SWRHL-96
PHEP-1
PUBLIC HEALTH EVALUATION
PROJECT RULISON
(PRODUCTION TESTING)
, iON AGENCY
P. 0 BOX 15027
IAS VEGAS, IM39114
Roy B. Evans
David E. Bernhardt
Programs and Plans
Southwestern Radiological Health Laboratory
U.S. Department of Health, Education and Welfare
Public Health Service
Environmental Health Service
Environmental Control Administration
Bureau of Radiological Health
May 1970
-------
DEPARTMENT OF HEALTH. EDUCATION. AND WELFARE
PUBLIC HEALTH SERVICE
June 17, 1970
NAOlOLOOlCAi. HCAl TM L*MHATORV
r.6. 00* lion
M VtOA». NfVAQA §»t!4
Our Reiference: PPCrDEB
To The Distribution:
The following corrections should be made in SWRHL-96, "Public Health
Evaluation Project Rullson (Produption Testing):"
1. Page iv, Item 2, line 3 - The line which was partlajty
eradicated should read "... are tritium and krypton-85. The
average airborne concentra- ..." 4
2. Page 8, Figure 2 - Insert the following information for the
-2-! to 4-mile sector.
40° 03
-------
SWRHL-96
PHEP-1
PUBLIC HEALTH EVALUATION
PROJECT RULISON
(PRODUCTION TESTING)
by
Roy B. Evans
David E. Bernhardt
Programs and Plans
Southwestern Radiological Health Laboratory
U.S. Department of Health, Education and Welfare
Public Health Service
Environmental Health Service
Environmental Control Administration
Bureau of Radiological Health
Drafted - October 1969
Published - May 1970
-------
ABSTRACT
Project Rulison is a Plowshare experiment to investigate
the feasibility of nuclear explosive stimulation of natural
gas production. The detonation of the explosive took place
on September 10, 1969. Production testing activities will
be initiated six months or more after the detonation and will
entail flaring of natural gas containing radioactivity. The
radionuclides of primary interest which will be released by
production testing are tritium and krypton-85.
This report presents an analysis of the public health impli-
cations of the radioactivity releases associated with Project
Rulison production testing. Concentrations and possible
movement of radionuclides in ground water near the chimney
are estimated, and potential human doses from radioactivity
released to the atmosphere during production testing are
postulated.
The analysis presented indicates that Project Rulison produc-
tion testing operations can be conducted well within the
radiological safety guides of the Federal Radiation Council
and Atomic Energy Commission. The postulated dose to members
of the public is on the order of one-tenth mrem.
-------
FOREWORD
PLOWSHARE
The mission of the Plowshare program is to develop technology
for the peaceful use of nuclear explosives for the benefit of
man. Plowshare is attempting to obtain additional return
from national resources expended in developing nuclear weapons,
Benefits can be derived from the peaceful application of
nuclear explosives, but risks from radioactivity and other
detonation effects are inherently associated with such appli-
cations. Nuclear explosives, like any other engineering tool,
must be shown to yield benefits that outweigh the costs and
r i sk s involve d.
PUBLIC HEALTH EVALUATION OF PLOWSHARE PROJECTS
In July of 1969, a project was initiated at the Southwestern
Radiological Health Laboratory (SWRHL) in Las Vegas, Nevada,
to evaluate public health aspects of peaceful applications
of nuclear explosives. The SWRHL, under a Memorandum of
Understanding between the Atomic Energy Commission (AEC) and
the U. S. Public Health Service (PHS), conducts the off-site
radiological safety program for nuclear tests. Although
this project to evaluate health aspects of peaceful appli-
cations of nuclear explosives uses competency developed
through work performed under the AEC-Memorandum of Under-
standing, the project is separate from that endeavor and is
solely a PHS function.
11
-------
The purpose of this, effort is to identify and evaluate the
radiological health implications of Plowshare projects. The
intent is to review and summarize the AEG safety evaluations
and to supplement them if necessary. If such AEC evaluations
are not available on a timely basis, independent evaluations
will be performed using preliminary information.
The AEC safety evaluation, NVO-61 (38), "Project Rulison
Post-Shot Plans and Evaluations," was not released until
December 1969. This report was written prior to release of
NVO-61. Pertinent parameters from NVO-61 are summarized in
Appendix H for comparison with those used in this report.
Differences between parameters used in NVO-61 and parameters
used here do not change the conclusions of this report.
111
-------
SUMMARY AND CONCLUSIONS
This report uses preliminary information available to the
public to analyze the public health implications of Project
Rulison. Concentrations and possible movement of radio-
nuclides in ground water near the chimney are estimated, and
environmental levels of radionuclides and doses to humans
resulting from radioactivity released to the atmosphere from
production testing are hypothesized.
The following summarizes the estimated environmental effects:
1. Contaminated ground water is not expected to migrate away
from the chimney during gas production since gas inflow
will tend to inhibit radionuclide migration. When move-
ment occurs, it will be so slow as to allow decay to
concentrations below the radiation protection guides (39)
within a fraction of a mile.
2. The only radionuclides expected to be present in detect-
able quantities in flared gas during production testing
are tritium and X^ypy#oJ <$£, ~JH jW&fr(*6 ftiRJbtW^ CA^C&J&J
tions are expected to be three or more orders of magni-
tude below the concentration guides for the general popu-
lation at the nearest populated locations (three miles
from the site).
3, Tritium concentrations in vegetation moisture at locations
a few miles away may range up to several hundred pCi/ml of
water and concentrations in milk to several tens of pCi/ml
IV
-------
The postulated dose to members of the public is on the order
of one-tenth millirem. The doses resulting from ingestion
of foodstuffs (milk, vegetables, etc.) may be an order of
magnitude or more above doses from inhalation or drinking
water. The chronic dose, due to the residual tritium in
the environment after the peak production testing phase,
is estimated to be an order of magnitude greater than the
acute dose during the three-week test period.
The analysis presented in this paper indicates that the Project
Rulison production testing operation can be conducted well with-
in the radiological guides of the FRC and the AEC standards.
In addition, the environmental surveillance program proposed
by the SWRHL should be capable of detecting levels of environ-
mental radioactivity well below the guides. This surveillance
program will provide information to initiate necessary pro-
tective actions should unexpected circumstances occur. The
use of surveillance data to detect the need for protective
action will require timely analysis and reporting of results.
-------
TABLE OF CONTENTS
Page No.
Abstract i
Foreword ii
Summary and Conclusions iv
Table of Contents vi
List of Appendices vii
List of Figures and Tables viii
I. Introduction 1
II. Report Objectives 4
III. Rulison Environment 6
IV. Production Testing 10
V. Environmental Radioactivity 13
Radioactivity in Ground Water 15
Biosphere Concentrations 15
Tritium 17
Krypton-85 21
Total Dose - Acute Plus Chronic 23
VI. Environmental Surveillance 27
References R-l
vi
-------
LIST OF APPENDICES
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Appendix F
Appendix G
Appendix H
Appendix I
Major Basins of the Rocky Mountain States
Gas Deliverability
Radioactivity in Gas
Radioactivity in Ground Water
Meteorology and Diffusion
Tritium Dose Calculation
Krypton-85
Summary of Parameters from NVO-61
Acute and Chronic Doses from Ingestion of Tritium
VII
-------
LIST OF FIGURES
Page No.
Figure 1 Index Map of Project Rulison Site
Figure 2 Rulison Population Summary
Figure 3 Rulison Site Wind Regimes and
Elevations
7
8
9
LIST OF TABLES
Table 1
Rulison Dose Estimates
25
Vlll
-------
I. INTRODUCTION
Project Rulison is the second of a series of Plowshare exper-
iments to investigate the feasibility and develop the tech-
nology for nuclear explosive stimulation of natural gas
production. Several gas fields in the U. S. contain large
reserves which cannot be economically recovered with conven-
tional we11-completion technology because the gas-bearing
formations are of relatively low permeability. Several
methods have been used in attempts to stimulate gas pro-
duction in these low permeability regions, including hydraulic
fracturing, fracturing with conventional explosives, and the
use of chemicals to increase formation permeability. These
stimulation techniques have met with varying degrees of
success and have contributed significantly to usable national
reserves of natural gas. However, gas in a number of reser-
voirs in Colorado and neighboring states (see Appendix A)
cannot be recovered economically with conventional stimula-
tion techniques, and nuclear stimulation is being investi-
gated for possible use in these reservoirs.
Nuclear stimulation, like stimulation with conventional explo-
sives, is intended to increase the effective well diameter
by fracturing the gas-bearing formation in the vicinity of
the drill hole. Since a nuclear device is capable of releas-
ing much more energy than a conventional device of comparable
physical size, the nuclear explosive can produce a much larger
effective well radius than is obtainable with a conventional
explosive. Gas deliverability should increase as the effec-
tive well radius increases.
-------
Concepts involved in the engineering use of nuclear explo-
sives are largely based on experience with underground
nuclear testing at the Nevada Test Site. Safety programs
for this testing have been described in references 1, 2, and
3.
In Project Gasbuggy, the first nuclear gas stimulation exper-
iment, a 29-kt explosive was detonated 4,240 feet below the
surface of the Carson National Forest in New Mexico on
December 10, 1967t The objective of this joint government-
industry experiment (sponsors were the Atomic Energy Commis-
sion, Department of the Interior, and the El Paso Natural
Gas Company) was to examine the general feasibility of the
technique and to provide an indication of possible problems
which might result from the radioactivity in the gas. The
device used was too small to produce a commercially profit-
able well in the chosen medium and the experiment was intended
only as a feasibility study (5, 12).
Project Rulison, a joint venture of the Atomic Energy Commis-
sion, Austral Oil Company, and the Department of the Interior,
was detonated on September 10, 1969. The project can be
thought of as a logical successor to Gasbuggy. A higher
yield device (40 kt) was used in a more promising gas field,
and if gas produced from the stimulated well were meant for
sale, the project would be commercially more attractive than
Gasbuggy. There have been no announcements of plans for
consumer use of gas from the Project Rulison well.
The first phase of Project Rulison consisted of exploding the
device to fracture the gas-bearing Mesaverde Formation. The
*Memo: Fred Holzer (LRL) to Distribution, 1/5/70; Subject:
"Yield of Gasbuggy Experiment."
-------
second phase of the experiment will begin about six
months after the detonation. This delay will allow decay
of most of the radioactive fission and activation products.
At this time the cavity will be re-entered by "drill back"
operations, and a period of gas analysis and controlled pro-
duction testing will begin. The primary objective of produc-
tion testing is to determine the degree of stimulation
achieved. Another objective is to determine the fate of the
radioactivity remaining from the nuclear detonation, the
fraction available for release, and the rate of release. A
subsequent phase, not planned for Rulison would involve dis-
tribution of the natural gas and consumer use of the gas
after possible processing and/or mixing with gas from non-
nuclear wells. The Public Health Service and several other
agencies are investigating possible implications to the
general population from the use of such gas.
-------
II. REPORT OBJECTIVES
This report presents an analysis of the public health impli-
cations of radioactivity produced by the Rulison explosion,
both that introduced into the hydrologic environment by the
explosion and that introduced into the biosphere by flaring
gas during production testing. The report includes a brief
description of proposed production testing operations and a
review of existing plans for an environmental surveillance
program for production testing. It is hoped that the report
will provide information to assist officials in state and
local health departments and other agencies in planning for
public safety operations related to production testing.
Much of the information necessary for predicting environ-
mental effects was not available to the public when this
report was written, and many of the predictions in this
report are based on preliminary data or assumptions.
The radionuclides of primary interest from production test-
ing operations are tritium and krypton-85. Strontium-90 and
cesium-137 were also considered in the ground water analysis,
Low levels of radionuclides such as carbon-14, argon-37 and
argon-39 may also be present in the flared gas, but the quan-
tities produced and their biological significance are such
that they are not considered to be of primary concern to ' '
public health.
The expectation that concentrations of radionuclides other
than tritium and krypton-85 will be small is largely based
on experience from Gasbuggy. The reasons for the absence of
-------
various radionuclides from the natural gas include ill, 30, 31):
A. Non-gaseous radionuclides are essentially trapped in
the melt at the bottom of the cavity.
B. The six-month delay prior to cavity re-entry allows
for decay of the short-lived radionuclides.
However, Gasbuggy data are based on non-detection in samples
that were generally collected for purposes other than quan-
titating fission and activation products. Thus, while there
is reasonable certainty that concentrations of fission and
activation products such as Cs and Sr will be very
small compared to concentrations of tritium and krypton and
will consequently be of little significance to public health,
the possibility of their presence should not be ignored.
Samples of the chimney gas should be analyzed for fission
and activation products and the environmental surveillance
program should not assume their absence. Because of the very
small chance of the presence of fission and activation prod-
85
ucts, other than tritium and Kr, tin
further consideration in this report.
85
ucts, other than tritium and Kr, they will not be given
The likelihood of a massive well blow-out during redrilling
is very low, but a very small possibility does exist. The
consequences of such an event will not be directly evaluated
in this report. However, the postulated release during nor-
pc
mal production testing operations amounts to 70% of the Kr
and tritium in the cavity and the AEC has prepared dose esti-
mates for a "maximum hypothetical accident" (38), which is
summarized in Appendix H. The environmental surveillance
program and the off-site safety program must be predicated
on the possibility of such an event.
-------
III. RULISON ENVIRONMENT
Figure 1 indicates the geographical location of the Rulison
site. Figure 2 is a map giving the population distribution
of the section of western Colorado surrounding
and centered on the Rulison site. Morrisania Mesa, the
populated area closest to the site, is located about three
miles northwest of surface ground zero (SGZ). Grand Valley,
a town of about 300 residents, is about six miles to the
northwest. Morrisania Mesa is primarily an agricultural
community. Agricultural production includes several types
of fruit, sheep, beef cattle, and small quantities of milk.
Important terrain features are indicated schematically in
Figure 3. Elevations on the figure are given in thousands
of feet above mean sea level (IK = 1,000 feet). Movement of
air away from Rulison SGZ is primarily controlled by three
wind regimes. Valley drainage winds and daily upslope winds
in both the Battlement Creek Valley and the Colorado River
Valley comprise two separate wind regimes. Regional gradient
winds, the third regime, blow generally to the east-north-
east above the topographical features throughout the year.
The meteorology is described in greater detail in Appendix
E.
-------
R.96W.
R.95W.
R.94W.
R.93W.
DBRGW RR
U.S.HWY.6S24
GRAND
VALLEY
•VORULISON
SITE
GAR£IELD COUNT
"MESA COUNTY
••'.DENVER |
FIGURE 1
! OGRAND
| JCT.
i
L_.
OPUEBLO
INDEX MAP OF PROJECT RULISON SITE
(from reference 25)
-------
RUUSON POPULATION SUMMARY
000 Adulli
00 Children
0 Cows
FIGURE 2
Rulison Population Summary
* Blank Sectors Indicate No Populat
8
-------
FIGURE 3 Rulison Area Wind Regimes
Roan Cliffs (approx. 9OOO feet msl)
nighttime flow daytime flow
Valley tlow occurs
about 50% of the time.
Colorado River Valley (approx. 50OO feet msl)
Average daytime velocities are 10 • 15 mph;
average nighttime velocities are about 7 mph.
Morrisania Mesa (approx. 6OOO feet msl)
elevation approx. 7000feet msl
elevation approx. 8000 feet
Battlement Creek Valley
\nighttime fl
.
\ \ *
t
t
t
NORTH
flo
Regional flow above the topographical features is toward ENE
about 50 % of the time (67.5° sector). Average wind speed
for regional flow is about 15mph.
Valley flow
Average daytime
5 . 10 mph.
SGZ (approx. 82OO feet r
\ v,
i v/.
occurs about 70% of time. W/>
yitime and nighttime velocities ^K/4
nph. m/>
I V,
Battlement Mesa (approx. 110OO feet msl)
-------
IV. PRODUCTION TESTING
To estimate the ability of the Rulison chimney-fracture system
to deliver gas to a pipeline, production tests will be performed,
Gas will be released from the wellhead under controlled con-
ditions, and the relationship between the formation pressure,
the bottom-hole pressure in the well, and the flowrate of gas
will be used to estimate the stabilized deliverability of the
well. Production testing has the additional purpose of purg-
ing the chimney of radioactivity.
Project Rulison was intended to stimulate gas production
by creating a large effective well radius. For nuclear
stimulation, the effective well radius is approximately the
radius of intense dynamic fracturing. Equation B-l in
Appendix B is a simplified expression for the radial flow of
gas into a well under steady-state conditions. Examination
of the equation indicates that the gas deliverability should
increase as the effective well diameter increases. Pre-shot
predictions of geologic effects from Project Rulison indi-
cated that a spherical cavity with a radius of approximately
80 feet would be formed initially, followed by cavity collapse
to form a rubble-filled chimney approximately 370 feet high.
The rock surrounding the chimney would be fractured
to a radius of about 370 feet from the detonation point (20).
Available information on the reservoir characteristics of
the Rulison field and the predicted geologic effects of the
detonation are used in Appendix B to estimate the stabilized
deliverability of the chimney-fracture system from Equation
B-l. Use of the equation requires the assumptions that the
10
-------
lateral extent of the paying sandstone lenses penetrated by
the chimney-fracture system is large compared to the hori-
zontal extent of intense dynamic fracturing and that these
sandstone lenses are of uniform thickness. The assumption
must also be made that the reservoir parameters measured.in
the exploratory hole R-EX (285 feet southeast of the emplace-
ment hole) are characteristic of the paying sandstone lenses
for several hundred feet in all horizontal directions from
the detonation point. Data reported by the USGS indicate
that these assumptions are reasonable (11).
Based on available information and the above assumptions,
the stabilized absolute open flow* deliverability of the
chimney-fracture system is estimated to be approximately
a
,6
c
5.5 x 10 cubic feet per day, NTP**; the maximum possible
stabilized open flow deliverability is estimated at 15 x 10
ft /day, NTP.
Wells are not commonly tested by open flow measurements,
however, and the maximum flaring rate will be different from
expected stabilized open flow deliverability. Initial flar-
ing rates can be higher than the stabilized open flow rate,
predicted by Equation B-l, since the volume of gas contained
in the chimney void space is large compared to the flaring
rate. Initially at least, flaring gas from the chimney will
be analogous to releasing gas from a large tank.
Gasbuggy created a chimney with a void volume of approxi-
c.
mately 2.1 x 10 cubic feet. After chimney pressure equili-
brated with formation pressure, the volume of gas in place
g
in the chimney was approximately 1.2 x 10 cubic feet
*The term "absolute open flow" is defined in Appendix B.
**NTP is an abbreviation for "normal temperature and pressure,
60°F and 14.7 psia.
11
-------
(corrected to NTP) (12); the initial stabilized open flow
deliverability was estimated from production testing to be
6
2.8 x 10 cubic feet per day (NTP) (4). This was estimated
by measuring the rate of decline of bottom-hole pressure
fi o
while flaring at 5 x 10 ft /day (NTP), measuring the subse-
quent increase in bottom-hole pressure while flaring at 0.75
6 3
x 10 ft /day, and using a straight-line interpolation between
the two observations to predict the flow-rate at which bottom-
hole pressure would remain constant. A back-pressure curve
was then used to calculate the equivalent open-flow rate (4).
Plans for production testing of Rulison have not yet been
announced, but the assumption that Rulison testing will
follow the pattern adopted for Gasbuggy seems reasonable*
This would indicate an initial flaring rate of twice the open
flow deliverability, or an expected rate of approximately 11
63 63
x 10 ft /day with a possible maximum rate of 30 x 10 ft /day.
Tables of the quantities of radionuclides produced by the
Rulison explosion and calculations for predicting the result-
85
ing concentrations of tritium and Kr in the natural gas
are given in Appendix C. The concentrations are based on
the assumption that radioactivity produced by the explosion
will be diluted in the gas contained in the chimney. The
85
Kr concentration in natural gas is estimated to be about
-4 3
1.4 x 10 u.Ci/cm when production testing is initiated. The
tritium concentration in natural gas at the initiation of
-4
production testing is expected to range from 0.7 x 10.
3 -4 3
U.Ci/cm to 15 x 10 [iCi/cm based on the calculations of
Appendix C.
*Plans for production testing have been announced (reference 38,
see Appendix H) but do not affect the conclusions of this section,
12
-------
V. ENVIRONMENTAL RADIOACTIVITY
To evaluate the implications to man of releasing radioactiv-
ity to the environment, the resulting contamination to the
total environment must be examined and the subsequent trans-
port of the pollutants to man and associated radiation dose
to man must be estimated. The general approach used in anal-
ysis of environmental releases and their impact on man
corresponds to the diagram below:
SOURCE
TERM
ENVIRONMENTAL TRANSPORT
(including dispersion)
INTERACTION OF EFFLUENT WITH VARIOUS
PHASES OF THE ENVIRONMENT (e.g., DEPOSITION
AND UPTAKE INTO FOOD CHAINS)
1
INTAKE BY MAN
OF ENVIRONMENTAL/MEDIA
PRODUCT/SUMMATION OF ABOVE PARAMETERS
(external, air, water, food)
Dose—Relate to guide value
13
-------
To assess the health implications, the estimated dose must
be compared with a public health guide (e.g., radiation pro-
tection guides).* The evaluation must include a summation
of the various related doses (e.g., all doses to a particular
organ). For instance, possible pathways for tritium include:
1. inhalation and uptake through the skin from air,
2. ingestion of contaminated food products (vegetation,
milk, etc.), and
3. ingestion of water.
Doses from all pathways must be summed to determine the dose
to the organ of interest.
Radioactivity in the environment is considered in two sections:
1. Concentrations and movement of radionuclides in ground water.
2. Environmental concentrations resulting from surface releases,
*The philosophy of the PHS group for the evaluation of Plow-
share projects is to use the radiation protection guidelines
of the Federal Radiation Council (37). The FRC recommenda-
tions are generally given in terms of dose guides for partic-
ular organs rather than as concentration guides for specific
radionuclides in particular compartments of the environment
(milk is an exception where the FRC has given concentration
guides). Where concentration guides are needed, the
standards in AEC Manual,. Chapter 0524, are referenced.
The recommendations of Chapter 0524 are essentially the same
as those of Title 10, Code of Federal Regulations, Part 20.
Both generally follow the recommendations of the FRC. AEC
Manual Chapter 0524 is applicable to projects conducted
under contracts or agreements with the AEC, and 10 CFR 20
applies to operations conducted under a license arrangement.
14
-------
RADIOACTIVITY IN GROUND WATER
Isotopes, Inc., an AEG safety contractor, predicted concen-
trations and movement of radionuclides in water for the
Rulison detonation and concluded that the occurrence of
levels of radioactivity exceeding the applicable concentra-
tions guides was not likely at any use point (15). The
nearest use point for ground water is the rural community
of Morrisania Mesa, approximately three miles from the deto-
nation point, where the major sources of water for domestic
use are shallow wells penetrating alluvium (11). Morrisania
Mesa lies approximately 6,000 feet above the detonation
point, and almost the entire thicknesses of the Mesaverde
and Wasatch Formations separate the water-bearing alluvium
from the Rulison chimney. Movement of radionuclides from
the chimney to the alluvium is very improbable. Movement
of radionuclides to the Colorado River, which lies about
5,000 feet above the chimney and at least five miles distant,
is also unlikely.
While gas is being released from the well, hydraulic grad-
ients in the formation surrounding the chimney will drive
water toward the chimney because chimney pressure will be
lower than the pore pressure of the formation. Until gas
production from the well ceases, no migration of radio-
activity away from the chimney is expected.
The analysis given in Appendix D indicates that the trans-
port of radioactivity by ground water from the Rulison chim-
ney to any ground water use point is extremely unlikely.
BIOSPHERE CONCENTRATIONS
The scope of this section is limited to consideration of trit-
D C
ium and Kr. Other radionuclides which may be released
15
-------
14 37 39 '
include C, A, and A. However, because of the com-
bination of the small concentrations of these nuclides ex-
pected to be present in the gas and their relative biolog-
ical significance compared to ti
nuclides will not be considered.
85
ical significance compared to tritium and Kr, these three
The transport of airborne effluent from the Rulison site is
affected by three wind regimes; the Battlement Creek drain-
age and upslope winds and the upper level or regional grad-
ient winds. The Battlement Creek nighttime drainage winds
flow into the Colorado River Valley drainage wind regime.
The effective stack height or plume rise will determine
what fraction of the effluent is transported in each of these
regimes. Based on the information in Section III and Appen-
dix E, it is assumed that the majority of the effluent will
rise above the local terrain to be transported by the regional
gradient winds. Prevailing direction of the gradient winds
is toward the ENE. It is logical to assume that a fraction
of the effluent will also be entrained in the Battlement
Creek Valley flow (see Appendix E). This would result in
the shortest possible transport distance (about three miles)
to a populated location.
The many assumptions which must be made in estimating the
doses to humans which might result from Rulison production
testing make difficult any truly accurate estimates. Esti-
mates developed in the following section should be regarded
as rough dosage approximations. The discussion outlines the
possible pathways for doses to humans from 'radioactivity
released during production testing. The object is to esti-
mate doses which might be expected to occur rather than
maximum credible doses.
16
-------
TRITIUM
Downwind surface concentrations of tritium can be estimated
3
by assuming a flaring rate of 2 x 10 ft /day (NTP) and a
tritium concentration in the chimney gas of about 42 p,Ci/ft"
—3 3
(1.5 x 10 [iCi/cm ; see Appendix C), equivalent to an activ-
3
ity release term of about 9.7 x 10 (iCi/sec. Assuming a con-
stant dispersion coefficient for the range from two to ten
miles downwind, as discussed in Appendix E, the expected
average tritium concentration in this distance range would
—11 3
then be on the order of 3 x 10 |iCi/cm of air (short term
concentrations may be up to several orders of magnitude above
this).* The figure of 3xlO~ jiCi/cm is several thousand times
—8 3
less than the concentration guide of 7 x 10~ (j.Ci/cm recom-
mended by AEG MC-0524 (39) for a suitable sample of the popu-
lation (individual guides divided by 3 as suggested by FRC).
The figure of 3 x 10~ (iCi/cm is the estimated average
concentration over a period of days to weeks and is based on
an assumed continuous release at the maximum expected flaring;
rate and tritium concentration. The initial tritium concen-
tration in the gas could well be as little as five percent
of the average value assumed and will decrease with the
volume of gas flared.
An adult breathing air containing a tritium concentration of
3 x 10~ p,Ci/cm for 24 hours a day for 10 days would
7
receive an inhalation dose of 6 x 10 rem for an adult.**
*The nearest population to the site is three miles. The
-9 3
dispersion coefficient of 3 x 10 sec/m is based on cross-
wind averaging in the predominant wind sector.
**The present proposals involve flaring three days and then
shutting in the well for about a week. Three high volume
flaring periods are contemplated (9 days of flaring over approx-
imately a three-week period; considered to be 10 days of flaring
to include initial tests).
17
-------
This is 3 x 10 times less than the FRC guide of 0.17 rem
per year. If it is assumed that a person receives an equiv-
alent dose from tritium uptake through the skin (29,32) then
the dose would be about 1.2 x 10~ rem. The inhalation dose
for an infant is about half of that for an adult and the
combined inhalation plus skin dose is about 40% of that for
an adult.* It is unlikely that the exposure from inhalation
will approach a significant fraction of the guides.
The calculation for dose per |j,Ci intake of tritium is indi-
cated in Appendix F.
The following discussion develops an estimate of potential
tritium uptake through food and water:
Following the November 1968 production testing of Gas-
buggy, the tritium concentration in vegetation ranged
3
* Infant breathing rate 5 m /day (41) 3.1 day t ff infant
3 X £££ X
Adult breathing rate 20 m /day (33) 10 day t adult
err,
Adult body weight 70 kg (33)
Infant body weight 10 kg (40)
= 0.5
To estimate the relative uptake through the skin, assume
the relative uptake is proportional to the skin area and
that the skin area can be related to the body mass to the
2/3 power (mass is a function of length cubed, whereas,
area is a function of length squared). Therefore, the
2/3
tritium uptake through the skin of an infant is (10/70) '
= 0.27 times that of an adult.
18
-------
up to 36 pCi/ml of moisture.* Based on the proposed
flaring parameters for Rulison, it is estimated that
tritium concentrations in vegetation moisture in the
Rulison area may range up to an order of magnitude
higher. Assuming that a person daily consumes 220 grams
of vegetation grown at the point of maximum concentration
and that the average vegetation moisture content is 0.9
ml/g, a person would ingest about 0.07 p,Ci of tritium
per day. Ingestion at this level corresponds to a dose
for an adu,
ingestion.
for an adult of about 7 x 10~ rem for each day of
Concentrations of tritium in milk would be roughly an
order of magnitude less than concentrations in vege-
tation moisture because of dilution by the cows' drink-
ing water.** Thus, a person drinking one liter of milk
per day would ingest about 0.04 |j,Ci of tritium per day.
The estimated long-term average airborne tritium concen-
tration is 3 x 10~ (iCi/cm , whereas, the short-term
air concentrations (minutes to hours) may be several
7 3
*In June and July of 1968, approximately 6 x 10 ft (NTP)
3
of gas with a tritium concentration between 10-17 |iCi/ft
(concentration decreased with time) was flared. The well-
head was then shut in until November, when flaring began at
a flowrate of 3.5 x 10 ft /day, with a concentration of
3
10 nCi/ft (28). One vegetation sample contained 36 pCi/ml
of tritium in moisture; concentrations in all other samples
were less than half of this value.
**This assumes a cow receives only 0.1 of its water from
vegetation.
19
-------
orders of magnitude higher.* The long-term average con-
centration corresponds to a specific activity of tritium
in condensed atmospheric moisture of about 10 pCi/ml.**
The short-term average air concentration corresponds to
a concentration of 10 pCi/ml in moisture. Thus, the
previously estimated concentration of 360 pCi/ml of plant
moisture does not appear unreasonable.
It is difficult to conceive of a reasonable mechanism
whereby large volumes of water used for human consump-
tion could become contaminated to levels equal to the
specific activity of the tritium in atmospheric moisture.
If the water contained 1/10 of the average specific
activity, a person drinking 2,200 ml/day would ingest
about 0.002 |iCi per day of tritium.
The estimated tritium intake through inhalation and skin
absorption is about 0.001 p,Ci per day assuming an average
tritium concentration in air of 3 x 10~ [iCi/cm .
*Using Pasquill D conditions, 7 mph wind speed, and a ground
level release, the short term concentration at 3 miles (over
a period of minutes) at the nearest downwind populated
—8 3
location would be about 10~ p,Ci/cm ,
**Assuming temperature of 50° F, relative humidity of 40%,
and barometric pressure of 750 mb, and that all tritium'is
present as HTO.
Battlement Creek is used to periodically fill cisterns used
for potable water. The SWRHL surveillance program (Section.
VI) includes appropriate sampling. Should the creek become
contaminated by rain scavenging, appropriate action can be
taken.
20
-------
The total tritium intake (for an adult) from the previously
mentioned pathways is:
nCi/daya
Inhalation and skin 1
Vegetation 70
Milk 40
Waterb 2
This is equivalent to a dose of about 10~ rem per day of
exposure for an adult and about the same or slightly less
Q
for an infant .
KRYPTON-85
7 3
Assuming a flaring flowrate of 2 x 10 ft /day (NTP) and a
krypton-85 concentration in the chimney gas of 4 |iCi/ft ; (1
-4 3
xlO |iCi/cm ; see Appendix C), the equivalent source term is
2 —9
9.2x10 p.Ci/sec. Using the dispersion coefficient of 3xlO~
|j.Ci/m :y,Ci/sec for the range from one to ten miles downwind
the long-term average krypton-85 concentration in this range
These estimates are predicated on the maximum possible gase-
ous concentrations. Actual gaseous tritium concentrations
are expected to be as much as an order of magnitude lower.
T_
ICRP (33) indicates 2.2 I/day consumption for standard man.
If it is assumed that a man corsumes 1 liter of milk per day,
this off-sets the water consunr/o-ion and reduces the tritium
intake via water to 1 p.Ci/day. The water in vegetation also
off-sets liquid consumption.
An infant consumes less vegetation and water than an adult,
but this is off-set by the infant's smaller body mass.
21
-------
-12 3
would be 2.8 x 10, p,Ci/cm of, air. This concentration is
4
a factor of 4 x 10 below the AEG MC-0524 concentration guide
— 7 3
of 10 |iCi/cm for a suitable sample of the population
in an uncontrolled area. This is equivalent to a dose of
about 4 x 10~ mrem per day of exposure (see Appendix G for
-4
assumptions), or 4 x 10 mrem for over a period of ten days.*
The assumed concentrations and flowrates on which the above
calculations are predicated would result in the release of
about 70 per cent of the total source term for tritium and
p c
Kr during the initial high flowrate tests over a ten-day
period. Production testing operations will include not only
the initial high flowrate tests but also intermediate and
low flowrate tests which will last for months. Radioactivity
released during the high flowrate tests will probably be less
than the amounts assumed above, but radioactivity will also
be released during the subsequent tests. Doses estimated
above should be representative of the potential radiation
doses from radioactivity releases during the total flaring
operation.
*The assumption of an infinite spherical cloud instead of
p c
a semi-infinite spherical cloud in the Kr dose estimate
is intended to compensate for any dose resulting from re-
85
tention of Kr in body fat (see Appendix G). The calcu-
lation is for the surface dose rather than the dose at some
2
depth within the body. The depth dose at 7 mg/cm , the
approximate thickness of the outer epidermal layer, is
about one-half of the surface dose (34). The gamma dose
contribution for Kr-85 is about one-hundredth of the sur-
face beta dose.
22
-------
TOTAL DOSE—ACUTE PLUS CHRONIC
85
The previous sections on tritium and Kr dose have indi-
cated maximum exposures during the' high flow-rate production
testing period (acute exposure). Subsequent to the high
flow-rate tests there will be low-level releases associated
with additional tests (the dose from these additional tests
has been considered by the assumptions used for the high flow-
rate tests). But, of possibly more importance will be the
chronic exposure from the residual radioactive effluent from
the high flow-rate tests remaining in the environment. The
Q c
estimated Kr dose is several orders of magnitude less
p c
than that for tritium and Kr is basically inert and thus
dispersed in the atmosphere, whereas, tritiated moisture is
retained in plants and soil with a fairly long effective
half-life (tens of days; see Appendix I). Therefore, this
section will only deal with the chronic dose from tritium.
The model used to assess the potential dose from both the
acute and chronic tritium exposure is presented in Appendix
I. The model considers two phases of environmental concen-
tration:
1. Assume that the environmental vegetation moisture con-
centration reaches 360 pCi/ml instantaneously with the
initiation of high flow-rate production testing and is main-
tained at this concentration for three weeks (three 3-day
production tests over a 3-week period). Among other things,
this assumes that 70% of the upper source term estimate
for tritium is released during this period. The actual
quantity of tritium produced may be lower by an order
of magnitude, and it is likely that only 10% of that
23
-------
produced will be released*..
This approach is used to arrive at conservative estimates
of initial tritium concentrations in the environment and
also to account for the dose that may be accrued from
subsequent releases during lower flow-rate testing**.
2. At the end of the 3-week production test phase, the
vegetation moisture concentration of 360 pCi/ml of tritium
was allowed to decay with an effective half-life of 85
days (see Appendix I for discussion of the half-life).
The dose to man is then calculated based on the build-up
and eventual decay of the tritium concentration in body
water as a function of the tritium concentration in man's
intake and the biological effective half-life.
*This expected conservatism of two orders of magnitude is
relevant not only here, but is also implicit in the previous
tritium dose estimates.
**A low release fraction during initial production testing
results in a greater potential source term for subsequent
tests.
24
-------
Table 1 summarizes the estimated doses from Project
Rulison. The results are reported for three time periods;
(1) dose as a result of one day's exposure at the peak
production testing flow-rate, (2) dose from exposure
during the total peak production testing period (three
weeks, including three 3-day testing period) and the dose
from the total operation.
TABLE 1
RULISON DOSE ESTIMATES
Radionuclide Average Daily Dose Acute Period Chronic Period
Peak Flaring Rate First 3 Weeks (Residual)
mrem mrem mrem
85Kr
External*
Tritium
Inhalation
Ingestion
4 x 10~5
6 x 10~3
io-2
4 x IO"4
6 x 10'2
0.2
3
TOTAL DOSE FROM OPERATION - 3 mrem
The important conclusions to be drawn from this information
are:
1. The chronic dose from tritium in the environment may
be roughly an order of magnitude greater than the dose
during the actual flaring operation.
2. The potential tritium dose is greater than the
8 S
potential Kr dose; and ingestio
icant uptake pathway for tritium.
8 S
potential Kr dose; and ingestion is the most signif-
*Largely beta skin dose.
25
-------
3. The conservative total dose estimate for the Rulison
production testing is 3 rarem; considering the previously
mentioned conservative assumptions (tritium production
and release fraction) the actual dose will probably be
in the range of 0.1 mrem or less.
The population information (in Figure 2, page 8) indicates
there are at most several hundred people within the area
covered by this dose estimate. All of these people will
not be in the area of the highest concentrations and few,
if any, of them will be consuming foodstuffs containing
tritium at the concentrations postulated in this section.
26
-------
VI. ENVIRONMENTAL SURVEILLANCE
Average airborne concentrations of tritium .are expected to
be three orders of magnitude below the concentration guide
at off-site populated locations, and average airborne con-
85
centrations of Kr are expected to be four orders of mag-
nitude less than the concentration guide. Tritium concentra-
\
tions in vegetation moisture can be expected to be a few
picocuries per milliliter to a few hundred pCi/mtL. Tritium
concentrations in milk should be an order of magnitude less
than the concentration in vegetation moisture. No signifi-
cant environmental concentrations of any other radionuclides
are expected. Migration of radioactivity away from the chim-
ney through ground water is considered unlikely. If such.
migration does occur, velocities of movement will be small.
Occurrence of significant concentrations of radioactivity
N
at any ground water use point is highly improbable.
Although radionuclide intake by residents in the Rulison area
is not expected to approach the levels of the applicable
public health guides, environmental surveillance must be per-
formed to verify actual environmental concentrations of impor-
tant radionuclides since the estimates of environmental con-
centrations are based on assumptions. Radiation doses to
the area residents can then be estimated more accurately
from measured levels.
An adequate environmental surveillance program should also
obtain information useful in health evaluations of similar
projects in the future. Determinations of actual plume rise,
points pf maximum concentration at surface level, and tritium
concentrations in vegetation moisture and mi.lk should be made,
both for health considerations and to aid in predicting en-
vironmental concentrations of tritium which may result from
future gas stimulation projects.
27
-------
The surveillance program proposed by the Southwestern Radio-
logical Health Laboratory has been reviewed. This plan, pre-
pared under a Memorandum of Understanding with the AEC, was
presented in the "Off-Site Radiological Surveillance Plan—
Project Rulison Drill-Back and Flaring Program," October 1969,
and was supplemented by the program outlined in NVO-Slt The
program includes analysis for tritium, fission products, and
activation products in: (1) air samples; (2) samples of
foods and drinking water used by wildlife, domestic animals,
and humans; and (3) precipitation samples. The program
includes collection and analysis of background samples.
85
Although tritium and Kr are the only radionuclides expected
to be released in detectable quantities, a significant part
of the total environmental surveillance effort is oriented
towards other fission product radionuclides (radioiodines,
strontium, Cs, etc.). It is prudent to establish a
program to document the levels of these fission products,
even though their release is not expected. However, most
of the effort should be devoted to sampling for radionuclides
which are expected in the flared gas.
The rural community of Morrisania Mesa is located 3 miles
down-wind in the path of Battlement Creek Valley drainage
winds. There are fruit orchards (apples, peaches, and a few
plum and apricot trees), and milk cows in the area. Since
radiation doses from ingestion of tritium may be equal to or
greater than doses from inhalation, attention should be given
to food products grown in this area.
The proposed SWRHL program appears to be adequate not only
to determine environmental levels of concern to public health,
but also to document the environmental concentrations of
*Draft, November 1969.
28
-------
radionuclides caused by Rulison production testing. Sensi-
tivities for detection are several orders of magnitude
below the FRC and AEC guides.
The surveillance program for Rulison production testing
should be designed to obtain information concerning the
following:
1. Effluent plume rise; not so much because of its
effect on atmospheric diffusion, but rather because it
determines the transport direction and thus the popu-
lation at risk due to the various wind regimes at the
Rulison site.
2. Maximum concentrations, and changes with time, of
tritium in the water associated with various types of
vegetation (milk cow forage, domestic animal forage,
and vegetation for human consumption—assuming the
concentrations will vary in the different vegetation
species), milk and potable water.
3. The effective half-life of tritium in vegetation
in the Rulison environment. The proper determination
of the half-life will require a minimum sampling time
interval of one or more samples per half-life period.
4. If possible, bioassay samples (urine), should be
obtained to give an indication of actual doses.
29
-------
REFERENCES
1. Technical Discussions of Off-Site Safety Programs for
Underground Nuclear Detonations, NVO-40 Revision No. 2,
May 1969, U.S. AEC/NVOO.
2. Safety Involving Detonation of Nuclear Devices, NVO-28,
May 1966 (presently being updated), U. S. AEC/NVOO.
3. Rapp, E. G., "Containment of Buried Nuclear Explosions,"
UCRL-50604, October 1968.
4. Ward, Don C., and Lemon, R. F., "Status of Reservoir
Evaluation, Project Gasbuggy" (Presented at Annual Fall
Meeting; the Society of Petroleum Engineers, Houston,
Texas, September 29, 1968) PNE-G-13.
5. Holzer, Fred, "Gasbuggy Preliminary Postshot Summary
Report," PNE-1003, Jan. 1968.
6. Preliminary Site Climatology Western Colorado, ESSA/ARL,
Feb. 1969.
7. Addendum to Preliminary Site Climatology Western Colorado,
ESSA/ARL, June 1969.
References 6 and 7 are unpublished reports from ESSA/ARL,
Las Vegas, Nevada, Oct. 10, 1969.
8. . Briggs, Gary A., "Prediction of Plume Rise Heights," to
be published in the proceedings of the 8th Annual Environ-
mental and Water Resources Engineering Conference, Vander-
bilt University School of Engineering, Nashville, Tenn.,
June 1969.
R-l
-------
9. Turner, D. Bruce, "Workbook of Atmospheric Dispersion
Estimates," Environmental Health Series, Air Pollution,
PHS Pub. No. 999-AP-26, Revised 1969.
10. CER Geonuclear Corporation, "Project Rulison Definition
Plan," 26 March 1969.
11. Voegeli, Paul T. Sr., "Geology and Hydrology of the
Project Rulison Exploratory Hole, GarfieId County,
Colorado," 4 April 1969, U. S. Geological Survey,
Denver, Report No. USGS-474-16 or PNE-R-2.
12. Smith, C. F., and F. F. Momyer, "Studies of Chemical and
Radiochemical Composition of Natural Gas from the Cavity
Produced by the Project Gasbuggy Nuclear Shot," Radio-
logical Health Data and Reports, V..10, No. 7, July 1969.
13. Rawson, D. E., J. A. Korver, R. L. Pritchard, and W..
Martin, "Gasbuggy Postshot Geologic Investigations,"
November 1968, AEG Report PNE-G-11.
14. Knox, J. B., D. E. Rawson, and J. A. Korver, "Analysis
of a Groundwater Anomaly Created by an Underground
Nuclear Explosion," Journal of Geophysical Research,
70:823-835, 15 February 1965.
15. Nork, William E., "Final Pre-Event Prediction of Radio-
activity in the Hydrologic Environment—Project Rulison,"
13 August 1969, AEC Report No. NVO-1229-108.
16. Davis, S. N., and R. J. M. DeWeist, Hydrogeology, John
Wiley and Sons, New York (May 1967).
17. Higgins, G. H., D. D. Rabb, and H. C. Rodean, "Theoret-
ical and Experimental Studies Relating to the Purging of
R-2
-------
Radioactivity from a Gas Well Stimulated by a Nuclear
Explosion," 24 December 1968, AEG Report No. UCRLT50519.
18. Eakin, J. L., R. V. Smith, and J. S. Miller, "Estimation
of Well Capacities and Gas Reserves," Gas Engineers
Handbook, pp. 4/28-4/47, The Industrial Press, New York
(1966).
19. El Paso Natural Gas Company, USAEC, U.S. Bureau of
Mines, and LRL, "Project Gasbuggy," El Paso, 14 May 1965,
AEC Report PNE-1000.
20. Coffer, H. F., B. G. Bray, and C. F. Knutson, "Applica-
tions of Nuclear Explosives to Increase Effective Well
Diameters," Engineering with Nuclear Explosives, pp. 269-
287, 21-23 April 1964, AEC Report TID-7695.
21. Austral Oil Company, Inc. and CER Geonuclear Corporation,
"Summary Project Rulison Feasibility Study," July 1966.
22. Brundage, Robert, CER Geonuclear Corporation, personal
communication, September 1969.
23. Todd, D. K., Ground Water Hydrology, John Wiley and Sons,
August 1963.
24. Davis, S. N., "Hazards Evaluation—Groundwater," Nuclear
Civil Engineering, ed. Paul Kruger, Technical Report No.
70, Department of Civil Engineering, Stanford University,
September 1966.
25. Effects Evaluation Division, NVOO, AEC, "Effects Evalu-
ation for Project Rulison," AEC Report NVO-43, June 1969.
26. Environmental Research Corporation, "Prediction of Seis-
mic Motion and Close-in Effects," NVO-1163-180, PNE-R-5.
R-3
-------
27. Southwestern Radiological Health Laboratory, USPHS, "Off-
Site Safety and Environmental Surveillance Operation Plan
for Project Rulison," April 1969.
28. "Answers to Questions Posed by CCEIRFR," PNE-G-48, 8/27/69.
29. Protection of the Public in the Event of Radiation Acci-
dents, World Health Organization, 1965.
30. Smith, C. F., "Non-Gaseous Radioisotopes - Project Gas-
buggy Chimney Gas," UCRL-50634, 4/7/69.
31. McBride, J. R., Hill, D., "Off-Site Radiological Surveil-
lance Program for Project Gasbuggy," PNE-G-46.
32. International Commission on Radiological Protection,
Report 10.
33. Recommendations of the International Commission on Radio-
logical Protection, ICRP-2, 1959.
34. Meteorology and Atomic Energy, July 1968, U.S.A.E.G.,
Division of Technical Information.
35. "Project Gasbuggy," PNE-1000, May 14, 1965.
36. Nucleonics Week, 27 November 1969.
37. "Background Material for the Development of Radiation
Protection Standards,"' Federal Radiation Countil;
a. Report No. 1, May 13, 1960
b. Report No. 2, September 1961
c. Report No. 5, July 1964
d. Report No. 7, May 1965
R-4
-------
38. "Project Rulison Post-Shot'Plans and Evaluations," USAEC,
NVOO, December 1969, NVO-61.
39. "Standards for Radiation Protection," USAEC Manual
Chapter 0524, 11/8/68.
40. Cowser, K. E., et. al., "Dose-Estimation Studies Related
to Proposed Construction of an Atlantic-Pacific Inter-
oceanic Canal with Nuclear Explosives: Phase II, March
1967, ORNL-4101.
41. Lieberman, J.A., "The Respiratory Rate of a One-Year Old Child','
AEC, Operational Safety, Preliminary Draft of Paper, 11/29/66.
42. Martin, J. R., Koranda, Kline, Jordan, "The Movement
of Tritium in a Tropical Ecosystem," presented at
American Nuclear Society Symposium on Engineering
with Nuclear Explosives, Las Vegas, Nevada, Jan. 1970.
R-5
-------
LIST OF APPENDICES
APPENDIX A MAJOR BASINS OF THE ROCKY MOUNTAIN STATES
APPENDIX B GAS DELIVERABILITY
APPENDIX C RADIOACTIVITY IN GAS
APPENDIX D RADIOACTIVITY IN GROUND WATER
APPENDIX E METEOROLOGY AND DIFFUSION
APPENDIX F TRITIUM DOSE CALCULATION
APPENDIX G KRYPTON-85 DOSE CALCULATIONS
APPENDIX H SUMMARY OF PARAMETERS FROM NVO-61
APPENDIX I ACUTE AND CHRONIC DOSES FROM INGESTION OF
TRITIUM
-------
A-l
MAJOR BASINS OF THE ROCKY MOUNTAIN STATES
WYOMING
--J
UINTA BASIN
/ PARADOX
BASIN
WIND RIVER
BASIN
GREEN RIVER
BASIN
WASHAKIE
DAdIN I
- L
WASH
PICEANCE BASIN
COLORADO
NEW MEXICO
SAN JUAN
BASIN
r
APPENDIX A
THESE BASINS CONTAIN SUFFICIENT RESERVOIR THICKNESSES
TO MERIT CONSIDERATION FOR NE STIMULATION.
(from reference 35)
-------
APPENDIX B
GAS DELIVERABILITY
An expression for gas deliverability from a well can be de-
rived by assuming that the paying strata penetrated by the
well are of uniform thickness and infinite extent and are
isotropic and homogeneous, and that flow is compressible and
Darcian. With these assumptions, the radial flow of gas into
a well under steady-state conditions at a given point in
time is given by (19, 20):
10.320 kh(P - P )
Q = (Equation B-l)
u In (r /r ) T (14.65) z
e w f
WHERE:
Q = Rate of flow in cubic feet per day at a pressure
base of 60°F and 14.65 psia (NTP).
k =
Permeability of the paying strata, in millidarcys
h = Net thickness of the paying sandstone between two
confining layers, in feet
P = Formation pressure in psig at the effective radius
of drainage, r .
P = Flowing pressure at the face of the wellbore, in
psia, r is the effective radius of the wellbore.
Absolute open flow deliverability (AOF) is calcu-
lated by setting P = 1 atm.
u = Viscosity in centipoise, of the gas, approximately
0.015.
B-l
-------
Tf = Formation temperature, degrees Rankine
z = Gas deviation factor
Sufficient information has been published to permit use of
Equation B-l to predict the initial deliverability of the
Project Rulison chimney-fracture system. For the overall
Rulison field, average reservoir properties are given as
(10, 11, 21):
k = 0.5 millidarcys (sandstone lenses)
P£ = 2700 psia
z = 0.88
On the average, the Mesaverde Formation in the Rulison field
contains a net thickness of 500 feet of paying sandstones
out of a total formation thickness of approximately 2500
feet (21). Assuming the paying sandstone lenses are uni-
formly spaced throughout the thickness of the Mesaverde, a
chimney 370 feet high would intersect approximately 74 feet
of paying sandstones. The sandstones are not likely to be
uniformly spaced, and good design would locate the chimney
to intersect as much sandstone as possible, but no other
assumption is possible without more detailed geologic in-
formation.
Average reservoir properties observed in the exploratory
hoile R-EX are given as (10):
k = 0.11 millidarcys' (sandstone lenses)
Tf = 215°F = 675* Rankine (at 8400 feet subsurface)
B-2
-------
A net thickness of 375 feet of paying sandstone out of a total
formation thickness of 1162 feet was observed in R-EX; a
chimney 370 feet high would penetrate approximately 120 feet
of paying strata, assuming uniform spacing of the sandstone
lenses.
Nuclear stimulation is thought to increase the effective
well radius to the radius of intense dynamic fracturing,
estimated to be approximately 370 feet for Project Rulison.
The above data can be used with Equation B-l to estimate the
"absolute open flow deliverability" (AOF) of the chimney-
fracture system. The absolute open flow deliverability is
that flow which would occur if pressure at the well face
(i.e., in the chimney) were reduced to one atmosphere. Using
the reservoir parameters reported for R-EX and an assumed
drainage area of 160 acres, the "expected" initial AOF deliv-
6 3
erability is estimated to be 5.5 x 10 ft /day; an expected
"maximum" initial AOF deliverability can be estimated from
the average parameter values for the Rulison field as 15.3
6 3
x 10 ft /day, assuming that the permeability of a large
area is not likely to exceed the average permeability of the
paying sandstones.
The use of Equation B-l to predict AOF deliverability implic-
itly assumes that the sandstone lenses have large horizontal
dimensions compared to the radius of fracturing, are of
uniform thickness, and are homogeneous. Data reported by
the USGS (11) indicate that the lenses may be several thou-
sand feet long; if so, the initial deliverability may be
reasonably approximated by Equation B-l.
Equation B-l represents a very simple approach to reservoir
engineering. Techniques actually used by petroleum companies
to predict deliverability are more sophisticated. Such tech-
niques generally involve finite-difference computer solutions
B-3
-------
of time-dependent partial differential equations which take
into account variation of fracturing with distance from the
detonation point. Effective use of these techniques requires
better knowledge of the geology of the field than is presently
available to the public. Detailed descriptions of reservoir
characteristics are included in the "Project Rulison Feasi-
bility Study" but the information has not been released by
Austral Oil. Estimates of deliverability are included in
the "Pre-Shot Reservoir Evaluation" by Austral Oil/CER which
presently (as of November 1969) exists only in draft form and
is not available to the public (22).
Since the gas flow rates which will occur during production
testing will be as much the result of administrative decisions
as of engineering calculations, it is felt that use of Equa-
tion B-l will yield acceptable estimates.
B-4
-------
APPENDIX C
RADIOACTIVITY IN GAS
Table C-l gives estimated quantities of selected radionu-
clides produced by the Rulison device. The radionuclides of
chief interest in consideration of gas use are Krypton-85
and tritium, since these two isotopes will be found in gas
released from the chimney. Estimation of the concentrations
of these isotopes in the gas requires knowledge of the amount
of the isotopes produced by the device, the void volume of
the chimney, the pressure in the chimney at the time flaring
begins, the bottom-hole temperature, and the water content
of the formation.
The calculation for Krypton-85 is straightforward. From
Q C
Table C-l, approximately 960 curies of Kr were produced
by the detonation. The void volume of the chimney is
approximately equal to that of a sphere with a radius of
80 feet, about 2.1 x 10 cubic feet.
The bottom-hole temperature of the chimney at the time of
re-entry is estimated to be about 380°F, based on experience
with Gasbuggy.* The bottom-hole pressure at the time of
*The Gasbuggy detonation raised temperatures in the chimney
from the pre-shot formation temperature (130°F) to 247°F (4)
Cavity dimensions and materials for Rulison are similar to
those of Gasbuggy, so the temperature increase should be
proportional to the ratio of energy yields of the respec-
tive devices (29 KT for Gasbuggy—LRL Memo, 5 January 1970).
215°F + (40 KT/29 KT) (117°F) = 376°F or 380° F. There are
obvious uncertainties in this estimate.
C-l
-------
TABLE C-l
RADIONUCLIDE ACTIVITY AT T
+180 DAYS
RESULTING FROM DETONATION OF 40 FISSION KILOTONS
Nuclide
HaIf-Life
Curies
85
89
90
Kr
Sr
Sr
Y
Zr
95
95Nb
103
103
106
106
Ru
Rh
Ru
Rh
131.
133
137
137
140
140.
141
Xe
Cs
Ba
Ba
La
143
144
144
147
Ce
Pr
Ce
Pr
Pm
H
10.76 y
50.6 d
28.8 y
59 d
65 d
35 d
40 d
57 min
1.0 y
30 sec
8.05 d
5.27 d
30 y
2 . 6 min
12.8 d
40 h
32.5 d
13.7 d
285 d
17.3 min
2.6 y
12.27 y
0
0
0
; 1
1
0
0
0
1
1
1
0
0
0
0
0
0
0
1
1
0
1
.96
.91
.59
.01
.82
.32
.41
.41
.52
.52
.13
.86
.75
.69
.34
.40
.52
.63
.47
.47
.28
.0
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
103
105
104
105
105
108
5
105
105
105
ID'3
4
4
103
3
105
io3
105
io5
IO5
io4
From reference 15
C-2
-------
re-entry is estimated to be about 2640 psia.* The volume
o
of gas in the chimney is estimated to be about 2.4 x 10
85
cubic feet (NTP), and the corresponding Kr concentration
3 -4 "3
is estimated to be about 4.0 |iCi/ft (1.4 x 10 |iCi/cm ) .
This figure is less than a factor of two higher than the
AEG predicted Kr concentration of 0.8 x 10~" (iCi/cm (28).
Estimation of tritium concentrations is more difficult. The •
most conservative approach assumes that all of the tritium
will exchange with hydrogen attached to gaseous hydrocarbons.
In this case, the tritium concentration in the gas flared
g
from the well-head would be 10 kCi/2.4 x 10 cubic feet =
3 — 3 3
42 |iCi/ft (1.5 x 10 |j,Ci/cm ). This approach is excessively
conservative, since it implies that there is no exchange of
tritium with water-bound hydrogen. In Gasbuggy, approximately
95% of the tritium produced by the nuclear device appears to
have eventually exchanged with water-bound hydrogen in the
chimney to appear as HTO (12). A pre-shot estimate of
tritium in Gasbuggy gas using the same conservative assump-
tion would have predicted a concentration of about 320 |iCi/ft ,
but the highest concentration measured, 100 }iCi/ft , was
observed in a sample taken the day after the detonation (12).
The tritium concentration in the gas decreased to approxi-
mately 18 |j.Ci/ft within 36 days after the detonation (12).
Assuming a similar reduction of the tritium concentration in
Rulison gas would lead to a predicted value of about 2 (j.Ci/ft
*Re-entry well-head pressures are estimated at 2400 psia by
the AEG (Press Release NV-69-135). Assuming uniform gas
temperatures of 380°F and gas of pure methane, the bottom-
hole pressure would be about 2640 psia. Actual pressures
will be somewhat higher, since the gas will contain other
species of higher molecular weight.
C-3
-------
when flaring begins at least six months after the detonation.
However, this assumption is probably not reasonable either,
since water may have flowed into the Gasbuggy chimney from
the Ojo Alamo Formation (13), while it is unlikely that large
quantities of water will enter the Rulison chimney. Some
reduction will undoubtedly occur, since the Mesaverde sand-
stones near the chimney contain about 4% water by volume (11),
but there appear to be no overlying aquifers near the chimney.
The actual tritium concentration should be somewhere between
3 3
2 [iCi/ft and 40 |j,Ci/ft . An "expected" value can be com-
puted by assuming that the initial reduction of tritium con-
3 3
centrations in Gasbuggy gas (from 320 p,Ci/ft to 100 |j,Ci/ft )
was caused by tritium exchange with water in the immediate
vicinity of the chimney, and that water from the Ojo Alamo
did not enter the chimney until several days after the
detonation. Both the Gasbuggy and Rulison media have approx-
imately the same water content, so a similar reduction might
be expected for Rulison; this assumption would result in a
3 —4
predicted tritium concentration of about 13 u.Ci/ft ( 4 x 10
liCi/cm ) .
C-4
-------
PURGING OF RADIOACTIVITY FROM CHIMNEY
As gas is flared from the chimney, "clean" gas from the
surrounding formation will flow into the chimney to decrease
concentrations of radionuclides in the chimney gas. Two
models are available for predicting the reduction of radio-
nuclide concentrations as gas is flared; the simplest assumes
that as fresh gas enters the chimney, it mixes completely
and instantaneously with all the gas in the chimney. This
results in predicted concentrations of radionuclides that
decrease exponentially with the total volume of gas flared
(17).
The other model, developed by Higgins, et. al. (17), assumes
that no mixing takes place and that the fresh gas displaces
the radioactive gas in a piston-like fashion. In this case,
concentrations of radionuclides in the flared gas would
remain constant at early times, decrease rapidly at inter-
mediate times, and finally go to zero after flaring at most
four to five chimney volumes (NTP) of gas.
The actual decrease of radioactivity probably will not corres-
pond precisely to either model but will lie somewhere between
the two predictions. Higgins, et. al., indicate that measure-
O c
ments of Gasbuggy Kr concentrations were constant at
early times, conforming to the no-mixing model (18). A
conservative approach would be to assume that concentrations
will remain constant until approximately 0.1 to 0.2 chimney
volumes have been flared (conforming to the no-mixing model)
and will then decrease exponentially with the volume of gas
flared (perfect-mixing model).
C-5
-------
APPENDIX D
RADIOACTIVITY IN GROUND WATER
Little information is available concerning occurrence and
movement of ground water or magnitudes of hydraulic gradients
existing in the Mesaverde Formation near the Rulison chimney.
The USGS indicates that usable ground water in the Rulison
area is primarily limited to alluvium and terrace deposits,
and that underlying bedrock formations are generally imper-
meable and yield little water. From.observations made during
the drilling of the exploratory hole R-EX, the USGS con-
cluded that little mobile water occurs in the formations
penetrated by the hole (11). However, the USGS also reports
that the Ohio Creek Conglomerate has yielded water in other
gas wells drilled in the Rulison field and that some water
was found in a sandstone lens in the upper Mesaverde Forma-
tion (11). There presently seems to be insufficient evidence
for concluding that mobile water cannot be present in the
fracture zone surrounding the chimney.
The Mesaverde Formation consists of sandstone lenses inter-
bedded with shales of much lower permeabilities (10,11).
Permeabilities of the shales might reasonably be expected
to be two orders of magnitude lower than the permeabilities
of the sandstone lenses (16). Average sandstone lens porosity
and water saturation for the Rulison field are given as 9.7
per cent and 45 per cent, respectively (10). Ground water
movement in such unsaturated media is typically much slower
than movement in saturated media. Capillary forces and
other factors affect moisture movement in unsaturated media.
Prediction of moisture movement in unsaturated media requires
more data than are available for the Rulison field. However,
some estimate of possible radionuclide movement is desirable.
Though "worst case" estimates are often misleading, it is
felt that the development of a "worst case" estimate of
D-l
-------
possible ground water movement near the Rulison chimney will
serve to illustrate that the chances are very small that
significant concentrations of radionuclides will occur at any
ground water use point.
The simplest approach to a "worst case" estimate of ground
water movement is to assume that fractures from the chimney
will intersect a water-bearing formation and will introduce
water into the chimney. Ground water velocities which would
occur under the resulting saturated conditions would be much
greater than velocities which would occur under unsaturated
conditions. It is emphasized that radioactivity will tend
to remain in place unless mobile water is present and that
the assumption that water will be introduced into the chim-
ney is therefore conservative. The observations of water in
the Ohio Creek Conglomerate at other locations and in a sand-
stone lens in the upper Mesaverde Formation partially justify
'the assumption that the fracture system will contact an
aquifer.
There is no means of estimating how much water will be
present. Therefore it will be arbitrarily assumed that
all radioactivity from the detonation will be uniformly
mixed with a volume of water equivalent to the void volume
/• o
of the chimney (2.1 x 10 ft ). Using these assumptions and
the estimated tritium production given in Table 1, the result-
ing tritium concentration can be estimated as 0.17 (iCi/ml,
about 170 times the concentration guides given by the AEC
Manual, Chapter 0524 (39).
Initial reduction of concentrations of radionuclides such
as Cs and Sr by ion exchange will be neglected. This
assumption results in estimates of concentrations of these
radionuclides which are probably conservative, regardless
of the amount of water actually introduced into the chimney.
D-2
-------
137 90
Corresponding estimated concentrations of Cs and Sr
—2
are 0.12 (iCi/ml and 9.8 x 10 y,Ci/ml, respectively, which
4 6
are 1.8 x 10 and 1 x 10 times the Chapter 0524 concentra-
137 90
tion guides for Cs and Sr, respectively.
Factors which influence the velocity of ground water move-
ment include the degree of saturation of the pore space,
the permeability of the medium, presence of fractures, and
the existing hydraulic gradients. For the entire Mesaverde
Formation these factors are not well-known. Characteristics
of the sandstone lenses have been reported (10), but the
geometry of these lenses is not known. Conservative
assumptions must therefore be made to permit estimation of
ground water movement.
It will be assumed that the entire formation has the per-
meability of the sandstone lenses, and the presence of
interbedded shales will be ignored. Saturated flow condi-
tions will also be assumed. The Green River, Wasatch, Ohio
Creek Conglomerate, and Mesaverde Formations dip to the north
at about two degrees (11). In the absence of disturbances,
the hydraulic gradient might be assumed to follow the dip
of the formations, so that ground water flow would also
follow the dip of the beds.
An expression for the average velocity (tracer velocity) of
ground water flow under Darcian conditions is (23,24):
ks dp
v = — a£ Equatxon D-l
WHERE:
v = tracer velocity, feet/second
D-3
-------
3 3
s = specific weight of water, Ibs/ft =59.8 Ibs/ft
at 215°F
u = viscosity of water at formation temperature
= 0.59 x 10"5 lb-sec/ft2
n = formation porosity
= pressure gradient or hydraulic gradient, feet of
water per foot of distance
2
k = permeability of the saturated medium, ft
The reported average permeability and porosity of the sand-
stone lenses of the Mesaverde Formation are 0.5 millidarcys
and 8-10 per cent, respectively. Using these values in
Equation D-l and assuming the hydraulic gradient is approxi-
mately equal to the dip of the beds, the estimated ground-
water velocity would be approximately 0.6 ft/year. This
development assumes saturated flow conditions. The conser-
vative assumptions leading to this estimated movement should
be emphasized. The first assumption is that relative large
amounts of water will be introduced into the chimney from
some overlying aquifer, resulting in saturated flow through
the Mesaverde Formation away from the chimney. It is also
assumed that the entire Mesaverde Formation has the permea-
bility reported for the sandstone lenses, which may be con-
servative by two orders of magnitude. It is implicitly
assumed that the wetting front of the flow from the chimney
will move at saturated flow velocities, also a conservative
assumption.
Ignoring dilution by diffusion and mixing, tritium would be
reduced to the AEC concentration guide of 1 x 10~ |iCi/ml by
radioactive decay in about 7.4 half-lives. Since tritium
D-4
-------
moves with the velocity of ground water flow, tritium might
move a very few feet before decaying to the concentration
guide. Even at 200 times the conservative estimate of ground
water velocities, tritium would still move less than three
miles before being reduced to the concentration guide.
Tritium can be assumed to move with the velocity of ground
water, but the movement of Cs and Sr will be retarded
by ion exchange with the medium. The degree of retardation
can be predicted from a property of the medium called the
distribution coefficient, K,. Values of K, for rocks in the
d d
Rulison area were not measured, but Isotopes, Inc. has
reported values for sandstones near the Gasbuggy detonation
as approximately 100 and 1.4 for Cs and Sr, respectively
(15). These values will be assumed to be representative of
the Mesaverde Formation also. CER Geonuclear reported ah
average overburden density of 2.35 g/cm and a core grain
density of 2.67 g/cm for the Rulison site (10), but the
density of the sandstone may be somewhat greater than the
3
average. A value of 2.4 g/cm
to be a reasonable assumption.
•5
average. A value of 2.4 g/cm for the bulk density appears
The relationship between the average ground water velocity,
v, and the effective velocity, v', of a given radionuclide
in ground water is (24)
v' 1
— = i ^ v I / -\ Equation D-2
v 1 + K, (m/n)
WHERE:
k^ = distribution coefficient for the particular
radionuclide in a given medium, ml/g
m = bulk density of the rock or aquifer, g/cm
D-5
-------
Using the above values for bulk density and K,,
1 • v'/v = 4.2 x 10~ for Cs
v'/v = 0.029 for Sr-90
Ignoring dilution by diffusion and mixing, Cs and Sr
would be reduced to their appropriate AEC concentration
guides in about 14 half-lives and 20 half-lives, respectively.
Thus, both would decay to the concentration guides before
moving more than a few feet. At 200 times the conservative
90
estimate of ground water velocity, Sr might move about
one-half mile before decaying to its concentration guide.
While gas is being released from the well, hydraulic grad-
ients in the formation surrounding the chimney will drive
water toward the chimney because chimney pressures will be
lower than the pore pressure of the formation. Until gas
production from the well ceases, no migration of radioactivity
away from the chimney is expected.
The analysis presented above cannot be construed as an
accurate picture of the movement of radioactivity in ground
water, since the estimates given result from assuming a very
unlikely "worst case." However, the estimates serve to
illustrate that transport of radioactivity from the Rulison
chimney to any ground water use point is highly improbable.
D-6
-------
APPENDIX E
METEOROLOGY AND DIFFUSION
Climatological Data
This section summarizes climatological data needed to estimate
environmental effects of production testing. Local meteorol-
ogy is strongly affected by terrain features. The area can
best be described as a large plateau sub-divided into mesas
by drainage channels. The important features are:
1. Battlement Mesa - The Rulison site is located on the
north side of the mesa, which slopes generally from SE to NW;
2. Morrisania Mesa - located below the Rulison site on
the northern slope of Battlement Mesa;
3. Battlement Creek and its associated valley - The valley
passes near the Rulison site, sloping generally south to
north, and cuts across Morrisania Mesa to reach the Colorado
River. This valley creates the local wind regime (drainage
and upslope) for the Rulison site. The valley is about 1,000
feet deep.
4. Colorado River and its associated valley - The river
runs generally from NE to SW, passing about six miles north
of the site. Battlement Mesa and Morrisania Mesa form the
valley walls to the south and the Roan Cliffs form the valley
wall to the north.
Important terrain features are indicated schematically in
Figure 3 of the text. Elevations on the figure are given in
thousands of feet above mean sea level (1 K = 1,000 ft).
E-l
-------
Three prevailing wind regimes are indicated on the figure.*
Nightly drainage winds blowing toward the NNW and daily
upslope winds toward the SSE in the Battlement Creek valley
form one wind regime. These winds probably build up to
about 1,000 feet above the valley floor.*
Valley flow in the Colorado River Valley directly north
of the site, which generally flows east-west, forms another
wind regime. The mixing depth for this valley is not well-
known but has been noted to build up to around 1500 ft or
more (over Morrisania Mesa).
Regional gradient flow above the topographical features is
generally toward the ENE. Although this wind information is
based on air soundings at Grand Junction (elevation 4820
feet), measurements taken at about 10,000 feet on Battlement
Mesa are similar. The 500-mb** wind rose (18,500 feet) from
Grand Junction is similar, but the prevailing wind direction
is toward the east and wind speeds are greater than those at
the 700-mb level.
*Meteorological data presented in this section are based on
References 6 and 7 and on private conversations with person-
nel of ESSA/ARL.
**mb-millibar—1,000 mb is equivalent to 29.53 inches of
mercury.
E-2
-------
Diffusion Estimates
There are numerous equations for estimating plume rise and
atmospheric diffusion. Equations for plume rise are usually
based on empirical data or dimensional analysis. Briggs'
equation (8) was used on recommendations of personnel from
ESSA/ARL and the experience with Project Rover reactor tests
at NRDS (felt to be a good analogy).* Briggs' equation is
based on dimensional analysis and incorporates an empirically
derived constant (Equation E-l).
Gifford's modification of Pasquill's diffusion equation is
used (9) to estimate dispersion with distance. The model
is generally known as the Gaussian model because of its
inherent assumption of a Gaussian-shaped cloud.
Calculations for the plume rise (also referred to as effec-
tive stack height) and downwind dispersion are given at the
end of this appendix. The results of these calculations
are indicated in Figures E-l and E-2.
Each figure represents a different set of conditions. Figure
E-l is based on an assumed effective stack height of 2700
feet and transport by the regional (aloft) winds, whereas
Figure E-2 is based on an assumed effective stack height of
only 1,000 feet and transport by the local valley winds.
This latter case (involving drainage winds)' is unlikely
since thermal energy will probably cause a significant frac-
tion of the plume to rise above the terrain. It could occur
during a strong inversion (causing limited plume rise) or at
lower flaring rates (since plume rise is proportional to the
cube root of the energy and thus flaring rate). Should the
*Recommendation> of Mr. N. Kennedy, ESSA/ARL, Las Vegas.
E-3
-------
E-4
10
FIGURE E-2
Total (sum of curves)
-9
10
-10
10
Dispersion versus Distance
2700 FOOT STACK HEIGHT
wind speed = 15 mph
neutral conditions
99% released at stack height
1% released at ground level
Assumes average dispersion
in 67.5° sector with 45% wind frequency.
u
0)
in
•^
O
3
O
3
c
0
C
0>
a
-------
t-5
-7
10
10
-91
10
-10
10
FIGURE E-3
Total (sum of curves)
Dispersion versus Distance
1000FOOT STACK HEIGHT
drainage wind conditions(7 mph 5O%
of time, slightly stable)
99% released at stack height
1% released at ground level
Assumes average dispersion in
22.5° sector with 65% wind frequency.
Release at stack height
U
3
c
o
e
0)
a
(0
O
Ground Level Release
-u
10
3 miles
Distance (kilometers)
-12
10 I
10
50
100
-------
"inversion case" occur, the effluent might be caught in the
inversion and possibly not be transported to the ground until
the.inversion broke up.
For both dispersion calculations, 0.01 of the source was
assumed to be released on the ground and the remaining 0.99
at the effective stack height. This assumption is based
on empirical analysis of the previously mentioned NRDS work.
This concept does not really imply release from two points
but rather a "peeling-off" of part of the effluent during the
plume rise. Thus, even if the estimated effective stack
height of about 3,000 feet is correct, some of the effluent
will probably follow the valley drainage winds.
The dispersion curves in Figures E-l and E-2 are based on
idealized topography (flat plains) and assumptions concern-
ing transport winds such as the fraction of the effluent
released at each altitude. Uncertainties caused by the
complicated terrain and uncertainties in wind parameters
throw doubt on the accuracy of estimates of the variation
of downwind concentrations with distance. Therefore, a con-
—9 3
stant dispersion coefficient, 3 x 10 sec/m , will be
assumed for distances beyond three miles, the distance of
the population closest to the site. This assumes the con-
ditions of Figure E-l are more representative of the long-
term average than those of Figure E-2.
It is felt that
a small fraction of the effluent will be caught in the
Battlement Creek Valley flow;
at high flaring rates, the majority of the effluent will
rise above the terrain to be transported by the regional
winds.
E-6
-------
ATMOSPHERIC DISPERSION CALCULATIONS
EFFECTIVE STACK HEIGHT - PLUME, RISE. (From Reference 8)
Ah = 1.6 F1/3^1*2/3 (Briggs1 Equation) E-l
WHERE:
Ah = the effective plume rise due to thermal buoyancy
above the stack in feet. For an effluent with high
thermal energy, thermal buoyancy is the dominant
factor and the momentum or jet rise is negligible.
1.6 = Constant for fit of empirical data
F = Flux of buoyancy force from stack, divided by TT and
the mean atmospheric density - 4.3 x 10~ QR. The
coefficient varies inversely with atmospheric pressure,
A value of 3.18 x 10~ was used for Rulison (8200 ft
MSL). Q^ = cal/sec and F = ft4/sec3.
rl
U = Mean wind speed = 7 mph = 10.3 ft/sec.
X = Distance downwind from the stack. Used 5280 ft
(based on experience from NRDS).
F = 3.18 x 10~3 Qv
n
QH = 2 x 107 ft3/day x (1,000 Btu/ft3 natural gas)
(252 cal/Btu) x (day/8.64 x 104 sec)
QH = 5.82 x 107 cal/sec
rl
F = 3.18 x 10~3 x 5.82 x 107 = 1.86 x 105 ft4/sec3
E-7
-------
h - 1.6 F1/3 U"1 X2/3
= 1.6 (1.86 x 105 ft4sec"3)1/3(10.3 ft sec"1)"1(5280 ft)2/3
Units - ft4/3 sec"1 ft 1sec1ft2/'3 = ft
= 1.6 x 57.1 x 10.3 l x 301
= 2700 ft
ATMOSPHERIC DISPERSION (Reference 9)
There are numerous approaches for estimating downwind con-
centrations from releases. The most common is to estimate
short-term average (3-minute) concentrations along the center-
line of the plume (Equation E-2). It was felt that a more
representative value in this case would be the cross-wind
average for the sector of concern (Equation E-3). Equation
E-3 was used to calculate cross-wind average air concen-
trations based on the prevailing wind sector and the prob-
ability for winds blowing into that sector.
The basic equation for describing the short-term average
ground level downwind concentration from a continuous point
source is:
z ->
(page 6, Ref. 9) E-2
WHERE :
X = Downwind concentration in units of release/m
E-8
-------
Q = Source term in units/sec
a & = Cross wind and vertical dispersion coefficients
respectively (meters)
U = Wind speed (meters/sec)
exp |_ ^]: Indicates e raised to the power inside the brackets
H = Plume rise
The ground-level crosswind-integrated concentration is given
by:
E-3
cwl
2Q
/27 *
u
-1/2
(in
-a"
z
>2
WHERE:
X _ = The cross-wind integrated concentration in
cwl , 2
(units/m ) x m or units/m
The other symbols are as previously defined.
To obtain the concentration within a sector, the equation
is multiplied by the frequency f and divided by the arc at
the distance of interest 2irX6.
WHERE:
.X = distance downwind in meters
6 = fraction of the 360° arc being considered (e.g., .
if 60°, then 0 = 1/6)
E-9
-------
The source is considered to be composed of two releases;
(1) 0.01 of the total released at the surface, and (2)
0.99 released at the height of rise of the plume.
Two wind regimes were considered: (1) valley flow with
Ah = 1,000 feet, and (2) Ah = 2700 feet and transport
by the "aloft" or regional winds.
The following illustrates the calculation for the concen-
tration ( Ah = 2700 ft) at 10 km or 6.2 miles downwind
assuming the regional wind regime prevails, and ignoring
topography effects:
X —
cwl
2Q
/IT" a
7 u
2 If'
f
xe
-1/2 (
Ho
^
a '
z _
Q = 0.99 unit (iCi/sec at H; 0.01 at surface
U =15 mph = 6.72 m/sec
X = 10 km
0 = (67.5°/360°)
2 * Xe = 2* 104 (67.5/360) = 1.18 x 104 m
f = 0.45
H = 2700 ft = 822 meters
2
a = (C, neutral) = 5 x 10
z
E-10
-------
X
cwl
2 x 0.99 . 0.45 ovn
_ _ exp
2ir 5 x 10 x 6.72 1.18x10
822 2
'
500
= 2.34 x 10 4 3.82 x 10~5 exp
-1.3
= 2.4 x 10 jiCi/m ; jj-Ci/sec (67.5° average)
E-ll
-------
APPENDIX F
TRITIUM'DOSE CALCULATION
Dose =
liCi intake 3 .7 x
g of ^id-sec
tissue
104:_d 0.01 Mev 1.6xlO~(
d rad/rem Mev
'ergs rad-g
100 erg
8.64 x 10 sec
0.693
= rem
WHERE:
[iCi intake
g of
tissue
0.01 Mev
8.64x10 sec -
is the quantity ingested or inhaled
4
whole body mass; 7 x 10 g for an adult, or
4
10 g for an infant (33,40)
effective energy for tritium (32,33)
effective biological half-life of tritium
for adults; 10 days (32)
For an adult this gives:
—4
1.06 x 10 rem per u.Ci intake
Infant:
ICRP has not specified a biological effective half-life
for tritium in an infant, but, considering the relative
body masses and daily liquid intake of an adult and
infant, it appears that the effective half-life in an
infant would be much shorter than for an adult. Assum-
ing a liter per day liquid intake and 10 kg body mass,
the intake equals the body mass after one adult effec-
tive half-life (10 days).
F-l
-------
The following is proposed as an estimate for the effective
biological half-life for an infant:
Assume:
a. Tritium is primarily associated with the body water.
b. The body water is relatively the same fraction of
total mass in an infant as in an adult. Thus, the
effective half-life for an infant can be scaled to
that for an adult by using the analogy of a tank mixing
mode1.
The concentration in a tank at t=0, where the in-
flow equals the outflow and the inflow concentra-
tion is zero can be expressed by the following
equation:
K(t) = KQ exp-C(Q/W)t
WHERE:
K(t) is the concentration at time t.
K is the concentration at time t = 0.
o
C is a constant to compensate for non-
ideal mixing.
Q is the inflow rate, or human consumption.
W is the tank volume, or human body mass.
The combination of the parameters C(Q/W) are analo-
gous to an effective decay constant. Assuming C is
essentially the same for both an infant and an adult
(assumptions a and b above), the infant effective
half-life may be scaled from the adult effective
half-life by the ratio of the decay constants:
F-2
-------
/Adult effectiveN ' Ca^a^Wa^ Infant, effective
[half-life J X Ci(Qi/Wi) = half-life
Using the following values:
C. = C
i a
Q. and Q are 1 and 2.2 liters per day respectively
W. and W are 10 and 70 kg respectively
1 a
Therefore:
Infant bio-
logical
effective
half-life
,10 aays^l^ay/-1^=3.1 aays £££ve
For a 10 kg infant:
-4
2.3 x 10 rem per uCi intake
The following indicates the derivation of the dose conver-
sion parameter for tritium concentration in |j,Ci/ml of body
water (hydrogen equivalent) integrated over a period of days
i.e., (|iCi/ml) x days).
4 -6
liCi-day 3.7x10 d 0.01 Mev-rem* - 1.6 x 10 ergs rad-g
ml of |j,Ci-sec d-rad Mev 100 ergs
water
4 4
6.3x10 g(H equivalent water/total body)* 8.64x10 sec
7x10 g/total body* day
0.46 rem
i-day/ml
*Reference 33.
P-3
-------
APPENDIX G
KRYPTON-85 DOSE CALCULATIONS
Cloud Submersion Dose (Ref. 33 and 34)
Assumptions:
1. Spherical cloud of uniform concentration. The cloud
is considered to be spherical and infinite with respect to
the range of the beta particles.
2. The receptor volume does not perturb the radiation field,
85
3. Any Kr in the lungs or dissolved in the body fat is
compensated by assuming an infinite spherical or 4 ir cloud.
The ICRP (Ref. 33) assumes a 2^ cloud.
Given:
p c.
1. Kr effective energy 0.24 Mev.
2. Temperature - 50° F, pressure 750 mb
D(rem/sec) = E(Mev) 1.6 x 10 6(ergs/Mev)3.7 x 1010(d/Ci-sec)(Ci/m3)
1293(density of air-g/m )100(ergs/g-rad)(rad/rem)
= 0.457 EX
Adjusting the constant for the ambient air con-
ditions, and changing the dose rate to rems per
day—
G-l
-------
D(rem/day) = 5.5 x 104 E(Mev) x (Ci/m3)
5.5 x 104 x 0.24 x 2.5 x 10~12
—8
Dose = 3 x 10~ rem/day
G-2
-------
APPENDIX H
SUMMARY OF PARAMETERS FROM NVO-61 (38)
Rulison Cavity (H+180 days)
Parameter NVO-61 This Report
Cavity Void - 3.05 x 106 2.15 x 106
Volume (ft )
Cracking radius (ft) 485 370
Cavity radius (ft) 90! 80
Cavity pressure (psia) 2,940 2,640
Cavity temperature (°F) 375 380
Flaring Schedule and -Parameters
Parameter NVO-61 This Report
High rate flow test 2 x 10 2 x 107
(ft3/day)a
Duration 3 3—day
Intermediate flow testa 5 x 10
(ftVday)
Duration 2 months
Long term production 5 x 10^
(ft3/day)a
Duration 6-8 months
alndicated as maximum flow rates
H-l
-------
Cavity Inventory of Radionuclides (H+180 days)
Nuclide NVO-61 This Report
Curies Curies
oc 2 9
ODKr 9.6 x 10 9.6 x 10
90Sr 5.9 x 103 5.9 x 103
1 "5 7 3 -a
'Cs 7.5 x 10 7.5 x 10
3H 103 - 104 104
14C 0.01-0.1
37A 10 - 100
39A 2-20
Radionuclide Concentration In Gas (H+180 days)
Nuclide NVO-61 This Report
3H 0.8 - 8 x 10~4 0.07 - 1.5 x 10~3
QC: _c: A
Kr 8 x 10 D 1.4 x 10
\ "7 O
Based on cavity gas volume of 1.27 x 10 m corrected to 10.8
psi and 0° C (4.55 x 108 ft3); this report 2.4 x 108 ft3
(NTP) .
H-2 :
-------
Dose Estimates
Dose Pathway
Inhalation
Adult (at 2 miles)
Food Chain
NVO-61
Max.
Credible
Dose(mrem)
0.3
4 - 100
Estimates in
this Report^
(mrem)
Conservative
Estimate
Expected
10
-3
3
0.1
•n
'Based on maximum credible accident type approach. It was
assumed 94% of the gaseous radioactivity (primarily tritium
o c
and Kr) was released in a 24-hour period.
This calculation was based on flaring at 20 x 10 ft /day
pr
for 10 days. Two thirds of the tritium and Kr in the
cavity are assumed to be released during this period.
H-3
-------
APPENDIX I
ACUTE AND CHRONIC DOSES FROM INGESTION OF TRITIUM
FIGURE 1-1
TRITIUM INGESTION MODEL
Release of
Tritium
from Well
— *
Atmospheric
Moisture
Tritium
Concentration
— +•
Concentrations
of Tritium in
Environmental
Moisture, C.
(Average of
Milk, Vegeta-
tion, and
Water)
Q
Concentration
of Tritium in
Man ' s Body
Water,-. Cm, V
Q
WHERE:
Q
Ci
m
V
Volume of intake and excretion of water by man;
2.2 I/day
Tritium concentration in man's intake water
Tritium concentration in man's body water.
Volume of man's body water; water equivalent of
total hydrogen.
Figure 1-1 indicates the model used to describe the exposure
of man to tritium by the ingestion pathway.
Methods are available for predicting atmospheric concentrations
which are likely to result from flaring natural gas from the
well head. Presence of tritium in the atmosphere will result
in uptake of tritium by vegetation but insufficient data
1-1
-------
are available to directly convert atmospheric tritium levels
to the corresponding levels in vegetation moisture. Conse-
quently experience from Gasbuggy will be used to estimate
tritium levels in vegetation. As discussed in the text, the
highest concentration of tritium in vegetation moisture observed
during environmental sampling for Project Gasbuggy was 36 pCi/ml
of moisture. Allowing for both ,a conservative assumption of
higher levels of tritium in flared gas from Rulison (42 u.Ci/ft
as compared with 18 y,Ci/ft from Gasbuggy) and the higher
expected volumes of flared gas (20 million cubic feet per
day from Rulison as opposed to 5 million cubic feet per day
from Gasbuggy), it is assumed that peak levels of tritium in
vegetation moisture might range up to a factor of ten greater
than occurred with Gasbuggy. The assumed maximum level in '
vegetation moisture might then be 360 pCi/ml of moisture.
Based on information in the text, it is assumed that the con-
centrations of tritium in milk and water will be 36 and 1
pCi/1 respectively.
Assuming a daily human intake of 2.2 liters per day of
moisture composed of 200 ml of vegetation moisture, 1,000
ml of milk, and 1,000 ml of water, the highest expected
weighted average tritium concentration of the total daily
intake would then be about 50 pCi/ml (360 x 200/2200+ 36 x
1000/2200 + 1 x 1000/2200 = 50). The majority of this is
from vegetation, about 1/3 is from milk, and the contribution
from water is over an order of magnitude less than the vege-
tation and milk fraction.
In order to calculate possible internal doses from tritium
ingestion, three simplifying, conservative assumptions will
be made. Assume first that tritium levels in vegetation
moisture equilibrate essentially instantaneously at the peak
level of 360 pCi/ml as soon as high-rate flaring begins.
Next assume that tritium levels in milk rise and fall with
1-2
-------
tritium levels in vegetation moisture without any delay.
Tritium levels in milk therefore are assumed to equilibrate
at 36 pCi/ml as soon as high-rate flaring begins. Also
assume that the levels of tritium in vegetation moisture
remain constant at the peak value of 360 pCi/ml throughout
the three-week high-rate flaring period and then decay
exponentially with an 85-day half-life. The following
indicates the reasoning used in selecting this value:
Tritium Environmental Half-Life
The assumption of an 85-day effective half-life is
based on the following information:
1. The effective half-life for tritium observed
for the previously mentioned sample from Gasbuggy
(peak 36 pCi/ml of water in vegetation) was 85
days (two results with an increment of time of 260
days).
2. Preliminary information from Schooner for a
high desert environment (about 5,000 ft MSL)
indicated values of 70-100 days for desert plants
such as Mormon Tea and Galleta Grass. The 70-100
day half-life occurred subsequent to shorter half-
lifes within the first several months after depo-
sition. A 10-day half-life accounted for about 3/4
of the total depletion*.
3. Effective half-life for tritium used in AEC
evaluation; 28 days (38).
*Personal communication with Dr. B. Mason of SWRHL (presently
employed with BMI, Las Vegas, Nevada).
1-3
-------
4. Martin, et. al. (42) indicated that the mean
residence time in soil for a surface application of
tritium was 37 days.
5. The ecological half-life for 'radionuclides
deposited on vegetation is generally accepted to be
about 14 days (37). This half-life is largely due
to plant growth and would be longer in an arid
environment.
For dry deposition, the predominant case for Rulison,
the. predominant vegetation uptake path should be through
deposition on the vegetation. Deposition on soil and
subsequent root-uptake should result in some dilution by
soil water. It would appear the effective half-life of
tritium in vegetation, contaminated by deposition, would
be equal to or less than the ecological half-life (14
131
days) for other radionuclides such as I, etc. It
is difficult to evaluate the significance of tritium
uptake via the soil-root pathway for the Rulison release
situation and environment, but this would increase the
half-life. The 85-day half-life observed for Gasbuggy
probably reflects the effect of the arid environment.
The Rulison environment is semi-arid. Vegetation for
human or'domestic animal consumption is largely irrigated,
and thus the effective half-life would be less than that
for Gasbuggy.
But due to the lack of specific information and in order
to be reasonably conservative, a value of 85 days for
the effective half-life of tritium will be used. This
may be conservative by a factor of about four.
The discrepancies between these assumptions and actual physi-
cal and biological processes are obvious. Tritium will be
1-4
-------
gradually deposited on vegetation, probably reaching a peak
value at the end of the high-rate flaring period. Levels
of tritium :i.n milk will be affected by the biological half-
life of tritium in cows and will therefore lag behind
increases in environmental levels. These assumptions
result in calculated doses which are conservative*, but
they allow comparison of the acute dose received during the
flaring period with the chronic dose resulting from residual
tritium concentrations in vegetation. More complex models
could be formulated, but the data available do not justify
more complex formulations.
Tritium concentrations in the human water intake and in
human body water which would result from the above assump-
tions would vary with time as shown in Figures 1-2 and 1-3.
The dotted line on Figure 1-2 indicates the general form of
expected actual variation with time of tritium levels in
human water intake.
*The estimation of the acute dose may be conservative by a
factor of two to three (rectangle phase in Figure 1-3),
but this is only a fraction of the total dose.
1-5
-------
I
C G
O rO
U g
3
E.C
D
•H C
-P -H
•H
n d
-P o
•H
(U -P
tn rt
(0 V-i
^ -P
QJ C
> cu
< u
QJ
E
•H
-P C
C -H
U
\
50 pCi/ml = C.(0)
k.
0
Assumed variation of C. with time
Probable actual variation of C.
i
t> 21 days, Ci(t) =
Ci(0)exp- *Jt-21 days)
21 days
TIME (days)
to TD
!^ O
-P A
C,
0) fi
o m
fi g
o a
e c
3 -H
-H H
•P C (U
-H O -P
M -H (0
EH -P £
U
0
21 days
TIME (days)
Using these assumptions, the tritium concentration in man's
body water, C , during the period 021 days, the average tritium concentration in man's
intake water is given by
1-6
-------
C. = C. (Ojexp-X^ (t-21) Equation 1-2
WHERE
\. = decay constant for tritium in vegetation moisture =
ln(2)
85 days
For the period t>21 days, an expression for the tritium con-
centration in man's body water is derived by considering the
body to be a tank or reservoir in which the intake water is
assumed to be completely and instantaneously mixed. Where
Q is the intake water (2.2 liters/day) and V is the volume
of body water (water equivalent of hydrogen), an activity
or mass balance of the tritium flow through the body results
in the differential equation
dC
(Q/V) (C - C^) + ~^ = 0 Equation 1-3
for t>21 days
This equation can be solved through normal techniques to yield
(Q/V)C. (0) r -,
C (t.. ) = exp-ji.t, - exp-t. (Q/V) +
m X (Q/V)- A. L x 1 l J
C (21 days) exp-t (Q/V) Equation 1-4
WHERE
t = t-21 days
C (21 days) = level of tritium in body water at the end
m
of the three -week high-rate flaring period
The quantity (Q/V) in Equation 1-4 is equivalent to a bio-
logical constant. If values of Q=2.2 liters/day and V =
1-7
-------
63 liters are substituted to calculate the corresponding
biological half-life, the value 13.5 days results. The
development above assumes that tritium interacts only with
water, but tritium in the body is also associated with hydro-
carbons in food and is excreted with solids as well as with
liquids, resulting in a biological half-life of 10 days (32).
Substituting (Q/V) = Am = IQ davs ^n E<3uation I~4 simplifies
to
fcm
C (21 days)exp-x tn „ , . _ c
m •* ^ m 1 Equation 1-5
The internal dose resulting from tritium in body water can
be calculated by integrating C over the period of interest
and multiplying the result (in units of pCi-days/ml) by an
appropriate dose conversion factor (4.6 x 10~ millirem/day)/
pCi/ml)*. It is of interest to compute the dose received
during the high-rate flaring period from t = 0 through t = 21
days (the "acute" dose) and compare it with the dose received
from residual tritium left in vegetation moisture after the
end of the high-rate period (the "chronic" dose).
Using C. (0) = 50 pCi/ml and a = " — = 0.0693 days"1
m _LU
and integrating Equation 1-1, the "acute" dose received
during the first 21 days is estimated to be about 0.2 milli-
rem. Using :\. = " . — = 0.0231 days" and integrating
Equation 1-5, the "chronic" dose received from t = 21 days
to infinity is estimated to be about 3 millirem. The
important conclusion of this exercise is that the "chronic"
dose is more than an order of magnitude greater than the
"acute" dose.
*Derived in Appendix G.
1-8
------- |