EPA/ORP 73-1
OBSERVATION OF AIRBORNE
TRITIUM WASTE DISCHARGE
FROM A NUCLEAR FUEL
REPROCESSING PLANT
t ENVIRONMENTAL PROTECTION AGENCY
Office of Radiation Programs
-------
OBSERVATION OF AIRBORNE TRITIUM
WASTE DISCHARGE FROM A
NUCLEAR FUEL REPROCESSING PLANT
\
vjy
Joseph A.Cochran
William R. Griff in, Jr.
Emilio J. Troianello
February 1973
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Radiation Programs
Field Operations Division
Washington.D.C. 20460
-------
FOREWORD
The Office of Radiation Programs carries out a national program
designed to evaluate the exposure of man to ionizing and nonionizing
radiation, and to promote development of controls necessary to protect
the public health and safety and assure environmental quality.
Within the Office of Radiation Programs, the Field Operations
Division conducts programs relating to sources and levels of environ-
mental radioactivity and the resulting population radiation dose.
Reports of the findings are published in the monthly publication, Radi-
ation Data and Reports, appropriate scientific journals, and Division
technical reports.
The technical reports of the Field Operations Division allow
comprehensive and rapid publishing of the results of intramural and
contract projects. The reports are distributed to State and local
radiological health programs, Office of Radiation Programs technical
and advisory committees, universities, libraries and information serv-
ices, industry, hospitals, laboratories, schools, the press, and other
interested groups and individuals. These reports are also included in
the collections of the Library of Congress and the National Technical
Information Service.
Readers of these reports are encouraged to inform the Office of
Radiation Programs of any omissions or errors. Comments or requests
for further information are also invited.
W. D. Rowe
Deputy Assistant Administrator
for Radiation Programs
iii
-------
PREFACE
The Office of Radiation Programs of the Environmental Protection
Agency, in cooperation with the New York State Department of Environ-
mental Conservation, Nuclear Fuel Services, Inc., and the Atomic Energy
Commission, have, over the past several years, conducted studies at the
Nuclear Fuel Services, the nation's first commercial nuclear fuel re-
processing plant. The overall purpose of this study was to determine
the requirements of an environmental surveillance program for fuel re-
processing plants. Specific study objectives included (a) characteri-
zation of both the gaseous and liquid waste effluents from the plant,
(b) measurement of the environmental concentrations of the discharged
radionuclides, and (c) delineation of the critical exposure pathways
and estimation of the radiation doses to the population living near the
plant. Earlier results of this study were published in a series of
reports:
(1) BRH/NERHL 70-1
(2) BRH/NERHL 70-2
(3) BRH/NERHL 70-3
(4) BRH/NERHL 70-4
(5) ORP/SID 72-5
An Estimate of Radiation Doses Received by
Individuals Living in the Vicinity of a
Nuclear Reprocessing Plant
Liquid Waste Effluents from a Nuclear Fuel
Reprocessing Plant
An Investigation of Airborne Radioactive
Effluent from an Operating Nuclear Fuel
Reprocessing Plant
85
Calibration and Initial Field Testing Kr
Detectors for Environmental Monitoring
Iodine-129 in the Environment Around a
Nuclear Fuel Reprocessing Plant
This report presents the results of followup studies on the tritium
discharges from Nuclear Fuel Services, on the concentrations of this
radionuclide in the environment around the plant, and resultant doses to
the population in the vicinity of the plant.
Charles L~. Weaver
Director
Field Operations Division
-------
ACKNOWLEDGMENTS
The assistance and cooperation of the following individuals in
planning and carrying out this study is gratefully acknowledged.
John Russell
Office of Radiation Programs
Environmental Protection Agency
Floyd Gal pin
Office of Radiation Programs
Environmental Protection Agency
Charles Weaver
Office of Radiation Programs
Environmental Protection Agency
Thomas Cashman
Bureau of Radiological Pollution Control
New York State Department of Environmental Conservation
Gene Michaels
Bureau of Radiological Pollution Control
New York State Department of Environmental Conservation
Robert Wozm'ak
Bureau of Radiological Pollution Control
New York State Department of Environmental Conservation
Edward North
Nuclear Fuel Services
Terry Wenstrand
Nuclear Fuel Services
Dave Wilcox
Nuclear Fuel Services
VI
-------
CONTENTS
Page
FOREWORD ill
PREFACE v
ACKNOWLEDGMENTS vi
ABSTRACT ix
INTRODUCTION 1
SAMPLING DESCRIPTION 1
In-plant sampling operations 1
Field sampling operations 3
RESULTS AND DISCUSSION 3
In-plant sample results 3
Weekly water vapor samples 3
Dissolution samples 6
Campaign samples 8
Tritium inventory 10
Discharge estimates 11
Field sampling results 12
Dose estimates 15
CONCLUSIONS 16
REFERENCES 19
APPENDIXES
Appendix A. Sampler description 23
Appendix B. Plant data 25
FIGURES
Figure 1. Field tritium sampling locations.
Figure 2. Comparison of weekly tritium stack output with weekly
burnup inventory of fuel processed
vii
-------
TABLES
Page
Table 1. Tritium stack output in the form of water vapor for
campaign periods 6
Table 2. Tritium in stack gas and water vapor for Humboldt Bay
fuel batch numbers 7 , 8, and 9 7
Table 3. Tritium in gas evacuated tank samples, Humboldt Bay batch
numbers 7 and 8 9
Table 4. Tritium stack output during campaign periods 9
Table 5. Tritium inventory for Humboldt Bay campaign 10
Table 6. Summary, stack tritium effluent results 11
Table 7. Weekly air concentrations of tritium around NFS during
1971 13
Table 8. Average tritium in air concentrations around NFS during
summer of 1971 14
Table 9. Specific activity of airborne water vapor at NFS during
summer of 1967 14
Table 10. Average tritium in produce composites around NFS during
summer of 1970 15
Table 11. Whole body dose from airborne tritium 16
Table B-l. Campaign data for fuel inventories processed during
tritium study at NFS during 1971 25
Table B-2. Average X/Q values for the NFS site during 1971 27
Vlll
-------
ABSTRACT
A study was conducted at Nuclear Fuel Services, Inc. (NFS) to:
(a) characterize the stack tritium effluent in the gaseous and water
vapor forms, during normal plant operations, (b) determine the weekly
average tritium air concentrations at five selected sites around the
plant over a 4-month period, (c) evaluate methods used for sampling at
the stack and in the environment, and (d) estimate the dose to the
population in the immediate vicinity of the plant from tritium stack
effluents.
Characterization of tritium release from the stack was carried out
during three reprocessing campaigns assumed to be typical of the NFS
operation. Evaluation of the data indicates that the tritium release
rate from the stack is 6.75 x 10"1* curies per megawatt-day burnup.
Approximately 25 percent of the activity is in the gaseous state and
75 percent is in the water vapor state. The average annual release of
tritium via the stack, considering the total fuel processed from 1966
to 1971, is estimated to be 500 curies per year. Data for the five air
sampling stations around the site indicate that the plant contribution
to the atmosphere is small.
ix
-------
OBSERVATION OF AIRBORNE TRITIUM WASTE DISCHARGE
FROM A NUCLEAR FUEL REPROCESSING PLANT
INTRODUCTION
During the reprocessing of nuclear fuel, tritium, a fission
product originating in the fuel elements, is released to the environ-
ment 01). The tritium may be released in the gaseous effluent via the
stack and/or in the low-level liquid wastes. If discharged princi-
pally to the atmosphere, tritium is expected to be a predominant
contributor to the local population whole body radiation dose ascrib-
able to a fuel reprocessing plant. The objectives of the present study
were to:
(a) determine the nature of tritium released from a fuel
reprocessing plant,
(b) understand the relation of this release to population
dose, and
(c) develop specific methods for detecting tritium onsite
and in the environment around a fuel reprocessing plant.
There have been many studies dealing with the sources of tritium
and transport of this nuclide in the environment (1-3). These studies
have added valuable information to the general subject of tritium
impact in the environment. Studies to date at Nuclear Fuel Services,
Inc. (NFS), a fuel reprocessing plant located at West Valley, N.Y.,
have derived information concerning tritium in liquid waste discharge
from the plant (b) and levels of airborne tritiated water vapor con-
centration (activity/volume of water) in the environment surrounding
the plant (5) . In addition to these studies, some preliminary work
was performed to determine the amount of tritiated water vapor dis-
charged through the NFS stack during the fuel dissolution cycle (6).
The present study was undertaken at NFS to supplement the studies
cited above.
SAMPLING DESCRIPTION
In-plant sampling operations
In-plant sampling was carried out to characterize the tritium
stack effluent. A sampler was designed to selectively collect the
water vapor and the gaseous components of tritium from the stack ef-
fluent. The tritium sampler was operated in parallel with the NFS
stack monitoring system which allowed isokinetic sampling at the
80-foot level of the 200-foot stack (6)-
-------
The tritium sampler was a flow-through system consisting of a
desiccant which removes water vapor, a catalytic burner which oxidizes
hydrogen compounds to water, and a second desiccant which removes the
oxidation products in the form of water vapor. Carrier hydrogen was
used in the burner stage to provide an adequate volume of combustion
products for collection in the second desiccant stage (7) -
The first desiccant stage of the system was operated when only a
water vapor sample was of interest. The burner and second desiccant
were put on-line when either a gaseous tritium sample or simultaneous
water vapor and gaseous tritium samples were desired. The sampling
train contained taps located after the first desiccant stage which
enabled 2.8 liter evacuated tank grab samples of the gaseous tritium
components to be obtained. A detailed description of the basic tritium
stack sampling system and analytical technique appears in appendix A.
Stack sampling was carried out over a 6-month period from June to
November 1971. During this sampling period, the plant processed four
separate fuel inventories or campaigns. These campaigns are considered
to be representative of the normal type of fuel reprocessing carried
out at NFS. A listing of campaigns and associated information on the
fuel inventories are given in appendix B.
The specific samples that were obtained are listed below:
(a) Weekly stack water vapor samples were obtained from June 7,
1971 to November 30, 1971. Twenty-five samples were collected
in this period. The purpose of this sampling was to obtain
weekly stack concentrations for correlation with fuel through-
put and offsite tritium air concentrations during the same
period.
(b) Simultaneous water vapor and gaseous tritium samples were ob-
tained during the sixth, seventh, and eighth dissolution runs
of the Humboldt Bay fuel campaign in order to measure total
tritium in-stack during a dissolution cycle.
(c) Simultaneous water vapor and gaseous tritium samples were
also obtained during the complete campaigns of the Parr and
Big Rock Point fuel inventories. These data were necessary
in order to determine the relative tritium concentrations
present as gas (HT) and water vapor (HTO) during a complete
reprocessing campaign.
(d) Gas samples were obtained during the three Humboldt Bay dis-
solutions mentioned above, using evacuated tanks. These
samples consisted of dried stack gas containing the gaseous
tritium component. The purpose of these samples was to cross-
check the values obtained from the desiccant samples obtained
during the same time period.
(e) A low-level liquid waste composite sample was obtained at the
interceptor for the entire Humboldt Bay campaign period. This
sample was used with the stack samples obtained during the
Humboldt Bay campaign in order to estimate the total tritium
inventory during a campaign period.
-------
Field sampling operations
The field tritium air samplers collected water vapor by use of a
desiccant and operated similar to the first stage of the stack tritium
sampler. A detailed description of the field sampler appears in
appendix A.
Five samplers were located around the perimeter of the NFS site.
Figure 1 is a map showing the location of each station in terms of
azimuth and distance from the plant. The average station distance was
approximately 2.6 kilometers (1.6 miles) from the plant whereas the
average property line distance is approximately 1.9 kilometers (1.2
miles). A control station, located in Winchester, Mass., was operated
on a weekly basis as were the stations around NFS.
Samples were collected weekly from June 7 to October 4, 1972, a
total of 17 weeks. During the sampling period, wind-rose data were
accumulated in order to determine if any correlation existed between
these data and the observed tritium levels at the field stations.
RESULTS AND DISCUSSION
In-plant sample results
Weekly water vapor1 samples
Tritium water vapor samples were obtained from the stack during
the period of June 7 to November 30, 1971. Four campaigns were proc-
essed during this period. Figure 2 shows a comparison of the weekly
tritium stack concentrations in the vapor form with the weekly plant
fuel processing load expressed in terms of burnup. The tritium ef-
fluent data were displaced forward in time by 1 week to account for
the lag between dissolution of the fuel and the effluent discharge as
water vapor. Since the amount of tritium present in the spent fuel is
essentially constant when expressed in terms of burnup, the quantity
discharged per megawatt-day (MWD) should be relatively constant. This
comparison (figure 2) indicates that the quantity of tritium vapor dis-
charged per MWD is not constant and may be dependent on other factors
such as fuel cladding, tritium inventory in the process solutions from
previous campaigns, and specific processing operations at any given
time. However, there is a general correlation where the tritium vapor
effluent increases with an increase in the burnup of the fuel being
processed.
The tritium water vapor yield for each campaign is given in
table 1 in terms of curie per MWD burnup. These values were derived
by summing the weekly activity values over the entire campaign period.
turnup is an expression of megawatt-days of energy generated by
the fuel.
-------
Figure 1. Field tritium sampling locations
-------
32
ffi 24
CC
D
U
HI
0
16
r
u
0) 8
HTO STACK DISCHARGE
I JUNE I JULY I
AUG
SEPT I OCT
NO V
10
s
Q.
D
Z 4
o:
D
CD
FUEL PROCESSED
JUNE I JULY I AUG I SEPT I OCT
1971
NOV
Figure 2. Comparison of weekly tritium stack output
with weekly burnup inventory of fuel processed
-------
Table 1. Tritium stack output in the form of water vapor
for campaign periods
Campaign
Humboldt Bay
Q
Yankee Rowe
Big Rock Point6 —
Release
per unit
burnup
(Ci/MW(t)-day)
3.1 x 10~4
2.0 x 10~4
-4
4^, -a- 1 A
. J X -LU
7.9 x 10~4
Cladding
material
Stainless steel
Stainless steel
Zircaloy-2 and
Iconel
Fuel age
prior to
reprocessing
(years )
3.5
.8
A Q
1.6
Based upon weekly tritium water vapor composites throughout the
entire campaign periods. The combined analytical and sampling la errors
for the weekly samples that made up the composites are less than 12 per-
cent. Burnup, cladding type, and age data from plant records (90.
A. 240 MW(t) General Electric Boiling Water Reactor operated by the
Pacific Gas and Electric Company.
£
A 600 MW(t) Westinghouse Pressurized Water Reactor operated by the
Yankee Atomic Electric Company.
The Carolinas-Virginia Tube Reactor, a 64 MW(t) pressure tube,
heavy water reactor operated by Carolinas-Virginia Nuclear Power Associ-
ates , Inc., from 1963 to 1967.
Q
A 240 MW(t) Boiling Water Reactor operated by Consumers Power
Company.
This yield varies by a factor of four and appears to be subject to
other factors, as discussed in the previous paragraph. Table 1 also
lists the fuel cladding material and the delay time between removal
from the reactor and reprocessing. The data shows a greater yield of
tritium as water vapor from the Zircaloy clad fuel than from the
stainless steel clad fuel.
Dissolution samples
The results of the simultaneous gas and water vapor tritium
samples obtained during three consecutive dissolutions of the Humboldt
Bay campaign are shown in table 2. The burnup (MWD) for the three dis-
solution runs are sufficiently close for comparisons to be made between
the three dissolution runs.
-------
Table 2. Tritium in stack gas and water vapor for Humboldt Bay
fuel batch numbers 7, 8, and 9
Total burnup
("MUYM rlava^
Sampling time
Chemical form
Water vapor
concentration
(pCi/cm3 of air)
Total activity
as water vapor
Gas concentration
(yCi/cm3 of air)
Total activity as
Batch number3
7
9.7 x 103
8.3
1.6 x 10~7
.9 x icf1
2.0 x 10"6
1.1 x 10+°
8
9.8 x 103
8.0
2.2 x 10~7
1.2 x 10"1
b1.9 x 10~6
b1.0 x 10+°
9
8.8 x 103
19.4
2.2 x 10~7
2.8 x 10'1
.9 x 10~6
1.2 x 10+°
Background
0
7.0
l.l x 10~7
5.2 x 10~2
4.0 x 10~9
1.9 x 10~3
The combined analytical and sampling la errors are less than 12
percent except where noted by (b) the errors are less than 22 percent
of the stated value.
The dissolution cycle is defined as the period of time starting
when the chopped fuel is immersed in the acid bath and ending when the
dissolved acid mix is transferred from the dissolving container. This
time period varies from 18 to 48 hours and is generally around 24 hours.
Sampling commenced at the beginning of the dissolution cycle and lasted
8 hours for the dissolution batches 7 and 8, and 20 hours for dissolu-
tion batch 9. The tritium stack background sample was obtained during
a non-dissolution period, but fuel chopping operations were carried
out during the background sampling run.
The tritium water vapor concentration measured during the three
dissolutions averaged 2.0 x 10~7 pCi/cm3. The evolution rate of trit-
ium as water vapor is fairly constant throughout the dissolution cycle
as shown by comparing the 8-hour runs with the 20-hour run in terms of
curies per hour.
-------
Batch 7 (8 hours) 1.1 x 10~2 Ci/h
Batch 8 (8 hours) 1.5 x 10~2 Ci/h
Batch 9 (20 hours) 1.4 x 10~2 Ci/h
The rate of release of the tritium from the fuel is not well
defined but is assumed to occur during the early stage of the dissolu-
tion cycle. Once released from the fuel, the tritium as water becomes
part of the dissolver solution. The evaporation of water vapor from
the dissolver solution into the off-gas line is dependent upon the
temperature of the solution. Since the temperature of the solution
is maintained at a constant level throughout the dissolution cycle, the
evaporation rate of the tritium as vapor is also constant.
The concentration of gaseous tritium measured during the three
dissolutions of the Humboldt Bay campaign averaged 2.0 x 10~6 yCi/cm3
air for the 8-hour runs and 9.0 x 10~7 for the 20-hour run. The vari-
ation in concentration between the 8- and 20-hour observations indi-
cates that the evolution of the gaseous tritium is not constant through-
out the dissolution cycle. The larger fraction of the gaseous tritium
is released from the fuel within the first 8 hours of the cycle. Un-
like the tritium in the vapor form, the gaseous tritium does not become
part of the dissolver solution but goes directly to the off-gas system.
Therefore, the tritium release in the gaseous form was related to fuel
burnup of the dissolution batch since its release is not a function of
time.
Batch 7 (8 hours) 1.1 x 10~^ Ci/MWD
Batch 8 (8 hours) 1.0 x lO"4 Ci/MWD
Batch 9 (20 hours) 1.3 x 10"^ Ci/MWD
Tank samples were obtained during the dissolution of fuel batches
7 and 8 to determine the gaseous tritium evolution rate. The sampling
consisted of constant low flow-rate withdrawal into evacuated tanks for
periods of approximately 3.7 hours. Two samples were obtained during
each dissolution. The results of these samples are shown in table 3.
Sampling commenced when dissolution started and the sampling times
relative to dissolution startup are given. The concentrations obtained
were used to project a normalized output over the sampling period. For
batches 7 and 8, approximately 80 percent of the total activity output
of gaseous tritium occurred during the first one and one-half hours of
the sampling period. This type of output is similar to the krypton-85
gaseous output observed previously (6) .
Campaign samples
Simultaneous water vapor and gaseous tritium samples were obtained
during the Parr and Big Rock Point campaigns. These samples permitted
a comparison to be made between the two tritium forms over an entire
campaign period. The samples were continuous and yielded the results
shown in table 4. The ratio of tritium gas to water vapor was 0.66
for the Parr campaign and 0.09 for the Big Rock Point campaign. The
8
-------
Table 3. Tritium in gas evacuated tank samples,
Humboldt Bay batch numbers 7 and 8
Dissolution
batch
7
7 _
8 - - - -
Q
Sampling time
relative to
dissolution
startup
(hours)
01 A
-L • O
1 f. T C
n i A
1 A 07
Concentration
(yCi/cm3 of air)
i i "; •«• i n~
-L . J. J X J.U
i f. ,, i n~
1 01 „ i o~
on v 1 0
Percent of
normalized
stack output
Of.
i /,
fin
?0
Table 4. Tritium stack output during campaign periods
Campaign
p.--.-. PVMPA
Big Rock Point
Total burnup
(MW(t)-days)
34,408
80,483
Activity released
(curies)
Vapor
15.6
63.2
Gas
10.3
5.4
Total
25.9
68.6
Per MWD
7.5 x 10~4
8.5 x 10~4
Based on composite samples over the entire campaign period. Sam-
pling and analytical la errors < 15 percent.
-------
reason for this variation is unknown. It is interesting to note that
even though there is a large variation in the gas to water vapor ratio,
the total tritium yield per MWD for each campaign is of the same
magnitude.
Tritium inventory
The samples collected during the Humboldt Bay campaign enable a
projection to be made of the total tritium inventory for this campaign.
The previously mentioned gaseous stack samples obtained during three
dissolution cycles and the weekly water vapor samples obtained during
the final 5 weeks of the campaign are the basis for the projected air-
borne tritium inventory. The liquid waste composite sample collected
throughout the entire campaign enabled projection of tritium inventory
of the liquid waste effluent. Some of the tritium remains in the
stored high-level liquid waste. This is assumed to be 5 percent of
the total tritium available in the fuel at reprocessing based upon
volume ratio (8) .
Table 5 shows the Humboldt Bay campaign fuel inventory in terms
of curies per megawatt-days. Based upon total tritium yielded from
the fuel, the stack tritium contribution is 7 percent while the
liquid effluent contributes 88 percent. These values differ from the
safety report projection of 25 percent to the stack and 65 percent to
the liquid effluent (9) .
The last column of table 5 shows the tritium inventory as a per-
cent of the total tritium produced in the reactor per megawatt-day
(1 x 10~2 Ci/MWD) (9). The inventory shows a deficit of 40 percent
Table 5. Tritium inventory for Humboldt Bay campaign
Source
"H-i o-Vi 1a-iTaT 1 -i /-Mi-i A
Assumsd loss** ~
Ci/MWD
c oo ,, i n
bo -10 ... lfl-4
J. J-O X J_U
-L.
3 no -a- i n
. uy x j.u
i i a v i n
3d o „ i n~
• y j x j_u
Percent of
total from fuel
fiB
oo
Percent of
predicted
c o
J J
An
40
aBased upon a total yield of 1 x 10 Ci/MWD for 20,000 MWD/tonne
burnup (9)-
Assumed based upon plant records (8).
10
-------
which was not accounted for during reprocessing. The deficit consists
primarily of tritium which diffused through the cladding while the
fuel was in the reactor and then stored prior to reprocessing. Some
of the deficit can also be contributed to tritium off-gassed from the
nitric acid recycle storage tank, buildup in the uranium product, and
hydride associated with the cladding.
Discharge estimates
The in-plant data have been summarized in table 6. The data
presented give actual measured values for stack releases of tritium in
both the gaseous and vapor forms for four campaigns considered repre-
sentative of typical fuel types normally processed at this plant.
Based upon the data in table 6, the average tritium stack output
is 6.75 x lO"1* Ci/MWD which consists of 66 percent in the water vapor
state and 24 percent in the gaseous state.
The total fuel burnup processed by NFS over the past 6 years is
4.1 x 106 MWD, or 6.8 x 105 MWD/year (8,10). The projected tritium
stack output for the same periods are 2.8 x 103 curies total or
4.6 x 102 Ci/year based on the Ci/MWD factor derived from table 6.
Likewise, using the data derived from the Humboldt Bay inter-
ception sample (table 5), the liquid discharge rate of tritium amounts
to 5.33 x 10~3 Ci/MWD. Using this discharge rate, the total liquid
discharge of the plant for 6 years of operation is 2.2 x 101* curies or
3.6 x 103 Ci/year. This value correlates closely with present pro-
jections of the liquid tritium plant discharge (10).
Table 6. Summary, stack tritium effluent results
Campaign
Humboldt Bay —
Yankee Rowe
Big Rock Point
Total
burnup
(MW(t)-
days)
151,328
224,730
1A "}fift
80,483
Gas
(Ci)
a!7.7
-
i n "}
5.4
Vapor
(Ci)
46.8
45.8
1 5 fi
63.2
Total
(Ci)
64.5
-
?S Q
68.6
Gas
(Ci/MWD)
1.2xlO~4
-
1 fWlfl
.6xlO~4
Vapor
(Ci/MWD)
3 . lxlO~4
2.0xlO~4
-4
A s-y-1 n ^
7.9xlO~4
Total
(Ci/MWD)
4.3xlO~4
-
7 ^-vl fi
8.6xlO~4
Based upon three dissolution cycles. All other values based on
measurements throughout the entire campaign periods.
11
-------
Field sampling results
The tritium field samplers were placed at five locations around
the plant boundary. The exact location of each sampler is shown in
figure 1. Weekly samples were obtained during June 5 to October 4,
1971. During this time period, the plant processed a portion of the
Humboldt Bay fuel and the entire Yankee Rowe fuel campaign. The weekly
results of all tritium concentrations from each field station are shown
in table 7. Station numbers 1 through 5 correspond to the locations
shown in figure 1 and station number 6 is the control station located
at Winchester, Mass.
The air concentrations of tritium observed at each station
represent background levels plus plant-related contributions. Plant
contributions include evaporation from the liquid waste lagoon system,
receiving streams, and solid waste disposal area as well as stack
discharge. On a weekly basis, the field stations do not show any
significant contributions of airborne tritium that can be related to
the plant. The plant contribution is not observed because it is small
compared to the normal week-to-week fluctuations of the background
tritium.
The weekly data have been compared with weekly wind-rose data (11)
in order to determine if a trend of concentration variation could be
associated with the general wind patterns at the plant site but no
correlation can be demonstrated. Relative humidity effects are minimal.
Some dilution is evident at high humidity levels but the air tritium
concentration remains the same.
The average tritium concentrations for the entire study period at
each station are shown in table 8. These average values are derived
by compositing the weekly concentrations. Values are given both in
terms of activity per volume of water vapor sampled and activity per
volume of air sampled. In general, the levels observed at the five
stations are within the background range observed at the control station
The ratio of air concentration to water vapor concentration is constant
for all stations including the control, with a mean value of 1.2 x 10~5
yCi/cm3 air per yCi/cm3 water.
The average air concentrations of tritium in terms of activity
per volume of water vapor can be compared with observations of Daly
et al. (5) made at equivalent locations around NFS during plant oper-
ation in 1967. Daly's data consists of 3-day samples using bulk desic-
cant at multiple stations. Table 9 is a presentation of the average
levels of tritium observed from July 17 to September 23, 1967. Dis-
tance from the plant is stated as a range since multiple stations were
used in each direction shown. Daly's results are very close to our
1971 values. Both the 1971 and 1967 data indicate higher average
tritium concentrations northwest of the plant when compared to the
other sample locations.
12
-------
Table 7. Weekly air concentrations of tritium
around NFS during 1971
Week of:
Timo 7R —
Tll1-\7 ^ _ _
Till v 1 7 —
Till IT 1 Q —
Til ITT 9 A _ _
September 13 ~
oepLeniDer L\J ~
beptemoer LI
Concentration9
10"5 pCi/cm3
Station
1
4.3
1.9
2.2
3.4
2.6
2.2
1.9
2.9
2.3
2.2
NS
NS
1.8
2.3
.7
.8
1.6
Station
2
2.0
1.8
2.2
3.2
2.8
1.9
2.2
1.8
1.9
2.4
2.6
1.5
1.4
1.6
.6
.7
1.0
Station
3
1.2
2.3
2.1
2.4
1.2
1.4
1.6
2.6
1.9
2.5
2.7
2.0
1.6
1.3
1.0
.7
1.0
Station
4
1.9
1.5
1.8
2.0
NS
NS
NS
NS
3.1
2.0
1.8
1.0
2.1
8.6
.9
.6
1.0
Station
5
0.9
4.4
8.1
5.9
5.3
4.1
7.8
3.4
3.0
4.1
4.3
3.6
3.7
6.7
2.9
1.7
4.1
Station
6
'control )
0.9
9.2
6.6
6.7
2.4
2.2
2.1
1.8
1.3
.8
1.3
1.6
1.6
.4
.6
.4
.9
ihe combined analytical and sampling la errors are < 15 percent.
NS, no sample collected.
13
-------
Table 8. Average tritium in air concentrations
around NFS during summer of 1971
Station
(EPA no.)
1 _
9 —
/._ .
^ _
fi -
Direction
from plant
NNE
NE
SE
SW
NW
Control
(km)
3.4
2.0
2.8
2.4
2.8
(a)
Average concentration
/ pCi \
\cm* H2(y
1.9 x 10°
1.8 x 10°
1.7 x 10°
1.3 x 10°
3.3 x 10°
2.4 x 10°
/ pCi \
Vcm3 air/
2.2 x 10~5
1.9 x 10~5
2.1 x 10~5
1.6 x 10~5
4.4 x 10~5
2.4 x 10~5
Concentration
ratio of
air to water
1.2 x 10"5
l.l x 10~5
1.2 x 10~5
1.2 x 10~5
1.3 x 10~5
1.0 x 10~5
Control located at Winchester, Mass.
Table 9. Specific activity of airborne water vapor
at NFS during summer of 1967 (5_)
Station
F
Direction
Ml?
QT7
QTftT
"MTxT
Distance
(km)
o /, f, /,
9 n ^ i
z . u j . j
O /, c c
8f\ L
Q /, c:
Average
specific
acti vi ty
(pCi/cm3 H20)
9 7 -v i n
10 __ T n
.OX XU
i 7 -v i n
i A v i n
S ^ v 1 0
14
-------
Further evidence of elevated tritium in this direction is available
from produce composite samples (12) obtained at five locations around
the NFS plant, corresponding to the five tritium air sampling stations.
The total tritium concentrations from the farm vegetable composites
are shown in table 10. The farm sample northwest of the plant shows
an elevated tritium concentration relative to the other four locations.
The tritium level northwest of the plant cannot be explained in spe-
cific terms because necessary detailed pathway data are not available
to describe this phenomena. However, we can assume that the northwest
direction is influenced somewhat by evaporation from the Buttermilk
Creek, the main pathway for tritium discharged as liquid waste from
the NFS lagoon system, and beyond this we can only speculate on the
micrometeorology in this sector that possibly influences this phenomena.
Dose estimates
The tritium data obtained from this study enables the projection of
dose directly from the field measurements as well as projecting the
contribution to dose from the stack tritium discharge.
The dose calculations are based upon the International Commission
on Radiation Protection (ICRP) recommendations (13) with a correction
factor of 1.5 to account for the increase in dose due to organic label-
ing from chronic exposure (14). A quality factor of 1.0 for tritium
is assumed in these calculations (15).
Table 10. Average tritium in produce composites
around NFS during summer of 1970
Station
T7o t-m "}
T?OY-TTI 9 —
r drm j-
Di rection
NNF
NF
CTT
cy
MU
NF
Distance
from plant
(km)
-3 f.
9 0
9 ft
? u
i n
24
Tritium
concentration3
(pCi/g wet weight)
n f. + n -i
f\ + ^
7 + i
U + 9
11+ L
S + ~\
aError is the 2a error due to counting statistics
15
-------
Total body dose = 1.7 x 103 (X) yrem/year
Where
X = average tritium air concentration (pCi/cm3 air).
Table 11 shows the calculated whole body dose from tritium around
NFS. The annual dose is based on the average tritium concentrations
observed at each station location (table 8) projected to an annual
average. The stack dose contribution is based upon the estimated annual
stack output of 460 curies of tritium per year and is projected to the
property line in the direction of each sampling station using current
wind-rose data obtained at the NFS stack (11). (See appendix B-2.)
These dose estimates show that the NFS stack contribution was 0.5 per-
cent to 3 percent of the total dose from environmental tritium.
It is not possible from the present study to ascertain the air-
borne environmental dose contribution from evaporation of the liquid
tritium effluent.
Table 11. Whole body dose from airborne tritium
(microrem per year)
Station
i
? - -
0
A -
5__
f. _
Direction
"MTJF
NF
QTT
QLT
MTJ
From all
sources
fin
ou
<^n
f,r\
DU
An
tu
iin
-L-LU
f.r\
DU
NFS stack
contribution
1 C
i c;
CONCLUSIONS
Tritium in airborne effluents consists of both the water vapor
and the gaseous form. The water vapor form is discharged continuously
during a campaign period, and originates from dissolution and various
evaporative processes ongoing during a campaign. The discharge rate
of the water vapor tritium is 5 x 10-*+ Ci/MWD of fuel processed. The
16
-------
gaseous tritium is primarily discharged during the initial dissolution
of fuel although a small amount is discharged during the fuel chopping
operation. The discharge rate of the gaseous tritium is approximately
2 x 10~4 Ci/MWD. Based upon a reprocessing production schedule of
6.8 x 105 MWD per year, the total tritium stack discharge rate is ap-
proximately 15 yCi/s.
The impact of tritium as stack discharge is minimal when compared
to the normal environmental level of airborne tritium. Based upon
measurements from this study, the average tritium background around NFS
is approximately 2.4 x 1CT5 pCi/cm3 air and the stack contribution is
estimated to be from 0.5 to 3 percent of the background tritium concen-
tration.
In terms of dose, the stack discharge contributes a total body
dose of less than 2 yrem annually outside the plant boundary. The air
concentration from evaporation of liquid discharges has not been esti-
mated but is assumed to be less than the stack contribution with the
possible exception of the sector northwest of the plant. In this sec-
tor, due to the close proximity of the Buttermilk Creek, evaporation
from the creek may contribute to the elevated tritium air concentra-
tions observed. This cannot be confirmed from the present study since
the levels in the sector, although elevated, are still within the back-
ground range.
The samplers used during this study proved to be both dependable
and adequate for the designed purpose of sampling tritium in both the
vapor and gaseous states. The desiccant approach to the collection of
water vapor is straight-forward and simple, which is advantageous under
field conditions. The combustion-desiccant device used for the col-
lection of both gas and vapor components of tritium in the stack proved
dependable over long periods of sampling (30 to 40 days) which permitted
long-term samples to be collected and also permitted continuous sampling
of the stack effluent over an entire campaign period.
17
-------
REFERENCES
(1) JACOBS, D. G. Sources of tritium and its behavior upon release to
the environment, AEC Report TID-24635 (1968).
(2) OSBORNE, R. V. Monitoring reactor effluents for tritium: Prob-
lems and possibilities, AECL-4054. Chalk River Nuclear Labora-
tories (September 1971).
(3) OSTLUND, H. G. Tritium in the atmosphere and ocean. Paper pre-
sented at the International Tritium Symposium, Las Vegas, Nev.
Rosensteel School of Marine and Atmospheric Science, University of
Miami (1971) -
(4) MAGNO, P. J., T. REAVEY, and J. APIDIANAKIS. Liquid waste ef-
fluents from a nuclear fuel reprocessing plant, BRH/NERHL 70-2.
Northeastern Radiological Health Laboratory, Winchester, Mass.
(1970).
(5) DALY, J. C., et al. Tritiated moisture in the atmosphere sur-
rounding a nuclear fuel reprocessing plant. Radiol Health Data
Rep 9:341 (1968).
(6) COCHRAN, J. A., D. G. SMITH, P. J. MAGNO, and B. SHLEIEN. An
investigation of airborne radioactive effluent from an operating
nuclear fuel reprocessing plant, BRH/NERHL 70-3. Northeastern
Radiological Health Laboratory, Winchester, Mass. (July 1970).
(7) GRIFFIN, W. R., J. A. COCHRAN, and A. A. BERTUCCIO. A sampler for
non-aqueous tritium gases, presented at the International Tritium
Symposium, Las Vegas, Nev. Environmental Protection Agency, ORP,
SID, Winchester, Mass. (1971).
(8) NUCLEAR FUEL SERVICES. Plant records of fuel processing data and
fuel burnup data. Nuclear Fuel Services, Inc., Health and Safety
Staff, West Valley, N.Y.
(9) NUCLEAR FUEL SERVICES. Spent fuel processing plant - preliminary
safety analysis report, Part B, AEC Docket 50-201. Nuclear Fuel
Services, Inc., West Valley, N.Y. (July 1962).
(10) MAGNO, P. J. and J. A. COCHRAN. Environmental aspects of radio-
active waste discharge from a nuclear fuel reprocessing plant.
American Public Health Association Proceedings, 99th Annual Meeting
(October 1971).
(11) ATOMIC ENERGY COMMISSION. Record memorandum placed in AEC Docket
50-201 file of telecon between W. H. Ray, Irradiated Fuels Branch,
AEC, and Dr. E. D. North, Nuclear Fuel Services on April 7, 1972.
AEC Public Document Room, Washington, D.C. 20545.
19
-------
(12) MAGNO, P., R. KRAMKOSHI, T. REAVEY, and R. WOYNICK. Studies of
dose pathways from a nuclear fuel reprocessing plant, environ-
mental surveillance data at NFS, summer 1970, field operations.
Environmental Protection Agency report (to be published).
(13) INTERNATIONAL COMMISSION ON RADIOLOGICAL PROTECTION. Report of
Committee 2 on permissible dose for internal radiation, ICRP Pub-
lication 2. Pergamon Press, London (1959).
(14) EVANS, A. G. New dose estimates from chronic tritium exposures.
Health Phys 16:57-63 (1969).
(15) NATIONAL COUNCIL ON RADIATION PROTECTION AND MEASUREMENTS. Basic
radiation protection criteria, NCRP Report 39. National Council
on Radiation Protection and Measurements, Washington, D.C. 20008
(January 15, 1971).
20
-------
APPENDIXES
-------
APPENDIX A. SAMPLER DESCRIPTION
Tritium sampler description
The following two sections of this appendix describe the tritium
samplers used both at the stack and in the environment during the
tritium study at NFS.
A more detailed discussion of the sampler and its operation, in-
cluding the laboratory testing and calibration will be found in refer-
ence (_7_) , "A Sampler for Non-Aqueous Tritium Gases" by Griffin, et al.
Stack tritium sampler
The sampler consists of the following items:
(a) Two desiccant traps each with a capacity of 60 grams of water.
The desiccant consists of approximately 1,000 grams of anhydrous
calcium sulphate, 8 mesh (Hammond Drierite Company).
(b) One catalytic burner consisting of a trap containing 0.5 per-
cent platinum on aluminum oxide pellets (Englehard Industries,
Inc.). The trap is inserted into a heating jacket and the
system operates at 400° C.
(c) One tank of 2 percent hydrogen in air-carrier gas.
(d) Various flow meters, manometers, valves, pump, and other hard-
ware.
The system samples filtered stack effluent at a flow rate of 0.2
liters per minute. The stack gas initially passes through a flow meter
and then into the first desiccant. Moisture is removed in this trap to
an equivalent -79° C. dewpoint (0.003 mg/liter). The dried stack gas
existing from the desiccant trap is mixed with carrier hydrogen, and
this mixture is injected into the catalytic burner.
The burner oxidizes the gaseous hydrogen compounds contained in
the stack gas, as well as the carrier, to the water vapor state. The
hot gases leave the burner and pass through a copper tubing coil which
allows the gas to cool. The cooled gas passes through the second
desiccant trap where water vapor produced in the burner is collected.
The stack gas and carrier gas flow rates can be varied so that
adequate sample size can be obtained over short-term periods of several
hours to periods of several weeks. In situations where only water
vapor is to be sampled, the carrier gas, burner, and second desiccant
trap are removed from the system.
23
-------
Water vapor collected in the desiccant traps is removed using a
specially designed dehydration unit. The dehydration process is very
efficient (97 percent recovery of activity) and yields a water sample
which is then filtered and distilled to remove any suspended desiccant
material and insure a neutral pH. The sample is analyzed using stand-
ard liquid scintillation technique.
Field atmospheric tritium sampler
The field sampler operates essentially as the first stage of the
stack unit, sampling only water vapor. Unlike the stack sampler, this
sampler operates over weekly periods under varying relative humidity
conditions. To accommodate for this variability, three desiccant traps
are utilized in series at a flow rate that allows for a reasonable
sample during a dry week and the capability to collect a large sample
without saturating the total trap capacity during a wet week.
The sampler consists of the following:
(a) three desiccant traps each with a capacity of 60 grams of
water,
(b) a flow meter, and
(c) an air pump.
During any given sampling period, only traps that contain sampled
water vapor are dehydrated and analyzed. The remaining traps are re-
used during the following sampling period.
24
-------
APPENDIX B. PLANT DATA
Table B-l. Campaign data for fuel inventories processed
during tritium study at NFS during 1971 (8j
Campaign
Humboldt Bay
fort?}
{r(j£)
V *i n lr /"• r* T? OT T/*
POT-T- f rVNTPA^
Big Rock Point
Total
burnup
(MW(t)-
days)
mT9Q
, J/o
77/, 7 -in
^i'+ , / JU
OA TOR
JH y jU o
80,438
Tonnes
total
i n
J.U
o A
j • 4
5.9
Fuel age
(years)
3=; T
. j j
77
nil
L. sn
H • OU
1.58
Fuel
cladding
•7i-
/:r ,
SS-304
oc
OO
7*. A
ZiT- 4
Zr-Z,
Iconel
Number
of
dissolu-
tions
9 7
Z J
i n
J.U
7
Process-
ing
period
A /OC
Q / Z.D
7/02
7/19
// I/
8/24
0/7-1
y/ /i
10/04
10/10
11/08
25
-------
Table B-2. Average X/Q values for the NFS site
during 1971a (11)
Wind direction
N_ _
NNF —
NTT — —
VNV —
E_
TTCW
qir _ _
QQF
S__
QQTJ _ _ _
OTJ _ _ _
UQU —
W_
TJMTJ—
MTJ _ _
WNTTJ — _
Boundary
distance
(meters)
9 9nn
/ , zuu
2r>nn
, uuu
2nnn
,uuu
i Ann
-L ,DUU
i Ann
i ?nn
i onn
3^nn
, JUU
9 Ann
i Ann
i ftnn
i ar\n
9 Ann
9 Ann
9 onn
9 Ann
Average
(s/r
Summer
7 no „ i n~"
/ . Uo X 1U
6AA v ~\ r\~
.DU X XU
6 TO „ i n~
. J/ X 1U
61 o „ i n~
. 1O X -LU
8AA Y in
_Q
i 7i TT i n
Q
o SA Yin
9 SA •»• in
—ft
9 i Q Y i n
Q
•3 QQ Yin
Q
•3 07 Y in
Q
i QS Y i n
Q
i An v i n
Q
i A9 Y i n
Q
1 01 Y 1 0
_o
Q i ^ Y in
i X/Qb
n3)
Annual
—ft
197 •«• in
. Z / X XU
_Q
80 Q ... in
7 T o „ i n~
/ . LL X JLU
Q
i co „ T n
j. . jy x xu
0
i AI Yin
Q
i A7 vin
Q
991 Yin
Q
i i R Y i n
o
i fi^ Y in
o
A i n Y i n
—ft
•3 q-\ Yin
Q
9 RA Y 10
Q
1 SO v 1 0
Q
1 Oft -x 10
-Q
a AA Y 10
Q
1 01 Y 10
Based upon hourly averages obtained from instrumen-
tation located at the top of the stack.
X/Q = air concentration in Ci/m3 per Ci/s release.
27
-------
THE ABSTRACT CARDS accompanying this report
are designed to facilitate information retrieval.
They provide suggested key words, bibliographic
information, and an abstract. The key word con-
cept of reference material filing is readily
adaptable to a variety of filing systems ranging
from manual-visual to electronic data processing.
The cards are furnished in triplicate to allow
for flexibility in their use.
28
-------
OBSERVATION OF AIRBORNE TRITIUM WASTE DISCHARGE FROM A
NUCLEAR FUEL REPROCESSING PLANT, EPA/ORP 73-1. J. A.
Cochran, W. R. Griffin, Jr., and E. J. Troianello
(February 1973) .
ABSTRACT: A study was conducted at Nuclear Fuel Services,
Inc. (NFS) to: (a) characterize the stack tritium ef-
fluent in the gaseous and water vapor forms, during
normal plant operations, (b) determine the weekly aver-
age tritium air concentrations at five selected sites
around the plant over a 4-month period, (c) evaluate
methods used for sampling at the stack and in the en-
vironment, and (d) estimate the dose to the population
in the immediate vicinity of the plant from tritium
stack effluents.
(over)
OBSERVATION OF AIRBORNE TRITIUM WASTE DISCHARGE FROM A
NUCLEAR FUEL REPROCESSING PLANT, EPA/ORP 73-1. J. A.
Cochran, W. R. Griffin, Jr., and E. J. Troianello
(February 1973).
ABSTRACT: A study was conducted at Nuclear Fuel Services,
Inc. (NFS) to: (a) characterize the stack tritium ef-
fluent in the gaseous and water vapor forms, during
normal plant operations, (b) determine the weekly aver-
age tritium air concentrations at five selected sites
around the plant over a 4-month period, (c) evaluate
methods used for sampling at the stack and in the en-
vironment, and (d) estimate the dose to the population
in the immediate vicinity of the plant from tritium
stack effluents.
(over)
OBSERVATION OF AIRBORNE TRITIUM WASTE DISCHARGE FROM A
NUCLEAR FUEL REPROCESSING PLANT, EPA/ORP 73-1. J. A.
Cochran, W. R. Griffin, Jr., and E. J. Troianello
(February 1973).
ABSTRACT: A study was conducted at Nuclear Fuel Services,
Inc. (NFS) to: (a) characterize the stack tritium ef-
fluent in the gaseous and water vapor forms, during
normal plant operations, (b) determine the weekly aver-
age tritium air concentrations at five selected sites
around the plant over a 4-month period, (c) evaluate
methods used for sampling at the stack and in the en-
vironment, and (d) estimate the dose to the population
in the immediate vicinity of the plant from tritium
stack effluents.
(over)
-------
Characterization of tritium release from the stack
was carried out during three reprocessing campaigns
assumed to be typical of the NFS operation. Evaluation
of the data indicates that the tritium release rate
from the stack is 6.75 x 10"^ curies per megawatt-day
burnup. Approximately 25 percent of the activity is in
the gaseous state and 75 percent is in the water vapor
state. The average annual release of tritium via the
stack, considering the total fuel processed from 1966
to 1971, is estimated to be 500 curies per year. Data
for the five air sampling stations around the site in-
dicate that the plant contribution to the atmosphere is
small.
KEY WORDS: Analysis; fuel reprocessing; nuclear; popu-
lation dose; sampling; tritium; waste discharge.
Characterization of tritium release from the stack
was carried out during three reprocessing campaigns
assumed to be typical of the NFS operation. Evaluation
of the data indicates that the tritium release rate
from the stack is 6.75 x 10"1* curies per megawatt-day
burnup. Approximately 25 percent of the activity is in
the gaseous state and 75 percent is in the water vapor
state. The average annual release of tritium via the
stack, considering the total fuel processed from 1966
to 1971, is estimated to be 500 curies per year. Data
for the five air sampling stations around the site in-
dicate that the plant contribution to the atmosphere is
small.
KEY WORDS: Analysis; fuel reprocessing; nuclear; popu-
lation dose; sampling; tritium; waste discharge.
Characterization of tritium release from the stack
was carried out during three reprocessing campaigns
assumed to be typical of the NFS operation. Evaluation
of the data indicates that the tritium release rate
from the stack is 6.75 x 10 curies per megawatt-day
burnup. Approximately 25 percent of the activity is in
the gaseous state and 75 percent is in the water vapor
state. The average annual release of tritium via the
stack, considering the total fuel processed from 1966
to 1971, is estimated to be 500 curies per year. Data
for the five air sampling stations around the site in-
dicate that the plant contribution to the atmosphere is
small.
KEY WORDS: Analysis; fuel reprocessing; nuclear; popu-
lation dose; sampling; tritium; waste discharge.
------- |