outside and inside the

-------
                                 75
       1.0
                             Release Time, TR » 2 hr
                             Exposure Time, T  - 1 hr
         01—L
J	,	L
                                                                '
                                                           J	L
                   0.2       0.4       0.6       0.8
                                   T   (hours)
                        1.0
1.2
Fig.  28--WB DRF versus TI?  case B,  (Ta=l,T2=0),  SS (1=0.5,1.0,1.5),  LS (L=2)

-------
                                    76
                                                      —1-71—I	1	1
      0
                0.2
0.4
0.6       0.8
(hours)
                                                       1.0
1.2
Fig.  29—Thyroid DRF versus T^ case B, (Ta=l,T2=0), L = 0.5, 1.0, 1.5, 2.0

-------
                                 77
at
o
   0.02
              0.2     0.4
0.6     0.8


    (hours)
1.0     1.2      1.4
    Fig.  30--WB and thyroid DRF versus TI§ case B, (T =1,T2=0,L=0.125)

-------
                                 78
   2.0
   1.0
   0.1
 0.02
        J	I	L
0      0.2     0.4
                             0.6     0.8
                              T   (hours)
 J	L
1.0     1.2      1.4
Fig. 31--WB and thyroid DRF versus Tj, case B, (T »1,T2=0.25,L»1.0)

-------
                              79
        0.1 -   •
                   0.2      0.4      0.6      0.8      1.0

                          Iodine ingress fraction
1.2

Fig.  32--WB DRF versus iodine ingress fraction, case B, (T,=O.T9=0,T =1)
                                                          icct

-------
                                  .80
 shelter.  As indicated above, a value of 0.51 was used in the calcula-
 tions for estimating the DRF.  If iodine ingress were 100 percent, the
 DRF may be from about 1.4- to 14-percent higher for SS and from about
 16- to 46-percent higher for LS.  Of course, for the thyroid dose, the
 DRF could be nearly double (assuming ideal shelter-access timing).
      Figures 33 through 35 are based on calculations for the combined
 protective action of sheltering and evacuation.  The results shown are
 for the shelter time, Tg, and the evacuation transport time, TT, which
 together would provide protection equal to that of sheltering alone
 during the period of cloud exposure, T .   The conditions of sheltering
 are consistent with ideal timing;  i.e.,  individuals are assumed to be
 in the shelter at the time of cloud arrival (T =0)  and exit immediately
 after cloud exposure (T2=0).   The combined protective actions of shelter-
 ing and  evacuation assume that individuals exit the shelter after a
 period,  TS,  and evacuate  during  the period TT,  while exposed to the
 airborne radioactive cloud material during its transit away from the
 shelter  area.   Accordingly,  if the structure  were  exited  after a
 shelter  period,  Tg,  evacuation out of  the  vicinity  of  cloud  exposure
 should not  exceed  the time period,  TT»  to  effect a  dose  protection at
 least  equal  to  that  provided  by  staying in  the  shelter.   Therefore,
 time  combinations  (Tg,TT) that lie between  the  curves  and the axes
 would  give  rise  to greater dose  protection  from sheltering plus  evacua-
 tion  than from  sheltering only.  For example, considering WB-dose  pro-
 tection  in the  SS  shelter for  low  air-change  rate conditions  and  a
 cloud  exposure period, Tg = 3 hr,  evacuation  from the shelter  vicinity
 should be accomplished in no more  than about  0.75 hr if the  shelter  is
 abandoned after 1 hr of cloud exposure; if  exit takes place  after  2 hr
 of sheltering, the evacuation time that should not be exceeded is
 shortened to about 0.4 hr.  Under  the higher  representative air change
 rate of L = 1 hr   , the maximum allowable evacuation times increase
somewhat to about 1 and 0.5 hr for respective shelter-exit times of
1 and 2 hr,  assuming a 3-hr cloud exposure period.   The increase in
allowable transit time, TT>  is primarily due to a larger dose incurred
in  the shelter structure with the higher air change rate.

-------
                                   81
     10
1-
3
O
    1.0
c
«J
    0.1
                                                      L =
                               0.125 hr

                               1.0 hr"1
                                                                  -1
                                                            \
       0.1
                                                             \
                                    \
           1.0


Shelter time,  TS (hours)
10
  Fig.  33—Sheltering with  evacuation,  WB,  SS--transit  time  versus  shelter  time

           (Ta=0.5)

-------
                                82
      10
    1.0
 -
-
.
    0.1
   0.01

             L i  042$ hi


             L »  i.o hi

               T   •   t

                        ,   ,  .
                                           *     *
                                                            4-


      0.1
1.0
10
                       Shelter  time, T«.  (hours)
Fig.  34--Sheltering with evacuation,  MB,  LS--transit time  versus  shelter

         time (T =0.5)
                3

-------
                           83
      10
     1.0
  -

  '
     0.1
    0.01
                  :   .
                  •   '
                  »   •
                  i   ;
                         ,  .
[ilm-_
L »• OJ125 Hr;

L f 1.0 hr. .
               1    '.
       0.1
      1.0
                                                      10
                     Shelter time, TS (hours)
Fig. SB—Sheltering with evacuation, thyroid—transit time versus shelter

        time (T =0.5)
             Q

-------
                                  84
     Figure  34 gives  shelter evacuation break-even  times for  the LS
 shelter.  For low air-change rate conditions, much  less time  is allowed
 for evacuation from the LS shelter area than from the SS for  a given
 shelter-exit time, Tg, because of the significantly greater margin of
 protection (lower DRF) offered by the LS.  For the  higher representative
 air change rate of L  - hr~ , the allowable transit  time from  the LS
 shelter again is less than that for the SS shelter; but the time
 difference is not as  great as compared with that for the low  air-change
 rate situation.
     Figure 35 gives  the shelter evacuation break-even time points for
 thyroid dose protection.  The lower maximum allowable evacuation transit
 times for the lower air change rate as compared with the higher representa-
 tive air change rate are due to the larger margin of protection provided
when air change rates are low and accordingly less  time is required for
 the accumulation of the break-even dose during evacuation from the
 shelter.

-------
                                 85
                   IV.  CONCLUSIONS AND RECOMMENDATIONS

     Shelter protection provided by a large variety of public structures
can provide a significant reduction in WB and thyroid dose from ex-
posure to radioactive gaseous fission products that might be released
during a nuclear power plant accident.  Protective sheltering is
attractive if shelter-access timing is ideal, but its effectiveness
diminishes almost linearly with access delay time after cloud arrival.
     Sheltering protection against inhalation exposures that result in
thyroid dose depends on the number of air changes taking place over
the period of exposure to airborne radioactive cloud material.  Shelter-
ing protection for WB exposures depends on the attenuation of gamma
radiation originating from the airborne cloud source, the number of
air changes during cloud exposure, and (to a lesser extent) the
attenuation of gamma radiation originating from the ground fallout
about the shelter structure.  Accordingly, optimum ventilation control
(low air-change rates during cloud passage) is more effective for
reducing thyroid dose than WB dose.  Albeit, ventilation control is
relatively more effective for reducing WB dose in LS than in SS.
     Large structures such as office buildings, multistory apartment
complexes, department stores, etc., generally would provide significantly
more sheltering for WB exposures than smaller structures such as single-
family dwellings—a factor of about 4.5 more during low air-change rate
conditions and 3 more for nominal air change rates.  That is, WB doses
would be reduced by a factor of 2.5 to 3 for SS sheltering; whereas
for LS sheltering, WB doses would be reduced by a factor of about 12
during low air-change rate conditions.  For representative air change
rate conditions, WB dose would be reduced by about 2.3 for SS and from
6 to 9 for LS.  Wli dose can be further reduced in a shelter structure
through use of expedient filtration; e.g., by stuffing cracks and open-
ings with cloth or paper materials, which would reduce radioactive material
ingress (discussed above, p. 10 ff.) and/or  the natural ventilation rate.
Similarly, another means of respiratory protection is to cover the nose

-------
                                   86
 and mouth area with such common items as towels, handkerchiefs, or toilet
 paper:  e.g., a crumpled handkerchief (or one with eight or more folded
 layers), a towel of three or more folded layers, or toilet paper of three
 or more folded layers can reduce inhaled radioactive material (particulate
 iodine in this study) by a factor of about 10 [35].  The reduction of
 WB dose in a SS, however, is not appreciable—about 2.5 percent for low
 ventilation rates and about 15 percent for representative ventilation rates.
 The reduction in WB dose in a LS would be more appreciable—about 13
 percent for low ventilation rates and about 70 percent for representative
 ventilation rates.
      The difference in thyroid dose protection between SS and LS shelters
 is not as apparent  as for WB dose,  because of the more nebulous correla-
 tion of building air change rate than gamma radiation-attenuation pro-
 perties with  the general type of structure.   The degree of variability
 in the air change rate—an important parameter affecting the thyroid
 exposure—prevents  meaningful estimates  of the thyroid DRF for SS as
 opposed to LS  shelters.   Accordingly,  LS may not necessarily have any
 protective advantage for thyroid dose  reduction  over  SS or vice versa,
 due  to any number of factors—open  portals,  filtering action,  air con-
 ditioning,  structural  integrity,  etc.  Sheltering protection for either
 SS or  LS,  however,  can result  in thyroid  dose  reduction by a factor  of
 from about  20  to  70  for  low air-change rates,  and from 4  to 10 for repre-
 sentative  air  change rates.  These ranges are  primarily due to the
 corresponding  range  of cloud-exposure periods  of  from 0.5  to  3 hr, where
 the DRF increases, although not  linearly with  the air  change rate  (or
 number of air changes).  Another important parameter affecting the thyroid
 DRF value  (also the WB DRF to a  lesser extent) is the  ingress  fraction,
 which is treated like  an effective filtering action in  this study.  For
 that parameter, a value of 0.51 was assumed for the radioiodines, based
 on review of limited experimental work discussed above  (p.  10  ff.).  The
 thyroid DRF values given would then scale linearly with whatever value
 is assumed.  The use of expedient filtration discussed  above for WB dose
 can be even more effective in reducing thyroid dose (i.e.,  reducing radio-
iodine ingress and/or ventilation by stuffing openings and cracks or

-------
                                 87
usini; such common items as handkerchiefs and towels for respiratory pro-
tection).   Such expedient filtration could reduce thyroid dose by a
factor of  about 10 [35].
     The protection against WB dose decreases linearly with the amount
of radioiodine penetrating to the occupied spaces of a shelter structure.
The decrease is more apparent for LS than SS, because of the relative
differences in the gamma ray attenuation from sources outside the shelter,
and is also related to the number of air changes that take place during
the cloud-exposure period.  For this analysis, an ingress fraction of
0.51 is assumed for making DRF calculational estimates.  This assumption
implies that radioiodine sources collect at certain locations in the
shelter structures.  Therefore, insofar as these locations could repre-
sent "hot spots," local exposure of individuals who may be adjacent to
these collection points could result in dose increase.  No attempt has
been made here, however, to deal with that problem other than to make
note of it.  In view of current uncertainty regarding penetration of
radioiodine into structures that could be used as shelters, the need
for more experimental results must be emphasized.
     The degree of WB dose protection afforded by shelter structures
as a function of cloud-exposure time depends largely on the relative
contributions of the exposure modes.  The larger the relative external
dose contribution from penetration of gamma radiation into the shelter
as compared with WB-inhalation dose, the less the effect of cloud-exposure
time on shelter effectiveness.  For example, for the SS where gamma ray
penetration is relatively more important, the DRF would remain relatively
constant for cloud-exposure periods up to several hours.  For low ventila-
tion rates, the sheltering protection may even increase somewhat—only
about 15 percent or so—because of changes in the radioisotope source
mix as a result of decay.
     For LS shelters, where the WB dose component from gamma ray pene-
tration is relatively less important than in SS shelters, the degree of
protection still remains nearly constant for cloud-exposure periods up
to several hours for low ventilation rates; but for representative
ventilation rates, the relative protection for sheltering diminishes

-------
                                  88
 significantly—e.g., a factor of about 1.7 for a 3-hr cloud-exposure
 period as compared with a 0.5-hr period.  The utility of ventilation-
 rate control in minimizing the number of air changes during sheltering,
 especially for LS, is strongly supported by the results of this analysis.
 Maintaining low ventilation rates is even more important from the stand-
 point of thyroid dose reduction for either LS or SS, as the loss in
 protection for the same cloud-exposure periods mentioned above would
 amount to a factor of about 2.5 for a representative ventilation rate
 of one air change per hour during sheltering.
      Small-structure shelter protection for WB doses tends to increase
 somewhat with cloud arrival time because of radioisotope decay and
 corresponding changes in radionuclide proportions.   For LS shelters,
 protection remains nearly constant with cloud arrival time, because
 of the relatively larger inhalation dose component;  this holds true
 even more so  for thyroid dose  protection.
      Shelter  protection for WB dose diminishes for  LS to a greater
 extent than for SS with increasing ventilation rates.  For a low
 ventilation rate (L=0.125 hr"1)  as compared with a  high ventilation
 rate (L«=4 hr   ),  SS shelter protection diminishes by a factor of
 ^1.32,  whereas  LS shelter protection diminishes  by  a factor of ^2.7;
 thyroid dose  protection decreases  by a factor  of ^6.
      The  attenuation  of  gamma  radiation  from airborne radioactive  material
 outside the shelter structure  is more  important  to  the WB  DRF  than that
 of ground  fallout  about  the  shelter.   Also, the  effect of  gamma  ray
 attenuation on the DRF  from  sources  outside the  shelter .is  more  signifi-
 cant  for  the  SS than  the  LS, whereas the converse holds  for the  ventila-
 tion  rate.  That  is,  a  factor-of-two increase  in  gamma attenuation
 results in about  an 80-percent increase  in shelter protection  for  the
 SS, whereas a factor-of-two  reduction  in the air  change  rate results in
 only about an 8-percent increase in  shelter protection for  WB  dose.
 For the LS, a factor-of-two  increase in cloud-gamma attenuation  results
 in a 50-percent increase  in  shelter  protection, whereas  a  factor-of-two
reduction in the air change rate gives rise to a 20-percent increase
in shelter protection.

-------
     The penalty in shelter protection for remaining in the shelter
after the cloud-exposure period depends on the number of air changes
taking place during cloud passage coupled with the relative contribution
to the dose from inhalation.  When air change rates are low, no signi-
ficant loss of protection for the WB dose in either the SS or LS occurs,
regardless of how long individuals remain in the shelter after cloud
passage.  WB-dose sheltering protection is not affected very much when
remaining in a SS after cloud passage; for a LS, shelter effectiveness
may be reduced from 10 to 20 percent by remaining in the shelter for a
period up to about an hour after cloud passage.  The sheltering pro-
tection penalty is much more pronounced for the thyroid dose, which can
amount to a factor of about a 1.2 to 3 increase in the DRF, as compared with
ideal shelter-timing conditions, should individuals remain in the shelter
for a period up to about one hour after cloud passage.
     The extent to which sheltering is attractive depends on the ratio
of the projected dose to the protective action guide  (PAG).  Generally
speaking, when that ratio is comparable to the reciprocal of the DRF,
sheltering is effective as  an emergency protective action.  Also,  for
conditions where the projected dose is so large as to  cause acute
injury, and the predicted time of cloud arrival prevents effective
evacuation, a reduction in  dose by even a factor of  2  to 3 may be
quite important.
     The combined protective actions of sheltering followed by evacua-
tion during cloud exposure  (as opposed to only sheltering) can be an
attractive option from the  standpoint of total dose reduction.  The
advantage becomes increasingly more attractive as the  degree of pro-
tection offered by a shelter structure decreases and/or the cloud-
exposure period increases.  That is, for WB DRF considerations, the
shelter/evacuation option is generally more attractive  for  SS  than LS
and also for high air-change rate conditions  than low  ones.  The air-
rate change considerations  are more important  for  the  LS  than  the  SS
as far as the option advantage is concerned,  and most  important for
thyroid dose protection.  Logistically,  the option  can be  attractive
for cloud-time arrival conditions that would  preclude  effective evacua-
tion coupled with increasing periods  of  cloud exposure.

-------
                                  90
     The extent  to which the results  for  shelter effectiveness developed
 in  this study can be applied to  the release of particulate airborne
 radioactive material from a nuclear incident can not be quantitatively
 estimated here for two reasons:  1) the relative contribution that
 radioactive particulates make to the  total dose depends on the extent
 of  their release; 2) the ingress of particulate fission-product material
 into shelter structures may be different from that assumed here for
 gaseous radionuclides.   Overall, however, shelters would tend to offer
 more protection in varying degrees than that indicated here for the
 gaseous fission product.   Therefore, application of the DRF values to
 particulate release material would be conservative.  Further mention
 should be made for some specific considerations.
     Shelter structures would be increasingly more effective in reducing
 dosages from inhalation exposures, for increasing proportions of partic-
 ulate release, simply because of effective filtering action.  For WB
 dosages, shelter structures would tend also to be somewhat more effective;
 however, the extent to which that may be the case is complicated by varia-
 tions in the dose component contributions.  In general, however, when the
 WB dose for nonshelter conditions (unprotected) becomes progressively
 more attributable to particulates, the more effective sheltering becomes.
Also, LS shelters would offer more protection than SS shelters for equi-
valent particulate release situations.
     Both experimental and analytical work is needed to more accurately
 and specifically assess the protective advantage of sheltering.
     In the experimental area,  the extent of radioactive ingress into
 potential shelter structures still remains uncertain.  Therefore, some
 effort using representative structures (or models) under controlled
 shelter-structure conditions and a variety of correlated meteorological
 conditions should be undertaken to obtain reliable measurements.  If
possible,  the experiment should also address representative particulate
 ingress.
     Another experiment that could yield useful information for shelter-
ing protection prediction is the measurement of WB external gamma dose
from airborne cloud material for shelter structures on an inside/outside
dose basis.   Of course, such an undertaking may be difficult in view of
the intentional controlled release of radioactive airborne material.

-------
                                 91
Such measurements, however,  could possibly be obtained in conjunction
with experimental programs carried out for verification of computer
codes used to predict off-site doses (e.g., the ERDA Health and Safety
Laboratory programs).
     In the analytical area, it would be useful to make additional esti-
mates of shelter protection for specific cases based on more definitive
shelter characteristics that might correspond to specific locations.
The principal specific parameters would be gamma ray attenuation, finite-
source geometry-correction factors, air change rate, fallout deposition,
and cloud arrival time.  Also needed is model improvement regarding
radionuclide source components.  To that end, it would be useful to
assess the effect on the shelter DRF when parent-daughter decay is
considered along with specific attenuation and finite source-geometry
correction actions for each radionuclide.  Finally, additional analytical
attention should be given to include estimates of  sheltering pro-
tection for radioactive airborne releases that contain particulate
material.  Such a research effort would focus on the extent and nature
of the particulates and their ingress into shelter structures.  DRF
estimates would also be made using the type of model for the gaseous
fission-product release addressed in this study.

-------
                                  92
                               Appendix A
                           FALLOUT GAMMA SOURCE
 FINITE  GEOMETRY CORRECTION
      Consider  the  following sketch for the dose calculated at a vertical
 distance  d from a  plane source of isotropic gamma-emitting material of
                              2
 source  strength S   (gammas/cm /sec):
                 3.
R2, is
     The dose rate at P  from  an  annular  source,  radially bounded from R  to
                                    4irr
                                                                    (1)
where k is a dose conversion constant, B(ur) is a gamma-ray  dose buildup
factor, and jj is the gamma-ray absorption coefficient.  The  ratio  of  the
dose D(0,R) to D(0,°°) is defined here as the finite plane-source geometry
correction factor given as

-------
                                93
                                           D(R>°°)
                                                                  (2)
                                                   222
Assuming the Berger buildup  factor  form, and since r  » p  + d ,
                   m
                 /
         kS   f               -yr
D(R,«0  = -~ I   (1+Cyr  eUwr) ^	 dr
              kS
    f  ^
   J     r
VR^?
                                dr  + Cy
                                          +d
                                                              (3)
Substituting u = yr for the first integral,  and evaluating:
              kS
                    V
                   fe-»
                  J  ~
                               du -
                                    Cy e
                     -(l-D)yr
                                       (l-D)y
                                                      7  7
             + _C	 e-(l-D)y
                                                 fR2-,d2l
                                                              (4)
where E^x) is the first-order exponential integral function.
since D(0,«>) = DU,"),
               lim
               R -> 0
                                                          Then,
                         ks
           0(0,00) . -a.  Ei(yd) + _£_ e-(i-D)ud
                                                                   (5)

-------
                             Appendix B




                        DOSE REDUCTION FACTOR






 DOSE COMPONENTS --UN SHELTERED




 WB Fallout Gamma Source



      The outside-fallout deposition rate is assumed to be
                              e"xt - AF(t>   •                       CD
 Multiplying (1)  by the integrating factor eAt,
which  can be written as  the  total  differential,
                                                                    (2)
                    	                O  **







Then, integrating








                                       t
              f d  [eXt1F(t')] = \XQfdt'    .
we have
                        F(t) = V x
                                go
and then
                F(t>n,,r - V X t e'U       Ci/in2
 Terms used in this appendix are listed on p. 102.

-------
                                95
     The fallout  dose  during cloud passage where x  1* 0»  over the interval
(0,Te),  is
          K,  /   F(t)     dt - V K,x  f
           A J    ^ 'out       g 4Ao./
          -Xt .
       t e    dt
                                                               (4)
Integrating by parts,
/•e      \f               0~xt
/   t  e    dt - -(Xt+1) -S-y-
                                     - * [*•
            -XT
e   w (XTe+l)|
                                  (5)
     After cloud passage, the residual ground fallout is
                 F' (t)    = F(T )    e
                     'out    v e'out
                                      -Xt
The WB gamma dose accumulated over the period  (T-+T  ) after  cloud  passage


from residual fallout is
      K,  /   F'(t)  ^ dt « K.F(T )     /"
       4 J        out       4   e out y

          0                            0
               -Xt .
              e    dt
^outlL1-'
                                                                    (6)
The outside  reference  fallout WB  gamma  dose (unprotected)  due to ground-


fallout deposition as  given by  Eqs.  (5)  and (6)  above is

-------
                                  96
                                          -AT
           FD  = V
             o
                  -  (AT +1) e
                     -XT
                 T  e
                  e
                           1 -  e
                                                 rem
                                                        (7)
 DOSE  COMPONENTS—SHELTERED
 Airborne  Source—Inside
      The  rate  of change  of the  airborne  concentration  in the  shelter
 structure during cloud passage  is
                                       -  IC(t)
                                                       (8)
               Kt
     Choosing e   as the  integrating  factor and  rewriting  as  the  total
differential,
        d   Ktp, .,
          te  C(t)l
                                         (R-A)t
                                         v
                                                                    (9)
where K = L+A+Kf and K-A * L+K-.  Integrating over  the interval  (0,t)
where C(0) = 0,
C(t)
                                     dt'
                                                 e
and the concentration is
               _, .    CXOL  , -At    -Kt.
               C(t)  - -r-  (e    - e    }
Ci/ra
                                                       (10)
     The dose in the shelter structure is given by integrating the con-
centration (10) over the interval (T ,T ) and multiplying by the appropriate

-------
                                 97
dose conversion and finite-source correction factors designated here
by K:
            T
            •2
            r
     D - K /   C(t) dt
           T,
            1
             L
                                                                   (11)
After cloud passage, the concentration in the shelter structure as a
function of time is
               ex L / -XT     -KT \
                 ol    o       e\  -Kt          T
                                   1           -" J    ,             (12)
and the dose accumulated in the shelter structure over  the  period T
after cloud passage is
            *2
     D = < J   C'(t) dt
           0
          (L+Kf
 Surface  Source — Inside
      In  the shelter structure,  the rate of change of surface-fallout
 deposition (assumed on  the floor space) is


-------
                                  98
      Again,  choosing  the integrating factor e   and rewriting as the

 total differential,
         d  , At
        dF [e
                              (L+Kf)
                                  (15)
 where  K1  =  (L+Kf).

     Then,  integrating where F(0) = 0,
                          Vlex L  r
                          V'eX L
                           8  °
                          (L+Kf)
['-* a-.*',]
 giving the inside fallout deposition as
WM    ,-  SCX°   rf-a~Xt    x  ^ ~At    -Kt,
F(t)in  T&HCj  te    -F"  (e    ' e   ^
                                                     Ci/m
                                 (16)
     The WB external gamma dose  accumulated over (T,,T ) is
                                                  1  e
   FD,
         G'V'ex  LK,
            e o  A
           (L+Kf)
 dt
  1   f& .  -At     -Kt,  .
~ K7" 7    ^6    ~  6   ^
                                                         .(17)
Integrating Eq.  (17) above  (first integral by parts)  gives

-------
                                 99
     FD
       1     (L+Kf)
           -XT             -AT
          e   x - (ATe+l)  e
                                                 -KT
                     - e
                                                           rem   .  (18)
     After cloud passage, the accumulated fallout level at T  is given by

Eq. (16) evaluated at I& O^Vin^ which diminishes by radioactive decay.
The WB external gamma dose accumulated over interval T. after cloud
passage is
      FD
                                 dt
V'G'ex LK.
g Ao 4
(L+Kf)
-AT
T e e ,
e 1
A K'A
, -At -Ktv
(e - e )
                                                                X
                                                                 1
                                - e     |    rem

                                          (19)
     After cloud passage, the continuing fallout rate in the shelter
structure due to residual airborne radioiodine is
                         - V'C'(t) - XF(t)   .,
                                           (20)
where C'(t) is given by Eq. (12).  Choosing eU as  the integrating
factor and rewriting as the total differential,
               At
               ^
V
                                  -AT
          dt
                                           (21)
Integrating, where F(0) - 0,

-------
                                 100
            it
                     V>X L / -AT     -KT.
                     (L+K,
                                   - e
                                            dt
             Vex L
              go

            
-------
                                101
Integrating by parts,
          FD'
            o
     r             -A*ii
      1 - ( Ij+iy e   *
                rem
(24)
     Similarly, the WB external gamma dose inside the shelter structure


accumulated over interval (TlfT ) from outside-fallout deposition is
                        T
          FD'
^'v X K, T

  S2°    (ATj+1)
                                    -XT
                     -XT 1
          - (XTe+l) e
                                                            rem  . (25)
SHELTERING AND EVACUATION—VEHICLE AIRBORNE CONCENTRATION


     The rate of concentration change in the vehicle is
                 -  Kvc(t)    ,
                                                                   (26)
where K  = L  + X.
               K t
                v
     Choosing e    as the integrating factor and rewriting,
                 _

                dt
 r K  t     i

 U V C(t)
= CX L
    o v
                                        L t
                                         v
Integrating where C(0) = 0,
                                                  (27)
                  K t

                 e V  C(t)

-------
                                 102
and the concentration is
                           /  _U    -K t\                1
                 C(O - ex I e    - e  V  )            Ci/mJ   .    (28)
DEFINITION OF TERMS



     F(t) = fallout (per unit area)



     C(t) = inside airborne concentration (per unit volume)



       X  = outside airborne concentration (per unit volume)



       V  = deposition velocity outside
        c»


       V = deposition velocity inside
        O


        X = radioactive decay constant (per unit time)



       K, = fallout dose conversion constant



        K = dose conversion constant



       T  = cloud exposure period



       T  = shelter entrance delay period



       T  = shelter period after cloud passage



       T  a evacuation period away from shelter



        L = ventilation turnover rate  (per unit time)



       Kf = V'/£ (per  unit time)
        1    8


        SL « mean fall  distance for iodine inside


        K = L +  A + K,



        e = ingross fraction



       G1  = finite-source  correction factor  for fallout



       LV » ventilation turnover rate  for vehicle  (per unit time).

-------
                                103
                            REFERENCES


1.  Reactor Safety Study, Appendix VI:  Calculation of Reactor Accident
      Consequences, WASH-1400 (Draft), U.S. Atomic Energy Commission,
      August 1974.

2.  Bell, M. J., ORIGEN-'The ORNL Isotope Generation and Depletion
      Code, ONRL-4628, May 1973.

3.  Nucleonics, Data Sheets No. 1-30, March 1955-February 1959.

4.  The Potential Radiological Implications of Nuclear Facilities in
      the Upper Mississippi River Basin in the Year 2000, WASH-1209,
      U.S. Atomic Energy Commission, January  1973.

5.  Radiological Health Handbook, PB-1211784R, U.S. Department  of Health,
      Education and Welfare, Public Health Service, September 1960
       (Rev. Ed.).

6.  Sabersky,  R. H., D. A. Sinema, and F. H.  Shair, Environ. Sci.
      Technol., 7,  1973, p.  347.

7.  Handley, T. H., and  C. J. Barton, Home Ventilation Rates:   A
      Literature Survey, ONRL-TM-4318, September 1973.

8.  Private  communication with  Dennis Latham, Los Angeles  City  Build-
       ing and  Safety Department,  Health  Division.

9.  Yocum,  J.  E.,  W. L.  Clink,  and W. A.  Cote, "indoor/Outdoor  Air
       Quality  Relationships," J.  Air Poll. Contr. Assoc.,  22,  May 1971,
       pp.  251-259.                                               y

10.  ASHRAE Handbook of Fundamentals,  American Society of Heating,
       Refrigeration,  and Air Conditioning Engineers,  New York,  1967.

11.   Coblentz,  C.  W.,  and P.  R.  Achenbach, ASHRAE Journal,  5, 1963, p. 69.

12.   Megaw,  W.  J.,  Intl.  J.  Air Vat.  Poll.,  6, 1962, pp. 121-128.

13.   Kabat,  M.J.,  "Iodine Forms  and  Removal Efficiencies," panel
     Discussion (Radioiodine Behavior Related to Power Reaction),
     Trans.  Amer.  Nucl.  SOQ.,  21,  1975,  p. 86.

14.   Private communication with M. J. Kabat,  Ontario Hydro, Pickering,
       Ontario, Canada.

15.   Hawley, c. A., Jr., et al.t Controlled Environmental Radioiodine
       Tests at the National Reactor Testing Station.  1965 Progress
       Report,  IDO-12047, February 1966.

-------
                                 104
 16.   Bigger,  M.  M.,  R.  J.  Crew,  and R.  K.  Fuller,  Non-Ingested Dose
        Associated with  Particulate  Ingress into  a  Prototype Shelter
        Via  the  Ventilation System,  USNRDL-TR-815,  January  1965.

 17.   Burson,  Z.  G.,  and A.  E.  Profio, Structure  Shielding  from Cloud
        and  Fallout Gamma-Ray Sources for Assessing the  Consequences
        of Reactor Accidents (Draft), E.G.&G., Las Vegas,  Nevada,
        March  1975.

 18.   Chabot,  G.  E.,  and K.  W.  Skrable,  "A  Simple Formula for  Estimation
        of a Surface  Dose and Photons Emitted  from  a Finite Cloud,"
        Health Physics,  27,  July  1974, pp.  153-155.

 19.   Design and Review  of  Structures for Protection from Fallout Gamma
        Radiation, PM-100-1, Department  of  Defense,  Office  of  Civil
        Defense,  February 1965.

 20.   Auxier,  J. A.,  J.  0.  Buchanan, C.  Eisenhauer,  and  H.  E.  Menker,
        Experimental  Evaluation of the Radiation  Protection Afforded by
        Residential Structures  Against Distributed  Sources, CEX-58.1,
        January 1969.

 21.   Burson,  Z.  G.,  D.  Parry,  and H. Borella, Experimental Evaluation
        of the Fallout Radiation Protection Afforded by  a Southwestern
        Residence, CEX-60.5, February 1962.

 22.   Burson,  Z. G.,  Experimental Radiation Measurements in Conventional
        Structures, Parti,  CEX-59.7B, July 1966.

 23.   Strickler, T. D.,  and  J.  A. Auxier, Experimental Evaluation of
        the Radiation Protection Afforded by Typical  Oak Ridge  Homes
        Against Distributed Sources, CEX-59.13, April 1960.

 24.   Summers, R.  L., and Z. G. Burson, Experimental  Evaluation of
        Techniques for Improving Fallout  Protection in Home Basements,
        CEX-65.5,  December  1966.

 25.  DCPA Attach  Environment Manual, Chapter 6,  "What the  Planner Needs
        to Know About Fallout," CPG-2-1A6,  Defense Civil Preparedness
        Agency, Department  of Defense, June 1973.

 26.  Burson,  Z. G., Experimental Evaluation of the Fallout Radiation
       Protection Provided by Structures in the  Control Point Area of
        the Nevada Test Site, CEX-69.5,  October 1970.

 27.  Burson, Z. G.,  Experimental Evaluation of the Fallout-Radiation
       Protection Provided by Selected Structures in the Los Angeles
       Area, CEX-61.4, February 1963.

28.  Borella,  H., Z.  G.  Burson, and J.  Jacovitch, Evaluation of the
       Fallout Protection Afforded by Brookhaven National Laboratory
       Medical Research  Center, CEX-60.1, October 1961.

-------
                                  105
29.  Stevens, P. N., and D. K. Trubey, Weapons Radiation Shielding
       Handbook, Chapter 3, "Methods  for  Calculating Neutron and
       Gamma-Ray Attenuation," DNA-1892-3,  Rev.  1,  March 1972.

30.  Ramsey, W., and P. R. ReeC, Land Use and Nuclear Power Plants,
       Case Studies of Siting Problems, WASH-1319  (no date).

31.  Manual of Protective Action Guides and Protective Actions for
       Nuclear Incidents, U.S. Environmental Protection Agency, Office
       of Radiation Programs, Environmental Analysis Division. September
       1975.

32.  Dickerson, M. H., and R. C. Orphan,  Atmospheric Release Advisory
       Capability  (ARAC):  Development and  Plans for Implementation,
       UCRL-51839, June 1975.

33.  Burson, Z. G., Health Physics,  26,  January 1974, pp. 41-A4.

34.  Petersen,  G.  A., and R.  H.  Sabersky, J. of the Air Poll. Contr.
       ASSOQ.,  25, October 1975, pp.  1028-1032.

35.  Guyton, H. G., et al., A.M.A.  Arch.  Indus.  Health,  20,  August
       1959, PP.  91-95.
                         * U.S. GOVERNMENTPDmniKOFFICE Mn-720-335/6098

-------