NERC-LV-539-6
NRDS NUCLEAR ROCKET EFFLUENT PROGRAM
1959-1970
by
D. E. Bernhardt, R. B. Evans, R. F. Grossman,
F. N. Buck, and M. W. Carter
National Environmental Research Center
U. S. ENVIRONMENTAL PROTECTION AGENCY
Las Vegas, Nevada
June 1974
This report was written under a Memorandum
of Understanding No. AT(26-l)-539
for the
U. S. ATOMIC ENERGY COMMISSION
-------
This report was prepared as an account of work sponsored by the
United States Government. Neither the United States nor the United
States Atomic Energy Commission, nor any of their employees, nor any
of their contractors, subcontractors, or their employees, 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.
Available from the National Technical Information Service,
U. S. Department of Commerce,
Springfield, VA. 22151
Price: Paper copy $5.45 Microfiche $1.45
-------
NERC-LV-539-6
NRDS NUCLEAR ROCKET EFFLUENT PROGRAM
1959-1970
by
D. E. Bernhardt, R. B. Evans, R. F. Grossman,
F. N. Buck, and M. W. Carter
National Environmental Research Center
U. S. ENVIRONMENTAL PROTECTION AGENCY
Las Vegas, Nevada
June 1974
This report was written under a Memorandum
of Understanding No. AT(26-1)-539
for the
U. S. ATOMIC ENERGY COMMISSION
-------
ABSTRACT
This report reviews the health implications of radioactive effluent released
during nuclear rocket engine tests at the Nuclear Rocket Development Station
(NRDS), Jackass Flats, Nevada, prior to 1970.
During this period, nuclear rocket engine concepts incorporated an open-cycle
hydrogen-cooled, reactor with core operating temperatures of approximately
4,000°F, which cause small quantities of fission products to migrate from the
fuel and to be released in the rocket engine exhaust as gases or micrometer-
sized particulates; radioactivity released in this manner is called the
"aerosol effluent." Core operating conditions caused minor fuel element
erosion and corrosion, releasing particles with high radioactivity content
(roughly 1011 fissions per particle) and diameters of tens of micrometers;
these particles are referred to as the "particulate effluent."
NRDS adjoins the Nevada Test Site and the Nellis Air Force Range and is located
about 80 miles northwest of Las Vegas, Nevada, in an area of low population
density and limited agricultural usage. Lathrop Wells (population less than
100), the nearest town, is about 15 miles from the testing locations.
Estimates of doses, both potential and actual, resulting from exposure of the
off-site population to aerosol effluent from past tests have been 15% or less
of the appropriate Federal Radiation Council (FRC) guides and Atomic Energy
Commission (AEC) standards.
Possible modes of interaction of the particulate effluent with humans which
have been investigated include deposition of particles on skin, in or near
the eye, in the respiratory system, and in the GI tract. The most limiting
mode of interaction, considering biological consequences and probability of
occurrence, appears to be deposition on skin. There are no AEC or FRC guides
or standards applicable to this type exposure, but the testing organization
with the assistance of consultants has established an operational guide. The
guide stipulates that the probability of a deleterious interaction (defined
-------
-4
as a certain dose to the skin) should be less than 10 . Several tests have
resulted in potential probabilities of interaction near or above this guide
in off-site unpopulated areas. However, there are no known instances of
interaction of the particulate effluent with individuals in the off-site
population.
11
-------
PREFACE
This report is a compendium of information about the Project Rover ground
testing at the Nuclear Rocket Development Station, Jackass Flats, Nevada,
prior to 1970. The testing program was directed by the Space Nuclear Systems
Office (SNSO), formerly the Space Nuclear Propulsion Office, a joint admin-
istration of the Atomic Energy Commission (AEC) and the National Aeronautics
and Space Administration.
Quantities and characteristics of effluent for future tests cannot be directly
inferred from the material presented here because of continued efforts to im-
prove reactor fuel element operating integrity and to reduce the quantity of
radioactive effluent. The effluent from the nuclear furnace test conducted in
June 1972, the first test since 1969, was passed through a liquid scrubber
which significantly reduced the quantity of radioactive effluent released to
the environment. In addition, the nuclear furnace was fueled with a new type
fuel with fission product release characteristics varying from those dis-
cussed in this report.
This report refers to a number of Federal Agencies, several of which have
been reorganized in the last several years. The report was written at the
request of SNSO, under the auspices of a Memorandum of Understanding between
the Environmental Protection Agency (EPA) and the AEC. Per this Memorandum
the National Environmental Research Center-Las Vegas (NERC-LV) is responsible
for conducting an off-site radiological safety program for Nevada Test Site
and NRDS activities. This Memorandum was originated between the AEC and the
Public Health Service, Department of Health, Education, and Welfare, but,
with the formation of the EPA, was subsequently transferred to the NERC-LV.
Similarly, the Air Resources Laboratory, Las Vegas, Nevada, (ARL-LV), which
provides meteorological services for the testing program, is under the
administration of the National Oceanic and Atmospheric Agency, but was formerly
under the Environmental Sciences and Services Administration.
111
-------
TABLE OF CONTENTS
Subject
ABSTRACT i
PREFACE iii
LIST OF FIGURES vi
LIST OF TABLES vii
I. INTRODUCTION ' 1
II. REACTOR DESCRIPTION AND EFFLUENT DEFINITION 4
A. Reactor 4
B. Reactor Effluent 7
1. Effluent Release -- Normal Operation 8
(a) Aerosol Effluent 8
(b) Particulate Effluent 11
2. Effluent Release — Accident Condition 14
C. Source Term Measurements 16
1. Aerosol Effluent 16
2. Particulate Effluent 19
III. EFFLUENT TRANSPORT AND DISTRIBUTION 23
A. Effluent Transport 23
1, Test Location and Site Geography 23
2. Factors Affecting Transport and Deposition 25
3. Plume Rise 26
4. Aerosol Effluent Transport 28
5. Particulate Effluent Transport 34
B. Effluent Distribution 38
IV. DOSE MODELS 40
A. Aerosol or Gas-Cloud Model 42
1. External Gamma Exposure 43
2. Other Approaches to External Gamma Exposure 45
Predictions
3. Evaluation 46
IV
-------
Subject Page
4. Thyroid Dose via Inhalation Pathway 47
5. Thyroid Dose via Ingestion -- Milk-Food Chain 52
B. Particle Dose Prediction Models 53
1. Particle Skin Dose Model 57
V. RADIATION PROTECTION GUIDES 61
A. Guides for Total Body and Internal Organ Doses 62
1. Normal Operation 62
2. -Accident Conditions 63
B. Guides for Exposure to Discrete Particles 64
VI. ENVIRONMENTAL LEVELS AND DOSES 68
A. Aerosol Effluent 69
1. Off-Site 69
2. On-Site 75
B. Deposition of Particulate Effluent 79
VII. SUMMARY AND CONCLUSIONS 88
REFERENCES 92
APPENDICES 102
DISTRIBUTION
-------
LIST OF FIGURES
Figure
Number
Title Page
1. Conceptual Reactor Schematic. 5
2. Fraction of Isotope Chain Assumed Released During
Full Power Run. 12
3. Population Centers Near NRDS. 24
4. XE Prime, EP-IXA, Normal Run Centerline Dose
Estimates. 33
5. XE Prime, EP-IXA, Particle Prediction. 37
6. Locations Where Fresh Fission Products Were Detected
(Phoebus-IB, EP-IV). 39
7. Dose Models. 41
8. Off-Site Whole-Body Gamma Exposures and Infant
Thyroid Doses Resulting from Reactor Engine Tests
from CY 1959 to 1969. 72
9. Pre-Event Dose Predictions and Dose Estimates from
Surveillance Results. 76
10. Relative Quantity of I On Natural Vegetation. 78
11. Cloud Centerlines for Selected Reactor Tests. 80
12. Estimated Hypothetical Whole-Body External Gamma
Exposures from Activity Deposited by Selected
Tests Indicated on Figure. 11. 81
13. Cloud Centerlines for Reactor Engine Tests for which
Particle Deposition was Documented in Detail. 83
14. Three Dimensional Representation of Particle Deposi-
tion for Pewee I, EP-III. 85
15. Probability of Receiving Critical Dose at Kreb's
Depth for Four Tests. 86
16. Normalized Maximum Particle Concentration. 89
VI
-------
LIST OF TABLES
Table
Number
Title Page
1. Summary of NERVA Reactor Development. 3
2. Estimated Migration Releases. 20
3. Estimated Corrosion Releases. 22
4. Effluent Parameters. 27
5. Parameters for Radioiodine. 48
6. Biological Parameters for Radioiodine Isotopes. 50
7. Occupational Dose Standards. 62
8. Comparison of Maximum Hypothetical Whole-Body Gamma
Exposures and Infant Thyroid Doses with Radiation
Protection Standards. 73
9. Summary of NRDS Nuclear Rocket Testing Results. 74
10. Typical Effluent Prediction Parameters. 77
vn
-------
I. INTRODUCTION
From the conception in 1958 of Project Rover, the program to develop a
nuclear powered rocket engine, through June 1970, nineteen nuclear reactors
or engines were tested at the Nuclear Rocket Development Station (NRDS).
About 80 miles northwest of Las Vegas, Nevada, NRDS is part of a large
complex of government-controlled testing reservations known as the Test
Range Complex and including the Nevada Test Site, the Nellis Air Force
Range, and the Tonopah Test Range. The distance between testing facilities
at NRDS and the outer boundaries of the access-controlled Test Range Com-
plex varies between ten and 80 miles. The closest populated area beyond
the Test Range Complex is Lathrop Wells, about 15 miles to the south.
This report summarizes information concerning the public health implica-
tions of the radioactive effluent released from past Project Rover Reactor
tests. The report includes descriptions of the mechanisms of production
and release of the radioactive effluent, the mechanisms of environmental
transport of the effluent and its potential interaction with man, and the
implications of this interaction with man. Emphasis has been placed on
the public health versus occupational health aspects of the program.
Engines tested recently have utilized a single-pass open-cycle hydrogen-
cooled reactor. During rocket engine tests, airborne radioactive gases
and particles have been transported to and detected in areas beyond the
Test Range Complex. As part of this testing program, considerable effort
has been devoted to determining the physical, chemical, and radiological
characteristics of the radioactivity releases, evaluating the potential
biological effects of the releases, and maintaining a radiological safety
program to insure the safety of on-site workers and off-site residents.
Table 1 contains the highlights of the development of the nuclear rocket
engine. *•»»•' The development of the nuclear rocket began in early 1955
C2")
under joint sponsorship of the Air Force and the Atomic Energy Commission.
-------
The development of nuclear rocket engine technology was the responsibility
of the Los Alamos Scientific Laboratory (LASL), Los Alamos, New Mexico.
The Air Force was responsible for the non-nuclear portion of the project.
The National Aeronautics and Space Administration (NASA) was formed by
executive order in 1958. In 1960 a joint AEC-NASA agency, the Space Nu-
clear Propulsion Office (now known as the Space Nuclear Systems Office),
was created to administer development of an operational nuclear rocket.
The first reactors of the program were called KIWI's after the flightless
Kiwi bird of New Zealand. Initially the KIWI reactor was intended as a
low-power ground test engine of roughly 1,000 Mw; it was proposed that a
larger version, the Phoebus class of reactors with a power of 5,000 Mw,
would be developed for flight engines. The Phoebus program was cancelled
after several tests, and the present concept is based on development of a
rocket engine of lower power and thrust than the Phoebus or Kiwi designs.
The success of the first nuclear reactor test, KIWI-A, on July 1, 1959,
encouraged SNSO to initiate industrial participation in the program. In
July 1961, Aerojet General Corporation was chosen to develop a Nuclear
Engine for Rocket Vehicle Application (NERVA) with Westinghouse Astronuclear
Laboratory (WANL) as the subcontractor to provide the nuclear reactors. The
NRX or NERVA experiments were aimed at developing the basic KIWI reactor
into a nuclear rocket engine. The NERVA program includes reactor develop-
ment and development of equipment for and testing of a breadboard system.
The NERVA program progressed to the "Experimental Engine" (XE) phase with
the test of the XE Prime in the winter and spring of 1968-69. The engine
was tested in the down-firing configuration (as opposed to the up-firing
attitude of previous tests), with auxiliary components in a semi-flight con-
figuration and in a partial vacuum to simulate the environment to be encoun-
tered in space. The Pewee Reactor tested in 1968 and the Nuclear Furnace
Reactors are scaled-down models for testing changes in technology in a
nuclear environment.
-------
Table 1. Summary of NERVA Reactor Development.
Year
Reactor Test Series
Configuration
1957
July 1959
KIWI-A
Sept. 1962 KIWI-BIB
Aug. 1964
Sept. 1964
Jan. 1965
June 1965
KIWI-B4E
NRX-A2
KIWI-TRANSIENT NU-
CLEAR TEST
PHOEBUS-1A
Feb.-Mar.
1966
NRX-A4/EST
Jan. 1967
June 1968
Nov.-Dec.
1968
Feb.-Sept.
1969
NRX-A6
PHOEBUS-2A
PEWEE-I
XE PRIME
Program started.
Gaseous
First power reactor test.
H_ coolant, 5 min. test, 70 Mw power
used micrometer-sized U0? particles,,,,
in a carbon matrix--plate geometry.
First liquid H? cooled reactor, 800 Mw.
Internal core structure failed.(2)
First all bead (UC-) fueled reactor which
contained a redesigned core support
structure. 900 Mw, run of 5 x 105Mw-sec.
(Ref.2)
First NERVA reactor test, 1100 Mw,
3.5 x 105 Mw-sec. Demonstrated flow
control characteristics.
Reactor nuclear safety test experiment.
First Phoebus reactor test. 1100 Mw,
7 x 105 Mw-sec. Ran out of run tank hy-
drogen and had loss of coolant incident.
(Ref.3) Some of the fuel elements had
exterior cladding.(Ref.4)
*
First experiment with breadboard design
and first bootstrap startup to power.
Maximum power 1140 Mw, peak integral
power during a test 1.1 x 106 Mw-sec.
Demonstrated that a nuclear rocket system
could start and operate on its own power
and operate over a wide range of condi-
tions. The total series nominal full
power operating time was 30 min.(Ref.2)
Operated at full power (1100-1200 Mw) for
60 min., 4.5 x 10* Mw-sec.(Ref.2)
Nominal full power 4200 Mw. Several short
runs, 15 min. run.
Scaled down KIWI type reactor for fuel
development, 500 Mw, 20 min. run time.
Exterior surface of all elements coated.
First test of a NERVA reactor in a flight
type configuration. Reactor in down-firing
position with components clustered in a
semi-flight configuration.
*A breadboard engine contains the principal components of a flight-test system
with the components arranged for test convenience rather than flight.
-------
II. REACTOR DESCRIPTION AND EFFLUENT DEFINITION
A. Reactor
The Kiwi/NERVA reactor design uses highly enriched uranium in a graphite-base
core cooled by hydrogen (the propellant of the rocket engine) which is ex-
pelled to the atmosphere. The reactor operates at high power densities
(Megawatts per litre) with core and coolant exit temperatures of about 4,000°F,
well above the melting point of most metals. The core structure is largely
(2)
composed of graphite and metal carbides. By its design, the reactor is
under-moderated and consequently operates on an intermediate energy neutron
spectrum.
Figure 1 is a general schematic of a conceptual design for the NERVA engine.
The hydrogen coolant flowrate is roughly 90 Ib/sec at 1,500 Mw. For ground
testing the hydrogen is ignited as it exits the nozzle to minimize explosion
hazards.
The reactor can be controlled either by the control drums in the periphery
of the core or by changes in the coolant flowrate. Power is measured by
neutronics measurements and by temperature, pressure, and flowrate measure-
ments at power.
Three different cryogenic systems are used in the ground test program. Hydro-
gen is used as the coolant and propellant. The specific impulse of a pro-
pellant is inversely proportional to the square root of the molecular weight
of the exhaust gases, making non-oxidized hydrogen the most preferable pro-
pellant. (In actual rocket operation in space the hydrogen would not be
oxidized.) Nitrogen is used in various phases of reactor cooldown and for
providing an inert atmosphere for some of the test areas. Helium is used
to purge the coolant system of nitrogen prior to the use of hydrogen.
The pressure vessel is the only containment for the reactor. The only housing
is a three-sided shed set up around the reactor during non-testing periods.
-------
THE HOT BLEED CYCLE
LIQUID
_.HYDROGENN
X_JANK /
100%
•3%
97%
TURBINE POWER
CONTROL VALVE
D
OVERALL HYDROGEN
FLOW PATTERN
Q»— GIMBAL
PUMP
3%
OF REACTOR-]
EFFLUX |(
BORON CONTROL DRUMS
IN PERIPHERY OF CORE
XTURBOPUMP
TURBOPUMP
EXHAUST
BERYLLIUM REFLECTOR
NOZZLE COOLANT PIPE
(CARRIES ENTIRE
HYDROGEN FLOW) (100%)
97%
Figure 1
Conceptual Reactor Schematic1
-------
The core of recent reactors, about 4 feet long and 4 feet in diameter, has
been composed of clusters of fuel elements, hexagonal rods composed of UC?
235
beads (containing oralloy uranium—93% enriched in U) in a graphite
matrix. The core structure design limits the core coolant flow to the
coolant channels in the fuel rods; coolant does not flow between the rods
or rod clusters.
High reactor operating temperatures cause fission products to migrate from
the fuel beads through the graphite matrix into the coolant. To reduce
fission product release, coating or cladding materials (not cladding in
the sense of metallic claddings for other types of reactor fuels) are used
on the beads and the elements to protect the elements and to reduce the
migration of fission products from the fuel.
There have been continual changes in the design of the reactor core support
structure, the means of incorporating the uranium into the fuel, and the
coating materials for the fuel beads and elements. Additional changes are
contemplated for the future. The changes are made to improve core integrity
and thus increase the design operating power and potential run time. These
changes have resulted in significant decreases in the fraction of the fission
inventory released by present day reactors as compared to the early Kiwi
reactors.
The reactor engines are tested at test stands or test cells, which contain
coolant storage and distribution systems and auxiliary equipment needed for
testing the reactors. The reactors are, so to speak, plugged into the test
stands or cells.
The initial tests were conducted at Test Cell A. When the reactor power and
run time were increased, Test Cell C, with its increased coolant storage
capacity, was used. Reactors were tested at Cells A and C in an "upside
down" configuration, with the exhaust expelled vertically upward rather than
downward. Structural support for the test vehicle was easier to construct
with the engine in this position.
-------
The XE Prime test was the first of the Engine Test Stand Experiments (ETS-1).
The ETS-1 tests positioned the reactor "normally," with the gases exhausting
vertically downward and being ducted out at an angle above the horizontal.
The component clustering for the ETS-1 tests approached flight configuration,
and the engine was operated in a partial vacuum to simulate a space environ-
ment.
Reactor engine tests are termed Experimental Plans (EP's). The first EP is
normally a criticality checkout of the general neutronics of the reactor
and its control systems and instrumentation. Other tests prior to the full
power tests include intermediate power tests. Subsequent tests include the
full power tests, operating systems tests, restarts, etc. As discussed
later, the amount of effluent released is a function of the operating tem-
perature as well as core design and total operating time. Significant
effluent release is largely limited to the full power tests, the primary
subject of this paper. (Radioactive effluent has been released not only
during the actual tests but also during reactor cooldown after full power
tests.) Coolant must be passed through the core to remove the fission pro-
duct decay heat. The required cooldown period depends on the experimental
plan operating power and time integral. A general rule-of-thumb for the
basic reactor is one day of cooling per 106 Mw-sec power integral.
After the Phoebus-IB test in February 1967, LASL developed a filtration
system to be placed over the reactor nozzle during reactor cooldown. Called
the "FROG" (Filter Reactor Outlet Gas), this filter is part of a rail-
mounted shed which can be placed around the reactor during non-testing periods,
This system has not been used on the test stands.
B. Reactor Effluent
The full power operating parameters for reactors (temperature, flowrate, etc.)
result in the release of a small fraction of the fission products. This
fraction has varied from test to test from about 10% to less than 1% of most
of the fission products (50% for Cd). ' The releases have occurred
through two modes: migration of fission products from the fuel elements into
-------
the coolant stream, and actual release of fuel fragments through erosion
and corrosion of the fuel. ' ' Since some of the fuel fragments have
disintegrated into aerosol-sized particles, field measurements have not
been able to fully distinguish between the two modes.
Fission products released by migration tend to condense or agglomerate on
to atmospheric dust particles or small particulate matter released from
the reactor, and on surface dust entrained by the reactor plume forming a
composite gaseous and small particulate (micrometer size or less) effluent. '
This composite will be called an aerosol since it acts as suspended material
and essentially follows the air stream. ,
The core fragments are generally in the size range of tens of micrometers.
(Ref. 13,14,15,16,17,18) Because of their size and density they act as
particles and fallout along the effluent track. Some of the material re-
leased as UC_ may be oxidized to U02> an
-------
Prediction of migration releases is based on two empirically
supported assumptions: (1) the rate of loss of any fission
product, dN/dt, is proportional to the amount of that fission
product (N) in the portion of the element or core being con-
sidered; and (2) the migration coefficient (K) for the
fission product of interest is a function of temperature (T).
(9 10 1Q)
Based on the previous assumptions ' ' "J
dN/dt = -KN II-l
and •
K = K exp(-F/RT) I1-2
WHERE:
N - amount of fission product in the fuel.
t - time
K - rate constant or migration coefficient;
K , migration coefficient at reference
temperature
F - activation energy based on laboratory data
R - gas-constant
T - absolute temperature of the fuel being considered
K,K , and F are empirical constants that are established
from data from past reactor tests and labora-
tory experiments.
WANL has noted that the release of some of the radionuclides can
be described better by two release constants. They relate this to
a ''slow migration" for the coating material and a "fast migration
parameter" for the graphite matrix. This interpretation
results in equation II-3.
dN/dt = -(K^ + K2N2) where N = 1^ + N2 II-3
The parameters for these equations vary according to the design
and are empirically determined.
In estimating the releases of various fission products, the
yield and decay sequence of the complete chain must be con-
sidered. For example, the I chain may be represented as:(20,21)
-------
131
mTe(30 hr)
i ?i -I-?-! X --
Sn(1.32m) + 10-LSb(23m) * 4- 0.20
*<%>
v. •*• sr
** '
° ' J Thus the important mechanism for
131 !31
release of I from the reactor is not the migration of I from
beaded fuel, but rather the migration of Sn and Sb. This brings
about two additional considerations:
1. If a reactor with beaded fuel is re-started several days
after a previous test, the primary source of I release is
the Sn and Sb iodine precursors from the second test. The
actual iodine release fraction is less than its precursors.
2. The release fraction has not been the same for all of
the iodine isotopes because of differences in their produc-
tion chains.
10
-------
Figure 2 indicates release fractions predicted for various isotope
chains for the Phoebus-2A reactor test. The curve was determined
from measurements on the Phoebus-1A. The curve was utilized for
subsequent tests by estimating the release for several mass chains
(based on changes in design) and then transposing the curve to fit
f4 7)
those points. ' In general the mass chains with the highest
fractional releases are those of relatively low fission yield.
b. Particulate Effluent
Release of core material through corrosion and/or erosion is more
difficult to describe and measure than the migration of fission
products from the core. The hydrogen coolant reacts with graphite
at operating temperatures to produce methane and other hydro-
(22)
carbons. The release mechanism has been described as chemical
corrosion which in its later stages is accompanied by physical
erosion of the fuel. ' The release mechanism has been related
to both minor breakage of fuel elements and reaction of hydrogen
with the fuel as a result of defects or failure of the fuel element
cladding, in some cases absence of cladding on sections of the fuel.
(Ref. 4,9,22,23). The fuel element losses have been related to
thermal and/or mechanical stresses. The quantity of material
released by corrosion has been found to increase approximately expo-
nentially with fuel core operating time ' and has been related
to the number of operating cycles.
The primary purpose of the cladding material is to protect the graph-
ite core matrix material. Graphite loss has been the predominant
factor in the loss of core reactivity from corrosion. It has been
estimated that roughly 10 grams of graphite per element could be lost
K £ . .... «. ,. . , . (10,22)
before a significant amount of uranium is lost. '
Mode? of corrosion can be separated into three categories:
1. Bore corrosion—from the coolant channel. This can be
further subdivided into mid-band (horizontal plane centered
on the vertical axis) or hot-end (exit end of core) corrosion.
11
-------
140
150
MASS NUMBER
Figure 2
Fraction of Isotope Chain Assumed
Released During Full Power Run
(Phoebus 2A)7
12
-------
2. External corrosion of elements (external surface, not
coolant channel). The reactor design minimizes, but does
not completely prevent, the flow of hydrogen around the
exterior of the fuel elements.
3. Corrosion and erosion of broken fuel elements.
The relative significance and extent of these mechanisms has varied
from reactor to reactor according to the design and type of fuel.
Corrosion can also create a feedback effect. Reactivity loss and
possibly gain as a result of corrosion is compensated for by ad-
justing the control drums. Adding reactivity by withdrawing the
control drums from their optimum position results in greater power
peaking in the periphery of the core, thus accentuating not only
corrosion, but also diffusion losses. The normal variation of
fission density (fission/gram of U) with physical position in the
core has been as high as a factor of 15.
Bore corrosion takes place as a result of flaking and/or cracking
and possibly imperfections of the bore coating. Hydrogen attacks
the graphite fuel matrix through these breaks in the fuel surface
protection causing either pinholes or holes in the bore surface. '
If the breach in coating is small, a cavity may be produced within
the matrix leaving only a pinhole at the surface. However, the
cavity wall may also collapse, resulting in a hole.
(22)
Lyon indicates that uranium release from the Pewee-I reactor
was primarily from broken and massively corroded fuel. The type
of corrosion was not identified, but it was probably extensive
bore corrosion. Lyon concluded that about 170 grams of U from
broken or massively corroded fuel elements was released, but
that about 100 grams from exterior and pinhole corrosion was
f221
probably retained in the reactor.^ 3
Quantitative projections for releases have been based on releases
from past reactors, releases from laboratory tests if there has
13
-------
been a significant change in fuel design, and scaling factors for
variations in reactor operating parameters and fuel changes.
The following outlines the WANL prediction technique.
1. Postulate reactivity loss for the proposed reactor tests
(considering temperature, time, fuel channel coating, pro-
posed number of operating cycles).
2. Estimate the uranium loss associated with the reac-
tivity loss. This estimate is evidently based on
inspection and analysis of the NRX-A5 reactor. It
assumes analogous modes of fuel loss and similar axial
fuel loss distribution.
3. Estimate the fission product loss associated with the
uranium loss. There is a significant variation in fission
product production per gram of uranium in the core because
of the axial and radial variation of power density. Also,
fuel may be lost prior to the end of the reactor test.
For the NRX-A5 reactor, it was estimated that 1.2% of the
fission product inventory was released per percent of
uranium lost from the core.
From this information, the uranium release for the NRX-A6 was estimated
to be 650 grams. This was three times the amount that Pan American
Airways (PAA) accounted for by an integration of ground deposition.
LASL estimates have been based on a similar approach where past
results were extrapolated to future tests. The pertinent factors
included operating temperatures, number of temperature cycles, and
. . _ . , . (22)
changes in fuel design.
2. Effluent Release—Accident Condition
The parameters for evaluating accident releases for nuclear rocket
engine reactors have varied with time. The changes have been largely
due to availability of information and changes in reactor design.
14
-------
The maximum design accident, related to uncontrolled drum roll-out,
is assumed to occur without coolant flow.
The following criteria have been used to evaluate the control-drum
roll-out accident:
a. LASL - For Kiwi and Pewee reactors:
2 x 1019 fissions have been postulated to occur as
•a result of a drum rotation rate of 45° per second.
(Ref. 3,4,13,24) The associated energy release is
near the threshold of dynamic destruction, but little
(25)
if any dynamic destruction has been expected.
(*? f\~\
The release assumptions of TID-14844 have been
used: 1% solids, 50% halogens, and 100% of the
volatile fission products, the cumulative total of
which is equivalent to 15% fission product release.
(Ref. 3,13,27) An excursion of 2 x 1020 fissions
was estimated to be possible for the Phoebus-2A.
b. WANL - For NERVA or NRX type reactors:
7,000 Mw-sec, which is equivalent to about 2 x 1020
fissions, has been postulated to occur as a result
of a drum rotation rate of 55° per second. WANL
has applied a general release assumption of 50%
for all of the fission products. The release
assumption is based on results from Kiwi-TNT, la-
boratory experiments, and predicted reactor core
temperatures.
The difference in the energy release assumptions appears to be due
to the different rates of control drum rotation. The estimates have
been based on the reactor criticality neutronic calculation com-
puter codes of LASL and WANL, respectively. The difference in re-
lease assumptions is minor.
15
-------
The loss-of-coolant accident is limiting for a reactor with a
significant fission product inventory remaining from a power
test. For conservatism the loss of coolant was assumed to occur
at the end of the power run, the time when the reactor would
have a substantial fission product inventory. An analogous
accident would be a control drum roll-out (to provide the
energy to melt the core) subsequent to a power test.
The following assumptions have been used to predict the conse-
quences of such an accident:
C *) f\\
LASL - The previously referenced TID-14844 assumptions.
(Ref.3,13,24)
WANL - 50% of the fission products. ('8'10-)
Experience to date has indicated that fuel bead material should not
be encountered in effluent under accident conditions. Material of
this nature was not noted for the KIwi-TNT test or the Phoebus-lA
test (partial destruction of core due to loss of coolant and over-
, .. , (3,13)
heating). ' J
C. Source Term Measurements
1. Aerosol Effluent
Three methods have been used to document reactor releases:
Aircraft sampling
Elephant Gun sampler
Post mortem analysis of the reactor.
The results from these documentation techniques have been compared
to the pre-event predictions and used to improve the estimates for
future reactor tests. The first two techniques are based on:
a. Evidence that approximately 50% of the Cd produced has
been released. The fraction released was due to migration.
This fraction has been fairly constant for the normal oper-
ating limits. Thus, the quantity of Cd in a sample has
been compared to the estimated total produced and from this
16
-------
ratio the fraction of the total effluent sampled has been
ely
(4)
estimated. The release of Cd is largely due to the
rapid migration of its silver precursor.
99
b. Mo has not been released by diffusion. Thus, the
99
quantity of fission product Mo present in effluent has
been taken to be an indication of the amount of corrosion.
The techniques of post mortem analysis have not been adaptable to
quantitating the small fractions of release of most of the fission
products. Also the time increment between the reactor test and
analysis has been such that quantities of most of the fission pro-
ducts have been significantly decreased by radioactive decay. The
primary objective of post mortem analysis has been to obtain reactor
design information, but it has also been possible to quantitate the
Cd release by analysis of the core for Cd. Thus, the Cd
release fraction has been refined and used to scale the quantities
of other radioisotopes to the total quantity released.
Both LASL and NERC-LV have used aircraft sampling to estimate efflu-
ent releases. The LASL technique, used on NRX/EST and subsequent
tests prior to XE Prime, involved aerial sampling of the plume at
numerous heights and transects within about six miles of the reactor.
Particulate samples taken during the plume transects were analyzed
for specific radionuclides and the quantities related to total
effluent based on the Cd results. Analysis of the samples was
performed by LASL and/or WANL. The results of this technique have
served as the official source term for most of the reactor tests
prior to XE Prime. The LASL aircraft sampling program has been
largely discontinued since 1967.
The NERC-LV aircraft sampling technique is based on sampling on a
pre-fixed coordinate system to obtain an estimate of the cloud size
and concentration profile within the cloud. The sampling equip-
ment aboard the aircraft includes a mass particulate sampler and
17
-------
charcoal bed (to obtain continuous samples), sequential sampler
(for obtaining profile), cryogenic sampler or a "Grab Sampler"
to obtain samples of gases in the cloud, electrostatic precipi-
tator, Anderson impactor (particle size), and instruments for
measuring external radiation exposure to obtain real-time posi-
tion relative to the cloud. The LASL sampling has taken place
at about one to five miles from the test location prior to
significant depletion of the cloud due to fallout. The NERC-LV
sampling is done at tens of miles from the test stand by which
time depletion may have occurred. Reports have been written
concerning the results for each event. Additional information
concerning the sampling procedure and examples of results are
given in references 11, 12, and 28.
The initial source term estimates based on radiochemistry and
inspection of the core material suffered from lack of sensitivity
and from inability to resolve the effluent release into compo-
nents for specific experiments within a test series. Aircraft
sampling improved the quantitation of effluent releases. However,
variations of the effluent composition with time and difficulties
in accurately locating the cloud (caused by wind shear with
height and variations in wind speed and direction with time) have
made accurate measurement difficult. Also, the LASL program was
based on particulate sampling and thus obtained no samples of the
noble gases and probably only a fraction of the sub-micron size
particles.
In 1967, LASL developed a new sampling apparatus known as the
Elephant Gun. The device is composed of a probe which can be
extended into the effluent, directly over the nozzle, to obtain
samples which can be stored in sampling tanks. Samples equiv-
alent to several moles of hydrogen effluent may be taken at
selected times throughout a reactor test. The sampling mechan-
ism is sealed when it is removed from the effluent, thus
18
-------
retaining any of the gases. The fission products are removed from
the sampling tanks by acid leaching and the solution is then ana-
lyzed for appropriate radionuclides. Problems concerning sampling
with the Elephant Gun relate to the concentration gradient across
the plume and insuring proper and unbiased sampler operation at the
effluent temperature and pressure.
Table 2 gives the pre-event predictions and post-test estimates of
the I and gross gamma releases for selected reactor tests. The
data indicate a definite decrease in the fractional release from
recent tests versus tests such as NRX/EST.
WANL and LASL have provided estimated source terms for such fission
, . 89C 91V 91C 95 and 977 111, 115r. 132_
products as Sr, Y, Sr, Zr, Ag, Cd, Te,
131"135I, 235U, 99Mo, and 144Ce.(6>8>9) The NERC-LV reports present
values for most of the isotopes indicated above (normally excluding
Cd and U) in addition to results for the noble gases. ' ' '
2. Particulate Effluent
The presence of particulate matter in the effluent in the size range
of tens of micrometers was observed prior to the early NRX tests. '
However, there was only limited evaluation of the released material
until the NRX/EST test in March of 1966.^ Not until the Phoebus-IB
reactor in February 1967 was emphasis placed on before-the-fact pre-
diction and after-the-fact quantitation of particulate effluent
(14,15)
releases.
There are several ways of quantitatively estimating the amount of
material lost as large particles:
a. Relate reactivity losses, as measured by relative
control drum position, to the uranium loss.
b. Relate the quantity of uranium or number of particles
collected by aerial sampling to the total release.
19
-------
TABLE 2. ESTIMATED MIGRATION RELEASES
(5)
ro
o
Reactor Test
Date Full Power
Mw
Power Integral
(106 Mw-sec)
Max . Chamber
Temp. (°R)
Percent Release*
Pre-Event
Predictions
Post-Event
Estimates
I Gamma I
NRX/EST-A4, EP-IVA
Phoebus-IB, EP-IV
NRX-A6, EP-III
Phoebus-2A, EP-IV
Phoebus-2A, EP-V
Pewee-I, EP-III
03/25/66
02/23/67
12/15/67
06/26/68
07/18/68
12/11/68
1200
1340
1140
4010
3430
503
1.1
2.6
4.5
4.5
2.5
1.53
4150
4000
4150
4000
4000
N/A
3.9
4.0
2.7
4.2
5.7
10.0
2.3
N/A
2.1
4.0
4.0
7.0
5.0
1.0
0.6
0.1
0.1
0.8
Gamma
3.0
0.5
0.8
0.2
0.1
0.5
*This includes only releases due to migration Cc°rrosion fraction subtracted from total).(5)
N/A = Not Available
-------
c. Data from Elephant Gun: Although the Elephant Gun was
originally intended only for aerosol sampling, it has been
noted that the number of beads or large particles impacting
on the probe can be estimated from autoradiographing the
probe. ' This estimate can be summed with time and
integrated for the total nozzle cross-section to indicate
the total release. The total release can also be estimated
from the amount of uranium in the Elephant Gun samples.
d. Survey the fallout pattern to obtain the number of
particles per unit area. Based on the particle size dis-
tribution and area distribution, the results can be integrated
to obtain the total mass deposited.
e. Corrosion losses have also been estimated from post
mortem data, but difficulties exist in distinguishing the
quantity of material lost in the effluent from the quantity
disturbed in reactor disassembly. This procedure does not
allow losses to be related to specific tests.
In essence, a combination of the previously indicated techniques has
been used. Not all of the techniques have been used for each of the
tests; i.e., neither the Elephant Gun nor aerial sampling by LASL,
(NERC-LV sampled) were used for the XE Prime reactor test series. The
Elephant Gun has only been used at the test cells, not at the test
stand. Table 3 indicates the release estimates for previous reactor
tests. Both pre-test and post-test estimates are given.
21
-------
TABLE 3. ESTIMATED CORROSION RELEASES
C5)
NJ
K)
Reactor Test
Phoebus-IB, EP-IV
NRX-A6, EP-III
Phoebus-2A, EP-IV
Phoebus-2A, EP-V
Pewee-1, EP-III
Date
02/23/67
12/15/67
06/26/68
07/18/68
12/11/68
Full Power
Mw
1340
1140
4010
3430
503
Power Integra]
(106 Mw-sec)
2.6
4.5
4.53
2.5
1.53
L Grams of
Pre-Event
Estimate
500
Not made
125
N/A
150C
Uranium
Post -Event
Estimate
200
210a
60
55b
140d
The reactivity loss corresponded to 650 grams of uranium. The PAA integration of the
ground deposition indicated the value of 210 grams. WANL concluded that this indicated
that only 0.3 of the uranium loss was related to particulate effluent.(10)
Van Vleck*- •* indicates 450 grams for Phoebus-IB, and 70 grams for Phoebus-2A, EP-IV.
C SNPO Operations Plan for Pewee-I.
d 170 grams/22) 139 grams, (-3°-) and 100 grams. (-16')
-------
III. EFFLUENT TRANSPORT AND DISTRIBUTION
A. Effluent Transport
1. Test Location and Site Geography
The NRDS test cells and stand are located southwest of the Nevada
Test Site, approximately 80 miles northwest of Las Vegas, Nevada.
Test Range Complex boundaries are approximately 10 miles to the south
and west of the test cells and over 80 miles to the north. Lathrop
Wells, Nevada, about 15 miles SSW of the test cell, population less
than 100, is the nearest community. Beatty, Nevada, population
about 500, lies about 30 miles to the west. The terrain surrounding
the test cells is largely mountainous except for a long, flat valley
extending for about 40 miles to the southwest.
The Nevada Test Site and the Nuclear Rocket Development Station
encompass an approximately rectangular area 50 miles north to south
and 30 miles east to west within a region of generally north-to-south
trending ridges and valleys. The Nellis Air Force Range provides an
additional buffer zone. The terrain is extremely irregular, with
elevations ranging from a high of 7700 feet on Rainier Mesa in the
north to lows of 3100 feet in Frenchman Flat in the southeast and
2700 feet in the extreme southwest corner on the edge of the Amargosa
Desert. The test locations are at about 3800 feet MSL. There is a
general but frequently interrupted downward slope from north to south.
The areas adjoining NRDS are sparsely populated and the desert envi-
ronment limits agriculture. Locations and populations of communities
near NTS and NRDS are shown on Figure 3. The Figures in Appendix A
give a more detailed description of populated areas and the distri-
bution of milk cows in the off-site area.
23
-------
LAS VEGAS
300,000
16,000«-BARSTOW
500 OR MORE PEOPLE
Figure 3
Population Centers Near NRDS
24
-------
2. Factors Affecting Transport and Deposition
Quiring has summarized the climatological data for the NTS and
(321
the NRDS. Van Vleck presented a summary of diffusion param-
eters related to the NRDS testing program. Observations of meteor-
ological variables have been recorded in certain locations at the
NTS and NRDS since 1956, when the U. S. Weather Bureau began its
observation program. Upper-air soundings have been made at the
Yucca weather station since October 1956. Since that time more
than 100 sites have been instrumented for various periods to meas-
ure wind, temperature, relative humidity, precipitation, or com-
binations of these parameters. Observations of winds aloft have
been generally made at intervals of five or six hours, except during
test support, when the interval has been shortened to one hour and
occasionally to 15 minutes. Around-the-clock surface observations
were started at the Yucca station in December 1961 and have continued
with a few minor interruptions since that time. In addition to the
continuous record at the Yucca station, observations of the winds
aloft are made at various points around the NTS and NRDS for specific
operations.
What is generally referred to as "good weather" is usually experienced
at the NTS. Cloudy days are rare in summer and winter, and precipi-
tation is even more rare. During late winter and spring, severe
winds sometimes create dust storms. Operations can usually be sched-
uled months in advance with delays for weather conditions resulting
(331
only from undesirable wind directions. Fultyn reported atmos-
pheric temperature profiles taken very near to test times for ten
reactor tests. All profiles were nearly equal to the dry adiabatic
lapse rate through much of the atmospheric layer containing the plumes.
The Air Resources Laboratory, Las Vegas (ARL-LV), National Oceanic
and Atmospheric Administration, Department of Commerce, is responsible
for meteorological support (measurements and data evaluation) and
effluent radiation dose predictions.
25
-------
3. Plume Rise
The coolant gas in most tests has been expelled vertically upward
(XE Prime is an exception) at speeds of about Mach 3 and discharge
temperatures of about 4,000°F. The total thermal energy of the
plume has been about four times as great as the nuclear energy of
the reactor, because of the chemical energy generated in burning
the hydrogen. The total thermal energy release rate for the
Phoebus-IB, EP-IV plume while operating at a power of 1,500 mega-
watts was about 1.5 x 109 calories/second, of which about 3.6 x 108
calories/second were contributed by the nuclear energy of the
reactor. Vertical momentum of the jet is believed to contribute
negligibly to the height of rise of the plume; the controlling factor
is thought to be the thermal energy of the plume. ' For purposes
(34)
of comparison, Briggs indicates that thermal energy release rates
for stack gases from the largest commercial power plants are on the
order of 10 calories/second.
The effective plume rise has been important from several standpoints.
The higher the plume rise, the lower the maximum ground level concen-
tration will be and the further out from the point of release it will
occur. Thus, underestimation of plume rise leads to conservative
dose predictions. If the atmospheric wind velocity varies both in
speed and direction with altitude, the plume rise determines the winds
that provide the effective transport medium. Under atmospheric con-
ditions with a large amount of shear, an incorrect estimate of the
plume rise can result in gross errors in the predicted trajectory
and transport speed. Estimates of average wind speeds and plume
(^f\*\
heights from several tests are given in Table 4. The average
altitude for the peak concentration with hydrogen coolant flow rates
of about one hundred pounds per second has been about 10,000 feet MSL
or above.
A variety of techniques for estimating plume rise have been proposed
(34)
and used. Briggs critically reviewed many of these models and
26
-------
TABLE 4. EFFLUENT PARAMETERS
Reactor Test
Average Hydrogen
Flow at Full Power(5)
(Ibs/sec)
Power Integral(5)
(106 Mw-sec)
Wind Velocity(36)c
(m/sec)
Mean layer wind speed within the transport layer.
Effective
Plume Height(36)
(meters)
Kiwi-B4D, EP-IV
Kiwi-B4E, EP-V
Kiwi-B4E, EP-VI
NRX-A2, EP-IV
NRX-A3, EP-IV
NRX-A3, EP-V
Phoebus-lA, EP-IV
NRX-A4, EP-IV
NRX-A4, EP-IVA
NRX-A5, EP-III
NRX-A5, EP-IV
Phoebus-IB, EP-IV
NRX-A6, EP-III
Phoebus-2A, EP-IV
Phoebus-2A, EP-V
Pewee-1, EP-III
Kiwi -TNT
69
68
70
75
74
72
69
78
80
72
71
94
72
261
244
42
0
0.1
0.5
0.2
0.3
0.3
0.8
0.7
1.0
1.1
1.2
1.0
2.6
4.5
4.5
2.5
1.5
0.009
10.8
8.2
4.1
6.7
4.1
3.6
4.1
5.1
4.6
2.6
2.1
3.6
12.4
4.1
5.1
4.1
1000
850
2050
1200
900
1350
1500
1500
2250
1350
1800
1650
725
3475
2550
1200
-------
concluded that transitional plume rise could best be described
by the "2/3 law":
h(x) = 1.6 F1/V1x2/3 III-l
Where
h(x) = plume rise or height of the cloud above the terrain
• as a function of downwind distance (feet).
F = flux of buoyancy force from stack, divided by the
mean atmospheric density - 4.3 x 10~ Q..
43
0 = cal/sec and F = ft /sec
rl
U = mean wind speed (ft/sec)
x = distance downwind from the point of release (in feet)
1.6 = empirical constant chosen by fit of data.
The plume rise is not a single value (equation III-l) but rather is a
function of the distance downwind.* Measurements of plume rise should
be associated with the distance from the test point at which the
measurements were made.
In the past, plume rise predictions for NRDS tests have been based on
a combination of plume rise equations (Briggs1 and others), information
from past reactor tests, and atmospheric lapse rate measurements. The
occasional existence of a subsidence inversion at 8,000 to 10,000 feet
MSL and the uncertainties of predicting penetration of the layer by the
plume present additional difficulties in making accurate predictions.
4. Aerosol Effluent Transport
During the course of nuclear rocket testing, radioactive gases and
small particulates that behave aerodynamically essentially like gases
^Equation III-l becomes indeterminate for zero wind speed, and is
thus inappropriate for such conditions.
28
-------
have been released into the atmosphere. Initial concentrations of
these materials are diluted by atmospheric turbulence as they are
transported downwind. For each reactor test, the potential for
downwind doses must be assessed based on the expected exposure to
concentrations of ionizing radiation. Predictions of effluent
release, transport, and human interaction have been presented to
a review panel prior to each reactor test.
Radiation exposure levels downwind can be estimated by knowing the
nature of the contaminating source and the parameters of the at-
mosphere that determine the dilution of the contaminant cloud as
it moves away from the reactor. The initial properties of the
source that need to be determined are: (1) the source strength
and cloud composition, usually expressed in curies of activity;
(2) the duration of the release; (3) the initial geometry of the
source [i.e., whether it is a vertical line source, an elevated
volume source, or an elevated point source] and (4) the available
energy, both kinetic and thermal, for producing plume rise. The
source strength and the duration of the run have been estimated
by the laboratory conducting the test. The initial geometry of
the source is difficult to determine because the cloud is invis-
ible, but an elevated point source has been assumed because of the
nature of the rising plume.
Dilution of atmospheric contaminants is a function of atmospheric
turbulence related in part to seasonal and diurnal effects (e.g.,
solar heating), the influence of large scale weather regimes,
and the ground surface roughness. All of these have superimposed
effects that result in a wide spectrum of turbulent eddies which
are quite random in their formation, individual behavior, and
description. It is the random nature of turbulent eddies that
makes statistical techniques essential in any practical dilution
equation.
29
-------
The dilution equations that have been used at the NRDS are special
forms of the Gaussian distribution formulation adapted by Sutton
and Pasquill and Gifford. ' ' ' Since the first reactor test
at NRDS, the Sutton diffusion equation has been used as the basic
dilution model by ARL to predict air concentrations of gaseous
and fine particulate matter released during full power tests.
Sutton1 s diffusion equation integrated with respect to time for a
continuous elevated point source with anisotropic eddy diffusion
downstream can be written as follows:
xr 1 = 2QR exp -
iir r 2~n r 2 2~n
irUC C x C x
y z z
Where (any compatable set of units)
X(x) = Air concentration (Ci-sec/m3) as a function of distance
downwind
QD = Source strength (Ci)
K
U = Mean wind speed (m/sec)
C = Lateral diffusion coefficient (meters )
C = Vertical diffusion coefficient (meters )
it
x = Downwind distance (meters)
h = Effective stack height (meters) at a selected downwind
distance; e.g., 1 mile.
n = Stability parameter (dimensionless)
This equation yields ground level values of the air concentration
with the x-axis along the centerline. If isotropic eddy diffusion
is assumed then C = C = C, and the equation becomes:
2
X(x) = 2QR exp - h
IIf,2 2-n _2 2-n
irUC x C x
30
-------
The source strength, QR, has been estimated by the laboratory con-
ducting the test. The mean wind speed, U, has been determined from
the predicted wind structure of the atmosphere. C and n have been
assumed to be 0.10 (m ) and 0.25 (dimensionless), respectively, for
normal daytime cases. For normal reactor runs, the effective stack
height, h, has usually been estimated to be between 1,000 and 1,500
meters above the ground depending on the atmospheric temperature
profile (see Table 4).
Equation III-3 is based on several assumptions:
1. The cloud diffuses over a flat homogeneous surface
which acts as a perfect reflector to the effluent.
2. The cloud is formed of gases or particles which
do not fall out.
3. The meteorological situation is steady state, i.e.,
there are no variations in the various parameters in the
diffusion equation in space or time.
4. The diffusion is isotropic.
5. The release can be described as an elevated point source.
None of these assumptions is strictly valid for reactor tests at the
NRDS. The steady state assumption has been perhaps the most serious
disparity as far as NRDS operations are concerned. It has been
observed that the wind speed and direction both change, often quite
drastically, in both space and time. Effluent clouds from reactor
tests quickly mix through several thousand feet of atmosphere and,
through this depth, there is often a considerable amount of wind
shear. Wind shear is not quantitatively taken into account in the
diffusion equation although it may be compensated for in the choice
of eddy diffusion parameters. The effect of shear is to spread the
cloud laterally and thus reduce the centerline concentrations.
31
-------
Evaluations of past results have indicated that the effluent release
cannot be described as a point source at the effective stack height.
f33)
Fultyn proposed describing the process as being equivalent to
release of 1% of the effluent at ground level with the other 99%
released at the effective stack height. ARL used this technique for
a number of years, but more recently has adopted the practice of
reshaping the near-in diffusion curve (out to 5-10 miles) to approx-
imate past results. * Both of these techniques are methods of
taking into account the fact that part of the plume "peels off" as
it ascends; i.e., the entire effluent does not rise to the same
altitude.
Figure 4 presents examples of pre-event dose predictions for the
normal run accident dose estimates and for one of the XE Prime tests.
Meteorological and trajectory information were also presented to the
safety panel prior to tests. Dose prediction models are discussed
in Section III.
Although the Sutton equation has been the basis for the dilution
prediction model used by ARL, a generalized Gaussian equation with
Pasquill/Gifford values for the diffusion parameters has proven
f35 37)
valuable under unusual stability conditions. Gifford ' J gives
values for the diffusion parameters as a function of the distance
downwind for various atmospheric stability conditions.
Numerous attempts have been made to develop diffusion models to
improve effluent prediction capabilities for NRDS operations.
/"ro~i
Several of these models were reviewed by Wilson in 1968.
Wilson concluded that, from the standpoint of availability of
meteorological parameters and ease of calculation, the previously
described Sutton model was preferable. Models reviewed by Wilson
included:
1. Stigall and Galley ' ^ presented an empirical
model based on effluent data for tests between May 1964
32
-------
LU
gioo
:Q
cc
o
~o
i CO
10
5
<
O
x
LU
CC
O
LL
OC
0.01
T.I.P. = 6.O X 10 Mw SECS
PLUS INVENTORY
RELEASE FRACTIONS
tr -0.9% • SUTTON MODEL.
131 I =1.6%
132 I =O.9%
133 I =O.3%
134 I =O.2%
135 1 =0.2%
Cy=Cz=O.316
n=0.2S
h=1OOOm
OJ
O
CO
LU
LU
CO
O
r
co
7 10 20
DISTANCE (MILES)
30
50 70
100
200
BRIEF 0900 PDT
VALID 1100PDT 8/28/69
(ARL-LV)
Figure 4
XE Prime, EP-IXA, Normal Run Centerline
Dose Estimates (from ARL-LV)
33
-------
and 1967. They determined empirical parameters analo-
gous to the Gaussian model.
(7 13 33)
2. Fultynv ' ' developed an empirical model for
predicting downwind concentrations and doses from tests
at NRDS. Results from past tests were normalized to
yield an "average" diffusion curve of relative concen-
tration as a function of distance which could be
scaled to future tests. The dilution factor at a
particular distance was assumed to be a log normally
distributed statistical variable. The model not only
predicts the geometric mean concentration at a particular
point, but also yields an estimate of probable deviations
from the mean. Predictions using this model were pre-
sented in several of LASL Safety Analysis Reports '
although as of 1970, LASL adopted the Sutton approach for
the sake of standardizing prediction techniques within
the NRDS program.
5. Particulate Effluent Transport
Experience to date indicates that large particles have been encountered
only as a result of operating a beaded fuel reactor in a normal-run,
full-power mode for several minutes. Particles were not encountered as
a result of the Kiwi-TNT Test (a planned reactivity insertion experi-
ment of a modified reactor that resulted in violent disruptive dis-
assembly of the reactor) or as a result of the Phoebus-1A incident
(loss-of-coolant due to depletion of primary coolant supply—destructive
overheating of the core with subsequent ejection of 5% of the fuel
material from the nozzle). It has been considered appropriate,
therefore, to estimate the probability of particles interacting with
man only for the full-power normal-run tests and not for the accident
case.
Henderson reviewed several proposed particle prediction schemes
contained in the literature. These include diffusion models such as
34
-------
those used to predict exposures expected from gaseous or aerosol-
type effluent; e.g., Sutton, Pasquill, Gifford^ J as well as the
(41)
scaling model of Cluff and Palmer. Several of these models
have received limited use for NRDS evaluations, but none has been
used as extensively as the Cluff-Palmer scaling equation.
The Cluff-Palmer equation uses data from one experiment to predict
environmental concentrations expected from another event, taking
into account differences in wind parameters and source terms. For
prediction of ground deposition of particles, the appropriate form
of the equation is
r a Q< /h \2/u 2
N'(x') = N(x) [^ -
and
u1 h'
x' = x — jj- III-4(b)
where
N(x) = number of particles per unit ground surface area as a
function of the distance x, downwind;
x = the distance downwind;
a = the total wind direction shear in the fallout hodograph;
Q = the amount of the reactor core released as particulate,
generally stated in terms of mass of uranium;
h = the height of the plume top above the ground; and
— = the mean wind speed of the atmospheric layer containing
the plume.
In each case, the primed symbol refers to a variable for the event for
which deposition is to be predicted, and the unprimed symbols refer to
data from past experiments. Techniques for the use of the Cluff-Palmer
(42)
equation are more fully described by Hughes, although Hughes' dis-
cussion is related to estimates of external gamma exposure expected
35
-------
C131
from nuclear explosive cratering applications. Fultyn observed that
the predictions of the Cluff-Palmer model depend strongly on the particu-
lar reference event used for scaling. The meteorological data required
to use this model for prediction are routinely recorded by ARL. Some
input parameters required by other models are not well-documented, such
as atmospheric turbulence parameters, initial physical dimensions of the
plume, and the distribution of particles along the vertical extent of the
cloud or plume..
The particle concentration N(x) used in the Cluff-Palmer model to predict
N'(x') may be taken from a set of results from a single test, from the
mean of all tests, or some other composite of past experience.
Figure 5 presents the pre-event prediction for one of the XE Prime tests as
an example of results from this technique.
(431
Altomare and Coleman developed another model for prediction of par-
ticle deposition on the basis of Van de Hoven's tilted plume formulation.
This formulation is based on the assumption that particles diffuse cross-
wind and vertically according to Button's equation and simultaneously
settle with a constant fall velocity. For the case of an elevated plume
such as the reactor effluent, the effect is to tilt the plume downward.
This can be expressed by replacing the constant height of the plume cen-
terline in the Sutton equation with a variable expression. Altomare and
(43")
Coleman adapted the modified Van der Hoven formulation for a computer
model, DIFOUT, to predict particle deposition for reactor plumes. This
technique approximates the distributed volume of the plume as a series of
stacked cylinders, each containing several size ranges of particles with
different fall velocities. The program accounts for the different volumes
and size ranges to predict diffusion and deposition downwind. This model
received some use in evaluating the Phoebus-IB EP-IV data and providing
preliminary predictions for the NRX-A6 tests. Because of difficulties
in adapting it to computers used by NRDS and a lack of the necessary
meteorological parameters, the model has not received extensive use.
36
-------
TO3
10
1.0
CO
111
_J
o
(-
IT
0-10*'
10"
10'
10
20
30
50
_J L_
70 100
200
DISTANCE (MILES)
BRIEF 0900 PDT
VALID 1100 PDT 8/28/69
(ARL-LV)
Figure 5
XE Prime, EP-IXA, Particle Prediction
(from ARL-LV)
37
-------
B. Effluent Distribution
This section briefly indicates the geographic distribution of the efflu-
ent. The effluent program and environmental concentrations are described
in Section IV.
NRDS effluent releases have taken place over a period of minutes to hours.
The actual power tests, fifteen minutes to one hour in duration, are re-
sponsible for the most important effluent releases, but reactor cooldown
has also produced effluent releases. The Filter Reactor Outlet Gas (FROG)
was installed on up-firing reactors subsequent to the Phoebus-IB to elim-
inate cooldown releases, though it was not used for the down-firing XE Prime
test.
Effluent from a given test can be distributed in several directions, as
illustrated in Figure 6. Several factors contribute to such widespread
distributions. Changes in wind direction, in addition to differences
due to wind shear with height, can occur during the period of effluent
release. The plume rise associated with reactor cooldown is usually less
than that of effluent from power tests, and the wind velocities at these
altitudes may be different. Also, as discussed earlier, a portion of the
power test plume "peels off" as the plume ascends, and wind shear can act
on this portion. ' The samples containing effluent, Figure 6, north
of the test site were primarily a result of effluent from the power test;
whereas those to the south were primarily a result of effluent from
reactor cooldown which was transported by the low-level valley flow
(drainage) winds during the evening following the test. ^45-*
The prevailing winds at NRDS cause most of the effluent trajectories to
be either to the northeast or southwest. At distances of a hundred to
several hundred miles, these trajectories tend to veer easterly or westerly,
respectively.
Particulate effluent (relative values of particles per square meter) has
(141
been detected out to about 80 miles, J ground level aerosol effluent to
38
-------
NEVADA
WELLS
ELKO
POSITIVE MILK SAMPLES
ALSO FOUND AT
gBUHL, IDAHO
O BOISE, IDAHO
®BLACKFOOT, IDAHO
CCOEUR d'ALENE, IDAHO
C BILLINGS, MONTANA
*
e*
•EUREKA
0
Q*
ROUND MTN.
*LUND
'CURRANT
MANHATTAN® B B|_UE JAY
WARM SPRINGS^ • NYALA
TONOPAH™ STA" " 'I"™'* SPR1NGS
DIABLO
\
GOLDFIELD" I NELLIS
| AIR FORCE
\ r-^^ RANGE
BEATTY
\
.
* ALAMO
UTAH
O GARRISON
RICHFIELD
CEDAR CITY
* TEST CELL "C" WARM SPRINGS RANCH
WELLS
^ DEATH VALLEY JCT.
SHOSHONE
VEGETATION
MILK Q
AIR •
Figure 6
Locations Where Fresh Fission Products
were Detected (Phoebus-IB, EPIV)
39
-------
(45 46 47)
hundreds of miles (doses of mrads or less normally.) * * J The effluent
cloud has been tracked to several thousand miles by aircraft utilizing sen-
sitive sampling and monitoring instruments. The effluent transport altitude
has ranged from the ground to better than 3 km above the ground.
IV. DOSE MODELS
Several models have been used to predict doses prior to reactor tests and to
calculate postulated doses from surveillance results after tests. The models
have included transport models for estimation of environmental concentra-
tions, and factors and techniques for converting these concentrations to
potential doses to man.
Figure 7 shows the major steps involved in each of the models. Since the
arid environment of the NRDS includes very few bodies of water, the models
are limited to atmospheric transport phenomena. Models for estimation or
prediction of internal doses to man have been primarily concerned with ra-
dioiodine isotopes >»»»•' anc[ to some extent with the inhalation of
i j- _- , c -A f «. * * (15,39,45,47,48,49,50,51,52)
large discrete particles. Evidence from past tests, »»»»»»>»•'
fission production ratios, release rates, ' factors of environmental
. (35, 39,55) , , . , . , . , , , _ . (56,57,58,59)
transport, and biological uptake and dose factors
have led to the conclusion that the radioiodines have been the most im-
portant components of aerosol effluents from reactor tests. For fission
product releases of the nature of past NRDS tests potential bone doses
from strontium-89 and -90, as well as doses to other organs from other
radionuclides, are lower than potential thyroid and/or external gamma
doses. Techniques used to control the potential thyroid and external gam-
ma dose also tend to control other potential doses.
The Aerosol or Gas-Cloud Model consists of steps, i, ii, iv, v, vi, vii, viii,
ix, and x in Figure 7. The particulate model includes steps i, ii, iii, ix,
and x. These two models will be treated separately; the gas-cloud model
in Section A and the particulate model in Section B.
Steps i and ii in Figure 7, the estimation of source terms and transport,
were discussed in Sections II and III. These steps yield environmental
40
-------
•fl
H-
OQ
c
i-S
CD
C
SOURCE
(
PARTICULATE MODEL
DEPOSITION OF \
DISCRETE PARTICLES )'
""" -H^TMOSPHERIC TRANSPORT^
/AEROSOL OR\
—V \GAS CLOUD MODEUJ
/
/
\
iv T
(INHALATION OF
EFFLUENT BY MA
>
N^
VII
/GAMMA EXPOSURE^
V FROM CLOUD J
INTERACTION WITH MAN
v T
^DEPOSITION ON^N
V COW'S FEED J
)
1
VI
f
C
POSTULATED DOSE
TO MAN
VIII
RADIOIODINE
IN COW'S
INHALATION
BY COW
Dose Models
-------
concentrations in Ci-sec/m3 for the aerosol and particles/m2 for the par-
ticulate effluent. The only difference between the normal run and acci-
dent predictions is in the source term.
A. Aerosol or Gas-Cloud Model
(35 39 521
The basic model that has been used by ARL ' ' ' considers three
pathways of exposure:
- External gamma exposure from the radioactivity in the
cloud during cloud passage (steps i, ii, vii, ix, and
x in Figure 7),
- Thyroid dose via inhalation of radioiodine isotopes
(steps ij ii, iv, ix, and x); and
- Thyroid dose via ingestion of radioiodine in cow's
milk (i, ii, v, vi, viii, ix, and x).
Although this model does not consider external gamma exposure from
fallout or ingestion of radioiodine via the vegetation pathway, these
sources of exposure have been considered by agencies within the test-
ing program. ' Off-site surveillance results have indicated
limited detectable external gamma exposure off-site (measurement
made 3 feet above the ground) from deposition of the reactor efflu-
ent. (Ref. 40,46-49,61,62,63) Calculations by LASL*-7-1 have indicated
that the infinite exposure from residual fallout may have been similar
to that from cloud passage.
Thompson noted that the I intake from non-milk food items
may be significant compared to that from milk. For adults this
is true, but for infants (the critical receptor for I via in-
gestion), the intake path by milk is controlling (about 90% or
more of total). Therefore, only the milk-food chain has been con-
sidered.
42
-------
1. External Gamma Exposure
Steps i, ii, vii, ix, and x are included in the external gamma
exposure model. The parameters for these steps are determined
as follows:
i. The source term has been estimated to be the
product of the gross fission product activity
in the reactor and the gross fission product
release fraction. Either the reactor inven-
tory at the appropriate time after fission or
the initial inventory decayed by the Way-
Wigner relationship has been used. The fission
product yield for instantaneous fission may
also be used (100 Ci/Mw-sec at H+l hour). This
simplification produces a conservative esti-
mate. Decay time has been assumed to be equal
to transport time, (x/U), so the Way-Wigner
decay factor may be presented as:
x/U
1 hr
-1.2
ii. The diffusion coefficient, X/Q, may be cal-
culated according to one of the variations
of Sutton's equation, as outlined in Sec-
tion II.
The product of steps i and ii yields
(release fraction) (Power integral,
Mw-sec) (100 Ci/Mw-sec)
x/U(hr)
1 hr
-1.2
(X sec\
Um3;
vii. An infinite semispherical cloud model has been used to
predict external gamma dose.(ref.35,51). The model
includes the following assumptions:
a. The effluent cloud has been assumed to have dimen-
sions large enough and concentrations uniform enough to
approximate the dose to an individual on the surface
from a cloud of uniform concentration and infinite
dimensions, (dimensions large compared to the range of
gamma photons). Several references outline techniques
which might be used to consider finite clouds and/or
the dose from a cloud passing aloft.(ref.37,65,66).
43
-------
b. Since the receptor is at ground level, the
dose has been assumed to be half that from a spher-
ical infinite cloud.(ref.37,59).
c. It has been assumed that the individual does
not perturb the gamma flux, so that the dose to the
individual is the same as that to a similar volume
of air except for differences in electron densities
between tissue and air.(ref.37,59).
Imai and lijima note that the maximum external gamma
dose does not occur at the point of maximum ground level
air concentration. The distance at which the maximum
ground-level gamma dose occurs is related to the effective
stack height or plume rise, as is the distance of maximum
air concentration. However, the maximum gamma dose occurs
at a point nearer the release point than the point of max-
imum air concentration because of the contribution from
the cloud aloft. Thus, although the homogeneous semi-
spherical cloud model is generally conservative, it tends
to underestimate doses at distances less than that of the
maximum air concentration.
The ARL model has compensated for this by shaping the dose
curve, to match near-in results from previous tests. ' '
Step vii yields the following conversion factor, when X is
the integrated air concentration of gross fission products:
3.7 x 1010 d/sec 0.7 MeV* ion pair
X(Ci-sec/m3) x x x
Ci d 34 x 10"6MeV
esu m3 97 ergs in tissue 1R - cm3
& x
x
2.08 x 109 ion pairs 106 cm3 R-g
rad-g 0.18 rad
x x 0.5 (for the semispherical cloud) =
esu
100 erg Ci-sec/m3
*An average photon disintegration energy for fresh fission products
of 0.7 MeV (ref.67,68) has been used. (ref.35,51).
44
-------
It is emphasized that the conversion factor that has been
used in the ARL-LV model, 0.18 rad per Ci-sec/m3, is
appropriate only for predicting ground-level gamma doses
from a fission-product cloud for which the average gamma '
photon energy is 0.7 MeV. It is not possible to use this
factor to estimate actual gamma exposures from measured
air concentrations. Several factors contribute to the
difficulty in relating measured air concentrations to
measured gamma exposures. A significant part of the gamma
exposure may be caused by shine from the cloud aloft and is
therefore unrelated to ground-level air concentrations. '
Clouds are generally non-homogeneous; concentration gradi-
ents are likely to exist within the cloud. Measurement of
air concentrations of noble gases and short-lived gross
fission products is difficult because air samplers that
have been used in the NRDS surveillance program have not
collected the noble gases and because many of the short-
lived nuclides have decayed prior to the sample analysis
(the NERC-LV aircraft program includes analysis of noble
gases).
2. Other Approaches to External Gamma Exposure Predictions
Fultyn developed a probabilistic model for estimating not only
the mean external gamma dose as a function of distance but also
the statistical deviation from the mean. Fultyn1s model assumed
that the gamma dose at a given distance was a lognormally dis-
tributed statistical variable. Fultyn described atmospheric
diffusion with an empirical curve of normalized concentrations
versus distance by normalizing I air concentrations observed
during ten reactor runs (1964-1966). A lognormal distribution
was fit to the data for each distance and a curve relating con-
centration to distance was then fit to the means of these distri-
butions.
45
-------
Fultyn used a conversion factor (integrated gross beta air con-
centration to external gamma dose) of 2 rads per Ci-sec/m3, roughly a
factor of ten greater than the ARL conversion factor. '
This empirical number related air concentrations to measured
gamma exposures observed for the NRX-A4, NRX-A5, and Phoebus-IB
reactor tests. Since the factor was based on measured air con-
centrations, the difference between it and the ARL factor may
be explained in part by gamma exposure from shine and noble
gases (not collected by air samplers) and by the non-homogeneous
nature of the clouds. Although this model has been used in LASL
reactor safety evaluation reports, ' its use was discontin-
ued in favor of adopting the ARL model for uniformity.
WANL has used a model similar to that of ARL-LV in its safety
( o->
evaluation reports.
Van Vleck proposed a gamma exposure conversion factor of 1.4 R
per Ci-sec/m3 for use with gross beta air sampling results.
This conversion factor requires the same qualifications as those
discussed previously for the factor by Fultyn.
From the Fultyn and Van Vleck factors, it can be seen that a
value greater than the theoretical value of 0.18 rad per Ci-sec/m3
is necessary for conversion of actual measured air concentrations
to gamma exposures. The NUS analysis of NRX-A4 data also bears
this out. (69,P. 20)
3. Evaluation
Stigall and Galley '"* ' noted in their analysis of effluent
data for 1964-1966 that 84% of the gamma exposure estimates
calculated from measured air concentrations were below the
corresponding gamma predictions (using post-run source term and
run-time meteorology). Wilson's analysis and the Fultyn '
and Van Vleck conversion factors indicate that the discrep-
ancy is caused not by any inherent conservatism of the model but
46
-------
more by the difficulty of converting air concentrations to gamma
(39)
exposures, Wilson ' compared model predictions with measured
gamma exposures and found a slight tendency to predict low
(actually close to 50%-50%).
4. Thyroid Dose via Inhalation Pathway
The inhalation pathway consists of steps i, ii, iv, ix, and x of
Figure 7. The following discussion outlines the calculation for
these steps.(Ref.35,39,51,52).
i. The predicted source term for a specific nuclide is
the product of reactor inventory , the release frac-
tion (discussed in Section II) and exponential decay re-
lations based on the radiological half-lives of the
specific radionuclide and its precursors. Table 5 indi-
cates the maximum activity of the various radioiodine
isotopes per Mw-sec for a reactor operating time of
1800 seconds. The maximum for several of the isotopes
does not occur until several hours after shutdown because
they build up as their precursors decay. Table 5 also
indicates the approximate time of occurrence for each
maximum. The ARL model has assumed that this maximum
activity is present at all times prior to its occurrence.
Thus, it also assumes that the radioiodine isotopes and
their precursors behave similarly during transport and
biological uptake.*
The inventory of radioiodine produced for a given run
(plus the remainder from previous tests) has been esti-
(53)
mated from Kochendorfer. Decay during transport
may be accounted for by use of Reference 58 or appro-
priate radioactive decay relationships.
*The treatment of _2Te-I has been an exception. Although
the half-life of Te was used during transport, the
implications of inhaling it were not considered by the
ARL model.
47
-------
TABLE 5. PARAMETERS FOR RAD10IODINE(53)
Iodine
Isotope
131
132
133
134
135
pa
Ci/Mw-sec
0.0200
0.115
0.448
6.05
1.29
Time When Max. Occurs
H + seconds*5
63,000
0
10,000
1,600
160
Radiological Decay
Half-Life
8.05 days
3.24 daysC
20 . 3 hours
52 minutes
6.68 hours
Based on an operating time of 1800 seconds.
H is the time of the end of reactor run.
Assuming 132Te is also released, the rate at which the quantity of
132I present is reduced is dependent on the 132Te half-life. The 1
half-life is 2.26 hours.
48
-------
ii. The atmospheric diffusion factor (X/Q) has been
calculated from one of the variations of Button's
equation given in Section II.
iv. The breathing rate recommended for the Inter-
national Commission on Radiological Protection(ICRP)
for standard man has been used— 2.32 x 10 m3/sec.
(Ref.35,39,51,70).
The intake (for each isotope) equals
Ci = X(Ci-sec/m3) x 2.32 x 10"4 m3/sec
ix. The interaction in man may be estimated by the
following assumptions and considerations:
a. 75% of inhaled radioiodine is retained
(59)
in the body and goes to the blood stream ,
and
b. 30% of the radioiodine in the blood stream
(59)
is transported to the thyroid.
Thus, 0.75 x 0.30 = 0.23 of the inhaled radioiodine
is assumed to be deposited in the thyroid.
x. The following indicates the dose to man based on
a given intake, assuming:
a. Radioiodine is uniformly dispersed in the
thyroid. (71-)*
b. The energy absorbed in a unit volume of the
thyroid is equal to the energy emitted by the
(71)
quantity of radionuclide in the unit volume.
c. The biological effective half-life and
effective decay energies (H ) from ICRP 2 are
(59)
given in Table 6.
"Hine and Brownell(71) indicate that iodine is not immediately taken up into the
thyroid, as is assumed by the model that has been used. An uptake half-time of
5 hours is indicated. The ARL model is conservative by up to a factor of two
for the shorter half-life radioiodine isotopes such as 13^I(Ref.71,p.837).
49
-------
TABLE 6. BIOLOGICAL PARAMETERS FOR RADIOIODINE ISOTOPES
Iodine
Isotope
131
132
133
134
135
Biological Effective
Half -Life (59)
Days (Seconds)
.7
0
0
0
0
.6
.097
.87
.036
.28
6.
8.
7.
3.
2.
6 x
4 x
5 x
1 x
4 x
10
10
10
10
10
5
3
4
3
4
Effective Energy
(20-g Thyroid) (59)
(MeV)
0
0
0
0
0
.23
.65
.54
.82
.52
Rad/Ci-sec/m
(20-g Thyroid)3
3
1
9
5
2
.4 x
.2 x
.2 x
.8
.9 x
io2
iolb
io1
io1
The inhalation dose for an infant with a two-gram thyroid has been assumed
to be 2.5 times that of a standard man.
This dose conversion constant is based on the biological half-life of 132I
versus the chain controlling half-life of 132Te. Assuming 132Te is inhaled
and deposited in the body according to the ICRP 2 assumptions(59), the dose
from inhaled 132I and 132I produced from decay of 132Te in the body indi-
cates a dose factor of 51 rad/Ci-sec/m3.
50
-------
d. The thyroid weight (m) for standard man is
20 grams.
The following expression gives the dose rate to *
the thyroid from a thyroid burden:
Ci in d (MeV) fi (er2s)
x 3.7 x 10 x E,, x 1.6 x 10
m(g) Ci-sec (d) (MeV)
rad-g
x (rads/sec)
100(ergs)
To obtain the integral dose, the dose rate is integrated
from the time of intake to infinity. Assuming the re-
moval of iodine can be represented by a single exponen-
tial, indicative of a biological half-life, the removal
of radioiodine can be represented by a biological
effective half-life (T __) which is based on both the
biological and radiological half-lives.
The approach to dose calculation parameters given in
Table 6 has been to calculate the inhalation dose for
an adult. Rather than actually using the parameters
for an infant, it is then assumed that the infant's
dose is 2.5 times that for an adult.
The calculation procedures used by the other organizations
are similar to that indicated above. Stocum (58) evalu-
ated the biological parameters and related them to statis-
tical distributions. The analysis is largely related
to the parameters associated with ingestion of radio-
iodine via the milk-food chain. In recent safety
evaluation reports, LASL has applied the geometric
standard deviation from this analysis to a mean based
on ICRP(59) parameters. C3
51
-------
5. Thyroid Dose via Ingestion--Milk-Food Chain
Steps i, ii, v, vi, viii, ix and x of Figure 7 apply to the dose via
ingestion. The calculations for steps i, ii, and x (transport and
dose parameters) are the same as presented for the inhalation dose
in the previous section although the only isotope that has been used
in predictions for ingestion dose is I. Iodinc-131 accounts for
the majority of the milk ingestion dose because of its longer radio-
logical half-life (with respect to other referenced radioiodine
isqtopes). The NERC-LV considers the presence of the other radioio-
dine isotopes in sample analysis, and when estimating the possible
, c , . , . ... (46,47,50)
dose from drinking milk. '
The following indicates the calculations for the steps pertinent
to this dose pathway not covered in the previous section:
v, vi, and viii. When cows are on fresh forage susceptible to
contamination by the effluent, the amount of radioiodine inhaled
by the cow (vi) is insignificant. Thus, for practical purposes
(vi) is considered to be included within (v). The conventional
technique for determining cow forage contamination from passage
of an effluent cloud is to use the product of the deposition
velocity (m/sec) and the reciprocal of the mass of feed growing
per unit area of the field (m2/g), i.e., m/sec x m2/g = m3/g-sec).
At the time the NRDS model was developed, insufficient data were
available to evaluate each variable. Thus, a single parameter
was used in the ARL model to relate the integrated airborne con-
centration of radioiodine to the resulting concentration in milk.
-4 3
The value now used in the model, 4 x 10 pCi/1 per pCi-sec/m ,
was based on a literature survey. The value is only intended for
I concentrations because of the short half-lives of the other
iodine radioisotopes.
(72 73")
Studies at NRTSV ' J noted that about half of the variability
in deposition velocity could be explained by the variation in
the density of vegetation on the ground. In accordance with
this, surveillance results from NRDS tests were fit to a log-
normal distribution based on the ratio of radioiodine concen-
(59)
trations in milk and vegetation to integrated air concentrations.
This study indicated a geometric mean of about 4 x 10 pCi/1 per
52
-------
pCi-sec/m and a geometric standard deviation of 2, which indi-
cates the appropriateness of the original value used.
ix. The deposition of effluent from NRDS tests has been an
\
acute or semi-instantaneous process. The concentration de-
crease subsequent to the peak for I can be approximated by an
exponential decay with an effective half-life in milk of five
days (Ref.56, Report 5). The integral concentration subsequent
to the peak has thus been estimated to be 7.2 times the peak;
'i.e., / (exp - Cln 2/5 daysDt)dt. The Federal Radiation Council
(FRC) Report 5k J estimates that 20-25% of the total intake of
I via milk, subsequent to acute deposition, may be due to in-
take prior to the occurrence of the peak. Thus, the total intake
has been assumed to be equal to about ten times the peak.
X. A convenient dose conversion parameter based on step ix
(Ci intake/peak cone, in milk) and the previously mentioned
dose conversion factors (rads/Ci intake) is 16 mrad thyroid dose
as a result of a peak milk concentration of 100 pCi/1. ' This
is based on an infant with a 2-gram thyroid consuming one liter
of milk per day.
The other organizations which have made effluent predictions have used
r CQ~\
similar techniques for estimating the ingestion dose. Stocum ana-
lyzed the biological parameters associated with ingestion. Based on
(74)
this analysis and a more recent evaluation, he suggested a dose
conversion factor of 12 mrad/100 pCi/1. Although this factor includes
the dose contribution from consumption prior to the peak, it is iden-
tical to the FRC Report 5^ ' value for the intake after the peak. The
difference stems from an assumption of a lower value for the combi-
(74)
nation of milk intake and thyroid uptake parameters.
(81
The doses estimated by the WANL model have been about 30-50% less
than those of the ARL model because the WANL model has not considered
the potential J I ingestion prior to the occurrence of the peak and
used a 4.8 versus 5-day effective half-life for I in milk.
B. Particle Dose Prediction Models
Models for prediction of biological hazards from particulate debris involve
steps i, ii, iii, ix, and x of Figure 7. Methods for prediction of the
53
-------
source term, step i, are discussed in Section II. For particle pre-
dictions, the source term may be expressed as a quantity of uranium,
in grams, expected to be lost from the core. Steps ii and iii, dis-
cussed in Section II, yield predictions of expected ground concentra-
tions in terms of particles per unit area (particles/m2). Predictions
at NRDS have generally used the Cluff-Palmer equation to scale ground
concentrations of particles measured on previous tests to predict
expected deposition from a future test, taking into account differences
in source terms and meteorological parameters. Steps ix and x consist
of evaluating various interactions of these particles with man and
estimating doses and/or effects from these interactions. Step x for
the particle model differs from that for the aerosol models in that
there are no official radiation protection standards or guides for
evaluating the implications of particle interactions; whereas, there
are for the aerosol effluent.
Interaction of the particles with man is treated statistically by the
model. Although there exists a possibility that a particle will fall
at any downwind distance out to several tens of miles, the probability
that a particle will fall at any particular location is small. The
model assumes that this probability can be predicted from the Poisson
distribution.
Modes of biological interaction of the particles with man that have
been considered include localized skin doses, eye cataracts (from
deposition of a particle in the eye), doses to the lung from inhalation
of particles, and doses to the GI tract from ingested particles. These
modes have been evaluated by several individuals and committees, in-
cluding the AEC Division of Biology and Medicine, Space Nuclear Safety
Committee, and it has been concluded that the localized skin dose has
been the limiting or critical mode of interaction. ' '
The hazards of inhalation of particles have been discounted largely be-
cause there is evidence that nuclear rocket engine particles small
enough to have a reasonable probability of being deposited in the deep
pulmonary region of the lung have not contained sufficient activity
• .... u- i - •, « * (76,77,78)
to produce significant biological effects.
54
-------
Most of the information on biological effects produced by highly
radioactive discrete particles in the lung is based on experiments
with plutonium particles. The doses from these particles are much
larger than doses which would result from inhalation of reactor
particulate debris, and biological effects have been observed only
at levels higher than expected from reactor debris. The two situations
are not totally analogous; however, this analysis cannot be regarded
as conclusive. Experiments with plutonium have emphasized the alpha
dose, in which there is a sharp demarcation between tissue receiving
a lethal dose and tissue receiving a sub-lethal but potentially
carcinogenic dose. Beta and gamma emission from reactor debris produces
a much more continuous depth dose distribution. Langham^ noted that
most investigations have given little attention to tissue outside the
range of the alpha dose which may however receive a significant dose
from x-rays. Biological effects in this tissue might be expected to
be similar to effects in tissue surrounding particles containing
fission products.
Another concept which deserves more investigation involves consideration
of the number of cells at risk and the quantity of tissue that must be
exposed to a sub-lethal but potentially carcinogenic dose to cause a
f79)
cancer. Dean and Langham in presenting a thought-provoking con-
cept rather than a final model, suggested that this problem be
approached by integrating the probability of cancer occurring for each
exposed cell to determine the total probability of biological effects.
The proper understanding of the relationship between the mass of tissue
exposed to a given dose (i.e., gram-rads) and the resulting biological
effect is also relevant to analysis of skin effects. But there are
differences between understanding the particle skin problem and under-
standing or evaluating the inhalation problem. First, micro-effects
in the skin can be observed; whereas those in the lung are difficult
to detect in-vivo. Past experience indicates that skin doses which
do not produce ulcers will not produce skin cancer. Second, skin
55
-------
f Rfl S1 8?1
ulcers or an indication of their formation1 ' ' J can be detected
soon after exposure (within about two weeks) and can therefore be
related directly with the exposure, while the time involved in lung
effects may be longer. Third, skin cancer is adaptable to successful
therapeutic treatment, whereas lung cancer may not be.
As an example to illustrate the plausibility of the argument that
inhalation hazards can be discounted, the activity in a particle of
respirable size can be compared to the activity required in the same
size particle to produce a dose of 750 rads at Krebs' depth.* Since
this requires the assumption that the dose-effect relationships for
the skin and the lung are similar, this exercise must be regarded
as an example only and not as conclusive evidence.
Henderson reported a value of 1.7 x 1018 fissions/gram of
uranium for the Pewee I, EP-III, which is higher than that reported
for other full-power runs. Data reported by the ICRP Committee II
f 831
Task Group on Lung Dynamics indicate that particles with aero-
dynamic diameters greater than 10 micrometers are not likely to be
deposited in the deep pulmonary region of the lung. The NERC-LV
reported an average particle density of 11 grams/cc from Phoebus IB
(14)
EP-IV , larger than average particle densities reported by PAA
and the NERC-LV for other full-power runs. ' ' ' Assuming a
particle with an aerodynamic diameter of 10 micrometers, a density
of 11 g/cc, and a specific activity of 1.7 x 1018 fissions/gram of
uranium, there would be 3 x 108 fissions per particle. The associated
radioactivity is roughly two orders of magnitude less than that re-
quired to produce 750 rads at Krebs1 depth for reasonable particle
travel times. This is based on a mean retention time of much
greater than 10 hours.
*Krebs' Dose is the dose at the periphery of a circular field of ,_...
4 mm in radius at a depth of 100 ym below the surface of the skin. '
This is discussed in Section V B.
56
-------
As discussed below, the figure of 750 rads at Krebs1 depth is a
skin dose particle guide used at NRDS and does not necessarily
apply to biological effects in the lung, especially since the
geometry involved in particle doses to the lung is unknown. How-
ever, the example serves to illustrate the plausibility of con-
clusions by some workers that the skin dose is a more important
mode of interaction than the lung dose. '•>»-'
r Q(T\
Sanders, ef.al., summarized the implications of respiratory
system irradiation by particles, in particular, UC~ particles
with a fission product inventory. They concluded that although
nonuniform irradiation of the lung is more carcinogenic than uni-
form irradiation, there is no justification for inferring numbers
of tumors incidence for doses less than 3000 rads, to the epithe-
lium 100 urn from the particle. Doses of this magnitude are only
related to particle sizes of about 10 ym or more (30 ym aerody-
namic diameter") and travel times of less than one hour. But,
Sanders, et.al., concluded that the data necessary for a compre-
hensive evaluation of inhalation probability and radiation effect
are not available.
The remainder of this section will be devoted to the skin dose
model. Discussion of models for the other modes of interaction
(lung, eye, and GI tract) is presented in Appendix B.
1. Particle Skin Dose Model
For this model, steps i, ii, and lii of Figure 7 consist of
predicting the source term in grams of uranium to be released
from the core as particles and applying the Cluff-Palmer
equation to predict ground deposition of particles in terms
of particles/m2 at points downwind, as discussed in Section II.
Step ix concerns the interaction of particles with man. This
interaction is evaluated by computing the probability that one
or more particles will impact on an individual in the effluent
trajectory and will produce a guideline skin dose. The NRDS
57
-------
guideline dose is 750 rads at Krebs1 depth/ as discussed
in Section V.
The model assumes that the probability P (n,X) of a certain
number of particles (n) impacting on an average individual
and inflicting the guideline dose can be described by the
Poisson distribution:
e"A An
P (n,X) = * L- IV-1
" n!
where X is the average number of radioactive particles per
individual. If n = 0, P is the probability that no
particles will interact with the individual, Eq. IV-1 becomes
P (0,A) = e~A IV-2
The probability P of at least one particle interacting with
an individual to produce the guideline dose is
P = 1 - P (0,A) = 1 - e~A IV-3
Here X is defined as the average number of particles impacting
on and retained by an individual with sufficient activity to
produce the guideline dose. A useful approximation for small
_2
values of X (X < 10 ) is
P % X IV-4
Only a fraction of the particles will contain a sufficient
inventory of fission products subsequent to a given transport
time to produce the guideline dose, even if they impact on
and are retained by the exposed skin of the individual. The
inventory which a particle must contain to produce the guide-
line dose is a function of particle size and transport time;
thus, both size and particle radioactivity inventory distri-
butions must be considered in computing X.
58
-------
The NRDS model divides the size distribution into i classes
and computes A from 0., the total number of particles in the
i_th size class which impact on and are retained by an indi-
vidual, and c^, the fraction of the total number of particles
in the rth size class with sufficient inventory to produce
the guideline dose:
X = E 0. a. IV-5
i
In practice, 0. and a. are estimated empirically from size
and radioactivity inventory distributions observed from
past tests, appropriately extrapolated to the proposed test
conditions. For reasons of convenience and the accuracy
of data, the size distribution is generally divided into
three classes: less than 15 micrometers, 15 to 50 ym, with
representative mean diameters for these classes of 10, 30,
and 70 ym, respectively.
The average number of particles, 0. (x), which impact on and
are retained by an individual at any downwind distance x is
the product of the predicted ground concentration of particles
in terms of particles/m2 in the rth size class (developed in
steps i, ii, and iii of Figure 7 and the exposed surface area
of the individual in m2. This assumes that the number of
particles which impact on and are retained by the exposed skin
of an individual is the same as the deposition on an equivalent
horizontal surface area. This is based on studies by
( 871
Booz-Allen which indicated that approximately the same
amount of mass was deposited on the exposed surface area of a
mannikin as was collected on an equivalent horizontal area.
These results imply that the fraction of particles which impact
but do not stick and the fraction which follows air streamlines
around an object compensate for the differences between particle
fluxes in the vertical and horizontal directions. This ap-
proach has the advantage of simplicity and bypasses more
59
-------
elaborate theoretical approaches proposed by Kochendorfer and
and (
(90)
Ulbergt88), DeAgazio(89), Mikhail and Collins <-76'), and the
Radiological Effects Working Group.
The total exposed skin area is used rather than the cross-
( 88)
sectional area as suggested by Kochendorfer and Ulberg
(91)
Henderson assumed winter dress styles similar to Model A
of Kochendorfer and Ulberg: long pants, long-sleeved shirt,
and hat, with a corresponding exposed skin area of 0.0702 m2
and summer dress styles similar to Model B of Kochendorfer
and Ulberg: long pants, short-sleeved shirt, no hat, with a
corresponding exposed skin area of 0.384 m2.
The fraction of particles,a.(x), at a given downwind distance,
x,with sufficient inventory to produce a guideline dose
depends on: (1) particle size distribution, (2) particle
inventory distribution, (3) reactor run history, and (4)
travel time from release of particles from the reactor to
impaction on the individual. The travel time has been
estimated by x/U, where U is the mean transport wind speed.
The number of fissions required in a particle to produce the
guideline dose may be determined from the beta tissue dose
model of Mikhail and Collins. This model is based on
(92)
the work of Ulberg and Kochendorfer.
The dose/fission calculations have been based on the reactor
power and run time profile and include self-absorption within
the particle. It has been assumed that all particles are
retained in the reactor until the end of the run.
The dose is obtained by integrating the dose-rate with time.
This integral has been based on arrival time and a mean
retention time. The mean retention time is estimated from
an empirical equation fit to experimental data ' ' which
yields the probability of P (t) of a particle being retained
60
-------
a period of time t:
Pr(t) = e-t/6 IV-6
where t is in hours and 6 is the mean retention time (hours)
for a particle of given diameter on the skin. The value of 6
for a given size class can be determined from references. ' '
The particle activity distribution has been predicted by scal-
ing a particle activity distribution observed on a previous test
in accordance with the ratio of the relative radioactive in-
ventories of the reactor cores for the two tests:
A' = A(F'/F) IV-7
where A is indicative of the radioactivity inventory in the
particle, in terms of fissions per particle, F is the number
of fissions/gram of uranium. The primed values are for the
proposed test.^
Pre-run safety evaluations of reactor tests at NRDS have involved
the use of Equations IV-3 and IV-6 to develop curves of the pro-
bability of an individual receiving 750 rads at Krebs' depth as
a function of distance.
V. RADIATION PROTECTION GUIDES
This section is subdivided into guides and standards for total-body, inhalation,
and ingestion doses and guides applicable to evaluating doses from discrete-
particles. Guides are recommendations of a committee, organization, or
individual, whereas standards have the impact of law (i.e., the AEC regula-
tions). Tests at NRDS have been conducted on a contractor basis with the AEC,
and thus have been conducted under the standards of AEC Manual Chapter 0524
(MC 0524) versus Title 10, Code of Federal Regulations, Part 20 (10 CFR 20 is
(57 93 94 95)
applicable to the licensee relationship). ' ' '
61
-------
A. Guides for Total Body and Internal Organ Doses
1. Normal Operation
The standards for normal operations have been those of MC 0524,
which specifies standards for both occupationally exposed people and
the general population. These standards generally follow the recom-
mendations of the ICRP/59-1 FRC, ^ and the National Council on Ra-
f97 98")
diation Protection and Measurements (NCRP). ' } The standards
given both as organ doses and as concentration guides for air and
water. Table 7 gives the occupational standards.
TABLE 7. OCCUPATIONAL DOSE STANDARDS
Organ Exposed Time Period Dose (rem)
are
Whole body, head and trunk, Accumulated 5(Age in yrs-18)
active blood forming organs, Calendar Qtr. 3
gonads, or lens of eye
Skin, thyroid, and bone
Other organs
Annual
Calendar Qtr.
Annual
Calendar Qtr.
30
10
15
5
The dose standards for the general population are generally one-tenth of
u i _ j j (56,59,96) „ „, _, . , _, „_„(56,Report 2)
the occupational standards. For the thyroid, the FRC ' F
recommended that a value of one-twentieth the occupational standard be
(57 93")
used. The AEC concurred with this recommendation. To compensate
for the variation of doses to individuals from the average of a suitable
sample, the FRC recommended the guides for individuals be reduced by a
factor of three when applied to the sample average.
A suitable sample of the population is defined as a sample of cohorts
representing the critical receptor; i.e., the group of the population
that receives the highest dose with respect to the standards or guides.
62
-------
For example, a one-year-old infant with a two-gram thyroid is the
critical receptor for ingestion of radioiodine in milk.
In January 1971 the NCRP issued an updating report, "Basic Radiation
Protection Criteria," NCRP 39. (-98'1 The NCRP recommended a basic dose
guide of 0.5 rem per year for the total body and/or critical organs
of individuals in the general population. The reduction of the
critical organ guide, from 1.5 to 0.5 rem, was based on administrative
simplification rather than biomedical need. The guide for an average
of the population is 0.17 rem per year.
(95)
The Manager, SNSCr ', in 1967 stipulated the guides to be used in
planning and approving the conduct of reactor tests at NRDS. The
applicability of AEC MC 0524 and the FRC guides^56-1 was noted. The
implication of the wording was that NRDS exposures would be con-
sidered separately from Plowshare exposures. This is similar to the
implications of AEC MC 0524 which specifically excludes Plowshare
activities, noting that guidance is "being developed." The FRC
indicates that the guides of 0.5 rem/yr and 1.5 rem/yr for total body
and thyroid doses, respectively, for individuals in the population are
applicable for the summation of all normal doses from peaceful uses of
(56)
nuclear energy.
2. Accident Conditions
AEC MC 0524 does not stipulate appropriate standards for approving
reactor tests based on hypothetical accident analysis. Although the
AEC has issued standards for reactor siting in Title 10, Code of
(99)
Federal Regulations, Part 100, they have not been considered
appropriate for evaluating the consequences of tests in the rocket
(56,99)
engine program. *•
(95)
In 1967 the Manager, SNSO,^ recommended emergency guides to sup-
plement the AEC standards. It was stipulated that accident
evaluation guides of 10 rem total body dose and 30 rem thyroid dose
63
-------
for individuals in the off-site area would be used in approving
the conduct of tests. Permission to conduct tests is also to be
denied if the predicted on-site accident doses are above 12 rera
and 3b rem to the total body and thyroid, respectively, of indi-
viduals. The dose guides for the general population are to be
reduced by a factor of three where exposures cannot be adequately
measured and evaluated on an individual basis.
The accident guides for the general population are based on the
protective action guides (PAG) of FRC Reports 5 and 7.*-56-' The
whole-body dose accident guide of 10 rem was evolved from the PAG
for 157Cs in FRC Report 7.*-56-1 Although the dose from 137Cs is a
whole-body dose, the intake pathway is the food chain for FRC
Report 7 versus external gamma dose for the NRDS guide. Actions
to prevent or control external gamma dose in a fallout situation
are generally more traumatic than those related to the food chain,
and this use of the PAG appears reasonable.
The accident guide for total-body exposure for on-site personnel was
f93)
based on MC 0524.^ MC 0524 stipulates a quarterly dose of 3 rem
for the total body. A year's summation of this standard would be
12 rem--the proposed accident guide. A similar treatment for an
accident thyroid guide would be 40 rem, whereas the proposed accident
is 36 rem or three times the total body guide. The factor of three
between the total body guide and the thyroid guide was based on the
ratio of the respective FRC guides.
B. Guides for Exposure to Discrete Particles
In 1967 there was increased concern about the biological effects from dis-
crete particles of fuel matter released from the reactor. This concern
related to the lack of applicable radiation protection standards or guides
and the increasing quantity and specific activity of the particulate matter
with the increase in reactor run time and power level. Some of the in-
creases—quantity, specific activity, etc.--were actually occurring whereas
64
-------
some were projected for the future. Thus, although specific guides were
not available for the Phoebus-IB test in February 1967, the effluent
trajectory was limited to minimize the possibility of effluent interaction v
with on-site and off-site population. The NRX-A6 EP-III test was con-
ducted in accordance with preliminary guides which were later adopted.
In May 1968, the Manager, SNSO, stipulated a particle probability-interaction
guide to be used for approval of future tests. ^ The guide was given in
terms of the probability of particles interacting with an individual and
producing a skin tumor or cataract. The guide stipulates that this proba-
bility should be less than one in ten-thousand per person. This probability
is to be based on the product of the probabilities of (1) a particle inter-
acting with a person and [2) the probability of a skin tumor or cataract
being produced from this interaction.
The guide allows credit for the protection afforded by buildings and/or special
clothing or equipment provided the actual use and implied protection can be
demonstrated. The probability of interaction is computed from the model in
Section IV. The probability of an effect from an interaction is based on
the Krebs1 depth concept and the distribution of particle sizes and specific
activities capable of giving doses exceeding Krebs' criteria. Although the
probability of an effect, given an interaction, is greater for a cataract
than for a skin tumor, the probability of deposition of a particle on skin
compared to that of deposition in the eye is such that the basic guide and
thus reactor safety evaluations have been limited to consideration of skin
tumor probabilities.
Krebs evaluated available information concerning biological response of
mammalian skin to irradiation with small, high-specific activity particles.
Krebs concluded that serious radiation-induced acute lesions are caused
primarily by the destruction of the germinal layer cells of the epithelium
(stratum germinativum) and that the viable germinal cells must be reduced
to a survival level of less than one-thousandth over an area large enough
that cell proliferation in the margin of the exposed field will not be able
to replace the dead cells.
65
-------
Based on considerations of the area of skin irradiated and the character-
istics of skin from the standpoint of biological effects, Krebs pro-
posed what has become known as Krebs1 Dose. This is the dose at the
periphery of a circular field of 4 ram in radius at a depth of 100 pm below
the surface of the skin. The center of the field is the point on the sur-
face of the skin where the particle rests.
Krebs postulated that at Krebs' Doses below 1500 rads, development of a
moist desquamation or ulcer type injury was improbable.* Thus, a Krebs'
Dose of 1500 rads was originally used as the criteria of producing an
"effect" in the aforementioned guide. This was based on the hypothesis that
"From a general medical viewpoint, the production of erythema or dry des-
quamation is probably inconsequential," whereas the development of an ulcer
and/or cancer is serious. In rats and mice the development of skin cancer
from radiation has consistently been associated with the prior occurrence
of acute lesions at the time of irradiation.
f O O~* f Q 1 "\
Experiments by Forbes and the biomedical research group of LASL in
1968 and subsequent work by Forbes indicated that doses of less than
1500 rads (Krebs1 criteria) could produce a moist desquamation or ulcer type
lesion.** The experiments were comprised of implacing irradiated UC~ particles
on monkey, pig, and human skin. Experimental dosages ranged from 1300 to
7,400,000 rads for the point basal layer dose (i.e., dose at 100 pm below the
particle skin interface) and roughly 3 to 35,000 rads Krebs' Dose.*- J
Preliminary results from these experiments indicated that moist desquamation
and/or ulcer type lesions could be produced at doses below 1500 rads Krebs'
Dose. Thus, based on the preliminary results, the dose guide was reduced
from 1500 rads to 750 rads Krebs' Dose.(103j I04)
*Krebs denoted four degrees of lesions; erythema, dry desquamation (peeling
or flaking), moist desquamation (moist discharge), anu ulceration (open
weeping lesion, which is slow to heal). Recent work by Forbes(ref.82,101)
and LASL(ref.81,102) indicates that moist desquamation and ulceration are
probably not separate entities for the type of lesions produced in this
case, but rather they merge together. Thus, they will be referred to as
ulcers.
**Preliminary information from these experiments was presented at a meeting
of the safety agencies in Las Vegas, Nevada, on November 8, 1968.
66
-------
The following summarizes the final results from the studies of LASI/81j
and Forbes and Mikhail1- and the status of the particle dose guide:
1. The present guide is a Krebs1 dose of 750 rads. (-103'104-) Based
on the original philosophy used by Krebs1 (i.e., prevent moist des-
quamation or ulcers)^ *, the work by Forbes and Mikhail^101-1 indi-
cates the guide should be reduced.
2. The work reported by Forbes and Mikhail*- ' using pigs indicated
that the lowest doses used, 400 and 440 rads Krebs' Dose, produced
ulcerous lesions; whereas the LASL studies using monkeys did not
indicate the formation of ulcers at doses of 470, 500, or 770 rads
„ , . n (81,102)
Krebs1 Dose. ' J
3. Additional studies are needed to determine the significance
of the difference between the LASL and Forbes1 results. The dif-
ferences may be due to the different animal species, the measurement
of the biological effect, etc. The highest human doses have been
roughly 130 rads Krebs' Dose.
4. Forbes and Mikhail postulated that ulcer size might be a mean-
ingful biological effects criterion. They noted that ulcers of
0.5 mm in diameter (400 rad dose) were probably of much less concern
than those of 1 to 2 mm produced by doses of roughly 750 rads Krebs'
Dose.
5. Krebs notes that the germinal cell layer surrounds the hair
follicle. That is, the layer of germinal cells extends down to the
depth of the hair follicle—roughly a millimeter. Krebs indicates
that this provides a reservoir of germinal cells that are relatively
protected against superficial injury. These cells would receive a
much lower dose than the cells in the main basal layer. Thus, other
things being equal, it is postulated that the dose effect relation-
ship might vary between hairless and hairy skin. This may partially
explain why the dose required to produce an "ulcer" on the sparsely
67
-------
haired pigs is less than that for the monkeys used by LASL. This
should also be considered in applying the results to man.
Appendix B includes a more complete review of the work by Krebs, Forbes
and Mikhail, and LASL.
VI. ENVIRONMENTAL LEVELS AND DOSES
Since the beginning of reactor testing at NRDS in 1959, the effluent docu-
mentation program has included the efforts of a number of organizations.
These have included:
1. EG§G - Long range aerial tracking—not discussed in this
report.
2. NOAA/ARL - Documentation of meteorology and effluent predictions.
3. LASL - On-site surveillance including air sampling, external
radiation monitoring, collection of discrete particles in fall-
out collecting mechanisms and fallout trays. Prior to 1966
most of the on-site sampling and monitoring was performed by LASL.
4. PAA - Primarily on-site surveillance with limited off-site
sampling to obtain information for effluent transport modeling
purposes. The program has included air sampling, gamma expo-
sure measurements, and ground surveys for discrete particulate
matter.
5. NERC-LV - Off-site surveillance program with limited on-site
sampling for inter-program cross-checks and to obtain near-in
samples to estimate potential implications in off-site popu-
lated areas. The program has included ground level surveillance
for external exposure results (mobile monitoring personnel with
survey instruments, fixed instrument stations, and thermolumin-
escent dosimeters); air sampling; sampling of milk, water and
vegetation; and ground surveys for discrete particulate matter.
68
-------
The aerial program has included sampling using equipment for
analysis of particulate, halogen, and gaseous constituents
of the effluent and monitoring equipment to measure gamma
exposure rates and to track the effluent.
6. Reynolds Electrical and Engineering Company - sampling and
monitoring on the Nevada Test Site—not discussed in this
report.
A. Aerosol Effluent
The fission products in the effluent are fractionated because of the different
migration constants in the fuel for the various radionuclides. The radioiodine
isotopes have been of primary importance with regard to potential health im-
plications because of their associated fission product yields and chain rates
of migration, environmental transport, and biological parameters. Of the
iodine isotopes, I has been the most important because of its fission
product yield, half-life, and potential transport to man through the forage-
cow-milk-food chain. '
The chemical states of the various elements composing the effluent have not
been fully documented. Several unsuccessful attempts have been made by
NERC-LV to determine the chemical forms of the radioiodine effluent. At
present, since methane is assumed to be present in the effluent (reaction
of graphite and hydrogen), it is postulated that iodine is present in the
effluent in numerous chemical forms.
1. Off-Site
Grossman summarized exposure history for the off-site areas for
the period 1959-1969, and the following discussion is taken largely
from his summary.
At the time of the first Kiwi reactor in 1959, NERC-LV operated a
network of 12 air sampling stations and a network of 28 film badge
stations in the immediate off-site area. Mobile monitoring personnel
69
-------
were used to supplement information from the networks. Prior to
each test, these monitors were positioned at populated locations
and on existing highways in the predicted effluent trajectory to
measure radiation levels and to collect environmental samples
(milk, water, cow feed, air) should airborne radioactivity be
released. The monitors were also in two-way contact with a control
center which followed reactor test operations and meteorological
conditions, and could be repositioned, as required.
In subsequent years several changes in monitoring techniques and
expansions in NERC-LV routine monitoring networks were made. Be-
ginning in 1961, self-powered air samplers were included with the
monitor's equipment, making the air sampling coverage for tests
more adaptable. In the same year, the monitoring of NRDS test
effluents by aircraft was begun, aiding in the locating of effluent
trajectories and in the positioning of ground monitors.
In 1963 gamma-rate recorders were placed at 16 of the air sampling
locations and in 1965, Model TL-12 thermoluminescent dosimeters by
EG§G, Inc., were included in the film badge network and used off site
by mobile monitors.
The film badge and air sampling networks were expanded after the
latter part of 1961, due to the resumption of nuclear weapons
testing. ' These off-site networks were gradually expanded
through the years so that they now number about 100 air sampling
stations in Nevada and the Western United States, and include 30
gamma-rate recorders, 100 fixed TLD stations, and selected off-site
volunteers who wear dosimeters. Use of film badges in the off-site
dosimetry program was discontinued in 1970 because of the greater
reliability and sensitivity of TLD's. The precise number of the
various sampling stations varies with time.
Grossman estimated whole-body gamma exposures (cloud passage
plus infinite deposition exposure) from reactor effluent and calculated
70
-------
hypothetical infant thyroid doses from inhalation of airborne
radioactivity and from the assumed ingestion of milk contaminated
with radioiodine from reactor effluent. Grossman's exposures and
infant thyroid doses were hypothetical in almost all cases, since
air samples were often taken at unpopulated locations and infants
were seldom present at the locations where air and/or milk samples
were collected. Whole-body gamma exposures were integrated
from G-M survey instrument data because no measurable exposure was
ever detected with TLD's or film badges. 0-*
Figure 8 taken from Grossman, •* indicates the totals for exposures
and hypothetical doses for reactor/engine tests for the period 1959
through 1969 for various sectors of the off-site area. Figures in
Appendix D indicate hypothetical exposures and doses for shorter
periods within the 10-year NRDS testing history. These figures were
compiled by summing the maxima for each test occurring within each
year and within a given sector and entering this value in that
sector. Blank sectors indicate that no radioactivity was detected
or that the hypothetical infant thyroid dose was less than 1 mrad.
The reactor engine tests which released effluent which was detected
off site are listed in Appendix D.
Table 8 [from Grossman ] compares the Radiation Protection Stand-
ards of AEC Manual Chapter 0524 and the FRC guides with the maximum
hypothetical whole-body gamma exposures and the maximum hypothetical
infant thyroid doses received by postulated receptors during the
history of NRDS reactor testing. Within any given year the postu-
lated whole-body gamma exposures were below 12% and 14% of the whole-
body and thyroid dose standards respectively, for a sample of the
population.
Table 9 presents some of the highest environmental concentrations
observed during past reactor tests and the potential hypothetical
doses from these concentrations. Higher values for the air con-
centrations were observed at close-in locations in some instances.
71
-------
WINNEMUCCA
BAKERSFIELD
o o c o BARSTOW
t-Lj \ X
ND: Not Detected
TOP NUMBER:
Hypothetical whole-body gamma exposure in mR
BOTTOM NUMBER:
Hypothetical infant thyroid dose in mrad
Figure 8
Off-Site Whole-Body Gamma Exposures and
Infant Thyroid Doses Resulting from
Reactor Engine Tests from CY 1959 to 1969
72
-------
TABLE 8. COMPARISON OF MAXIMUM HYPOTHETICAL WHOLE-BODY GAMMA EXPOSURES
AND INFANT THYROID DOSES WITH RADIATION PROTECTION STANDARDS.(50)
Radiation Maximum Whole-Body Gamma Exposure (mR)
Type of Exposure/Dose Protection and Infant Thyroid Doses(mrad)**
Standard* '59-'63 '64 '65 '66 '67 '68 '69
Whole-body gamma exposure
Thyroid dose
170 mrem/yr
500 mrem/yr
ND
<3
<1 6
24 72
20
36
2 <1
18 13
2
* Standards are for sample of population, AEC Manual Chapter 0524. (57)
**Units in mR and mrad are equivalent to mrem for this comparison.
73
-------
TABLE 9. SUMMARY OF NRDS NUCLEAR ROCKET TESTING RESULTS
Milk Results, 131I - Infant
Air Concentrations - Adults
Event
NRX/EST-A4,
EP-IVA 3/25/66 (46)
NRX-A5,
EP-IV 6/23/66 (47)
Phoebus-lA,
EP-IV 6/25/65(105)
Phoebus -IB,
EP-IV 2/23/67(45,48)
Distance
from Test
Cell (mi.)
95C
180
120
193
Peak Milk
Concentra-
tion (pCi/1)
140
240
180
iooe
Potentiala
Dose(mrera)
16
17d
20
156
Distance
from Test
Cell(mi.)
95C
60
65
78
131Z
(yCi-sec/m3)
7.4'
9.7
1.7
2.8
Potential
Dose (mrem)
4
7
0.1
6
Potential thyroid dose for an infant drinking 1-liter of milk per day assuming a 2-gram thyroid.
Potential thyroid dose for standard man; based on iodine-131, 132, 133, and 135.
C The meteorological conditions appeared to produce a subsidence and possibly rainout or washout
at this distance, resulting in higher concentrations than would have been suspected from those
near-in. The cows were receiving only about 1/10 of their feed from fresh forage.
FRC 5 type assumptions (5-day effective half-life in milk) would indicate a dose of 48 mrem; but
the observed half-life was shorter than 5 days.
6 There were not complete daily results. The peak concentration is based on extrapolating values
prior to the peak. The highest measured result was 60 pCi/1. Based on FRC Report 5 assumptions.
-------
It is apparent from a comparison of Tables 8 and 9 that most of the
totals for annual exposure and hypothetical thyroid dose given in
Table 8 results from one or two tests for the indicated time period.
Figure 9 ' compares estimated hypothetical doses with pre-run
dose predictions for Phoebus-IB EP-IV. The estimated doses were
calculated from environmental surveillance data and are about an order
of magnitude below the pre-event predictions. Estimated hypothetical
external "gamma exposures are not indicated on the graph. The off-site
gamma exposure-rate was near the limit of detectability, with the
highest measured exposure being less than 0.1 mR. Thus, all of the
gamma exposures fall off the scale of the graph. The predicted external
gamma exposure was based only on cloud passage (i.e., did not include
exposure from deposited activity).
The difference between the pre-event predictions and the post-event
estimates was in large part due to differences between pre-event
estimates of parameters and what actually occurred. Table 10 compares
some of these parameters and shows the effect of various assumptions
on predicted doses.
Figure 9 indicates that pre-event predictions were roughly an order of
magnitude above the hypothetical dose estimates based on surveillance
results. Table 10 indicates that the difference between pre-event
predictions of parameters and post-event estimates of these parameters
accounted for roughly a factor of 7 of the order of magnitude differ-
(48)
ence.
Figure 10 shows the crosswind distribution of the effluent and indi-
cates the relative quantity of I on natural desert vegetation
(relative pCi per unit wet weight of vegetation).
2. On-Site
This section is concerned with exposures produced by reactor test
effluent and does not consider occupational exposures to personnel
75
-------
THYROID DOSE ESTIMATES FROM
RADIOIODINE
OlNGESTION OF 131I
IN MILK (INFANT)
• INHALATION (ADULT)
10
50 100 200
MILES FROM TEST CELL
Figure 9
Pre-Event Dose Predictions and Dose
Estimates from Surveillance Results'*8
76
-------
TABLE 10. TYPICAL EFFLUENT PREDICTION PARAMETERS(48)
_ „,_„,. Pre-Event Predictions Post-Event Factor Change
Parameter Accident Normal Run Normal Run , . In Predicfon
(+ increase; - decrease)*
Power Integral
(Mw-sec) 2.7 x 106 2.7 x 106 ' 3 x 106 +0.1
Fission Pro-
duct Release a 0
r OU'o o-a 1-6 -5
Fraction for
1311 (percent)
Wind Speed
(mph) 6
Effective
Stack Height 300
(meters)
6 15 -2.5
1,500 1,200 Minor for off -site
area
Summing this factor indicates that the pre-event prediction was roughly a factor
of 7 above doses estimated by use of post-event parameters and the model.
77
-------
10 4
PHOEBUS 1B EP IV
10 3
o
LU
O
X
O
LU
10"
or
LU
Q.
6
z
o
o
103
ARC AT 90 MILES
ARC AT 20-23 MILES
LL1
>
103
102
340'
Figure 10
ARC AT 15-20 MILES
350° 0° 10° 20°
AZIMUTH FROM TEST CELL
30'
Relative Quantity of 131I on Natural
Vegetation (Sagebrush or Greasewood
Phoebus IB, EPIV)
40'
78
-------
involved in reactor disassembly or test operations. No compre-
hensive summary of surveillance data or actual absorbed doses in
the on-site area is available, and information presented here is
limited to hypothetical whole-body gamma exposures estimated from
exposure-rate data compiled by Van Vleck^ ' for several reactor
tests. The on-site program has been conducted by LASL and PAA.
Figure 11 indicates the cloud centerlines for the following tests:
NRX-A3, EP-V; NRX/EST-A4, EP-IV; NRX/EST-A4, EP-IVa; NRX-A5, EP-V;
Phoebus-IB, EP-IV; NRX-A6, EP-IIIa; Phoebus-2A, EP-IV; Phoebus-2A,
EP-V; and Pewee-I, EP-III.
Figure 12 presents estimated hypothetical whole-body external gamma
exposures (infinite exposures integrated from time of completion of
cloud passage) resulting from activity deposited by the tests whose
cloud centerlines are indicated on Figure 11. Figure 12 excludes
external exposure received during cloud passage. Past experience
at NTS indicates that exposure during cloud passage may be about
half of the infinite exposure from deposited activity. The inte-
grated exposures are based on the assumption that the deposited
-1 2
activity decayed with time according to t .
B. Deposition uf TVrticulate Effluent
(29)
Simens summarized the early history of environmental surveillance ef-
forts related to particulate effluent. Ground deposition of particulate
matter out to several miles from the test stand was noted commencing with
the first NERVA test (NRX-A2, 1964). Prior to the NRX-A5, EP-IV test in
June 1966, the particulate effluent was considered to be related to the
immediate environment within several miles of the test point. Subsequent
to the NRX-A5 test, particles were found on site 30 miles downwind of the
test stand^ ' . This initiated the concern for particles as a poten-
tial off-site safety problem.
Beginning with Phoebus-IB, EP-IV, LASL, PAA and NERC-LV mounted extensive
field efforts to measure the extent of particulate contamination and to
79
-------
NTS BOUNDARY.
AREA BOUNDARY
DIRT ROADS
PAVED ROADS.
NRDS (AREA 400)
w
il'UOG 0
Figure 11
Cloud Centerlines for Selected Reactor
Tests
80
-------
10
9
8
7
6
5
tr
o
<
Q
UJ
V)
2
UJ
Q
O
cc
u.
LU
o:
en
O
Q.
X
UJ
Q
UJ
CC
o
UJ
10
9
8
7
6
5
10
Q
8
7
6
5
10
9
8
7
6
\
10 15 20 25
DISTANCE FROM TEST POINT, IN MILES
30
Figure 12
Estimated Hypothetical Whole-Body
External Gamma Exposures from Activity
Deposited by Selected Tests Indicated
on Figure 11
81
-------
estimate particle size and activity distributions. The findings of the
organizations are described in their respective reports of surveillance
operations for the Phoebus-IB, EP-IV; NRX-A6, EP-IIIa; Phoebus-2A, EP-IV
and EP-V; and Pewee-I, EP-III reactor tests.(14,15,17,43,84,85,108.109,110,111)
Altomare and Coleman compared the preliminary particle data obtained
for effluent from Phoebus-IB, EP-IV, by PAA, LASL, and NERC-LV and found
the data to be inconsistent. For this test, PAA and LASL data indicated
a log-normal size distribution of particulates with a count median
diameter of about 67 micrometers and a geometric standard deviation of
2.0. NERC-LV particulate data indicated a log-normal size distribution
with a count median diameter of 11.5 micrometers and a geometric standard
deviation of 2.9. PAA reported an average particle density of 6 gm/cc
while NERC-LV reported an average particle density of 11 gm/cc. Altomare
and Coleman indicate that the PAA data showed a decrease in particle size
with distance and a correlation between particle size and activity.
Mikhail and Collins^ 6' also evaluated the Phoebus-IB, EP-IV, particle
data. They divided the particles into size groups, greater than and less
than 12 urn, and evaluated the associated inhalation, ingestion, and skin
exposure implications of human interaction. They concluded skin exposure
was the limiting mode of exposure.
The final NERC-LV report for the Phoebus-IB particle studies indicates an
f!4)
overall count median diameter of 12 ym. The count median diameter for
the particles over 10 ym in diameter and within 10 miles of the test cell
was 35 ym. This latter number is somewhat analogous to the data of PAA
and LASL, which only collected particles greater than 10 to 15 ym in
diameter, the majority of which were within 10 to 15 miles of the test cell.
It was also noted that the particle size generally decreased with distance
although the correlation coefficient was not significantly different from
zero at the 95% confidence level.
The "hot-lines" for the five reactor engine tests for which particulate
data were collected in the off-site area are indicated on Figure 13,
82
-------
CD
00
CH
CROSSBARS INDICATE AREAS WHERE
DATA WERE COLLECTED
36'00
8 0 8 16 12
KILOMETRES
ARTRO S ( f6
11600
11500
Cloud Centerlines for Reactor Engine
Tests for Which Particle Deposition
was Documented in Detail
-------
together with the arcs on which particle data were collected. Particle
sampling efforts for Phoebus-IB, EP-IV, were primarily devoted to lo-
cating the "hot-line" and determining concentrations along the centerline,
rather than establishing lateral distribution about the centerline. Fig-
ure 14 indicates the particle concentrations observed by NERC-LV for the
Pewee-I, EP-III test. Data for the other tests indicated in Figure 13
are in Appendix E.
NERC-LV data indicate particle concentrations for Pewee-I, EP-III, as high
as 0.4 particles/m at a distance of 38 miles, somewhat higher values
than NERC-LV reported for other tests at this distance. However, particle
concentrations in themselves are not sufficient to appraise the hazards
from particles. As outlined in Section IV, particle hazards have been
evaluated in terms of the probability of receiving a specified skin dose
at Krebs1 depth. Initially, for operational purposes at NRDS, proba-
bilities greater than one in ten thousand of receiving 1,500 rads at Krebs1
depth were considered unacceptable in the off-site area. This dose value
was later reduced to 750 rads. * Figure 15 shows the potential proba-
bility of receiving the specified doses at Krebs1 depth as a function of
distance for NRX-A6, EP-IIIa; Phoebus-2A, EP-IV; Phoebus-2A, EP-V; and
Pewee-I, EP-III; as estimated from particle data collected after the tests.
The curve for Phoebus-2A, EP-IV, is the probability of receiving 1500 rads
at Krebs' depth; the other three curves are calculated for a critical dose
of 750 rads at Krebs' depth.
These probabilities were estimated using the ARL model and methods.
Kennedy and Henderson performed the calculations for Phoebus-2A, EP-IV,
and Pewee-1, EP-III.'-112-' The curves for Phoebus-2A, EP-V and NRX-A6,
EP-IIIa, were estimated from data reported by PAA^ ' ^ using the ARL
model and values for the particle activity required to produce 750 rads
at Krebs' depth calculated by Mikhail for XE Prime. Comparison of
curves of particle travel time versus particle activity required to produce
"'See Section V on guides.
84
-------
H-
TO
C
^
0>
00
01
a
a.
Z
O
H;
175
O
a.
y
H-
oe
<
Q.
100 98 96
Three Dimensional Representation of
Particle Deposition for Pewee I, EP-III119
-------
1.0
0.1
CL
HI
a
CD
LU
cc
<
LU
„
Sib2
o
t
cc
o
o
LJJ
o
LU
u_
O
CD
<
CO
o
cc
CL
10
lO'5
SYMBOL TEST CRITICAL DOSE
A NRX A6 EP mA 750 rads
• PHOEBUS 2A EPEZ 1500 rads° (ref.m)
• PHOEBUS 2A EPS 750 rads
O PEWEEIEPH 750 rads0 (ref.iiz)
PHOEBUS 2A EP-ffi
PHOEBUS 2A EP2'
I I I M Mil |_l
PEWEE I EPH
345 678910 20 50 100
DISTANCE FROM TEST POINT IN MILES
Figure 15
Probability of Receiving Critical Dose
at Krebs Depth for Four Tests
86
-------
750 rads at Krebs1 depth for Pewee-I, EP-III, (-91-) (activity production
time about one hour) with the corresponding curves for XE Prime, EP-V,'-16-'
(activity production time about 6 minutes) indicated that the errors intro-
duced by using the XE Prime curves for NRX-A6, EP-IIIa, and Phoebus-2A,
EP-V, were relatively small, probably less than a factor of two.
For NRX-A6 (December 15, 1967) and Pewee-I, EP-III (December 4,1968), the
exposed area of an individual was assumed to be 7.0 x 10~ m , which cor-
responds to a dress style similar to Model A of Kochendorfer and Ulberg. '
For the two Phoebus-2A runs (June and July 1968), the exposed area of the
:o be
(88)
2
individual was assumed to be 0.384 m , corresponding to Model B of
Kochendorfer and Ulberg.
The data in Figure 15 indicate that the probability of receiving the critical
dose (either 750 or 1,500 rads) at Krebs' depth exceeded 0.0001 at some lo-
cations in the off-site area for NRX-A6, EP-IIIa, and Pewee-I, EP-III (see
Figure 16). The 0.0001 probability points for Phoebus-2A, EP-IV and EP-V,
occurred at about 20 and 30 miles, respectively, from the test point; Fig-
ure 13 shows these points to be within the NRDS-NTS complex. For NRX-A6,
EP-IIIa, off-site residents in the effluent trajectory within 25 miles of
the test point were advised to stay indoors for at least one hour after the
end of the test. There were no residents in the range between 25 miles
and 60 miles (the 0.0001 probability point for 750 rads). The 0.0001 proba-
bility point for Pewee-I, EP-III, occurred at about 40 miles and the effluent
did not pass over any populated areas prior to reaching this distance.
There are no known instances of interaction of particles with individuals in
the off-site area; although there are several instances of interaction of
particles with NRDS test related personnel. These include the exposure of
the individuals related to the Phoebus-IB, EP-IV, test^ . Measurable
biological effects from these interactions including erythema were not
observed.
Henderson^ ' analyzed particle deposition data collected for the five
reactor tests where there was significant documentation (Phoebus-IB, EP-IV;
87
-------
NRX-A6, EP-III; Phoebus-2A, EP-IV and EP-V; and Pewee-I, EP-III) by PAA,
LASL, and NERC-LV. Maximum particle concentrations for these tests at
each arc distance are plotted in Figure 16. Particle concentrations shown
in Figure 16 have been normalized to all parameters in the scaling equa-
tion: wind speed, shear angle, and source strength (grams of uranium).
The downwind distance was normalized to the height of rise of the plume,
which is an attempt to normalize the peak concentration for all runs at
a constant dimensionless distance x/h (distance divided by release
height). However, after normalization, there remains variations in the
data of approximately an order of magnitude. Henderson has indicated that
some or all of this variation might be explained by taking into account
the effects of terrain and variations in particle sizes. The lines on
Figure 16 are indicative of the range of expected results, based on past
data.
Values for the variables used in the normalizations shown in Figure 16
are given in Appendix F. Figure 16 can be used to scale particle depo-
sition by forecasting normalization parameters.
VII. SUMMARY, CONCLUSIONS
This paper has reviewed the history of nuclear rocket reactor engine tests
at the Nuclear Rocket Development Station, which adjoins the Nevada Test
Site. The nuclear rocket engines tested during the period covered by this
report, have primarily been based on single-pass hydrogen-cooled reactors
with fuel rods of UC_ (highly enriched uranium) beads in graphite matrices.
NbC and other materials have been used for coolant channel coating to both
protect the fuel and reduce the diffusion of fission products into the
effluent.
The subject tests included planned radioactive effluent releases. Many
of the changes and improvements in reactor design have been primarily
intended to maintain reactor integrity and improve the operational char-
acteristics of the reactor-power level, run-time, etc. In addition, these
changes were intended to and did reduce the radioactive effluent releases.
88
-------
100
10.0
Nua
••••HMBM
Q
1.0
0.1
0.1
O PHOEBUS 18 EP-IV
+ NRX A6 EP-III
A PHOEBUS 2A EP-IV
D PHOEBUS EP-V
• PEWEE I EP-III
\ \ \ I l (I I I I I I I I I I I I I I i i I I i I
Figure 16
1.0 10.0
x/h (DIMENSIONLESS)
Normalized Maximum Particle Concentration16
100
89
-------
The reactor effluent has been described under two categories: (1) fission
products which migrated from the fuel into the coolant and were exhausted as
gases and/or small particulate matter (micrometer size range)--denoted
"aerosol effluent," and (2) discrete fuel fragments which were exhausted
as a result of erosion, corrosion, and minor breakage of the fuel elements
denoted as "particulate effluent." The potential effluent from reactivity
insertion or loss of coolant accidents has also been considered. The
accident effluent has been assumed to be characterized by the aerosol
effluent, but has been postulated to be about an order of magnitude greater.
The biological or health implications of the effluent are evaluated by
comparing effluent predictions and/or surveillance results to the appro-
priate radiation protection guides. Actual biological effects have not
been observed at the low doses associated with the effluent from this
program.
The surveillance results indicate that postulated hypothetical thyroid
doses have been well below (roughly an order of magnitude) the FRC guides
and AEG standards ' ' ' ' . The term postulated hypothetical doses
refers to theoretical estimates which are based on environmental surveil-
lance results (air, milk, etc.) which assume biological parameters for
an infant the critical receptor or person who might receive the highest
potential dose.
Results from surveillance for discrete particles indicate that, although
the administrative guide (less than a probability of 10 of a person
v u I A £ -rrn A -,(80,100,103,104) , ,
receiving a Krebs" dose of 750 rads)v may have been
exceeded in unpopulated areas, there have not been any reported inter-
actions of particles with people in the off-site area. For the NRX-A6,
-4
EP-III test, the probability of 10 was exceeded in a populated area,
but there was a specific operational field effort to request and insure
that the off-site populace in the area of concern stayed indoors.
The guide^ ' ' ' includes the latitude for use of appropriate counter-
measures. It is concluded that off-site exposures or doses from nuclear
90
-------
rocket engine tests at NRDS have been below the applicable guides. In
general, it is felt that the program has been administered and conducted
in a creditable manner and that the results reflect favorably on the
management agencies. In reviewing the effluent program, several aspects
were noted where possible improvements could be made. Potential areas
for improvement include:
132
1. Consideration of the dose from I resulting from the decay
132
of • Te that is inhaled.
2. Additional investigation of techniques for estimating the
external gamma dose from sky shine.
3. Resolve the remaining uncertainties related to the biological
effects of human interaction with the particulate effluent.
Proposed changes in fuel design, indicated in the preface,
may cause these differences to be irrelevant to future nuclear
rocket engine tests. The effluent from the nuclear furnace
tests started in June 1972 was passed through a high-efficiency
scrubber/filter system.
91
-------
REFERENCES
1. Atomic Energy Commission and National Aeronautics and Space
Administration. Environmental Statement, Reactor Testing,
. FY-1972. WASH-1508. (December 1971)
2. Atomic Energy Commission. Nuclear Propulsion for Space. Library
of Congress 67-61704. Technical Information Center, P.O.Box 62,
Oak Ridge, Tennessee. (February 1968)
3. Los Alamos Scientific Laboratory. Pewee I Effluent Effects
Analysis. Memo P.V.Fultyn to H.T.Knight. (November 4, 1968)
4. Los Alamos Scientific Laboratory. Personal communication,
E. A. Bryant and Luther Lyon. (May 1970)
5. Space Nuclear Systems Office. NRDS effluent data bank. Unpub-
lished, undated manuscript received by the National Environ-
mental Research Center, Las Vegas, June 1971.
6. Bryant, E.A., and A.E. Norris. Pewee 2 Fission Product Release
Estimates. Los Alamos Scientific Laboratory, Los Alamos, New
Mexico. (January 15, 1970)
7. Los Alamos Scientific Laboratory. Phoebus-2A Safety Analysis
Report. JB-67-740, Sections 5.3 and 5.4. (July 1967)
8. Westinghouse Astronuclear Laboratories. NRX-A6 Safety Evaluation
Report. WANL-TME-1633, Chapter 7. (October 1967)
9. Los Alamos Scientific Laboratory. Fission Product Release from a
PH-2 Reactor. Memo E.A.Bryant to R.V.Fultyn. (August 8, 1967)
10. Westinghouse Astronuclear Laboratory. Source Term. Letter A. C.
Sanderson to J. P. Jewett, Space Nuclear Systems Office.
(October 28, 1969)
11. Southwestern Radiological Health Laboratory. Final Report of Aerial
Surveillance for the NRX/EST, Experimental Plans IIB through IVA.
Technical Information Center, P.O.Box 62, Oak Ridge, Tennessee.
(May 1966)
12. Southwestern Radiological Health Laboratory. Final Report of Aerial
Surveillance for the NRX-A5, EP-III and EP-IV. Technical Informa-
tion Center, P.O.Box 62, Oak Ridge, Tennessee. (February 1967)
13. Fultyn, R.V. Effluent Effects Evaluation. Part of Pewee 2 Safety
Analysis Report. Los Alamos Scientific Laboratory, Los Alamos,
New Mexico. (Undated)
92
-------
14. Southwestern Radiological Health Laboratory. Particulate Effluent
Study Phoebus-IB, EP-IV. SWRHL-46r. Technical Information Center,
P.O.Box 62, Oak Ridge, Tennessee. (April 1970)
15. Van Vleck, L. D. Summary of Results Effluent Monitoring Phoebus-IB,
EP-IV. PAA-33-12. Pan American World Airways,Inc. (July 1967)
16. Henderson, Donald. Predicting Potential Skin Dose Effects Result-
ing from Particulates Released by Nuclear Rocket Engines. ARLV-351-6.
Environmental Science Services Administration/Air Resources Lab-
oratory, Las Vegas, Nevada. (September 1969)
17. Van Vleck, L. D. Summary of Results Effluent Monitoring Phoebus-2A,
EP-IV and EP-V. PAA-33-31. Pan American World Airways, Inc.
(December 1968)
18. Fultyn, R. V. Large Particles in the Phoebus-IB Effluent. LA-3799.
Los Alamos Scientific Laboratory, Los Alamos, New Mexico. (June 1968)
19. Los Alamos Scientific Laboratory. Letter H.T.Knight to J.P.Jewett,
Space Nuclear Systems Office. (October 10, 1968)
20. Strom, P. 0., D. L. Love, A. H. Greendale, A. A. Delucchi, D. Sem,
and N. E. Ballou. Nuclear Charge Distribution of Fission Product
Chains of Mass Number 131-133. Phys. Rev. 144, 3, p. 984-993. (1966)
21. Katof, S. Fission Product Yields from Neutron-Induced Fission.
Nucleonics 18:201. (1960)
22. Los Alamos Scientific Laboratory. Contribution to Pewee 2 Safety
Analysis Report, transmitted L. L. Lyon to H. T. Knight. N-l-2537.
(January 27, 1970)
23. Los Alamos Scientific Laboratory. Release of Fuel Beads from
Phoebus-2A During EP-IV. Memorandum E. A. Bryant to distribution,
(July 3, 1968)
24. Los Alamos Scientific Laboratory. Private communication, W. Stratton
and R. V. Fultyn. (May 7, 1970)
25. Los Alamos Scientific Laboratory. Personal communication, W. Stratton.
(May 11, 1970)
26. DiNunno, J. J., F. D. Anderson, R. E. Baker and R. L. Waterfield.
Calculation of Distance Factors for Power and Test Reactor Sites.
TID-14844. (March 1962)
27. Los Alamos Scientific Laboratory. Phoebus-2A Experimental Plan IV,
Full Power Run. 64 J-17-393-68, pp 34. LASL, Mercury, Nevada.
(June 26, 1968)
93
-------
28. Coleman, R. D., and V. E. Andrews. Aerial Measurement of the Kiwi
Transient Nuclear Test. Radiation Data and Reports 8:558-566. (July 1967)
29. Aerojet-General Corporation. Untitled, unpublished manuscript,
Hugo Simens, Nuclear Rocket Development Station, Nevada. (1967)
50. Los Alamos Scientific Laboratory. The radioactive effluent from
Pewee-I. A memorandum, E. A. Bryant. (December 27, 1968)
31. Quiring, R. F. Cliinatological Data, Nevada Test Site and Nuclear
Rocket Development Station. ERLTM-ARL 7, UC-53, Meteorology.
TID-4500. Environmental Science Services Administration.(August 1968)
32. Van Vleck, L. D. Meteorological Variables Applied to the Prediction
of Dispersion from Nuclear Rocket Tests at the NRDS. PAA 55-25.
Pan American World Airways, Inc. (June 1968)
55. Fultyn, R. V. A probabilistic scheme for predicting and evaluating
hazards resulting from radioactive emissions to the atmosphere, and
statistical nature of the radioactive effluent from ground tests of
nuclear rocket engine prototypes. Proceedings of the Health Physics
Society Symposium, Augusta, Georgia, January 1968. Environmental
Surveillance in the Vicinity of Nuclear Facilities; edited by W. C.
Reinig. Charles C. Thomas, Springfield, Illinois. (1970)
54. Briggs, G. A. Plume Rise, AEC Critical Review Series. Technical
Information Center, P.O.Box 62, Oak Ridge, Tennessee. (1969)
55. Booth, H. G. The gas-cloud radiological dose prediction models
used by ESSA in support of testing at the Nuclear Rocket Development
Station. Environmental Science Services Administration/Air Resources
Laboratory, Las Vegas, Nevada. Unpublished. (May 1968)
56. Air Resources Laboratory-Las Vegas, National Oceanic and Atmospheric
Administration, Department of Commerce. Personal communication,
Donald Henderson. (September 27, 1971)
57. Slade, D. H. (editor). Meteorology and Atomic Energy. Atomic Energy
Commission. Technical Information Center, P.O.Box 62, Oak Ridge,
Tennessee. (July 1968)
58. Wilson, D. A. Application of Diffusion Models to NRDS Effluent
Released. A progress report. Environmental Science Services Adminis-
tration/Air Resources Laboratory, Las Vegas, Nevada. (Unpublished May 1968)
39. Wilson, D. A. An Evaluation of Radiation Prediction Models Used by
ARFRO at the NRDS. Environmental Science Services Administration/Air
Resources Laboratory, Las Vegas, Nevada. (March 1967)
40. Stigall, G. E., and H. W. Galley. Summary and Analysis of Effluent Data
from Nuclear Rocket Testing at the Nuclear Rocket Development Station.
NUS-599, Vol. 1 and 2. NUS Corporation. (September 1967)
94
-------
41. Environmental Science Services Administration/Air Resources
Laboratory. A Fallout Scaling Model for the Prediction of Gross
Gamma Dose Rates from Earth Cratering Detonations. An unpublished
manuscript, F. D. Cluff and T. Palmer. ESSA/ARL, Las Vegas, Nev.(1964)
42. Hughes, D. C. Nuclear Construction Engineering Technology. NCG
Technical Report No. 2. U.S. Army Nuclear Cratering Group,
Livermore, California. (September 1968)
43. Altomare, P. N., and J. R. Coleman. Study of Particulate Effluent
from Nuclear Rocket Engine Test. NUS Corporation. (September 1967)
44. Van der Hoven, I. A Diffusion Deposition Model for Particulate
Effluents from Ground-Tested Nuclear Engines. Int., J. Air, Water
Pollution 7:1025-32. Pergamon Press. (1963) ~
45. Southwestern Radiological Health Laboratory. Final Report of Off-Site
Surveillance for the Phoebus-IB Reactor, EP-I, EP-II, EP-III, EP-IV
Test Series. SWRHL-45r. Technical Information Center, P.O.Box 62,
Oak Ridge, Tennessee. (July 1969)
46. Southwestern Radiological Health Laboratory. Final Report of Off-Site
Surveillance for the NRX-A4/EST Test Series. SWRHL-30r. Technical
Information Center, P.O.Box 62, Oak Ridge, Tennessee. (September 1966)
47. Southwestern Radiological Health Laboratory. Final Report of Off-Site
Surveillance for the NRX-A5 Test Series. SWRHL-32r. (October 1968)
48. Bernhardt, D. E., and J. R. McBride. Normal Run and the Accident Case
for a Nuclear Rocket Engine. Proceedings of the International
Congress on the Diffusion of Fission Products, Saclay, France.(Nov.1969)
49. Southwestern Radiological Health Laboratory. Final Report of Off-Site
Surveillance for the NRX-A2 Experiment. SWRHL-16r, p. 15. (Feb.3, 1965)
50. Grossman, R. F. Summary of Hypothetical Whole-Body Gamma Exposures
and Infant Thyroid Dose Resulting Off-Site from Project Rover Nuclear
Reactor Engine Tests at the Nuclear Rocket Development Station.
SWRHL-92r. (August 1970)
51. Griffiths, P., and P. B. Erickson. Radiological Prediction and
Monitoring of Tests at the Nuclear Rocket Development Symposium of the
Institute of Electrical and Electronic Engineers, Space Nuclear
Systems Office. (October 1969)
52. Goldman, M. I. Safety Aspects of Ground Testing for Large Nuclear
Rockets. Nuclear Applications 2:94-98. (April 1966)
53. Kochendorfer, D. B. Calculated Activities of U-235 Fission Products
for Very Short Nuclear Reactor Operation. USNRDL-TR-757, Vols. 1 and 2
(June 1964)
95
-------
54. Westinghouse Astronuclear Laboratory. Rover Program Reactor Tests
p. III-l, Performance Summary, NRX-A1 through NRX-A6. WANL-TME-1788.
(July 1968)
55. Bemhardt, D. E. Milk-to-Air Ratios for Iodine-131 Concentrations
(Project Rover). Proceedings of the Health Physics Society Symposium,
Augusta, Georgia, 1968. Environmental Surveillance in the Vicinity
of Nuclear Facilities; edited by W. C. Reinig. Charles C. Thomas,
Springfield, 111. (1970)
56. Federal Radiation Council. Background Material for the Development of
Radiation Protection Standards, Report 1 (May 1960), Report 2
(September 1961), Report 5 (July 1964) and Report 7 (May 1965).
57. Atomic Energy Commission. Standards for Radiation Protection,
USAEC Manual Chapter 0524. (November 8, 1968)
58. Stocum, W. E. Variability in the Parameters Used to Predict the Dose
to the Thyroid from the Ingestion of Iodine-131 in Milk. Proceedings
of the Health Physics Society Symposium, Augusta, Georgia, 1968.
Environmental Surveillance in the Vicinity of Nuclear Facilities;
edited by W. C. Reinig. Charles C. Thomas, Springfield, 111. (1970)
59. International Commission on Radiological Protection. Report of Com-
mittee II on Permissible Dose for Internal Radiation. ICRP Publi-
cation 2. Pergamon Press. (1959)
60. Southwestern Radiological Health Laboratory. Ingestion of Iodine-131
from Food Products Other Than Milk. Memorandum to the Files,
D. E. Bernhardt. (June 5, 1968)
61. Fultyn, R. V. Environmental Effects of the Kiwi-TNT Effluent: A
Review and Evaluation. LA-3449. Los Alamos Scientific Laboratory.
(April 1968)
62. Southwestern Radiological Health Laboratory. Kiwi-B Report, U. S.
Public Health Service Off-Site Radiological Safety Organization.
Distributed December 7, 1961.
63. Fitzsimmons, C. K., and W. H. Horn. Environmental Monitoring with
Thermoluminescent Dosimeters. SWRHL-58r. Technical Information Center,
P. 0. Box 62, Oak Ridge, Tennessee. (1969)
64. Thompson, J. C. Comparison of Iodine-131 Intake from Milk and
Foods. Health Phys. 14_:483-489. (May 1968)
65. Imai, K., and T. lijima. Assessment of Gamma Exposure Due to a
Radioactive Cloud Released from a Point Source. Health Phys.
18^:207-216. (March 1970)
66. Alonso, A. Generalization of Holland's Method to Assess Gamma Doses
from Radioactive Clouds. Health Phys. 13:487-494. (May 1967)
96
-------
67. Glasstone, S. Principles of Nuclear Reactor Engineering.
D. Van Nostrand Co., Inc. (1958)
68. Glasstone, S., and A. Sesonski. Nuclear Reactor Engineering.
D. Van Nostrand Co., Inc. (1963)
69. NUS Corporation, Environmental Safeguards Division. Effluent Report
for the NRX/EST Test Series. NUS-311. (October 1966)
70. Space Nuclear Systems Office. Standard Effluent Concentration to
Dose Conversion Factors for Rover Reactor Operations at NRDS.
Letter R.S.Decker to B.P.Helgeson, April 1, 1966.
71. Mine, G. J., and G. L. Brownell. Radiation Dosimetry. Academic
Press, Inc. (1956)
72. Bunch, D. F. (editor). Controlled Environmental Radioiodine Tests,
Progress Report No. 2. IDO-12053, TID-4500. USAEC. (August 1966)
73. Hawley, C. A. (editor). Controlled Environmental Radioiodine Tests
at the National Reactor Testing Station, 1965 Progress Report.
IDO-12047. USAEC. (February 1966)
74. Stocum, W. E. Conversion Factors for the Child Thyroid Dose in the
H-8 Prediction Model. LASL H-8 Division letter, H-8 M1045. Los Alamos
Scientific Laboratory. (April 3, 1968)
75. Atomic Energy Commission. Conclusions of the Division of Biology
and Medicine/Space Nuclear Safety Advisory Committee. Letter
W. E. Lotz to D. S. Gabriel, June 6, 1968.
76. Mikhail, S. Z., and K. A. Collins. Analysis of the Particulate
Effluent Data from Phoebus-IB, EP-IV Nuclear Reactor Test.
USNRDL-TR-69-76. Naval Radiological Defense Laboratory. (July 15, 1969)
77. Sanders, C. L., R. C. Thompson and W. J. Bair. Carcinogenesis in the
Lung from Inhalation of Radioactive Particles. NRDL-TRC-69-8.
Battelle Memorial Institute, Pacific Northwest Laboratory (contract
with the Naval Radiological Defense Laboratory. (March 18, 1969)
78. Mikhail, S. Z., and K. A. Collins. NRX-A6, EP-III Propulsion Reactor
Test: Characteristics of Particulate Effluent and Comparison to
Phoebus-IB, EP-IV. NRDL-TR-69-76. Naval Radiological Defense
Laboratory. (July 15, 1969)
79. Dean, P. N., and W. H. Langham. Tumorigenicity of Small Highly
Radioactive Particles. Health Phys. .16_: 79-84. (1969)
80. Krebs, J. S. The Response of Mammalian Skin to Irradiation with
Particles of Reactor Debris. USNRDL-TR-67-118. Naval Radiological
Defense Laboratory. (September 1967)
81. Dean, P. N., J. Langham and L. M. Holland. Skin Response to a Point
Source of Fissioned Uranium-235 Carbide. Health Phys. l_9:5-7. (July 1970)
97
-------
82. Forbes, P. D. Acute Effects on Skin Produced by Radioactive
Microspheres: A Progress Report. Undated, unpublished manu-
script. The Skin Cancer Hospital, Temple University, Health
Sciences Center, Philadelphia, PA. (USNRDL contract No. NOO-228
68C571).
83. Deposition and Retention Models for Internal Dosimetry of the
Human Respiratory Tract. By the Task Group on Lung Dynamics for Com-
mittee - II of the International Commission on Radiological Protec-
tion. Healtl^Phys^ 12^:173-207. (1966)
84. Southwestern Radiological Health Laboratory. Particulate Effluent
Study - Phoebus 2A, EP-IV and EP-V. SWRHL-60r. (June 1969)
85. Pan American World Airways, Inc. Summary of Results Effluent Moni-
toring NRX-A6, EP-IIIa. PAA33-21. (April 15, 1968)
86. Sanders, C. L., R. C. Thompson and W. J. Bair. Lung Cancer: Dose
Response Studies with Radionuclides. CONF-691001. AEC Symposium
Series 18. Technical Information Center, P.O.Box 62, Oak Ridge,
Tennessee. (April 1970)
87. Booz-Allen Applied Research, Inc. Empirical Relationships Between
Deposition on Mannikins and Ground Contamination density from
Particulate Clouds. Special memorandum No. 4 to Systems Analysis
Division, Chemical Research and Development Laboratories, Edgewood
Arsenal, MD. (1963)
88. Kochendorfer, D. B., and J. C. Ulberg. Human Exposure to Particu-
late Debris from Aerospace Nuclear Applications. USNRDL-TR-67-59.
Naval Radiological Defense Laboratory. (June 8, 1967)
89. DeAgazio, A.W. Dose Calculation Models for Re-Entering Nuclear
Rocket Debris. NUS-229. NUS Corporation. (May 1965)
90. Space Nuclear Systems Office, Naval Radiological Defense Laboratory.
Rover Flight Safety Program Preliminary Review. IV. Safety Analysis
Report: Radiological Considerations in Nuclear Flight Safety.
USNRDL-TR-1010.(April 20, 1966)
91. Environmental Science Services Administration/Air Resources Lab.
Post-Run Probability Calculation for Pewee I, EP-III. Memoran-
dum D. Henderson to R. Nelson, Space Nuclear Systems Office. (Mar.26, 1969)
92. Ulberg, J. C., and D. B. Kochendorfer. Models for Estimating Beta
Dose to Tissue from Particle Debris in Aerospace Nuclear Applications.
USNRDL-TR-1107. Naval Radiological Defense Lab. (Dec. 12, 1966)
93. Federal Register. Standards for Radiation Protection, Title 10,
Code of Federal Regulations, Part 20.
98
-------
94. Atomic Energy Commission. Radiological Protection Criteria at
the Nevada Test Site. Memorandum N. H. Woodruff, Director,
AEC Division of Operational Safety, to J. E. Reeves, Manager,
AEC Nevada Operations Office. (August 1, 1962)
95. Space Nuclear Systems Office. Radiation Dose Guides for Reactor
Engine Test Approvals. Memorandum M. Klein, Manager, Space
Nuclear Systems Office, to J. P. Jewett, Space Nuclear Systems
Office-Nevada. (July 26, 1967)
96. Atomic Energy Commission Off-NTS Radiological Safety Criteria for
Reactor Runs at NRDS. Memorandum R. E. Miller, Manager, AEC
Nevada Operations Office, to M. B. Biles, Director, AEC Division
of Operational Safety. (June 26, 1969)
97. National Committee on Radiological Protection. Maximum Permissible
Body Burdens and Maximum Permissible Concentrations of Radionuclides
in Air and in Water for Occupational Exposure. National Bureau of
Standards Handbook 69. (June 5, 1959)
98. National Council on Radiation Protection and Measurements. Basic
Radiation Protection Criteria. NCRP Report No. 39. (Jan. 1971)
99. Federal Register. Reactor Site Criteria. Title 10, Code of
Federal Regulations, Part 100. (April 5, 1966)
100. Space Nuclear Systems Office. Additional Radiation Criteria. Memor-
andum M. Klein, Manager, SNSO, to J. P. Jewett, SNSO-N, (May 20, 1968)
101. Forbes, P. D., and S. Z. Mikhail. Acute Lesions in Skin Produced by
Radioactive Microspheres. Final Report, Contract SNSO-49. (April 1970)
102. Los Alamos Scientific Laboratory, Biological and Medical Research Group.
Some Biological Aspects of Radioactive Microspheres. LA-3365-MS.
(August 23, 1965)
103. Space Nuclear Systems Office-Nevada. Radiation Dose Guides. Memor-
andum J. P. Jewett to Milton Klein, Manager, SNSO. (January 1969)
104. Space Nuclear Systems Office. Particle Guides. Memorandum Milton
Klein, Manager, SNSO, to J. P. Jewett, SNSO-N. (January 1969)
105. Southwestern Radiological Health Laboratory. Final Report of Off-Site
Surveillance for the Phoebus-lA Experiment. SWRHL-19r. Technical
Information Center, P.O.Box 62, Oak Ridge, Tennessee. (Jan. 1966)
106. Pan American World Airways, Inc. Personal communication, L. D.
Van Vleck. (1970)
107. Southwestern Radiological Health Laboratory. Particulate materials.
Memorandum D. N. McNelis to D. S. Barth. (Oct. 18, 1966)
99
-------
108. Pan American World Airways, Inc. Summary of Results Effluent
Monitoring Pewee I, EP-III. PAA 40-1. (March 24, 1969)
109. Southwestern Radiological Health Laboratory. Particulate Effluent
Study of NRX-A6, EP-IIIA. SWRHL-57r. Technical Information
Center, P.O.Box 62, Oak Ridge, Tennessee.
110. Southwestern Radiological Health Laboratory. Particulate Effluent
Study, Pewee I, EP-III. SWRHL-66r. (in press)
111. Pan American World Airways, Inc. On-site Report for the Phoebus-IB
Reactor Test Series. PAA 33-13. (August 18, 1967)
112. Kennedy, N. C., and D. Henderson. NRDS Radiation Prediction Model
Evaluation. Environmental Science Services Administration/Air
Resources Laboratory. Report No. ARLV-351-7. (June 1970)
113. Southwestern Radiological Health Laboratory. Final Report of
Off-Site Surveillance for the NRX-A6 Test Series. SWRHL-78r.
(August 1971)
114. Space Nuclear Systems Office. TWX concerning particle guides,
J. P. Jewett, SNSO-N, to Milton Klein, Manager, SNSO. (Dec.2, 1968)
115. Spencer, L. V. National Bureau of Standards Monogr. 1. (1959)
116. Mikhail, S. Z. Tissue Beta Radiation Doses from Particulate
Fission Product Sources: Comparison of Model Predictions with
Experimental and Monte Carlo Values. ESA-TR-70-1. Environmental
Science Associates. (May 1970)
117. Space Nuclear Systems Office. Particle Study. Letter S. Z. Mikhail,
Naval Elect. Lab., to R. Nelson, SNSO-N. (Oct. 7, 1969)
118. Eve, I. S. A Review of the Physiology of the Gastrointestinal Tract
in Relation to Radiation Doses from Radioactive Materials. Health
Phys. 12^:131-161. (February 1966)
119. Dolphin, G. W., and I. S. Eve. Dosimetry of the Gastrointestinal
Tract. Health Phys. .12^163-172. (February 1966)
120. Coleman, J. R., and L. J. Perez. Consideration of Dose Models for
the Eye and Lung from Nuclear Rocket Particulate Effluent.
NUS-467. NUS Corporation. (May 15, 1968)
121. International Commission on Radiological Protection. Radiosensitivity
and Spatial Distribution of Dose. Committee I. ICRP Publication 14.
Pergamon Press. Glasgow. (1969)
122. Southwestern Radiological Health Laboratory. Personal communication,
R. E. Stanley. (1970)
100
-------
123. Los Alamos Scientific Laboratory. Personal communication, W. H.
Langham. (May 1970)
124. Coleman, J. R., and L. J. Perez. Radiological Safety Studies of
Space Nuclear Systems—Fourth Quarterly and Final Report.
SC-CR-67-2729. Sandia Corporation. (October 1967)
125. Pestaner, J. F., D. A. Kubose and W. R. Balkwell. Beta Dose Rates
from Neutron-Irradiated Uranium-235 Carbide Particles. USNRDL-TR-1082.
Naval Radiological Defense Laboratory. (June 30, 1966)
126. Richmond, C. R., J. Langham and R. S. Stone. Biological Response
to Small Highly Radioactive Sources. Health Phys. 18:401-408.
Pergamon Press. (1970)
127. Altomare, P. M., L. J. Perez and J. R. Coleman. Method Applied for
Calculation of Probability of Cataract and Probability of
Exceeding Krebs1 Depth Dose for Phoebus-2A, EP-IV. NUS Corporation.
(June 20, 1968)
101
-------
APPENDICES
Appendix Title Page
A. .Population and Milk Cow Distribution. 104
B. Models and Information Related to Potential
Biological Effects of Radioactive Particles. 105
B-l. Acute Lesions in Skin Produced by Particulate
Effluent. 106
Dose Calculation Techniques. 106
Results of Experiments. 108
Summary of Particle Skin-Bioeffects Experiments. 110
Summary. 113
B-2. Potential Dose to the GI Tract from Particulate
Material. 114
B-3. Mathematical Models for Predicting Effects On the Lung
From Inhalation and Deposition of Radioactive Particles.117
Introduction. 117
Deposition of Particles in the Respiratory System. 118
Estimation of Dose to the Lung from Deposited
Material. 120
Biological Effects of Doses to the Lung from
Particulates. 121
B-4. Model Used for Calculation of Probability of Cataract
from Deposition of Particle in Eye. 122
Schematic Drawings of the Model Eye. 123
Anatomical Constants for Model Eye. 124
Probability of Cataract Per Particle for Various
Locations and Residence Times in the Eye as a Function
of Total Particle Activity at Time of Deposition. 126
Average Number of Particles Deposited in Eye Region
Per Unit Air Concentration Versus Wind Speed. 129
102
-------
Appendix Title Page
C. Dosimetry and Air Surveillance Network Stations. 130
D. Reactor Engine Tests at NRDS from Which Airborne
Radioactivity was Detected Outside the Test Range
Complex. 132
E. Deposition Measurements and Survey Results. 140
F. Particle Concentrations and Normalization
Parameters. 145
103
-------
POPULATION AND MILK COW DISTRIBUTION
APPENDIX A, FIGURE A-1
POPULATION/MILK COWS
U.S. ENVIRONMENTAL
PROTECTION AGENCY
NATIONAL ENVIRONMENTAL
RESEARCH CENTER
LAS VE6AS, NEVADA
Figure A-1
Population and Milk Cow Distribution
104
-------
APPENDIX B
MODELS AND INFORMATION RELATED TO POTENTIAL
BIOLOGICAL EFFECTS OF RADIOACTIVE PARTICLES
1. Acute lesions in skin produced by particulate effluent.
2. Potential dose to the GI tract from particulate material.
3. Mathematical models for predicting effects on the lung from inhalation
and deposition of radioactive particles.
4. Model used for calculation of probability of cataract from deposition
of particle in eye.
105
-------
APPENDIX B - 1. Acute Lesions in Skin Produced by Particulate Effluent.
This appendix includes information which supplements the information in
Section V on guides for the particulate effluent.
Based on considerations of the area of skin irradiated and the character-
istics of skin from the standpoint of biological effects, Krebs pro-
posed what has become known as Krebs1 Dose. This is the dose at the
periphery of a circular field of 4 mm in radius at a depth of 100 ym below
the surface of the skin. The center of the field is the point on the
surface of the skin where the particle rests. He postulated that Krebs'
doses of less than 1500 rads would not produce serious skin lesions.
In 1968 preliminary results from experiments conducted by Forbes and
Mikhail(82>101) and the Biomedical Research Group of LASI/81'102-* indi-
cated the need for reconsidering the dose guide for the skin. The
experiments were based on implanting irradiated pyrolytic graphite-coated
UC,, beads on the shaved skin of monkeys and humans (LASL) , and pigs
(Forbes).
Based on the preliminary results of these experiments ' ' ' ' the
particle dose guide was reduced to 750 rads Krebs' dose. ' '
Dose Calculation Techniques
The following reviews briefly compares the dose calculation techniques
used by the investigators.
Three types of dose calculations were made:
1. Beta dose at the surface of the particle.
2. Beta dose at the pstulated depth of the basal or germinal cell
layer—100 ym.
3. The dose at the periphery of a 4 mm circular plane at a tissue
depth 100 ym below the particle, Krebs1 dose.
106
-------
The Krebs1 criterion has been used as a normalizing tool for presenting
the results from particles of various sizes for the various studies
referenced in this paper. The basal layer dose is also given. The
surface dose has received limited use. Since the surface skin is composed
of dead cells and the point dose gives no consideration to area distri-
bution of the dose, it has limited biological meaning.
The LASL dose calculation technique is given in References 81 and 102.
In summary the beta energy spectrum is measured using a plastic scintil-
lator, the dose rate is then computed based on energy loss values for
100 keV intervals and integrated. *• jl ^ The dose at 100 pm is calculated
based on exponential absorption.
The doses reported by Krebs and Mikhail are based on a determination of the
number of fissions that have occurred in the particle using a specially cali-
brated high pressure ionization chamber and a combination of computer codes
produced by the former Navy Radiological Defense Laboratory (NRDL) and known
as the Transmission-Degradation-Dissipation (TDD) beta dose model. The first
code estimates the build-up and decay of the fission products and a second
code computes the beta energy spectrum for each beta emitting nuclide. A
third code computes a composite spectrum weighted for the quantity of each
nuclide present. A fourth code calculates the energy loss by self-absorption
within the particle and a fifth code then gives the beta radiation emitted by
the particle. A sixth code computes the beta depth-dose-rate in tissue and
the integrated dose.
Mikhail^ ' has performed several evaluations of the NRDL model and com-
pared it to the LASL model. In summary, there is good agreement between
the LASL and NRDL models and between the NRDL model and empirical meas-
urements. The following summarizes some of the conclusions:
1. A comparison of the NRDL and LASL techniques for calculating
the basal layer dose (100 pm skin depth) for 11 particles
indicated the maximum difference (difference/average) was 33
and the average difference was 27%. The NRDL dose estimate
was always larger than the LASL estimate.
107
-------
2. An evaluation of the TDD NRDL program indicated that it gave
invariably higher results than empirical experiments for
shallow tissue depths (range of 100 ym). *• ' The TDD doses
ranged to roughly 50% higher than the measured doses (varied
with particle size, radioactive decay, tissue depth, etc.).
This gives reasonable confirmation for the LASL dose
estimates.
The various dose models are in reasonable agreement. The apparent bias of
the TDD model towards higher results is acceptable if it is used for both
estimating the doses versus biological effects and the evaluations for
hazards from reactor operations; as is presently done. Mikhail * dis-
cusses the possible biases in the model and these will not be discussed here.
The estimates for Krebs' doses for the LASL results were determined in sev-
eral ways:
f 811
1. For the third experiment with monkeys, Dean, et al., the
Krebs' dose was calculated by Mikhail using the TDD program.
2. For the human experiments the Krebs' dose was estimated from
a plot of particle size versus the ratio of the basal layer
dose to the Krebs' dose (based on Mikhail's data in item one
above).
3. For the first two experiments with monkeys, actual particle
sizes were not reported. ' The ratio of the LASL basal
layer dose versus Krebs' dose for a 150 ym (a range of 140 to
160 ym was given) particle was used per the information in
items 1 and 2 above. This ratio was 400.
Results of Experiments
The preliminary results of these studies indicated that the skin lesions
f 881
did not fall into the four categories indicated by Krebs. J Rather,
moist desquamation and ulceration did not appear to be separate entities,
but merged together. Thus, for simplicity the single term "ulcer-
ation" will be used.
108
-------
The following sections summarize the results of the respective studies.
Table B-l indicates the particle size used, the resulting doses, and the
lesion produced for all of the experiments. LASL, Experiments by Dean,
et al,: ^'102^
1. The study utilized UC2 beads of 100-200 urn in diameter which
were placed on the skin of humans and monkeys. The resulting
doses were calculated at a point 100 ym below the particle
and'at the skin surface. The LASL work is essentially com-
posed of four experiments: two in 1965 on monkeys (basal skin
layer dose ranged from 1,300 to 52,000 rads or 3 to 130 rads
Krebs' dose); one on human skin in 1965 (basal skin layer dose
ranged from 14,000 to 54,000 rads or 40 to 130 rads Krebs1
dose); and a recent one on monkey skin with basal skin layer
doses of 157,000 to 664,000 rads or 470 to 2,700 rads Krebs'
dose.*
2. Ulceration, or moist desquamation, was observed at all sites
with basal point doses of 261,000 rads or Krebs' dose of
850 rads or higher. The lesions remained open for approxi-
mately two days after exposure. Epithelization was complete
by 71 days after exposure, the residual effect being a dimple.
3. Skin sites receiving less than a 261,000 rad basal layer dose
showed only a dry desquamation over an area of 3-4 mm in
diameter.
4. The LASL results bracketed the minimum dose which produced a
moist desquamation and/or ulcer.
Forbes' and Mikhail's Experiments Using Pigs.
1. Forbes' experiments dealt with implacement of 12 irradiated
particles on pig skin. The particles ranged from 140-328 ym
*The LASL studies were not based on the concept of Krebs' dose. Estimates of
Krebs' doses for the LASL studies are based on calculations, using LASL data,
by Mikhail (Ref.117) or estimates by the authors of this paper using a plot of
particle size versus the ratio of the basal layer dose to the Krebs1 dose.
109
-------
TABLE B-l. SUMMARY OF PARTICLE SKIN-BIOEFFECTS EXPERIMENTS.
Particle
Size
(urn)
Forbes §
148
140
145
153
144
148
152
154
150
283
305
308
295
304
282
294
328
149
298
Dose
NRDL
..... -,(101)
Mikhail^ •*
240
286
488
514
563
569
642
717
1370
1230
1530
1510
2350
2320
3550
3800
3850
7400
1700
at Basal Layer
(KRAD)
LASL Calc by
NRDL*
Krebs '
Dose
(RAD)
405
444
770
771
921
944
060
210
310
400
320
490
10,980
11,100
15,100
17,300
19,800
35,200
28,900
Ulcer Type
Dia. Lesion
(mm)
0.5 Ulcer-
ation
0.5
1
2
0.5
0.5
2
2
3
4
5
5
5
5
6
6
7
8
4
Experimenta
Animal
Pig
LASL--First Monkey Experiment (102)
140-160
1.3 Not
Calc.
2.0
2.8
3.4
4.1
4.5
4.9
5.4
7.9
7.0
7.5
8.9
9.6
3
5
7
9
10
10
10
10
10
10
20
20
20
Erythema
Monkey
110
-------
TABLE B-l. Summary of Particle Skin-Bioeffects Experiments.(continued)
Particle
Size
(yra)
LASL- -Second
140-160
LASL --Human
174
159
140
LASL--Third
172
174
186
184
186
184
191
225
204
215
213
Dose at Basal Layer
(KRAD)
NRDL LASL Calc by
NRDL*
Monkey Experiment (102)
15.8 Not
Calc.
19.5
• 21.5
23.8
29.3
29.5
31.4
45.8
52.1
Experiment (81)
14.2 Not
Calc.
40.0
54.0
Monkey Experiment (81)
157 220
173 230
242 320
261 350
270 360
289 390
313 420
351 430
472 600
458 590
664 870
Krebs'
Dose
(RAD)
40
50
50
60
70
70
80
110
130
43
110
130
470
500
770
850
890
950
1,100
1,400
1,800
1,800
2,700
Ulcer Type
Dia. Lesion
(mm)
Erythemia
Possibly--
Dry Desquam
Erythemia
Erythemia
Erythemia -
Dry Desquam
Dry Desquam
Ulceration
Experimental
Animal
Monkey
.
Human
•
Monkey
Calculated from NRDL fission product beta dose calculation code. J The NRDL dose
estimates for LASL experiments are presented to allow normalization of all of the
data.
Ill
-------
in diameter. Based on the Krebs' dose criteria, the doses ranged
from 400 to 35,000 rads and the basal layer point doses from
240,000 to 7,400,000 rads.
2. The minimum dose used in the experiments, 400 rads (Krebs1 dose),
produced a small ulcer.
3. The authors suggested using ulcer size as a biological end-point
versus ulceration, etc.
4. The various biological responses versus dose were evaluated.
Straight line plots of lesion diameter versus the log of the
dose resulted in correlation coefficients of 0.94 and 0.98,
respectively, for dry desquamation and ulceration. Assuming
a straight line extrapolation is applicable, the authors noted
that the lines approached zero at 250 and 350 rads (Krebs'
dose), respectively, for dry desquamation and ulceration.
The following items denote the uncertainties that occurred in attempting to
correlate these experiments:
1. The doses were not uniformly distributed throughout the total
dose interval.
2. The experiments were conducted on both pigs and monkeys, with
limited experimental results on human skin. The range of ex-
perimental doses for each species was even more limited than
for all of the experiments.
3. In addition to the uncertainties of the dose estimates, dif-
ferent dose calculation techniques were used by the different
investigators. The uncertainties due to this were largely
resolved by the dose calculation correlation efforts of Mikhail.
4. The variation of the depth of the basal or germinal layer of cells
on an individual animal can cause biological variability in
addition to the biological variability between animals and
across specie lines. There may be other sources of biological
variability.
The AEC Division of Biology and Medicine, Space Nuclear Safety Advisory
Committee met on February 15, 1968, to consider the hazards associated
112
-------
with nuclear rocket ground tests --concerned primarily with particulate
effluent. They were asked to consider only the implications of various
modes of interaction of particulate material with humans--not the prob-
ability of the interaction. Based on their conclusions, and upon con-
sidering probabilities of interaction, the dose to the skin from particles
residing on the skin was considered to be the limiting situation.
This committee did not have the benefit of the previously mentioned studies
(Forbes and LASL final results) and thus their dose-effect relationships
were not in full agreement with the aforementioned information. But, one
of their recommendations is of special interest. The committee endorsed
the concept of Dean that the skin tumor hazard could be estimated from
an integration of the expected tumor yield for each exposed cell. The
tumor yield per cell would be based on the dose yield per cell and published
f 79)
dose-tumor response data for rats.
Summary
The present guide is a Krebs1 dose of 750 rads. ' •* This is based on
minimizing the probability of ulceration occurring. Moist desquamation
was not considered as a separate entity.
The work reported by Forbes and Mikhail^ ' using pigs indicated that
the lowest doses used, 400 and 440 rads Krebs' dose, produced ulcerous
lesions; whereas the LASL studies using pigs did not note the formation
of ulcers at doses of 470, 500, 770 rads Krebs' dose. (81>102}
Additional studies are needed to determine the significance of the
difference between the LASL and Forbes results.
Forbes and Mikhail postulated that ulcer size might be a meaningful bio-
logical effects criterion. They noted that ulcers of 0.5 mm in diameter
(400 rad dose) were probably of much less concern than those of 1 to 2 mm
produced by doses of roughly 750 rads Krebs1 dose.
Krebs ^83^ notes that the germinal cell layer surrounds the hair follicle.
That is, the layer of germinal cells extends down to the depth of the hair
113
-------
follicle—roughly a millimeter. Krebs indicates that this provides a
reservoir of germinal cells that are relatively protected against super-
ficial injury. These cells would receive a much lower dose than the
cells in the main basal layer. Thus, other things being equal, the authors
postulate that the dose effect relationship might vary between hairless and
hairy skin. This may partially explain why the dose required to produce an
ulcer on the sparsely haired pigs is less than that for the monkeys used by
LASL.
APPENDIX B - 2. Potential Dose to the GI Tract from Particulate Material.
Several investigators have evaluated the implications of ingesting discrete
radioactive particles either directly or through clearance from the respir-
atory system to the gastro-intestinal tract (GI). Several of these inves-
tigators and their general area of investigation are indicated below:
1. The LASL Bio-Medical group experimentally evaluated the solubility
of coated and uncoated UC? particles in the gastric juices of
monkeys. They also determined the clearance time for particles
of various densities for the GI tract of man.
fO Q •*
2. Kochendorfer and Ulberg evaluated various modes of intake.
They noted that a mouth breather had a higher probability of
exposure than a nose breather.
3. The Radiological Effects Working Group of SNSO developed a
complete model from potential intake to dose to the GI tract. '
The model considers both the average gamma dose to the GI tract
and the beta dose at a depth of 300 ym from the gut wall. The
depth of 300 ym is based on the depth of the dividing crypt
cells. The model also considers transfer of radionuclides from
the GI tract to the blood.
4. Mikhail and Collins^ evaluated the effluent data from the
Phoebus-IB reactor test. They used the intake model of item 2
above, and indicated intake probability and potential doses at
a 300 ym tissue depth. They apparently calculated the integral
dose from the passage of particles through the intestines versus
the dose to discrete volumes of tissue. They concluded that the
dose to the skin was more limiting than that to the GI tract.
114
-------
5. Altomare and Coleman*- J evaluated the ground deposition and
transport data from the Phoebus-IB reactor test and developed
dose prediction models for human interaction, including the
GI tract. Their dose calculation is based on the beta energy
loss to the GI contents during transit. This is the basic
model used by the ICRP^ ^ for GI tract dose calculations.
The transit times suggested by Eve^ ' were used.
6. At the request of SNSO, the AEG Division of Biology and Medicine
Space Nuclear Safety Advisory Committee evaluated the various
modes of potential particle interaction with humans. ' They
were requested not to consider the probability of interaction--
only the implications of an interaction. They concluded that
for the GI mode of interaction, "...the impact on a person's
health would be roughly the same or less than the effects
resulting from skin exposure." Mikhail and Collins, ' indi-
cated that the probability of a particle reaching the GI tract
was at least two orders of magnitude less than that for inter-
acting with the skin.
The above models have little in common other than that they indicate that
the probability of interaction with, and irradiation of, the skin is more
limiting than that for the GI tract. The following steps indicate a general
approach for estimating the dose to the GI tract. As with the skin and lung,
the basic uncertainty is the appropriate tissue, both depth from the surface
of the GI lining and tissue mass and/or distance along the GI tract that
should be used for the dose calculation.
1. The intake of particles can be based on evaluations in references
43, 76, 88 and 90. These references use similar approaches. It
would appear that the particles cleared from the respiratory
tract to the GI tract are of prime importance. The ICRP Task
f831
Group report on lung dynamics^ is a good basic reference from
which the transfer of material from the respiratory system to the
GI tract can be calculated.
2. The transfer time through the various segments of the GI tract is
important for all of the various dose calculation techniques.
115
-------
There are several values given in the literature:
Small Upper Lower
Stomach Intestine Large Large
Intestine Intestine
a. The ICRP in Report 2, 1959
recommends(ref.59) 1 hr 4 hr 8 hr 18 hr
b. Kochendorfer and Ulberg
(ref.88) indicate that
Snyder of ORNL indicates
that the ICRP will change
the values to 1 hr 4 hr 13 hr 31 hr
c. Kochendorfer and Ulberg
(ref.88) also refer to
work at Argonne that
indicates some retention
for several days
d. Eve (ref.118) recommended 1 hr 4 hr 13 hr 24 hr
3. Eve's work concurs with the SNSO Radiological Effects Working
Group in that the mitosing cells are at some depth below the
GI surface lining. Her literature review indicated that this
depth varies from 140 ym for the small intestine (SI) to possibly
420 pm for the large intestine (LI). But, she also notes that it
is not certain that some of the GI contents do not penetrate the
wall to some extent.
In summary, it appears reasonable to select the large intestine (upper or
lower--ULI or LLI) as the critical organ (due to the transit time) and cal-
culate the dose at 300 nm. The dose can be calculated by the techniques
indicated in references 102 and 116.
Dolphin and Eve^ recommend changing from the ICRP 2^ J concept of cal-
culating the dose at the entrance to a given section of the GI tract to
calculating the average dose for the section. They note that this would be
more consistent with dosimetry used for other body organs. Due to the rel-
atively short half-life of the gross fission products in the particles
(based on roughly 2 hours transport time prior to ingestion) the dose for
the various sections will change by roughly a factor of two between the
116
-------
beginning and end of a section. Considering the lack of information on
the bio-effects to the GI tract from irradiating various masses of tissue,
this factor of two is probably minor.
Due to the generally low solubility of the particles,(90>102) uptake of
radionuclides from the particles to the bloodstream does not appear to
merit further consideration.
APPENDIX B - 3.' Mathematical Models for Predicting Effects On the Lung
From Inhalation and Deposition of Radioactive Particles.
Introduction
Prediction of biological effects on the lung from inhalation of particulate
debris from nuclear rocket engine tests is best visualized as a four step
process: (1) prediction of debris transport; (2) prediction of inhalation
and deposition of the particles in the lung; (3) estimation of the dose to
the lung from deposited particles; and (4) prediction of biological effects
resulting from the estimated doses. The techniques for carrying out each of
these steps are not equally well defined; in particular biological effects
on the lung from radiation are not well understood.
Particle air concentrations at points downwind can be estimated with one of
the several available models. Examples include the ARL-LV model described
by Henderson^ ' and the Van der Hoven model adapted for computer usage
reported by Stigall and Galley.^
Several methods have been proposed to accomplish the second step, prediction
of inhalation and deposition of particles in the lung. Perez and Coleman
summarized and compared five deposition models contained in the literature.
From a knowledge of particle characteristics such as activity and size distri-
butions and the particle residence time, doses to the lung from deposited
particles can be estimated (the third step).
The fourth step, prediction of biological effects on the lung from inhaled
particles, is not well-defined. A question which is central to this problem
is the relative tumor-producing effectiveness of a relatively small dose to a
117
-------
large mass of tissue (kilograms) compared to a large dose to a small mass
of tissue close to the particle (grams or fractions of a gram). The
question is whether averaging the dose from the inhaled particle over the
entire lung is more conservative from a health standpoint than computing
the dose to the small mass of tissue which is actually exposed. The ICRP
in Publication 14 notes that "the same radiation energy absorption
might well be less effective when distributed as a series of "hot spots'
than when uniformly distributed." While there is uncertainty regarding bio-
logical effects from lung irradiation, it appears on the basis of present
evidence that for the particular particle size and activity distributions
expected from reactor engines, exposure to the skin is more limiting than
exposure to the lung. ' '
Prediction techniques such as those of the eye model for predicting cataract
probability have not been developed for the lung. Studies are in progress
to better define dose-effect relationships for the lung. These studies
include projects specifically dealing with reactor engine effluent such as
(122)
that of Stanley at NERC-LVV where activated UCL particles are implanted
(123) (77)
in rat lungs. The studies of Langham and Sanders, et_ al_. , are
concerned with alpha and x -radiation from plutonium particles. Sanders,
f Rfi*\
et_ al_. , summarized a number of lung radiological bio-effects studies and
evaluated the hazard of particles from nuclear rocket engine tests.
Deposition of Particles in the Respiratory System.
Several models for prediction of particle deposition in the respiratory
system have been proposed. The detailed description and comparison of these
models is beyond the scope of this appendix. Perez and Coleman reviewed
and compared five of these models, and further information can be obtained
from their work and references 59, 83, 88, 90, 124.
Coleman and Perez ^ developed a mathematical representation of the most
comprehensive deposition model, that of the Task Group on Lung Dynamics of
f 83^
Committee II of the International Commission for Radiological Protection. l
This model is distinct from the ICRP 2l J lung model.
118
-------
On the basis of purely geometrical considerations, the respiratory system
acts like a multicompartment selective sampler which filters out most of
the larger particulate material in the upper respiratory tract. The mech-
anisms and efficiencies of dust or particle clearance differ considerably
for various regions of the respiratory tract. Consequently the various
models for particle deposition generally separate the respiratory system
into three compartments for purposes of estimating particulate deposition
and movement. Coleman and Perez's^ ' adaptation of the ICRP Lung Dynamics
Task Group model uses the following separation scheme: Compartment I in-
cludes the nasopharynx and oral pharynx to the larynx; Compartment II
includes the oral pharynx starting at the larynx, trachea, bronchus, and
connecting and terminal bronchiole; and Compartment III includes respira-
tory bronchiole, alveolar duct and alveolar sac.
The model considers only the entrance of particles through the nose,
ignoring mouth breathing. It assumes that all particles, regardless of
size, have some probability of entering the system and that the size range
available for entrance may be approximated from inertial properties of the
particles.
One of the problems involved in evaluating biological hazards from inhaled
particles is that even though the probability of inhalation of particles
with aerodynamic diameters greater than approximately 10 vim may be very
small, the consequences of deposition of particles this large and containing
the associated quantities of radioactivity in the lung are so serious that
even this small probability should not be neglected.
The raction of particles deposited in each of the various compartments is
calculated from an empirical equation which Coleman and Perez fitted to
f 83)
deposition data reported by the ICRP Task Group on Lung Dynamics. The
model also attempts to incorporate data presented by the ICRP Task Group on
clearance of dust and particles from the various respiratory compartments.
Clearance actions by the blood, lymph system, and ciliary-mucus transport
to the stomach are estimated by the model, based on the assumption that
119
-------
material available for clearance is cleared according to first-order
kinetics (the rate of clearance is assumed to be proportional to the
amount of material present). The output of the deposition and clearance
models is used to integrate the lung activity burden with respect to
time, yielding a figure in curie-days or equivalent units.
Estimation of Dose to the Lung from Deposited Material
Doses to the lung from radioactive particles are usually computed using
either of two approaches. One approach is to assume that the energy from
particle radioactivity is deposited uniformly throughout the lung; this
is referred to as the "smeared dose" concept. The other basic approach
is to calculate the dose to the relatively small quantity of tissue near
the particle. Some combination of these two approaches may also be used,
analogous to the Krebs1 depth dose concept used for the skin.
Coleman and Perez use the smeared dose concept. The estimated time-integrated
organ activity burden is multiplied by a dose conversion factor; this factor
depends on (1) radiation type and average energy; (2) physical properties of
the particle material when activity is associated with a particle, since
self-absorption must be considered; and (3) physical properties of the
organ. This process yields what the authors describe as an average dose
which is satisfactory if the number of particles per person is large. That
is, if the number of particles inhaled is large, then the fraction of each
size range deposited in each region of the lung can be estimated with satis-
factory accuracy from the ICRP Task Group data. However, most cases
encountered in estimating hazards from nuclear rocket effluent involve
probable deposition of a very few particles or of only a fraction of a.
single particle per person. In such a case the NUS model gives an average
dose for a number of individuals. To overcome this shortcoming, the
authors also developed a method of generating dose distribution, using
the deterministic or average-dose model as a basis. Using this approach
estimates can be made of the probability of an individual receiving a
particular dose.
120
-------
The model by Coleraan and Perez yields dose estimates in terms of the
"Smeared dose" to the lung. Mikhail!?116-) Mikhail and Collinsf76^ Ulberg
and Kochendorfer, -1 and Pestaner, et_ al_. , -* developed techniques for
computation of point doses to tissue from deposited particles. Mikhail
and Collins^ J calculated energy deposition in the lung by integrating
the beta dose rate from each particle with respect to volume (equivalent
to mass-integration, since they assumed a lung density of 1 gram/cc). The
integration is performed by summing the dose rates absorbed in successive
differential volume shells. Dividing this mass integrated dose rate
(rad-g/sec) by the total mass of the lung and integrating over the particle
lung residence time produces the smeared dose to the lung.
Biological Effects of Doses to the Lung from Particulates
Two primary points of contention remain concerning the lung models. First,
which is appropriate, the smeared dose concept or the point dose concept?
If the point dose concept is selected, then the tissue volume of interest
(77")
must be defined. Sanders, Thompson, and Bair^ J indicate that most studies
have used, as a standard, the dose at 100 ym from the particle. Second, the
effect or biological endpoint must be defined (e.g., scar tissue or cancer).
The dose-effect relationship then determines what volume of tissue must
receive what dose to produce a given probability of incidence for this end-
point.
Sanders, et_al., ' ' indicate that non-uniform irradiation of the lung
from deposited particles is clearly more carcinogenic than uniform exposure
to the entire lung. ICRP Publication 14*- ' formally indicates that, based
on general considerations and limited experimental data, localized "hot spot"
doses are probably not as deleterious as average lung doses (based on gram-
rad integral doses).
Sanders, et al., summarized most of the pertinent literature and developed
a model for evaluation of pulmonary carcinogenesis from inhaled radioactive
particulates. They also pointed out that the necessary data for precise
evaluation of the hazards do not exist. They further state, however, that
121
-------
the doses required for an observable tumor incidence are very high, on the
order of several thousand rads if measured in tissue very close to the
particle. There are no data to establish the low incidence end of a dose-
effect curve and extrapolation of high-incidence data would be of ques-
tionable validity. While there is apparently no basis for postulating a
threshold dose, the doses apparently required for observable tumor inci-
dence (several thousand rads) could be delivered only by particles so
large (aerodynamic diameter greater than 10 ym) that their probability
of being respired would be extremely small. '
Several dose calculation approaches have been proposed, notably that of
(791
Dean and Langham, which was based on the observation that tumor inci-
dence exhibits a non-linear response with increasing dose, there being
an optimum tumorigenic dose beyond which tumor incidence decreases because
higher doses simply kill the cells, preventing cell mitosis and tumor for-
mation. Another recent paper by Richmond, Langham, and Stone reports
the results of experiments which indicate that biological response to
doses from highly radioactive particles is definitely influenced by the
number of cells at risk.
APPENDIX B - 4. Model Used for Calculation of Probability of Cataract
from Deposition of Particle in Eye.
Coleman and Perez developed a mathematical model for computing the pro-
bability of cataract production by radioactive particles deposited in the
eye. Use of the model was described by Altomare, et al. Only the
physical dimensions and the basic dose-response relationships will be
described here.
Figure B-l, views a and b, are schematic representations of the eye assumed by
the model. Coleman and Perez selected the values for anatomical con-
stants for the eye given in Table B-l as composites of compromises from
four medical references. In addition, they assumed that the eyelids could
be described by ellipsoidal surfaces with a thickness of approximately 1 mm
at the free margin when closed and that the conjunctiva fornix is completely
symmetrical at a distance of approximately 8 mm from the corneal limbus.
122
-------
OVERALL LENGTH^
(A) SIDE VIEW
ASSUMED LIMIT OF FORNIX
APPROXIMATELY 8mm FROM LIMBUS
SURFACE OF GLOBE OF THE EYE
RADIUS^12.1mm
ANTERIOR SURFACE OF UPPER LID
LENS: RADIUS OF CURVATURE
ANTERIOR«10mm
POSTERIOR«6mm
LENS EPITHELIUM ONE CELL THICK
(ASSUMED 12 MICRONS)
ANTERIOR SURFACE OF CORNEA
RADIUS OF CURVATUREaSmm
ANTERIOR SURFACE OF LOWER LID
CORNEA RADIUS«6.3mm
LENS RADIUS w 4.36mm
ANTERIOR BORDER OF UPPER LID
ANTERIOR BORDER OF LOWER LID
SURFACE OF GLOBE OF EYE
(B) FRONT VIEW
Figure B-l
Schematic Drawings of the Model Eye120
123
-------
TABLE B-l. ANATOMICAL CONSTANTS FOR MODEL EYE(120)
(all dimensions are given in millimeters)
Eye Area
Measurement
Eyeball
Cornea
Lens
anterior-posterior length
radius of curvature
radius of curvature
radius in vertical plane
radius of curvature anterior
surface
radius of curvature posterior
surface
radius in vertical plane
distance anterior cornea to
anterior lens
epithelium thickness, single
cell
25.5
12.1
8.0
6.3
10.0
6.0
4.4
3.5
0.012
124
-------
Apparently the lens epithelium of the lens equator is the location of
most cell mitosis within the lens and is consequently more radiosensitive
than the rest of the lens. Coleman and Perez accordingly incorporated
a weighting function into the model such that the epithelium at the lens
equator is considered to be 10 times more radiosensitive than epithelium at
the lens center. The figure of 10 for the relative radiosensitivity was
based on clinical observations.
A dose-response relationship was assumed such that the probability of
cataract formation with 150 rads uniform dose to the lens is 0.01% (essen-
tially zero) while 620 rads uniform dose to the lens produces a cataract
formation probability of 99.99% (essentially unity). Coleman and Altomare
then performed calculations assuming both normal and log-normal dose-response
curves between these limits and concluded that the log-normal distribution
more accurately represented reality.
Residence time of particles in the eye is another variable of importance to
the eye model. Opthalmologists consulted by Coleman and Perez recommended
use of a residence time of 5 to 15 minutes. The final form of the
model assumed residence times of 4 hours and 24 hours for particles on the
lid, 24 hours in the corner of the eye, and 5 minutes on the limbus.
From the physical model of the eye represented by Figure B-l, views a and b,
Table B-l, and the dose-response relationships outlined above, the probability
of cataract production as a function of total particle activity at the time of
deposition can be computed. Figures B-2 and B-3 ' show the computed pro-
babilities for particles resulting from a 30-minute reactor run and transport
times of 30 minutes and 2 hours.
For purposes of pre-run prediction, the ARL-LV scaling model described by
Henderson has been used with estimated particle fall velocities to
predict air concentrations of particles as a function of distance downwind
from the test point. Particle activity distributions are estimated for
various distances downwind, and the activity distribution is broken into
activity groups or ranges. For a representative activity within each group,
125
-------
UJ
_i
o
h-
cc
<
o.
cc
UJ
Q.
\-
o
•2
<
I-
<
O
LL
O
DQ
<
CD
O
CC
CL
*
/
CORNER
24 MRS
U-
LIMBUS
5 MIN
i I i
10
10-3
10-2
KT1
10'
101
ACTIVITY PER PARTICLE, CURIES
Figure B-2
Probability of Cataract Formation as
a Function of Particle Activity (Based
on 30-minute reactor run and 30-minute
transport time)120
126
-------
t-1 II llfl
KT1
h- LID 24HRS.
LLJ
_l
o
1-
cc
DC
LJJ
Q.
O
<
CL
o
u_
o
DO
<
GO
O
CC
CL
,o2
io3
104
IO5
,o6
10
,r7
I Illllll
1\ i i i mil] | i i HIW
LID 4HRS.
CORNER 24HRS.
LIMBUS 5 MIN. J
SINGLE POINT
ii
iiiil i i mini iimiiil
i 11
io6
io5
fo4
io2
io1
iov
10
ACTIVITY PER PARTICLE, CURIES
Figure B-3
Probability o£ Cataract Formation as a
Function of Particle Activity (Based on
30-minute reactor run and 2-hours trans-
port time)120
127
-------
the probability of cataract formation, Pi, for a particle of that group
deposited in the eye is determined from curves such as Figures B-2 and
B-3. The total probability, P, of cataract formation for a particle at
some location in the eye is then
P = Z P N Eq. B-l
i i i
WHERE
N. is the fraction of the particles in activity group i.
f 881
The NRDL deposition model for particle deposition, assuming a 5 square cm
area for the eye, is used to generate a curve such as Figure B-4, giving the
average number of particles deposited in the eye region per unit integrated
air concentration as a function of wind speed. These results are used to
estimate the average number of particles in the eye region, X , as a func-
tion of distance downwind. Assuming a Poisson distribution, the probability
of one or more particles being deposited in the eye, K, is given by
K = 1 - e~Ae Eq. B-2
The probability of cataract formation, P , is
Pc = (P)(K) Eq. B-3
128
-------
(2 gm/cc density)
CO
ft To2
op
LU
O
H-
DC
LLJ
z
O
a:
LU
o_
z
O
O
LU
DC
LU
>
LU
C/5
LU
_l
u
DC
<
D_
U.
O
DC
LU
CQ
LU
O
<
DC
LU
fo3
fo4
I
I
10 20 30 40 50
WIND SPEED, MPH
Figure B-4
Potential Number of Particles Deposited
In Eye120
129
-------
Figure C-l
Routine TLD Dosimetry Stations
130
-------
"^ XILT
J tJA^Q 1 •BtLLic,
»ovo °o
UTAH f
CANADA
M°NTANA
{ NORTH DAKOTA *
.ITI WYOMING J
SOUTH DAKOTA
• RAPID CITV
NEBRASKA
• NORTnPtAT
So, H •MNVEF,
% I ."«-'<»o '. COLORADO !
*\ „«££&:"•• s-^^^sr10"
«« \ 0"'"»»,s?:V» . .'MONT.CE •-WO I
-CEc^, \ l£^l:s"^ ? ""° I
•
•
W ----- Y
— - \.STJOSEPM V
X V
KANSAS 1 STLOOISI
»
! MSSOURI
'. J0«.~ \"'
*£??•• *\|. .S£IPOA,.N / I 1 OKLAHOMA \ If'"
- •-,»•!"*«*«. . j I .MABIUJ I -"=«">«• \ ARKANSAS /
W **"° f ARIZONA I J I N°RM*N« | "TTLE^ /
5a / "x>e^». I I ^X^ • /
£\— _i I i ^^^ _ —.I I
sa^ "3C f NEW MEXICO i ~" ^ I
^ C^-. ; CA I . I \
• ___1T_S1A1'—| *B"-£~E-
TEXAS
I LOUISIANA V ALE' )N, K|LOMETREs
MEXICO V
Figure C-2
Air Surveillance Network Stations
131
-------
APPEND IX-D
TABLE U-l. REACTOR ENGINE TESTS AT NRDS FROM WHICH AIRBORNE RADIOACTIVITY WAS
DETECTED OUTSIDE THE TEST RANGE COMPLEX(5)*.
Reactor/
Engine
Kiwi A
Kiwi A1
Kiwi A3
Kiwi B-1A
Kiwi B-1B
Kiwi B-4A
Kiwi B-4D
Kiwi B-4E
Kiwi B-4E
NRX-A2
NRX-A2
Kiwi
NRX-A3
NRX-A3
NRX-A3
Phoebus -1A
NRX-A4/EST
NRX-A4/EST
NRX-A4/EST
NRX-A4/EST
NRX-A5
NRX-A5
Phoebus -IB
Phoebus -IB
NRX-A6
Phoebus -2A
Phoebus -2A
Phoebus -2A
Pewee-I
XE Prime
XE Prime
Experimental
Plan
XVI
VII-116-B
VII-216-B
VI/A
IV
VI
IV
V
VI
IV
y***
(TNT)
IV
V
VI***
IV
IIB***
III
IV
IVA
III
IV
III
IV
IIIA
III
IV
V/ASB
III
V/C
IXA
Date
07/01/59
07/08/60
10/19/60
12/07/61
09/01/62
11/30/62
05/13/64
08/28/64
09/10/64
09/24/64
10/15/64
01/12/65
04/23/65
05/20/65
05/28/65
06/25/65
02/03/66
03/03/66
03/16/66
03/25/66
06/08/66
06/23/66
02/10/67
02/23/67
12/15/67
06/08/68
06/26/68
07/18/68
12/04/68
06/11/69
08/28/69
Maximum
Chamber
Tempera-
ture (°R)
—
--
--
--
--
—
4280**
4240**
4000**
3600
—
—
4900
3940
4370**
2576
4100
4000
4150
4000
4100
2900**
4500**
4150
2680**
4060**
3900**
4600**
4200
4200
Nominal
Power
(Mw)
70
85
100
300
800
500
915
914
882
1100
--
1110
1080
1070
442
1140
1100
1200
980
1030
588
1340
1140
1930
4010
3430
503
1070
680
Integrated Power
(106 Mw-sec)
0.02
0.06
0.06
0.03
0.01
0.04
0.11
0.5
0.18
0.3
-0.3
0.009
0.32
0.84
-0.5
0.74
0.88
1.0
1.1
1.2
1.0
0.14
2.6
4.5
0.63
4.5
2.5
1.5
0.42
0.34
* Data for experiments prior to 1964 are not included
unofficial estimates.
** Average maximum fuel temperatures.
*** Data not given in Reference 5.
132
in Reference 5 and are thus
-------
WINNEMUCCA
BAKERSFIELD
ND: Not Detected
TOP NUMBER:
Hypothetical whole-body gamma
BOTTOM NUMBER:
Hypothetical infant thyroid dose in mrad
Figure D-l
Off-Site Whole-Body Gamma Exposures and
Infant Thyroid Doses Resulting from
Reactor/Engine Tests from CY 1959 to 1963 50
133
-------
WINNEMUCCA
100_MIIES
TONOPA
St. GEORGE
90°
LONE PINE/-N
BAKERSFIELD
NDr Not Detected
TOP NUMBER:
Hypothetical whole-body gamma exposure in mR
BOTTOM NUMBER.
Hypothetical infant thyroid dose in mrad
Figure D-2
Off-Site Whole-Body Gamma Exposures and
Infant Thyroid Doses Resulting from
Reactor/Engine Tests During CY 1964 50
134
-------
BEATTY
/
IVLONE RINE
BAKERSFIELD
o BARSTOW
ND: Not Detected
TOP NUMBER:
Hypothetical whole-body gamma exposure
BOTTOM NUMBER
Hypothetical infant thyroid dose in mrad
ISTIMAIEO INFANT INHALATION DOSE
WHl >l BOD" (HUNT OF > f'* PERSONNEL INDICATED 3
Figure D-3
Off-Site Whole-Body Gamma Exposures and
Infant Thyroid Doses Resulting from
Reactor/Engine Tests During CY 1965 50
135
-------
WINNEMUCCA
St. GEORGE
90°
FRESNO ion
V-LONE PINE^-X
VI
LAS VEGAS
BAKERSFIELD
ND: Not Detected
TOP NUMBER:
Hypothetical whole-body gamma exposure in mR
BOTTOM NUMBER:
Hypothetical infant thyroid dose in mrad
Figure D-4
Off-Site Whole-Body Gamma Exposures and
Infant Thyroid Doses Resulting from
Reactor/Engine Tests During CY 1966 50
136
-------
100_MltES 3mrod
TONOPA
ND: Not Detected
TOP NUMBER:
Hypothetical whole-body gamma exposure in mR
BOTTOM NUMBER:
Hypothetical infant thyroid dose in mrad
Figure D-5
Off-Site Whole-Body Gamma Exposures and
Infant Thyroid Doses Resulting from
Reactor/Engine Tests During CY 1967 50
137
-------
WINNEMUCCA
\ I...
BAKERSFIEUD
\
- o BARSTOW
ND: Not Detected
TOP NUMBER.
Hypothetical whole-body gamma exposure in mR
BOTTOM NUMBER
Hypothetical infant thyroid dose in mrad
Figure D-6
Off-Site Whole-Body Gamma Exposures and
Infant Thyroid Doses Resulting from
Reactor/Engine Tests During CY 1968 50
138
-------
WINNEMUCCA
BEATTY
I
LONE PINE
BAKERSFIELD
NO: Not Detected
TOP NUMBER:
Hypothetical whole-body gamma expos
BOTTOM NUMBER.
Hypothetical infant thyroid dose in mrad
Figure D-7
Off-Site Whole-Body Gamma Exposures and
Infant Thyroid Doses Resulting from
Reactor/Engine Tests During CY 1969 50
139
-------
Warm Springs
Vegetation Arc -
Approximate Hotline
(As determined by
vegetation samples)
Tonopah to Coyote Summit |
350
-N-
SCALE IN KILOMETRES
O 5 10
SCALE IN MILES
Phoebus 1B EP IV
Reactor Test
4tO Locations surveyed
for particles. Number
indicates particles found
normalized to 1OO m2.
S/S/S/Ss Vegetation Arcs
Beatty
Highest A
O Sample Result
Yucca A.S.
Test Cell "C"
Figure E-l
Particle Deposition Measurements,
Phoebus 1B-EPIV1*
140
-------
NOTE: Dose rates shown
are net increase above
background.
Isodose Contours
(measured)
Isodose Contours
(interpolated)
Aircraft Flight
Path
.04 MR/HR
O4-.O8 MR/HR
.08-.12 MR/HR
.12-.2O MR/HR
.2O-.28 MR/HR
.2S-.4 MR/HR
.4-1 MR/HR
1-2 MR/HR
2-4 MR/HR
4-5 MR/HR
Figure E-2
Reactor Test Ground Depositon Pattern
for Phoebus IB, EPIV
(Data used by permission from
141
-------
KJ
(W
C
*i
(D
W
i
0.20
232° 230° 228° 226° 224° 222° 220° 218° 216° 214° 212° 210° 208° 206° 204° 202°
Three-Dimensional Representation of
Particle Survey Results for NRX-A6, EPIIIA109
-------
c
H
CD
ta
i
at
0)
O
D.
Z
O
U
Z
O
u
u
15 MILES
23 MILES
40 <
Three-Dimensional Representation of
Particle Survey Results for
Phoebus 2A, EPIV8"
-------
H-
OP
c
H
CD
W
i
Cn
CN
E
^s
M
_«
W
w
o
a.
O
U
z
o
u
0.2
15 Miles
25 Miles
30° 28° 26° 24° 22° 20° 18° 16° 14° 12° 10° 8°
AZIMUTH FROM FROM TEST CELL 'C
Three-Dimensional Representation of
Particle Survey Results for
Phoebus 2A, EPV8<*
-------
APPENDIX-F
TABLE F-l. MAXIMUM PARTICLE CONCENTRATIONS AT INDICATED DISTANCES
Reactor Run Arc Distance (miles)/Particles/m2
Phoebus-IB, EP-IV 2.5/2.4 5.0/3.3 15.0/1.1 25.0/1.4 40.0/1.2
NRX-A6, EP-III 1.5/0.6 3.0/0.9 6.0/0.9 10.0/1.4 25.0/0.9 40.0/0.2
Phoebus-2A, EP-IV 1.5/0.3 3.0/0.4 7.5/1.3 15.0/0.3 23.0/0.2 40.0/0.2
Phoebus-2A, EP-V 1.5/0.6 2.5/0.4 6.0/0.6 15.0/1.1 25.0/0.2 50.0/0.1
Pewee-I, EP-III 1.5/8.5 3.0/3.6 5.0/3.8 10.0/1.7 20.0/1.1 60.0/0.2
145
-------
APPENDIX-F
TABLE F-2. NORMALIZATION PARAMETERS
Reactor Test
Source
Q
(grams)
Cloud height
h
(feet MSL)
Mean Layer Wind
u
(knots)
Hodograph Shear
a
(degrees)
Phoebus-IB, EP-IV 450
NRX-A6, EP-III 210
Phoebus-2A, EP-IV 70
Phoebus-2A, EP-V 55
Pewee-I, EP-III 103
13,000
9,500
18,000
13,500
11,500
10
25
5
8
8
25
10
20
25
10
146
-------
DISTRIBUTION
1-15 National Environmental Research Center, Las Vegas, NV
16 Mahlon E. Gates, Manager, AEC/NV, Las Vegas, NV
17 Robert H. Thalgott, AEC/NV, Las Vegas, NV
18 David G. Jackson, AEC/NV, Las Vegas, NV
19 Arthur J. Whitman, AEC/NV, Las Vegas, NV
20 Robert R. Loux, AEC/NV, Las Vegas, NV
21 Mail g Records, AEC/NV, Las Vegas, NV
22 Technical Library, AEC/NV, Las Vegas, NV
23 Chief, NOB/DNA, AEC/NV, Las Vegas, NV
24 P. J. Mudra, AEC/NV, Las Vegas, NV
25 B. W. Church, AEC/NV, Las Vegas, NV
26 Harold F. Mueller, ARL/NOAA, AEC/NV, Las Vegas, NV
27 Howard G. Booth, ARL/NOAA, AEC/NV, Las Vegas, NV
28 D. T. Schueler, AEC/NV, Las Vegas, NV
29 C. P. Bromley, AEC/NV, Las Vegas, NV
30 K. M. Oswald, LLL, Mercury, NV
31 James E. Carothers, LLL, Livermore, CA
32 Ernest A. Bryant, LASL, Los Alamos, NM
33 Harry S. Jordan, LASL, Los Alamos, NM
34 Charles I. Browne, LASL, Los Alamos, NM
35 Jerome E. Dummer, LASL, Los Alamos, NM
36 Eastern Environmental Radiation Facility, EPA,
Montgomery, AL
37 Donald R. Martin, AEC/NV, Las Vegas, NV
38 Martin R. Biles, DOS, USAEC, Washington, DC
39 F. K. Pittman, WMT, AEC, Washington, DC
40 J. Doyle, EG?TG, Las Vegas, NV
41 Richard S. Davidson, Battelle Memorial Institute,
Columbus, OH
42 Carter D. Broyles, Sandia Laboratories, Albuquerque, NM
-------
43 Maj. Gen. Ernest Graves, AGMMA, USAEC, Washington, DC
44 Albert C. Trakowski, Act. Ass't Administrator for
Research $ Development, EPA, Washington, DC
45 William D. Rowe, Deputy Assistant Administrator for
Radiation Programs, EPA, Washington, DC
46 Ernest D. Harward, Act. Dir. of Technology Assessment,
Office of Radiation Programs, EPA, Washington, DC
47 - 48 Charles L. Weaver, Dir., Field Operations Div., Office
of Radiation Programs, EPA, Washington, DC
49 Gordon Everett, Dir., Office of Technical Analysis,
EPA, Washington, DC
50 Library, EPA, Washington, DC
51 Kurt L. Feldmann, Managing Editor, Radiation Data §
Reports, Office of Radiation Programs, EPA, Washington, DC
52 Regional Radiation Representative, EPA, Region IX,
San Francisco, CA
53 Arden E. Bicker, REECo, Mercury, NV
54 John M. Ward, President, Desert Research Institute,
University of Nevada, Reno
55 - 56 Technical Information Center, USAEC, Oak Ridge, TN
(for public availability)
------- |