Technical  Note
                                                  ORP/LV-80-3
     RADIOACTIVE  EMISSIONS FROM YELLOWCAKE
       PROCESSING STACKS  AT URANIUM MILLS
              Charles W.  Fort,  Jr.
               Richard  L.  Douglas
                William E.  Moore
                  October  1980
       U  S  ENVIRONMENTAL PROTECTION AGENCY
       EDISON, H. J-  0881Z
Office of Radiation Programs,  Las  Vegas Facility
      U.S. Environmental  Protection  Agency
            Las Vegas, Nevada  89114

-------
                                  DISCLAIMER

    The Office  of Radiation Programs-Las  Vegas  Facility, U.S.  Environmental
Protection Agency  has  reviewed  this  report and  approved  its  publication.
Mention of trade names  or commercial products does not  constitute endorsement
or recommendation for their  use.
                                      n

-------
                                   PREFACE
    This study  was  conducted  by  the Office  of Radiation  Programs-Las  Vegas
Facility of the U.S. Environmental Protection Agency.  The Office of Radiation
Programs carries  out  a  national  program  designed to  evaluate  population
exposures to radiation,  and  to promote the  development  of  controls necessary
to protect the  public  health and  safety.   The purpose  of  this study  was  to
evaluate the  emissions  of  uranium  concentrate (yellowcake)  from  processing
stacks at uranium mills.

    We would appreciate receiving any comments on this report that readers may
have to offer.
                                            Donald W. Hendricks, Director
                                            Office of Radiation Programs
                                            Las Vegas Facility

-------
                              TABLE OF CONTENTS

                                                                          Page

LIST OF TABLES	vi

ACKNOWLEDGEMENTS	vii

INTRODUCTION 	  1

SAMPLING EQUIPMENT AND METHODS 	  4

    Stack Sampling Equipment 	  4
    Stack Sampling Methods 	  6
    Yellowcake Grab Samples  	  9
    Sample Analysis and Data Calculation 	  9

DESCRIPTION OF MILLS AND YELLOWCAKE PROCESSING FACILITIES  	 11

SAMPLING AND ANALYTICAL RESULTS  	 12

DISCUSSION OF RESULTS  	 22
    Uranium Emission Rates 	 22
    Estimation of Annual Average Emission Rates  	 27
    Concentrations of Uranium Daughter Radionuclides in Yellowcake .... 28

SUMMARY AND RECOMMENDATIONS  	 30

REFERENCES	35
                                 «
APPENDIX A	37

-------
                                LIST OF TABLES
                                                                          Page
1.  Uranium Emission Rates from Yellowcake Processing Stacks ....... 14
2.  Radium, Thorium, Lead, and Polonium Emission Rates from
    Yellowcake Processing Stacks ..................... 16
3.  Ratios of Daughter Radionuclides to U-238 Parent for
    Stack Test Samples .......................... 17
4.  Inter-Laboratory Comparison of Uranium Analytical Results
    on Stack Test Results  ........................ 18
5.  Radionuclide Concentrations in Yellowcake Grab Samples ........ 19
6.  UsOs Drying, Packaging, and Emission Rate
    Parameters ................... .  .......... 24
7.  Variability of Measured Yellowcake Emission Rates  .......... 25
8.  Daughter Radionuclides in Yellowcake as Percent of
    U-238 Parent ............................. 29
                                       VI

-------
                               ACKNOWLEDGEMENTS
                                            «
   The approval  and cooperation  of mill management  and staff  made  possible
this study  of  uranium mill yellowcake  emissions.   We  gratefully acknowledge
the following key individuals:

    Mr. Bill Gray and Mr. El rod Leany, Anaconda
    Mr. Bill Shelley and Mr. Jim Cleveland, Kerr-McGee
    Mr. Ed Maurer and Mr. John Ritts, Sohio
    Dr. Noel Savignac and Mr. Todd Miller, United Nuclear
    Mr. Ed Kennedy, United Nuclear-Homestake Partners
    Messrs. Bob Beverly, Roger Jones and Ed Loshbaugh, Union Carbide

We thank again the  many  others  at these mills whose helpfulness made our work
that much easier.

    Our thanks  also to  Dr.  Walt  Kisieleski  and Dr.  Mike  Momeni  at  Argonne
National  Laboratory for  their interest  and  helpfulness  during the studies.  A
belated note  of  thanks  should  also go to  Mr.  Leonard Link  (formerly  of  the
Argonne  National  Laboratory),  who provided  the  original  leadership  for  the
program leading to these studies.
                                     vn

-------
                                 INTRODUCTION
                                                                «
    The  recent  expansion  of   the  uranium  milling  industry,  coupled  with
increasing awareness of the potential  health  effects of radioactivity releases
from uranium milling  sites, has created  increased  concern about  the current
and future radiological environmental  impact  of the  industry.   The regulations
and licensing  requirements  for  the  industry, issued by the Federal  and State
agencies, reflect this  concern.   For example, the  U.S. Environmental Protec-
tion Agency's (EPA)  uranium fuel cycle standard, 40  CFR 190,  limits the annual
radiation exposure  to any  member  of the  public,  from a  uranium fuel  cycle
facility, to 25  mrem.  This  standard becomes effective for uranium mills  in
December 1980.    It does not include  exposure from (1) radon and  its progeny,
including  lead-210   and  polonium  210,   (2)  radioactive  contamination  from
milling operations prior  to December 1980, or  (3)  radioactivity  from uranium
mining operations.

    Several   reports  have  described  the  sources  of radioactive  material  re-
leases in and  around a uranium mill, the quantity  of material  released  per
unit time (the source term),  the pathways of  the released material  to people
living near  the  mill,  and  the  dose  received  by these people.   (See,  among
others,  Glauberman  and Breslin  [1964],   Sears  et   al.   [1975],   EPA [1976],
Harward [1977], Sill  [1977].)   The source  terms and  pathways  are qualitatively
well  defined  by  these  reports.    They,   as  well  as  numerous  environmental
reports and environmental  impact  statements  prepared as mill  licensing  docu-
ments, also assign quantitative values to the  source terms and pathway trans-
fer parameters,  and  estimate radiation  doses  to  individuals   and population
groups.   However, these  quantitative values are necessarily  often  based  on
estimates and assumptions,, because few actual measurements of the source terms
and  transfer  parameters   have  been  made and  published  in  the  scientific
literature.

-------
    Considering the  shortage of documented  technical  data about  yellowcake*
emissions from uranium mills, and the increasing  need for such data,  the EPA's
Office  of  Radiation  Programs-Las  Vegas  Facility  (ORP-LVF)  has  undertaken
studies to provide such data.   Excepting tailings  piles, the  source  with  the
highest potential  for human exposure appears to  be particulate emissions from
the yellowcake  drying and packaging  stacks.    (The dose from radon and  its
progeny are not included in the 40 CFR 190 standard.)  Although emissions from
these stacks are small in terms of mass  release  rate,  the concentrated  nature
of  the  yellowcake  product  (usually  reported   to  be  about  85-90%   UsOg)
provides  a  relatively high  radionuclide release  rate.   Consequently,  EPA's
studies gave the yellowcake stacks high priority.

    Planning  for  these studies  began  in  late  1976.    We  then learned  that
Argonne  National  Laboratory  (ANL)  was  beginning  similar  studies  for  the
Nuclear Regulatory Commission.  After discussion, the two groups joined  forces
to reduce duplication of effort.  As  study plans  evolved, ANL  assumed the lead
role  in  defining  radon  source terms, while ORP-LVF  concentrated  on  measuring
releases  from the yellowcake  stacks  and on measurements  of ambient  levels of
radon and airborne particulates.  In early 1977,  we approached several uranium
milling companies to establish a cooperative basis for studies to be conducted
at their mills.  Several companies agreed to cooperate in these studies.

     This  report  describes  the results  obtained  from  a series of  tests on
yellowcake drying and packaging stacks at six mills.  The tests were conducted
from April 1977 to August 1978.
*   Yellowcake  is  the  generic  name  given  the  packaged uranium  concentrate
produced  by  uranium  mills.     The  chemical  form  of  the  initial  uranium
concentrate at all mills prior to drying or  calcining  is ammonium (or sodium)
diuranate which has a bright yellow color-hence the name yellowcake.  However,
the uranium concentrate as packaged  at  the various mills will  have a variety
of  chemical  forms  including  ammonium  (or  sodium)   diuranate,  U^OQ,  and
U02»  depending  on each mill's  process and  drying or  calcining temperature.
The color  of  these forms  will vary  from bright  yellow through  olive green to
black.   The  uranium  content of  yellowcake  is  typically  expressed  as  an
equivalent  mass  of  U30a  regardless  of  the   actual   chemical  form.    The
chemical forms  of  the yellowcake  reported  on are unknown and it is important
for the reader to  realize that these  materials may  not  be U30s  or exhibit
the solubility  characteristics of

-------
    The six mills that participated in this study are:
        Anaconda
        Kerr-McGee
        Sohio
        United Nuclear-Churchrock
        United Nuclear-Homestake Partners
        Union Carbide
Bluewater, New Mexico
Ambrosia Lake, New Mexico
Seboyeta, New Mexico
Gallup, New Mexico
Milan, New Mexico
Uravan, Colorado
    The ORP-LVF  staff  and the responsible staff at  each  mill  agreed that the
tests would be conducted  with full company approval  and cooperation and would
not interfere with mill operations.

-------
                        SAMPLING EQUIPMENT AND METHODS

     Stack sampling for  particulates consists of  extracting from within  the
stack samples  of stack  gas  and collecting  any  entrained  particulates  on  a
filter.   The air velocity through the opening of  the sample nozzle is equal  to
the stack gas velocity at the sample  point.   Sampling  this way  is  isokinetic
and  ensures  that all  particles approaching  the plane of  the sample  nozzle
opening will  be collected without bias due to size.

Stack Sampling Equipment

     The  stack testing  equipment used  for these  tests was  manufactured  by
Research Appliance Company.   Essentially, the equipment is  a meter console  and
a sample case connected  by sampling  hose,  power  cables and pitot lines.   The
meter  console   monitors  stack  gas   conditions   (temperature  and   velocity
pressure) and allows the  operator to maintain an isokinetic sampling rate  by
adjusting the sampling velocity through  the  sampling nozzle.   The sample case
holds the sampling probe  for in  situ  extraction  of  the stack  gas sample,  and
also contains a  filter in a  heated compartment with temperature control  and a
series  of four  glass  impingers in  an  ice bath  compartment  for  moisture
condensation and removal.

    The meter console consists of an oil-type fiber vane positive displacement
pump; a  0.1-cubic foot  per  revolution  totalizing  dry  gas  meter; an  orifice
meter  for sampling  rate determinations;  a  dual  0-10" H20  incline/vertical
manometer  (incline  portion  -  0 to  1"  H£0  with 0.01"  HgO minor  divisions)
to  measure  the  orifice meter  differential  pressure  (AH) -and   stack  gas
differential  velocity pressure (AP);  a vacuum gauge  to  indicate filter loading
as  testing  progresses,  or to indicate  the occurrence  of a major leak  in  the
sampling  train;  coarse  and fine  adjustment valves  to set   isokinetic  flow
rates; and a rheostat for adjusting the temperature of  the  sampling probe.   An
optional  feature of  this unit  is a  six-button digital temperature  indicator
using special  grade  (±2°F)  "J" thermocouples for temperature  measurements  of

-------
the stack  gas,  probe liner, filter  compartment,  silica gel  inpinger  outlet,
and inlet and outlet of the dry gas meter.

    The sampling probe consists of a three foot stainless  steel  liner  wrapped
with heat tape  protected  by a stainless steel sheath.  A  gooseneck  nozzle is
attached to  the inlet  end  of the  probe.    A thermocouple  to measure  probe
temperature  is  located  underneath the heat  tape  near the exit of the probe.
The heat tape lead  and  thermocouple lead are terminated with  connectors  just
outside the sheath.  A Type S pitot for AP measurements and a thermocouple for
stack  gas  temperature  measurements are  mounted  on  the  sheath  surface  and
extended to  a point in-line with the gooseneck sample  nozzle.   The  pitot and
thermocouple  are  positioned  to  conform  with   the  specifications  of  EPA
Method 2.

    A  Hewlett-Packard  programmable calculator (HP-65  or  HP-67) was used for
all tests  instead  of the  standard stack testing  nomograph.    The  calculator
offers  small  size,  accuracy,  speed, and  each of  the ten  variables   in  the
isokinetic  equation  are used  in  the isokinetic  sampling  rate  calculations.
This is  unlike  the  nomograph with  its  several  potentially  faulty assumptions.
Ultimately,  HP-67  programs  were  used  in  all  phases  of  a  stack  test,  and
reference materials were  unnecessary.  These programs  included pre-test stack
gas moisture  fraction  estimation, sampling  nozzle  size selection,  number of
sampling points, location  of sampling  points for a  circular  stack,  stack gas
moisture fraction  by condensation,  dew  point temperature  and the  post-test
determination of the isokineticity  of the test.   An EPA publication (Ragland,
1976) is an  excellent source  of  calculator  programs for use  in  air  pollution
measurements at stationary industrial sources.

     The pre-test estimation of  stack  gas moisture  fractions  calculated using
the  HP-67  program  proved  extremely  reliable, providing  care  was  taken  in
determining  stack  gas  wet  and  dry bulb temperatures  and stack gas  dynamic
pressure.  The  program  was  so  developed  that reference charts  or'tables  were
unnecessary.  Appendix A describes this program.

-------
Stack Sampling Methods

    All yellowcake stack tests employed the EPA Reference Methods  1  through  5
as  defined  in  "Standards  of Performance  for New  Stationary  Sources"  (CFR,
1977).  Reference Methods 1 through 4,  respectively,  define  the techniques  for
(1) locating  and  defining the number  of  sample or  velocity traverse  points,
(2) determining  stack gas  velocity  and volumetric  flow  rate,  (3)  analyzing
stack  gas  carbon dioxide  and  oxygen concentrations  and  determining the  dry
molecular weight of the stack gas, and  (4) determining the  stack  gas moisture
content.   Reference  Method  5 combines these techniques  in  determining  the
particulate emissions from a stack.

     Since yellowcake stacks are not subject  to the EPA standards of  perfor-
mance, these methods are not officially prescribed for these stacks.  However,
these  methods  (with slight  modifications  to  Method  5  reflecting the  unique
material being sampled) are  the best available.   Since  tests at  all  six mills
were conducted on a non-interfering and non-interrupting basis, they  were per-
formed under existing operating conditions without consideration  for a  stable
or  optimum operating  level  as  would  be  required   for  compliance testing.
Therefore, parameters affecting emission rates were not controlled except that
tests  were  usually  conducted when  yellowcake was  being  actively  dried  or
packaged.   Once a test  was started, it was  completed  even though  drying or
packaging  may  have   ceased  or  a  scrubber   malfunction  may  have  occurred.
Attempts were made to determine amounts of yellowcake dried  or  packaged  during
tests.

    The modifications of Method 5 referred to above consisted of (1) using dis-
tilled water  or 0.1N HNOs  as  the probe and  glassware  rinse solution  instead
of acetone, (2) using a  pressurized  spray  nozzle to clean  the sampling probe
instead  of  a  probe brush, (3) reducing the filter and probe heat  from   248 _+
25°F  to  20-30°F  higher  than stack gas  dew  point  when the  higher  temperature
could  not be  achieved,  (4)  deleting  the filter weighing requirements,  and (5)
removing  the  sampling  holder  filter  gasket.   These changes  are  minor  but
resulted in increased accuracy and ease of operation noted below.

    An early test used acetone, as prescribed by Method 5, as  a wash solution.

-------
However, the  acetone tended  to  leave a yellowcake  film that was  not  easily
removed, and  it was  therefore not acceptable as  a  washing agent.   Distilled
water or 0.1N HNOs  was  adopted as the wash solution and worked  very  well.   A
probe brush was initially used to clean the probe.   It  smeared the yellowcake
over the internal  probe  surface,  and  the  particles  adhering to the brush were
not easily removed.   To overcome this  difficulty,  a standard garden sprayer
nozzle, adjusted to  a  conical spray, was fitted  to  a  brass pipe and flexible
hose attached to a plastic  pressure tank  containing water or 0.1N  HMOs-   The
tank was  pressurized and  the nozzle was  inserted  into the  stain-less  steel
probe liner and run  its length for three passes.   A wash  bottle  was used  to
rinse the surface of the nozzle into the probe wash  after the final pass.  The
pressurized spray removed  all  particles from the probe wall and  washed them
ahead of  the  nozzle into  the collection  jar.   By advancing  the spray nozzle
slowly  on  the first  pass  through  the probe,  surface  contamination of  the
nozzle was minimized.

     The filter and probe temperature requirement of EPA Method  5 is directed
at compliance testing to define the particulate or gaseous nature of the stack
gas components.  If, at this temperature,  any of the stack effluent components
exist as a gas, they obviously will not be collected by the filter and there-
fore, at  least by  EPA standards, are not considered particulates.   Since the
purpose of  the tests reported herein was  not  to  determine  a   mass  emission
rate, this  temperature   requirement  was  not considered binding.    Initially
polonium-210  was considered as existing as  a  gas and  for several tests  the
Method 5 temperature was  rigidly  maintained  as  a  reference temperature should
there be volatilized polonium.   After determining that this  was  not  the case
(polonium was  not detected in impinger catches)  the Method 5 temperature was
used when  achievable.    Otherwise,  a filter  and probe temperature  at  least
20-30°F above the  stack  gas  dew  point  temperature was  maintained only  to
prevent moisture  condensation upstream of  the filter.

Deletion  of  the  filter  weighing requirement  was an  obvious and  timesaving
simplification of Method 5.  Since these tests  were  designed to measure radio-
activity emission  rates  instead of mass rates, there was no need for  a gravi-
metric determination.

-------
    The filter holder was modified  by  removing the  rubber  gasket that  normally
fits between  the  filter and the glass  inlet  cap,  and  adding  a  second glass
fiber filter.   This eliminated cleaning the gasket  after each test, and  the
tandem filters were less likely to  tear  at  the edges  when  removed.   Leaks were
never  detected using  this  technique  and  the  change  caused  only  a  slight
increase in flow resistance.

    For these studies, the stack testing team consisted of two people a probe
operator to position the probe at the  traverse  (or  sampling)  points  within  the
stack as testing  progressed, and a  control  meter operator to observe changes
in  stack  gas velocity  pressure  and temperature,  and  to  adjust and  maintain
isokinetic sampling rates as these  parameters changed.  The control meter  was
at  ground  level,  connected  to the  sample case at the stack by 50-100 feet of
cable and  tubing.   The probe operator  at the  stack  and the  control  meter
operator used portable radios to communicate.

    Basically, the stack testing routine consisted  of moving  the  sample  nozzle
from traverse  point to traverse point  at  predetermined intervals  (usually  5
minutes) until  all  points  across the  stack diameter had  been sampled.  After
the first  traverse,  the sample case with probe  was  repositioned  to a  second
port  at 90°  from  the  first  and  the  sampling  routine  repeated.    Velocity
pressures  and temperatures  were recorded  at each traverse point  and, when
different  from the  previous  point,  entered  into  the  calculator to  compute  the
adjustment needed  at the control meter  to maintain an  isokinetic flow rate
through the sample nozzle.

    At  the conclusion of a  test,  the sampling  probe was dismantled  from  the
filter  holder,  and  the  probe, filter  holder, and  intermediate glassware were
taken to  a clean  area for washing.  The filter was  removed and placed  into a
covered petrie  dish.   All  parts of the  sampling train  forward of the  filter
were then  washed  into a glass  jar.  The filter  plus  the wash solution consti-
tuted  the sample.   Water  condensed  in  the impingers  during the  test  was
weighed to calculate the stack gas  moisture fraction.

    All tests  were  conducted for a minimum of one hour.  Typically,  at least
30  dry standard  cubic  feet  of  stack  gas  (at   68°F  and  29.92  in. Hg) were

                                      8

-------
sampled during each test.

Yellowcake Grab Samples

    The  project  collected six yellowcake  grab samples,  not associated  with
stack  tests  from  four mills.   These  samples  were analyzed  to measure  the
concentration  of  the  uranium  daughter  radionuclides  in  yellowcake  from
different mills, and coincidentally to cross-check the analytical capabilities
of  different  laboratories.   Grab  samples were  used rather  than stack  test
filters  to provide  a  larger quantity  of  yellowcake  for  analysis, and  for
convenience in sending replicate aliquots to the different laboratories.

Sample Analysis and Data Calculation

    Several laboratories  analyzed  the stack test samples from  these studies.
Early in the study, the EPA laboratory in Las Vegas  analyzed the samples using
standard radiochemical procedures as  described by Johns (1975).   Later  in the
study, Eberline  Instrument  Corporation  analyzed  samples using  similar  proce-
dures.  EPA, Eberline, and LFE Corporation Environmental Analysis Laboratories
analyzed the grab samples  discussed above and  functioned  as laboratory  cross-
checks.  These  three  laboratories  analyzed the samples for isotopic uranium,
isotopic thorium, radium-226,  lead-210,  and  polonium-210.   Uranium-West  Lab-
oratories  analyzed  one set of stack  test  samples for  uranium,  using a tech-
nique counting delayed neutrons  from  the fission  of uranium-235 (Gale,  1967).
The  respective  company  laboratories  also analyzed  aliquots of  some  of  the
stack  test samples,  using fluorimetric  or colorimetric  procedures,  as  an
analytical  cross-check procedure between them and the EPA (or EPA contractor)
laboratories.

     The  stack  test  samples consisted of the 2.5-inch  glass fiber filter  and
one  to  three  liters  of  the probe wash  solution,  which was  either  distilled
water or dilute  nitric acid.   They were prepared for analysis  by evaporating
the probe  wash  solution  to near dryness and combining  the  remaining solution
with the filter. This  combined sample  was treated  with nitric and hydrofluoric
acids  to  dissolve  the  glass  fiber  filter   for   subsequent   radiochemical
analysis.   If  inter-laboratory cross-check analyses  were to be  made  on  the

-------
sample, aliquots of this solution were taken.

    The analytical laboratory reported the stack test  sample  results  as total
picocuries of each nuclide in the sample.   The radiochemical analysis  measures
the  three uranium  nuclides   (U-238,  U-234,   and  U-235)  separately  by  alpha
spectroscopy.  These  uranium nuclide  results  were converted to  total  uranium
(Utot) activity by  summing the  activities of the three  nuclides.   This total
uranium activity  was converted  to total  uranium mass  using the  conversion
factor  of  0.677  uCi  Utot/gram  Utot  (CFR,  1975).    The uranium  mass  was
converted  to   UsOs  using  the  conversion  factor   of  0.848  gm   Utot/9m
U308.
                                      10

-------
          DESCRIPTION OF MILLS AND YELLOWCAKE PROCESSING FACILITIES

   Five of the  mills involved in  this  study employ  the acid leach  process,
while  the United  Nuclear-Homestake Partners  (UNHP) uses  an alkaline  leach
process.  A  general  description  of each mill  and  the yellowcake  exhaust  air
cleanup system(s) in use at the time these tests were made is given below.

Kerr-McGee:    Nominal  ore processing capacity 6500 tons  per  day  (tpd).   Three
exhaust stacks  vent  the  plant, each served  by  a Joy Microdyne  wet  scrubber.
The three scrubbers  are referred  to  as the Main,  East, and West.   The Main
Microdyne scrubber and  stack  is  the primary exhaust pathway from the  dryer,
and the East and  West  systems primarily  serve the  packaging area,  although
there is some cross connection.

Union Carbide-Uravan:  Capacity 1300 tpd.   The exhaust  streams  from  both  the
dryer  and  packaging  area are   combined  and  discharged  through  a  single
rectangular stack.  The exhaust air is cleaned by a three-tiered  wet  scrubber.

Anaconda:  Capacity 2000 tpd at the time of the test, subsequently expanded to
6000 tpd.   The  exhaust  stream from  the packaging  area is  cleaned by  a  Joy
Micordyne; the  dryer  exhaust by  the  "Anaconda  scrubber,"  a  wet  impaction
scrubber.

Sohio  L-Bar  Mill:   Capacity  1500 tpd.   The  dryer  exhaust  is  cleaned by  a
venturi  scrubber;  the  packaging  area  exhaust  by  a  Mikro-Pulsaire  Dust
Collector bag filter system.

United  Nuclear  Churchrock:    Capacity  4000 tpd.    Dryer exhaust cleaned  by
Impi-Jet scrubber, packaging area exhaust by Joy venturi wet scrubber.

United Nuclear-Homestake Partners:  Capacity 2700 tpd.  Vanadium  roaster stack
has a  Hydrofilter followed by a  Type  W Rotoclone; the  packaging  stack  has  a
Hydrostatic precipitator followed by a Type W Rotoclone; the dryer stack has  a
Joy-Denver scrubber followed by a Type W Rotoclone.
                                     11

-------
                       SAMPLING AND ANALYTICAL  RESULTS

    Table 1 shows individual  stack  test results.   Each  sample  is  identified by
mill, date of sampling,  and  stack  sampled.   The calculated uranium  emission
rate  for  each  sample  is  shown,  expressed as  microcuries of  each  uranium
nuclide  released  per hour  (radioactivity  emission  rate),  and  as  pounds  of
UsOs  released  per  hour  (mass  emission  rate).     The   table  also  shows
volumetric flow rate of the stack at the time  of  sampling.

    Table  2  shows the  emission rates  for radionuclides other  than  uranium
(i.e., thorium-230, radium-226, polonium-210,  and lead-210)  for the stack test
samples  collected  at the  United  Nuclear-Homestake  Partners   and  the  United
Nuclear  Churchrock  mills.    Eberline  analyzed  these samples,  the only  stack
test samples analyzed for the nuclides other than uranium.  Table  3  shows the
ratios of these nuclides to the U-238 parent for  these  samples.

    Table  4  shows  the  inter-laboratory  cross-check  results  for uranium  on
several  stack  test  samples.    Since  the   industry  laboratories   used  a
fluorimetric analysis (United Nuclear-Churchrock  used a colormetric procedure)
with results  expressed  as  mass  per  unit volume, the  EPA,  Uranium-West,  and
Eberline   radionuclide   results  have  been  converted  to  mass   units  for
comparison.  The table reports the results as  milligrams of uranium per liter
of solution and refers to the solution used to dissolve the  stack test filter.
Since the  purpose  is  to  compare the results obtained  by different  laborator-
ies, the results have not  been  converted to total  uranium per sample or to an
activity or mass emission rate.

    Table  5 shows  the radionuclide concentrations in  several   grab  samples  of
yellowcake  from four mills.    EPA,  Eberline  and  LFE  analyzed  six  of  these
samples.   Due to  a  misunderstanding,  the Eberline  results were  reported  as
picocuries per total sample, rather than as picocuries  per gram of yellowcake.
Consequently, the  Eberline analytical  results  cannot be compared  directly  to
the EPA  or LFE results.

                                       12

-------
    Each  analysis  shows the percentage  of  each of the nuclides  of  the U-238
parent.   (Note the Eberline results  for  this value are comparable to  EPA and
LFE analyses.)   The right hand  column shows the  average  percentage  from the
three analyses for each nuclide.
                                     13

-------
TABLE 1.  URANIUM EMISSION RATES FROM YELLOWCAKE PROCESSING STACKS
Date
Kerr-McGee
4-15-77
4-16-77
4-17-77
9-08-77
9-09-77(am)
9-09-77(pm)
9-12-77
9-13-77
Stack
(samples analyzed
Main
East
East
West
West
West
East
Main
Union Carbide-Uravan (samples
10-13-77
10-14-77
10-15-77
10-16-77
10-17-77
10-18-77
10-19-77
Dryer-Package
Dryer-Package
Dryer-Package
Dryer-Package
Dryer-Package
Dryer-Package
Dryer-Package
Volumetric Flow
Rate 105 DSCFH*

E
M I
S S
I 0 N

uCi/hr
U-238

U-235

U-234

RAT
Pounds
U308/h
E
of
r
by EPA-Las Vegas)
6.7
2
.4
2.3
2
2
2
2
4
.0
.0
.4
.4
.5
39.
12.
9.
4.
7.
30.
6.
113
2
8
1
4
3
7
4

1.
0.
0.
0.
0.
1.
0.
4.
4
5
3
2
3
2
2
0
38.
12.
8.
4.
7.
29.
6.
109
4
8
4
4
4
7
4

0.
0.
0.
0.
0.
0.
0.
0.
30
10
067
034
058
24
050
87
analyzed by EPA-Las Vegas)**
2
2
2
1
1
1
1
.1
.0
.0
.9
.9
.9
.9
18
14.
22.
32.
24.
40.
36.

2
8
1
2
6
9
0.
0.
1.
1.
0.
1.
1.
8
6
1
1
9
7
4
17.
13.
22.
33.
22.
39.
34.
7
5
7
0
0
6
5
0.
0.
0.
0.
0.
0.
0.
14
11
18
25
18
31
28
Anaconda (samples analyzed by EPA-Las Vegas)
4-18-77
United Nucl
6-16-78
6-17-78(am)
6-17-78(pm)
6-18-78(am)
6-18-78(pm)
6-19-78
Package
1
.1
ear-Churchrock (samples analyzed by
Dryer
Dryer
Dryer
Package
Package
Package
3
3
3
1
1
1
.5
.6
.5
.0
.0
.0
9.
3
0.
3
9.
9
0.
075
Eberline)**'***
50.
29.
61.
30.
30.
5
2
9
0
,4
15.4
2.
1.
3.
1.
2.
6
6
7
6
.1
0.7
47.
27.
59.
28.
28.
15.
0
7
4
2
3
4
0.
0.
0.
0.
0.
0.
38
22
48
23
23
12
                                 14

-------
TABLE 1  URANIUM EMISSION RATES FROM YELLOWCAKE PROCESSING STACKS (continued)
                                                            EMISSION   RATE
                             Volumetric Flow                  uCi/hr        Pounds of
Date        Stack            Rate  1Q5 DSCFH *     U-238   U-235   U-234    U^Oa/hr

United Nuclear-Homestake Partners (samples analyzed by Eberline)**
3-13-78     Dryer                  0.4                5.2     0.2     5.2      0.041
3-14-78     Package                3.1                2.7     0.1     2.7      0.021
3-15-78     Vanadium Roaster       0.9                2.7     0.1     2.7      0.021
(samples analyzed by Uranium-West Labs)**,***
8-15-78     Vanadium Roaster       1.0                4.6     0.2     4.6      0.036
8-16-78(am) Vanadium Roaster       1.0                4.2     0.2     4.2      0.033
8-16-78(pm) Vanadium Roaster       1.0                2.1     0.1     2.1      0.016
8-17-78(am) Vanadium Roaster       0.9                4.3     0.2     4.3      0.033
8-17-78(pm) Dryer                  0.4                1.5     0.07    1.5      0.012
8-18-78(am) Dryer                  0.4                3.8     0.2     3.8      0.030
8-18-78(pm) Package                2.8              110       5.1   110        0.87
8-19-78(am) Package                2.8               82.6     3.9    82.6      0.65
8-19-78(pm) Package                2.8              118       5.5   118        0.93

Sohio (samples analyzed by Eberline)**
5-17-78     Package                0.3               45.2     1.9    49.0      0.37 ++
5-18-78     Package                0.4                0.6     0.02    0.7      0.0052
5-21-78     Dryer                  1.3                0.5     0.03    0.4      0.0039
5-22-78     Dryer                  1.0                0.2     0.008   0.2      0.0013
5-23-78     Dryer                  1.0                0.8     0.03    0.8      0.0061
5-24-78     Package                0.4                2.6     0.09    2.6      0.020
*    DSCFH = Dry Standard Cubic Feet per Hour (at 68°F and 29.92 in. Hg).
**   Cross-comparison analyses were also performed on aliquots of these samples. (Table 4)
***  See Table 2 for Ra-226, Th-230, Pb-210, and Po-210 emission rates.
++   One bag filter was ruptured during this test.  This condition occurred the day of the
     test and was corrected that evening.
                                           15

-------
      TABLE 2.  RADIUM, THORIUM,  LEAD,  AND POLONIUM EMISSION RATES FROM
               YELLOWCAKE PROCESSING STACKS*

                                 EMISSION   RATE, nCi/hr**
Date
Stack
Ra-226
Th-230
Pb-210
Po-210
United Nuclear - Homestake Partners
3-13-78
3-14-78
3-15-78
8-15-78
8-16-78(am)
8-16-78(pm)
8-17-78
8-17-78
8-18-78
8-18-78
8-19-78(am)
Dryer
Package
Vanadium
Vanadium
Vanadium
Vanadium
Vanadium
Dryer
Dryer
Package
Package


Roaster
Roaster
Roaster
Roaster
Roaster




8-19-78(pm) Package
18.
19.
10.
77.
15.
10.
20.
11.
41.
52.
54.
25.
5
1
3
4
9
8
2
9
2
3
4
7
37.
32.
10.
56.
52.
24.
77.
151
46.
3780
2580
4050
8
2
4
6
0
8
8

0



132
<50****
105
176
133
108
72.0
57.3
94.4
222
165
274
NA***
NA
NA
328
780
299
285
77.
55.
161
76.
93.






8
7

9
0
United Nuclear - Churchrock
6-16-78
6-17-78(am)
6-17-78(pm)
6-18-78(am)
6-18-78(pm)
6-19-78
Dryer
Dryer
Dryer
Package
Package
Package
4.41
<1
<1
1.41
3.49
0.59
142
79.1
185
102
46.4
49.7
114
36.1
29.4
13.6
12.6
< 5
NA
NA
NA
NA
NA
NA
     All analyses were by Eberline
**   nCi = 10-3 yci
***  NA = no analysis
**** < indicates "less than" value at 95% confidence level
                                    16

-------
     TABLE 3.  RATIOS OF DAUGHTER RADIONUCLIDES TO U-238 PARENT FOR
               STACK TEST SAMPLES*
                                        Percent of U-238 Activity
Stack
Date
Ra-226
Th-230
Pb-210
Po-210
United Nuclear-Homestake Partners
Dryer
Package
Vanadium
Vanadium
Vanadium
Vanadium
Vanadium
Dryer
Dryer
Package
Package
Package
3-13-78
3-14-78
Roaster 3-15-78
Roaster 8-15-78
Roaster 8-16-78(am)
Roaster 8-16-78(pm)
Roaster 8-17-78
8-17-78
8-18-78
8-18-78
8-19-78(am)
8-19-78(pm)
0.354
0.700
0.378
1.86
0.458
0.621
0.539
0.162
1.17
0.055
0.071
0.024
0.723
1.18
0.383
1.36
1.50
1.43
2.08
2.06
1.31
4.00
3.36
3.70
2.53
< 2 ***
3.87
4.21
3.83
6.21
1.92
0.779
2.69
0.235
0.214
0.250
NA**
NA
NA
7.86
22.5
17.2
7.62
1.06
1.59
0.171
0.100
0.085
United Nuclear-Churchrock
Dryer
Dryer
Dryer
Package
Package
Package
6-16-78
6-17-78(am)
6-17-78(pm)
6-18-78(am)
6-18-78(pm)
6-19-78
0.0087
< 0.004
< 0.002
0.005
0.015
0.004
0.282
0.271
0.290
0.341
0.153
0.324
0.226
0.123
0.046
0.045
0.041
< 0.035
NA
NA
NA
NA
NA
NA
*    All analyses were by Eberline
**   NA = no analysis
***  < indicates "less than" value at 95% confidence level
                                    17

-------


















__J
-
	 j
J
CO
Oi
"jg
|
^^



•M
•p—

s_
O)
Q.

E

«P-
c

BfB
^_>
!_»
•P- *
E i-
O CO
C i_ Ol
•p- S- CO
Ol >>
T7 r—
O) CT> IO
co c c
CO -r- IO
Ol 4->
i- C Ol
Q. 3 -M
X O IO
0) 0 0
Ol 10 r—
j- E a.
10 rj> 3

CO CO
4J 1 M-
p- CM O
3
CO Ol Ol
Ol JZ C7>
i- 4-> IO
t.
i— O» Ol
P— i- >
< 
-------
         c
         41
         u
         1_

         £
                         in  CM
                         co  to
                                  r»»  O   ^  fi
                                  -H  o   O  O
                               •   •    •    •   •
                              o  o  o   o  o
                                  +1  -I- 1
                                               -H
                                                                         CO  CO  id
                                                                             Cn  *o
                                                                             O  O
                                                                                      O  O
                                                                                          irt  •«   vo  O  cn
                                                                                          CM  CM   CM  ^«  o
                                                                                                   o  o
                                      in   r-»  »— I
                              CM  oo  co   en  r-
                              en  i-.  ro   O  o
                                                            3  *
                                                                —i  r—   PO
                                                                 •    •    *
                                                                    CM   O  O
                                                                                      o  o  o   o  o  o

                                                                                      +1 +1  +1  +1  +1  +1

                                                                                              «">   I*.  O  CO
                                                                                          in  V   r^  CO  *O
                                                                                      »  —•  co   CM  O  O
                 CM  CO  CM
                 O
                              O   O  O
                                                                                              OO  V  CM
                                                                                              en
                                                                                                   o  o  o
         O JJ
         41 OO
t/>   *  u PO
LU   «  (. CM
—I    CU  41  I
        U  O
         O.U
                              CO
                                  oo  co   en
                                  on  co   o
                                                                                                   — i  o
                                                                                          en  oo   en  r>.
                                                                                          CM  en   CM  o
                         oo
                         OS
                                  CM  O   O
                                                VI

                                                X
                                                    VI


                                                    V)
                                                                 VI   VI

                                                                 X  X
                                                                              VI


                                                                              V)
                                                                                      00
                                                                                      en
                                                                                                       —  o  o
                     O  O   IO  CM
                                                             in
                                                             c
                                                                                  
                                                                                            - o
                                                                                      CM  PO  O
                                                                                              in
                                                                                                                    VI
                                                                                                               O   —
                              •H  -M  -fl  -fl  ~

                                           o   S
                                           >C   !0
                              O  O  en
                              CM  fi  in
                     oo  <
-------
         s

         t.
         3
         a-.—
CO


CM
 I  C
3 IQ



^
         1)
         O
            >
        H- i—
         O **
   §3
«  S. CM
«  41  I
     .2
   ~u
   *lg

   OILbl
                                  to  10
                              CO   Cn  ^
                         o   o   a  o
                                     ro  ro
                                     O  o
                                     o  o
                                              -4  to   —
                                              t^  O   O
                                              000
                                                                        CM
                                                                        o
                                                                        o
                              o  o  o  o   o

                              +1  -M -t-l  +1  +1
                             O  -"  O  CD
                             co  *r  «r  m
                             r»  O  O  O
                              o
                         00   10
                         03
                                  O  O  O  O
   LO  LO


   CM  d  o   o  o  o

I   +.| 4-|  -f |  -f |  -f |  -t-|




   GO  00  U3   O  O  O


   r»  co  O   O  O  O
                                                                                             §
                                                                                CM  IO  O  «•
                                                                                LO  ,*  CM  —I

                                                                            en  a
                                                                                                     c   c  c
                                                                                                     a   ia  iq
                                                           00
         41 CO
         uco
         I. CM
         01  I
   cs.2
   •f |LU

   S2
   i- •«-
                            vo   en  o  vo
                            o   LO  v  co
                            oo   o  o  o

                            dodo'
                                              LO
                                              r^  ^   n*  LO
                                              LO  -<   —  —S
                                              LO  O   O  O

                                              d  d   d  d
                                                       o
                                                       o
LO

o
SB    <
O    a.
w    wJt  u

LU      CM  t.
        3S
         evu
                 I
••.   S   £  •»       .  "       "-  "S   4!  -o
LO   4^   4.1  o  v*-)  O  en       LO  +»  i^  *f
CMS-tCM^-—           CM  ^   L.  —•

•f|  H  S.  -f|  •+•! -f|  +1      -fl  a.   8.
     41   a.                  *—
O   t,   !„  O  O  ~  LO   4J
                                                                            SCM  —

                                                                            VH  VO
                                                                       CM   en  49-
                                                                                         v>   at  ut   (/>  v>     at   ut
                co
                LO
                CM
 0°  S   Z  CM               O

                              a_


                              |

                              i/i

                              "c
                                 SCM  P»  O
                                 CO  CO  •.!•  4«-
                    o  LO  co   o  es  o  CM

                    LO  co  d   d  d  o  co
                    en                       co
                                              an




                                             i^

                •fl  -fl  +1  +•!  +1  -fl -fl <


                    SO   o   ^   co  en  en

                      .  ''P.  1               i
                o*  o   ao   CM                41
                LO  *r                       *o
                CM  CM                       •—
                     CM  CM   —
                f^  CO  ^^  CO   ^|  ^^  •"•*   C
                CMCMCMCMCMCMCM   O
                 II    II    II    I   •*—
                =]  3  3  J=   Old   OC
                            f-   cea.o.3
                                                   —  —  -fl  -fl
                                                   41  V     '
                                                   (-  S.  O  <«•
                                                                                                      el
                                                                                                      a
                                              CM

                                              VO
                                              CM
                                                      £  g  __.  CM  *
                                                                            LO  CO

                                                                                CO
                                                           O  rr
                                                           en
                                                                                    •o
                                                                                     41
                                                                                     01
                                              CM  LO   LO  CO
                                              en  ^4   ^^  *^  LO
                                              vo  o   O  o  CM
                                                •    •    •   •  en
                                              O  O   O  O  CO
                                                       CM
                                                        •

                                                       VO
                                                       •fl  +1 -fl  fl  •*•!
                                  o  o
                                  o  o
                                   0   A

                                  O  O
                                  LO  »
                                  CM  CM
                                                               O  co
                                                               in  rv.
                                                                 *    •>

                                                               OO  ~4
                                                                             41



                                                                        °.   S
                                                                        LO   a
                                                                            to



                                                                        -'g
                                                                        VO  LU
                                                                   o  
-------
00
CO   +41
CM
 I    C!
3   >OI



*4
         c
         a>
         u

                           iT>
                            •
                           «r
                           o
                                        CM

                               2       g
    —   n  in  «•  o
r—  -*   o  o  O  o
 •    •    •  o  o  o
CM  o   o   •    •    •
             o  o  o

+ 1 -M  -H  +1  +1  +1

         03  TT  —  CO
    r**   cn  o  »^  O
Cl  O   O  O  O  O

^  CO   O  O  O  O
                                                                                                10

                                                                                                CM
                                                                                                o
                                                                                                a
                                                                                                     LO



                                                                                                 o  o  o   o

                                                                                                 d  (3  O   O

                                                                                             i   -H  + |   -H  +1
                                                                              •»  O
                                                                              8  §
                                                                              o  o
                                                                                                         O   ^
                                                                                                         °   =>
         s  s
      *  i.  CM
      at o.  a
 *'£
 II
 <  *
 O  g
 a,  u
               U>   Ul   I/I   Ul   Ul   I/I  Ul
                       Ul  Ul   Ul   Ul   Ul  Ul   Ul
                       >,>,>>>,>,>,>,
                                to   O
                                                                                           CMCM
                                                                 o  o
                                                                  •    •

                                                                 00  00
                                                                                                        CMO
                                                                                                          •   Ul
                                                                                                        i   >>   >»  >»  >»  >>
                                                S.
                                                         §
                                                                      cn  o  o   o
                                                                      O  O  O   O
                                                                      O  O  O   O
                                                                                   00
fe

i      «  ^
 o  *-•
z
s
UJ
O
 41

 S

.£
        CM
S
                                                 CM   t-  'C  CM
                                                      o   o
                                                  •H  a.   Q.  +|  +1  -"-I +•!
                                                      01   41
                                                 cs   L.   u  in   in
                                                 g           CM   cn  *  cn

                                                   •  o   o       CM  10  cn
                                                 r>.   z  z           "•  —

                                                 CM
                                                                                    O'O'O
                                                                                     *OIOI
                                                                                                O   g  ^»  «*>
                                                                                                LO   O  O  CM
                                                                                                r»»   O  C3  C3

                                                                                                dodo
                                                                                  m  oo  oo
                                                                                  d  06  in
                                                                                                        t

                                                                                                        &
                                                                                                        41
                                                                      o   g
                                                                 +    a.  &  +
                                                                      01   01
                                                                 §    w   «•  |


                                                                 5  *  I  *

                                                                 CM
                                                                                                                          a.

                                                                                                                          e
                                                                                                                             •  4->   4J  «C
    CM   O  IO   O   Q  Cv
         Z      2   Z  UJ
                                                                  cn  c^i  «*4
                   8
                   CM
oovmovoooz
PO  CO  (^  ?**  &4   <"^  ^^   I
CMCMCMCMCMCMCM   W
 I    I    I    I    I    I    I    U
333£   o  o
                                                     LO  CM  d   d  d
                                                                           03
                                                                           o
                                                                           o
                                                                                               CMV
                                                                                               o
                                                                                                        CM  O

                                                                                                        d  d
                                                   §  o

                                                   d  d
                                                                      CO  ^^

                                                                      -t-l  +|
                                                         •t-l -H
                              o  
                                                                                                         2  S
                                                                                                         m  CM
                                                                                  cn

                                                                                  CM
                                                                CO
                                                                <*i
                                                                CM
                                                                                       T  LO
                                                                                       
-------
                             DISCUSSION OF RESULTS

Uranium Emission Rates

    Table 1 shows the basic data of the study, and the uranium emission rates.
At the  outset  of the  study,  we hoped  to find  some  correlation between  the
uranium emission rate and one or more mill operating parameters.   The emission
rate  should  logically be  a function  of the  amount  of  uranium  concentrate
produced by the mill per unit time, with some adjustment for the efficiency of
the stack gas cleaning equipment.  Such analysis  is typical  in  the  scientific
literature and  in environmental impact  analyses.  Specifically, the uranium
emission rate is usually estimated using the following parameters:
         - Ore processing rate (usually tons per day or year)
         - Average ore grade (percent UsOs)
         - Efficiency of the mill process for recovering uranium
         - Appropriate assumptions of operating time per year
         - Efficiency of effluent cleaning equipment

An annual  average emission  rate,  usually  expressed  as curies  per  year,  is
desired for evaluating the environmental impact of the mill.

    Presumably, the emission rate  at  any  given time should  be proportional to
the mill operating parameters at that time.   When possible  during this study,
the best estimate of mill operating parameters during the stack testing period
was obtained, including  the feed rate of yellowcake  slurry to the  dryer  and
the quantity of yellowcake packaged.  However, the observed emission rates did
not correlate  with  these two  parameters.   This  is  likely  because  the  stack
tests  were  conducted  over  a  relatively  short  period (usually  one to  two
hours).    Consequently,  mill  operating  changes  are  not   likely to  produce
observable changes  in  emission  rate  using such short  measurement periods.  The
Union Carbide results of Table 6 are a good example of such inconsistency.  Of
the  seven test  results,  the  second  to  the  highest emission  rate measured
                                      22

-------

O)  rr"
r" i i
I— ( CO =3
oo co
OO -r- CO
1-1 e .a
SI UJ _l
UJ
o
z
<:
A
CD
Z
I-H
O
< (U CO S.
i<£ -M O -C
CJ fB CO — .
< E =5 -0
^ ._, fll
h>^ "^™ (X/
X CO 01
« O ~O -  s.
_J •!- J=
CQ X CO --^
< OT3-O
1— i- C CJ
0.3 -^
0.0 S-
< Q. Q










^~
!"•
•^
Z



co ^i-
CM r^ o co«— cncn *
CMOCMLO cocn^i- cno Ha- (Or- iior>. *
i— ir--^-i— i o o •— i o -HO OOOO C
OOOO OOOO OOOOOOOO OOOO C


II II II II

0) O) O) O)
Ol Ol Ol Ol
fO fO fO fO
S- i- i. S-
OJ tt) OJ OJ
> > > >
< < < <






en co r-i
CO ^-1  4^ t. T-
•i—3 £ T- CO -i-3 JT
CJT o c$_ cj= o
=5O 00 =5=3 =5CJ OO
J-!

cy
g
CM en co LO o
i— ( i— 1 r-l LO ^f •* 
II II O
o
ai CD . u.
01 01 .*:
 > COi—
< <: .a
•a 
O C3 O CD S_
c a>
§>
(O

O 0)
0 Si
4->

•r- -O
2 0)
CO
•»-> 3
C
ai 4->
s_ o
1_ C
o o o o o -i->
LO CO •«!• CO C r—
«J- O 00 p»» O 3
r-t r-i CM o CO
CD
0) C£
i_
fO
CO -(->
c co
o ai
,—m 1 %
^^ •<•*
4->

CO U O.
S- (O 3
OJ Q. i-
1 c
i- 4J -a s_
(O 1_ CO)
O) (O CO -M
•— O. ^>
CJ Ol-r-
3 (U C (f_
Z _* •!-
, Ol
"O ^J S- (O
o co a CQ
+•) d)
**— £
CO *
=53= * *
23

-------
(0.28  Ibs  UsOs/hr)   corresponds  to  the  lowest  production  rate  (159  Ibs
UsOs/hr).    Alternately,  the  test   resulting  in  the  lowest  emission  rate
(0.11  Ibs U30s/hr)  corresponds to  a  period  of  identical  production  rate
(159 Ibs
    The  study  demonstrated  that the  emission  rate at  a  given  time  (the
"instantaneous"  emission  rate)  is   controlled  by  a complex  interaction  of
several factors.   In addition  to the mill  operating parameters  listed  above,
these include,  as a partial  list, such things  as:
         - Drying and packaging ventilation design factors  which, under
           varying operational  conditions,  create  varying yell owcake
           entrainment  with  subsequent variations  in yellowcake concentra-
           tions  arriving at the stack gas  cleaning system.
         - The adjustment and "tuning" of stack  gas cleaning systems at a par-
           ticular time (i.e.,  the "instantaneous" efficiency of the system).
         - The general  condition  and  cleanliness  of the yellowcake processing
           and stack gas cleaning systems,  including ductwork.   There may be a
           considerable amount  of yellowcake deposited  in  these systems that
           may be available  for resuspension.
         - The nature  of operating  procedures  for yellowcake  packaging, and
           how well  these  procedures  are  being  followed  by  the  particular
           operator at the particular time.

    Undoubtedly,  the  instantaneous  yellowcake  emission rate at the  point  of
the stack  exhaust  involves  still other  factors,  which  may  be  subtle,  inter-
mittent,  and  highly variable  in  impact  on the emission rate.   These factors
are  also difficult  to  observe,  and  are  difficult or impossible,  at  least
within  the  context of  this  study,  to quantify  relative to  the  observed
emission rate.

    Table  7  shows  a measure of  the  gross variability of  the  test  results,
expressed  as  a  ratio of  the highest to the  lowest measured  values  for each
stack tested.  (The Anaconda result is not  included because  only one test was
performed  there.)   For  the UNHP and Sohio packaging stacks,  an "alternate"
calculation  was  made  to  include  additional   tests   made  under  operating
                                      24

-------
         TABLE 7.   VARIABILITY OF MEASURED YELLOWCAKE  EMISSION RATES
Mill
Kerr-McGee*
UC-Uravan
UN-Churchrock
UNHP*
No. Of
Stack Tests
Main
East
West
Dryer-Package
Dryer
Package
Vanadium
Roaster
Dryer
Package
2
3
3
7
3
3
5
3
3
Ratio-Highest To
Lowest Emission Rate Remarks
2.9
2.0
7.1
2.8
2.2
1.9
2.3
3.4
1.4 Yellowcake was being
Sohio
              Package
               (alternate)
Dryer
Package
3
2
Package       3
 (alternate)
                           23
4.7
3.9
                                         71
                           packaged.
                           Includes 3 tests above,
                           plus one when yellow-
                           cake was not being
                           packaged.
Excludes test made
when bag filter was
ruptured.
Includes test with
ruptured filter.
* The stacks at these mills were tested on two occasions at five-month
  intervals.
                                       25

-------
conditions that  were known to  be unusual or  markedly  different.   The table
notes these conditions.   Although usually the number of  tests  on each stack is
small, the  ranges  of emission rates  shown  in  this table give  an estimate of
the variability to be expected from a series of tests on a given stack.

    The alternate calculation for the UNHP  packaging  stack  shows the emission
rate  range from  a  packaging stack  that  can occur depending  upon  whether
yellowcake is or  is  not being packaged  at  the time  of  the test.   The Sohio
alternate calculation shows  the effect  of  a ruptured bag  filter.   Certainly
more data are needed to draw definitive  conclusions  about the  effects  of such
variables.    However,   by  just  these  two  tests,  the  emission rate  shows
variation by a factor of roughly 10 to 20.

    Aside from these major known variables, this study shows that the emission
rates  from a  given  stack, as  measured  by  an individual  test, can  vary by
roughly a factor of two to five.  Again,  the relatively  small  number of tests
involved should be kept in mind when evaluating this data.  The only exception
is the three  tests on  the Kerr-McGee west Microdyne stack,  which  vary  by  a
factor of  seven.   This may be  at least partially explained by  the  fact that
mill operating data were not obtained during these tests.   Consequently, some
unobserved  factor  (e.g.,   differences   in   the  amount  of  yellowcake  pack-
aged) may account for the larger range of these results.

    The range  of variation discussed above  includes sampling  and  analytical
error.  The Method 5 sampling  procedure is  estimated to have an overall prob-
able sampling error of  about  7% (Shigehara  et  al.,  1970).   Analytical errors
may vary  considerably,  depending on the method used  and the  expertise of the
laboratory performing the  analysis.   Estimated analytical errors for  routine
analyses at the uranium levels involved are as follows:
         - Delayed neutron counting:  5% or less
         - Radiochemistry and alpha spectroscopy:   5-20%
         - Fluorimetric analysis:  25%
         - Colorimetric analysis:   3%

The approximate  combined  sampling and analytical   error  could  therefore range
                                      26

-------
from less than 10% to about 25%.   Consequently,  the combined error in measur-
ing the emission rate is considerably less than the  minimum  variation of 100%
(factor of two) observed in the stack test results.

Estimation of Annual  Average Emission Rates
     Disregarding  the  variability and  the apparent  lack  of correlation,  we
calculated  emission  rates  of  U^OQ  as  a  percentage  of  UsOa  processing
rates for four of the mills where such data was  available,  and  averaged these
percentages over the number of tests conducted.  The results of these calcu-
lations, listed in Table 6, show  United  Nuclear  Churchrock  (UNC) releasing an
average  of  0.118%  of   the  1*303 dried  and  packaged.  Sohio  releasing  an
average  of  0.0066% of the   UaOs  dried  and  packaged,  and  Union  Carbide
releasing an  average of 0.094% of the  UsOs  dried  and packaged.   The Union
Carbide result is  notable  since it corresponds  to  an  identical  release rate
calculated  by  Momeni et  al.   (1979).   They were  provided  the  Union Carbide
     emission  rate data from  the tests reported  herein,  but  calculated  the
      production  rates  by   considering  ore throughput  and  grade,  extraction
efficiency, and mill operating  time.   Using Union Carbide data  and  that from
another mill,  Momeni  recommended  using  a value  of 0.1% for estimating U^QQ
released  from annual UsOs production  estimates.-   Note  also  that  the  UNC
data agrees with Momeni's estimation.  Sohio is obviously the exception with a
total  UaOs  release rate  of  0.0066%,  and based  on the  data  of  Table 6,  a
0.1%   release   rate  applied   here  would  over-estimate   their  total  UaOs
release by a factor of  15.

    The use of emission  rates (as normally reported from stack tests viz Ib or
pCi  U30a/hr)  in calculating  annual   source  terms must consider the operat-
ing periods of the several  sources.  Since yellowcake packaging operations  are
typically intermittent with respect to yellowcake  drying  operations,  it would
obviously be invalid to  simply add the reported hourly release rates for these
two  sources  to arrive  at  a total hourly  release rate in  calculating annual
source  terms.   How  intermittent  packaging operations  are,  at  least  at  two
mills,  can  be calculated based on  the data of  Table 6.   These calculations
show the  ratio  of  the  packaging time to drying time at  UNC and Sohio  to be
respectively about 0.14 and 0.08.  Therefore,  in calculating an annual source
term, the total  operating  times for each  source  must be defined after which

                                      27

-------
individual source terms  may be calculated for each  stack  for the year.   The
source term for each stack is  then  summed to  arrive at  the  total  annual  source
term.   Annual  source term  calculations  based  on  the summation of drying  and
packaging stack hourly release rates,  without time-weighting,  at  UNC  and Sohio
would overestimate  their annual  source terms by a factor  of  1.5  and  2.8,
respectively.

Concentrations of Uranium Daughter  Radionuclides in Yellowcake
    Table  5  shows the concentrations  of the  three  uranium nuclides  (U-238,
-234,  and  -235)  and  of  the  daughter  nuclides  (thorium-230,  radium-226,
lead-210, and polonium-210).  This  table includes  results for  six samples that
were  analyzed  by three  laboratories,  and for three  additional  samples  that
were analyzed only by the EPA  laboratory.  Each analysis shows the percent of
each daughter nuclide's activity relative to  the parent U-238  activity.

    The agreement between the  three  laboratories  is  generally good.   Usually
where comparable analytical results are  available (between EPA and LFE), they
are within the  two-sigma counting  error range.   (Remember that the Eberline
concentration results are not  directly  comparable to the EPA  and LFE results,
due to  reporting differences.)  All the samples  analyzed  by  EPA  and Eberline
show  U-238 and U-234  in equilibrium within the  limits of reported counting
errors.   (LFE did not report U-234 results.)

    For the daughter nuclides, the results reported  by all three laboratories
can  be  compared  by  referring to  the  calculated percentage  of the daughter
nuclides relative to the parent U-238.   The  differences between  these  results
are fairly randomly distributed between  the  three laboratories.   For thorium-
230,  EPA tends to  be  low  and Eberline high,  with  LFE in the  middle.   For
polonium-210,  EPA  is  generally  lower  than  LFE (Eberline   did  not  report
polonium-210 results).   While  inter-laboratory agreement does  not  necessarily
prove accuracy,  it  is  encouraging that  there are  no  strong biases between the
three laboratories.  Because  of  this,  the percentage  values  for  all analyses
of all  samples  from  each of the  four  mills have  been averaged to arrive at an
overall  average daughter percentage in  yellowcake  from each  mill.   Table  8
shows these values.
                                      28

-------
    Sears  (1975)  reported  that,  for  an acid  leach  mill,  about  5% of  the
thorium and  0.2%  of the radium are  carried over  to  the yellowcake  product,
although these percentages may vary according to the  details  of  the chemistry
used for  uranium  extraction and yellowcake  purification.   An alkaline  leach
mill is  reported  to carry over 1.8% of  the radium,  and none of  the thorium.
For  both  circuits,  no other  significant  radionuclides  are reported  to  be
present.  The results  in Table 8  show  that  thorium-230 in the yellowcake from
the three acid leach mills is considerably lower than  the  5%  stated above  and
that the  radium-226 is equal to  or  considerably lower than the  0.2% stated.
For  the  one alkaline  leach mill, significant  thorium-230 was  found in  the
yellowcake, while the  radium-226  was  about  17% of the value  quoted by  Sears.
Measurable amounts of  Pb-210 and Po-210  were found in  all  yellowcake samples,
with that  from  the alkaline leach mill  being  higher than that from  the acid
leach mills.  Always,  however, less than 0.09%  each  of lead and  polonium  was
carried over to the yellowcake.
                Table 8.  DAUGHTER RADIONUCLIDES IN YELLOWCAKE
                          AS PERCENT OF U-238 PARENT*
                               Mean ± one standard deviation (percent)
Nuclide/Mill    UNHP        Union Carbide    Kerr-McGee
             (3 samples)    (3 samples)      (2 samples)
  Sohio
(1 sample)
Th-230      2.58 ± 0.309    0.689 ± 0.110  0.0980 ± 0.0360    0.284 ± 0.0731
                (7)**            (7)             (3)               (3)
Ra-226     0.312 ± 0.036   0.0266 ± 0.0165 0.0038 ± 0.0042  0.00033 ± 0.00006
                (7)              (7)             (4)               (3)
Pb-210    0.0876 ± 0.0117  0.0271 ± 0.0119 0.0107 ± 0.0040  0.00467 ± 0.00208
                (7)              (7)             (3)               (3)
Po-210    0.0688 ± 0.0095  0.0246 ± 0.0124 0.0080 ± 0.0000  0.0170  ± 0.00849
                (5)              (5)             (2)               (2)

 * Calculated from results in Table 5.
**  Number  in parenthesis  is  the number  of  analyses used in  calculating  the
   mean.
                                      29

-------
                         SUMMARY AND RECOMMENDATIONS

    The Nuclear  Regulatory  Commission  is  currently considering  the monitoring
and sampling  procedures  that  should be  required  at  uranium  mills to  demon-
strate compliance with environmental radiation  protection  standards.   In order
to estimate  dose to  people  near the  mill,  a  source term,  or  radioactivity
released per unit time, must be known or estimated.   This  is  usually expressed
as curies per year  in order to average over a suitable  period.   The  problem
with determining  uranium mill  source  terms  in  general,  and yellowcake stack
source terms in particular,  is  in realistically estimating  or  calculating this
annual  average emission rate.

    This study shows that measured emission rates  from a yellowcake processing
stack can routinely vary by a  factor of  two  to  five  when tests  are  conducted
without any  controls  on the  mill  operating parameters.   In addition,  these
tests show that major anomalies (e.g.,  a ruptured  filter  in the  exhaust clean-
up system) can cause  variations of as much as a factor of 20 in the measured
emission rate.  Based on our experience, the major operational  parameter that
affects the  emission  rate  is  simply  whether  yellowcake  is  or is  not being
dried or packaged during the testing period.  Typically, the  cleaning systems
run continuously  while  the  operations  may be performed intermittently.   This
is especially true  for many packaging operations  and to  a lesser  extent  for
drying operations.  As discussed earlier, results  from a  single  test at United
Nuclear-Homestake  Partners   show   measurable  amounts  of  yellowcake  being
released  from the  operating  packaging  stack  (0.02  Ib  UsOs/hr)  even  though
yellowcake was not being packaged during the test.   Later tests  on this stack
show a significant increase  in the emissions during periods when packaging was
under  way.    Consequently,  an  emission  rate  measurement  (or  measurements)
representing only one operating condition would bias  the  annual  average source
term considerably.

     Major events that affect  the emission rate would also have  to be consid-
ered  for  a  calculation as  described  above.   A serious  malfunction  of  the

                                       30

-------
exhaust  air  cleanup   systems  is  the most  obvious  of  these,  as  discussed
previously for the  Sohio  sample.   Maintenance or  repair work on  the exhaust
system  might also  increase  the  emission  rate  significantly by  dislodging
yellowcake  from  the  system,   although  this  would  probably  be  short-lived.
Conversely,  a complete shutdown  of  the yellowcake  systems would  reduce the
emission rate to zero during the down period.

      Based on the  results of this   study  and on  other  general  information
obtained  during  the  course  of the  study,  we can  recommend a  general  stack
testing and  emission  rate  calculation  scheme that  should provide a reasonable
estimate  of  the  annual source term  from yellowcake  stacks.   Basically, this
scheme would  consist  of sampling  each stack several  times  under various mill
operating conditions and time-weighting these various emission rates according
to the  proportion  of  time per year  during  which  the stack is  operated under
each particular condition.  Our general recommendations are as follows:

    1. A  number of samples would be collected from each stack over a period of
       a year to establish the range of emission rates from that stack.  These
       tests  would be  made  while the  yellowcake  processing  operations are
       underway  with  the operating conditions  maintained  at  steady-state
       conditions.

    2. During this  year,  each  stack  would  also be  tested  several  times when
       yellowcake operations are  not  underway,  but  with the air cleaning sys-
       tems  in  operation,  if  this  condition  occurs  normally  at  the  mill.
       This  would establish the range of emission rates under this condition.

    3. The tests  should  be conducted  according  to the  EPA Reference Methods
       using  the  appropriate modifications  for  yellowcake  stack  testing
       described earlier in this report.

    4. For each stack, a record of the operating time under each of the above
       conditions would be maintained  for a year,  and summarized on an annual
       basis.
                                      31

-------
5.  The total annual  source  term for each  stack  would then be  calculated
   by multiplying the hourly emission rate  for each condition  by the  hours
   per year the  stack operates under that  conditon,  and summing for  all
   known operating conditions.   The annual  source term for the  mill  would
   be obtained  by summing the source terms  for each stack.

6.  The above calculation  is  intended to  apply  to uranium emission  rates.
   The  uranium  daughter  nuclides  are a  small  fraction of  the uranium
   in yellowcake,  and the analyses for them are expensive and  time con-
   suming.   Since  the ratio of the daughter nuclides  to uranium appears
   to be fairly constant for a given mill, we suggest  that they be  deter-
   mined once  a  year  from  a blended sample  of the  yellowcake  produced
   during the year.

7.  An alternate method of estimating the source  term is  to relate the em-
   ission rate  from the various stacks to the amount of yellowcake proces-
   sed (e.g., dried or packaged) in the areas served by  each  stack.   This
   technique requires an  additional piece  of  data,  the  amount  of yellow-
   cake processed during a stack  test.  One can then calculate the uranium
   emission (not emission rate)  normalized  to  the yellowcake  production,
   expresed as pCi  Ut0t/pound  of yellowcake processed.  Since  the  amount
   of the yellowcake  produced  per  year  by  the mill is known,  the  annual
   source  term  can  then  be  estimated  by  multiplying  this  normalized
   uranium emission by the annual yellowcake production.

   We  feel  that  this  calculation  method  would  be  inherently   less
   satisfactory than  the  time-weighted method discussed  above because (a)
   it does  not  account for  releases which occur when yellowcake is  not
   being processed;  (b)  in  some  mills,  the exhaust stacks do  not serve a
   single process  area,  thus complicating the  relationship  between yel-
   lowcake  processing and uranium  releases;  and (c) as  shown previously,
   there is very poor correlation  between the measured emission rates  and
   the  concurrent  yellowcake processing  rates.   A  comparison  of the  two
   methods  may be more  feasible  when  further  data   and  experience  are
   available.
                                  32

-------
    Technically the number and frequency of  samples  should  be determined by a
statistical  analysis  of  the  data, and  would depend  primarily on the  varia-
bility and the distribution of the results.   These parameters would determine
the  reliability  of the  calculated annual  emission  rate  and  the  degree  of
confidence associated with it.   However, from a  regulatory viewpoint,  it may
be simpler to specify a fixed number  and frequency of samples.

    We think it  important  to remember the  variability between mills and the
uniqueness of  each mill.   Also, changes  in operating parameters  can  change
emission rates over a period of time.   Consequently,  a single, fixed sampling
regime  is  not  necessarily always  the  optimum.    It  may not  provide  the most
useful technical data, and  it  may not be the least costly way to  obtain the
data, if costs are to be considered.

    Our approach  in  the above sampling  scheme has been to outline  a general
program that should provide a  realistic estimate of  the annual  emission rate
for a mill.  Logically, this would involve a larger number of samples and more
frequent sampling the first year.  This  will  allow one to  develop a  feel for
the  range  of the  results  for the particular stacks and,  hopefully, for the
causes of these variations.  Once this range  is  defined and the variables are
at least partially controlled, the sampling program could be modified and used
primarily  to follow  trends  and  to  confirm  that  no major variations  are
occurring.

    All the  above  discussions of  stack  sampling techniques  and  programs are
based on the EPA Reference Method 5 sampling technique used in this  study.  As
mentioned previously, the  major  shortcoming of this  method for measuring the
annual yellowcake emission rate is the short  duration of the test period (one
to two hours) compared to the  time frame over which most of the variations in
the  emission rate  occur.   As a  result,  a single  test, or a  few tests, will
likely  give  misleading  results.   As  more  and  more  tests are  conducted  to
average out  these  variables,  one is, in effect,  approaching continuous stack
sampling.  This  ideal situation would  eliminate  the need  for, and uncertainty
involved in, averaging and manipulating the  results  from a  number of individ-
ual  tests to obtain  the annual average.  Such a  continuous monitor should be
reliable, reasonably  accurate, and inexpensive.    Argonne  National  Laboratory
                                       33

-------
(Momeni et  al.,  1979)  has developed, but  not  yet tested, such a  concept  and
prototype.  We recommend the further development  of such  equipment, with  field
tests and evaluations of its suitability in this  area.

    As an aside, we  feel  it important to mention  a dramatic  reduction of  the
packaging stack  yellowcake  release  rate  we observed at  the United  Nuclear-
Churchrock  mill,  after minor  modifications.     During   the  emissions  tests
reported herein,  the average emission rate for  the packaging  stack  was 0.19 Ib
UaOs/hr  (Table 6).  Subsequently,  a series of  particle-sizing studies were
performed on the  same stack.  The emission rates  calculated from these studies
averaged  0.005  Ib  UsOs (Fort  et  al.,  1980).    Between  these two  studies,
minor  modifications  had been made  on  the  packaging stack  ventilation  and
cleaning  systems (Todd  Miller,  United Nuclear-Churchrock, personal communica-
tion, 1978).  While emission rates determined from particle-sizing  studies  are
not  necessarily  as accurate as  those determined by the  Method  5  procedures,
the  magnitude of  the  reduction  indicates a  definite  improvement  in  the
efficiency  of the  exhaust  air cleaning  system.    Whether cleaning systems at
other mills could  be "fine-tuned"  to achieve this magnitude  of  emission rate
reduction is unknown and was not a goal  of these  studies.
                                       34

-------
                                  REFERENCES

CFR  (1975),  Code  of Federal  Regulations,  Title 10,  Part 20.  "Standards  for
Protection Against Radiation."

EPA  (1976),  "Environmental  Analysis of  the  Uranium  Fuel  Cycle,  Part  IV  -
Supplementary Analysis-1976," EPA 520/4-76-017, July 1976.

CFR  (1977),  Code  of  Federal  Regulations,  Title  40.  Chapter  1.  Part  60,
Appendix A, "Standards of Performance for New Stationary Sources - Revision to
Reference Method 1-8."

Fort, C.  W.,  Jr., R.  Douglas,  R. Gauntt, A.  R.  McFarland  (1980),  "Particle
Size  Distribution  of  Yellowcake  Emissions  at the  United Nuclear-Churchrock
Uranium Mill," ORP-LV-80-1, January 1980.

Gale, N.  H.  (1967),  "Development  of  Delayed Neutron Technique  as  a  Rapid and
Precise Method  for  Determination  of Uranium  and  Thorium at Trace  Levels in
Rocks  and  Minerals,  with   Applications  to   Isotope  Geochronology,"  IAEA
Symposium Proceedings,  Radioactive Dating and  Methods of Low-Level  Counting,
August 1967.

Glauberman, H., and  A. J. Breslin  (1964),  "Uranium  Mill  Tailings Study," HASL
Technical  Memorandum 64-14,  Health and  Safety  Laboratory, U.S.  AEC,  New York
Operations Office, July 31, 1964.

Harward,  E.  David,  editor (1977), "Workshop on Methods  for  Measuring Radia-
tion in and Around Uranium Mills," Atomic  Industrial  Forum,  Inc. August 1977.

Johns, F.  B.,  editor (1975), "Handbook of  Radiochemical  Analytical  Methods,"
EPA 680/4-75-001,  February 1975.
                                      35

-------
Momeni, M. H., W. E.  Kisieleski, D.  R.  Rayno,  and  C.  S.  Sabau (1979), "Radio-
isotopic  Composition  of Yellowcake  An Estimation  of Stack  Release  Rates,"
NUREG/CR-1216,  ANL/ES-84, December 1979.

Ragland, J.  W.,  K.  M. Gushing, J. D. McCain, and W.  B.  Smith (October 1976),
HP-65  "Programmable Pocket Calculator  Applied to  Air Pollution  Measurement
Studies:  Stationary  Sources."    Ind.  Env.   Res.   Lab.,  U.S.  EPA,  Research
Triangle  Park,  N.C.  and Southern Research  Inst., Birmingham,  AL,  Contract
Number EPA 68-02-2131.  EPA Report Number 600/8-76-002.

Sears, M.  E.,  R.  E. Blanco,  R. C. Dahlman, G. S. Hill, A. D.  Ryan,  and J. P.
Witherspoon  (1975), "Correlation of  Radioactive Waste  Treatment  Costs and the
Environmental Impact  of  Waste  Effluents in the Nuclear  Fuel  Cycle for Use in
Establishing  'as  Low as  Practicable'  Guides  -   Milling  of  Uranium  Ores"
ORNL-TM-4903, Vol. 1, May 1975.

Shigehara, R. T., W.  F.  Todd,  and  W. S. Smith (1970), "Significance of Errors
in  Stack   Sampling  Measurements,"  APCA  No.   70-35,  presented  at  the  Annual
Meeting  of  the  Air  Pollution Control  Association,  St.  Louis,  Missouri,
June 14 -  19, 1970.

Sill,  C.   W.  (1977)  "Simultaneous  Determination   of U-238,  U-234,  Th-230,
Ra-226, and Pb-210 in Uranium Ores, Dusts, and Mill Tailings," Health Physics.
Vol. 33, No. 5, November 1977.
                                      36

-------
                                  APPENDIX A

    HP-67/97 CALCULATOR PROGRAM FOR ESTIMATING STACK GAS MOISTURE FRACTION

The basic equation for estimating moisture fraction is:  BWs = D\
    Where: BWS = Moisture fraction
           4a  = Dew point saturation vapor pressure (in. Hg)
               = a»   (pabs-*")(Td-^w)        (The Carrier Equation)
                    " 2800 - 1.3 * Tw
           i"  = Saturation vapor pressure at Tw (in.  Hg)
          pabs = Absolute stack pressure (in. Hg)
          Td   = Stack gas dry bulb temperature (°F)
          Tw   = Stack gas wet bulb temperature (°F)

The  HP-67/97  program  calculates   i"   for   a  given  Tw  using  a  series  of
equations derived to define saturation vapor pressure (SVP) in inches of Hg as
a function of  wet  bulb  temperature over the range 0-159°F.   These equations
are given below:

    for Tw =    0- 29°F,  SVP = 0.03821 e°-04911 * Tw
        Tw =  >29- 59°F,  SVP = 0.05290 e0.03840 * Tw
        Tw =  >59-105°F,  SVP = 0.07590 eO-03250 * Tw
        Tw - >105-159°F,  SVP = 0.14210 e0.02651 * Tw

SVP  values  calculated  using  these equations  will have a  maximum error  of
approximately 3%.  For  those  interested, more  precise  results may be obtained
by reducing the temperature intervals and using the HP-67/97 exponential  curve
fit  program  to  define  the equation  of the new  data  group.   In this  same
manner,  SVP  equations   for  temperatures  greater than  159°F  may  also  be
obtained.  Use  the goodness of fit test (r2)  in the program to  determine the
appropriate grouping for the accuracy desired.
                                       37

-------
      The equations have  been found to  be more  than adequate  in  estimating
moisture  fraction  for  calculating isokinetic  AH.   The  maximum  difference
observed  between  BWs  calculated  using  this method  and  Bw$ as  determined by
the condensation method for the 40 tests described in this  report was approxi-
mately 10%.  Typically, the agreement was  a few percent  or less.  This method
has proven  sufficiently accurate for calculating  AH with  dew  point tempera-
tures as high as 141°F.

    To obtain the accuracy described,  care must be taken  in making the wet and
dry bulb  temperature  measurements.   This  amounts  primarily  to  allowing the
thermometer to  reach  equilibrium conditions.   Of course,  when  using separate
thermometers for the two measurements,  they must be well  matched.

     This program is easily adapted to  the HP-65 by one of two  methods: (1)
Perform  the SVP  calculations for  the  range  0-1S9°F on  one  magnetic  card,
storing  SVP and  Tw for  use  with  a  second magnetic card to complete  the
calculation  for Bws.    (2)   Incorporate  on one  magnetic  card the complete
program for Tw = 0-59°F, using a second magnetic card for Tw s >60-159°F.

    It should be pointed out that this method will  also  work to determine the
moisture fraction in ambient air.
                              USER INSTRUCTIONS
    STEP                      INSTRUCTIONS             .     DISPLAY
     1          .         Input TW(°F), Touch "A"          SVP,(in. Hg)
     2                   Input Td(°F), Touch HB"             T
-------
STEP  . KEY    CODE
002   .::«;
Sow   /•, v .
004   STC4
005      2
006       •
.007   RCLC
MS   n'>V?
305   5TC1
010
011      O
012      4
013      S
314      I
815      :
017
313
.015
320
321
822
323
324
025
           5
? * *«
tat A A *
~e fln
ww fcl
f - "f
1C Tw
»" •>.»
^t. wf
   02
   25
"V i"P
ww ww
f i".?4
* w wt
22 01
  -62
   00
   04
   05
   01
   01
   02
  ,7V
   77
   WW
  -62
   00
   03
   CS
   02
   01
  -35
7f pi
ww W *
027
82£
825
830
031
832
833
834
835
836
837
836
835
040
841
842
843
344
345
846
847
848
845
P.T,V
tLBLi
f
w
5
PCI?
;^y?
STC2
•
0
«*
I
4
X
e*
•
0
r
w
.^
£
C
V
CTrtf
5y?
6TC5
*
c'
•1
i.
6
r
w
1
A
X
ex
0
t
•
4
2
4
A
X
STC1
KTH
+Lr:e
STC2
R7N
*LBLC
STJ3
                                             JCODE
                                               STEP   KEY   CODE
                                                    C5
                                                  o
                                                  : or
                                                    Cw
                                                    02
                                                    35
                                                    •c^
                                                    05
                                                    CS
ww Ui
   24
21 £3
   dl
   05
   05
36 C?
16-34
22 05
"-62
   00
   02
   06
   05
   01
                                                    .***
                                                    04
                                                    02
                                                 35 01
                                                    24
                                                 21 12
                                                 35 02
                                                    •* «
                                                    »;t
                                                 21 13
055
183
181
182
103
184
185
106
187
108
185
110
111
112
113
114
115
116
117
118
115
128
121
122
123
124
125
K-12
RCLQ
-
RCL3
RCil
-
X
KCL3
A
•
W
X
2
S
0
0
xrv
-
~
RCLl
xry
-
STC4
RCL3
T
D5P3
RTK
36 02
Jc 20
-45
3o 33
36 31
-45
.7*
WW
3c 00
31
-62
03
-35
32
3£
30
30
-41
-45
-24
33 31
-41
-45
35 34
36 03
-24
-63 33
24
                                          39

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO 2.
ORP/LV-80-3
4. TITLE AND SUBTITLE
Radioactive Emissions From Yellowcake
Stacks at Uranium Mills
7. AUTHOR(S)
C. W. Fort, Jr., R. L. Douglas and W.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Radiation Programs -Las Veqas
U.S. Environmental Protection Agency
P.O. Box 18416
Las Vegas, Nevada 89114
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Radiation Programs-Las Vegas
U.S. Environmental Protection Agency
P.O. Box 18416
Las Vegas, Nevada 89114
Processing
E. Moore
Facility
Faci 1 i ty
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
October 1980
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
14, SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
1S. ABSTRACT
       A sampling program was  undertaken to evaluate the releases  of uranium concentrate
  from uranium mills.  This  concentrate, known as yellowcake,  is the final  product of
  the mill.  It is routinely released to the environment through stacks  which exhaust
  air from both the yellowcake drying and packaging operations.  During  this study,
  samples were taken  in  these  stacks at six mills to evaluate  the  total  yellowcake
  emission rates.  This  paper  describes the sampling and analytical  methods used, and
  presents the results obtained.   Considerable variation in  the emission rates was
  observed, both between mills and at the same mill over time.  A  total  of 40 emission
  rate tests were made,  with the  measured rates ranging from 0.001  Ib U^0g (0.4 yCi
  Uf t -,) oer hour to 0.9  Ib U~0p (220 yCi U.   . ,) per hour.   Data are also presented
  oirtfie levels of uranium daughter radionucTTdis (Th-230, Ra-226,  Pb-210,  and Po-210)
  found in the yellowcake.   The implications of the results  of this  study for monitoring
  to meet State and Federal  licensing requirements are discussed.   Measurements such as
  these can provide specific and  realistic input data to models used to  estimate radia-
  tion doses to people living  in  the vicinity of the mill.   However, the variations
  observed indicate the  need for  relatively frequent and site-specific measurements.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS  C.  COSATI Field/Group
  Uranium Isotopes
   Yellowcake
   1802
18. DISTRIBUTION STATEMENT

  Release to public
19. SECURITY CLASS (ThisReport)
  Unclassified
21. NO. OF PAGES
                                              20. SECURITY CLASS (This page)

                                                Unclassified
                           22. PRICE
EPA Form 2220-1 (R«v. 4-77)   PREVIOUS EDITION is OBSOLETE

-------

-------