Robert A. Taft Sanitary Engineering Center

TECHNICAL REPORT   W62-17
      PROCESS  AND WASTE
               CHARACTERISTICS
      AT SELECTED URANIUM MILLS
                                            •V
                       U. S. DEPARTMENT OF HEALTH
                         EDUCATION, AND WELFARE
                               Public Health Service

-------
 SEC TR W62-17
PROCESS AND WASTE CHARACTERISTICS
      AT  SELECTED URANIUM MILLS
                      Prepared by

            Radiological Pollution Activities Unit
                 Field Operations Section
                Technical Services Branch
       U. S. Department of Health, Education, and Welfare
                  Public Health Service
         Division of Water Supply and Pollution Control
          Robert A. Taft Sanitary Engineering Center
                    Cincinnati, Ohio
                        1962

-------
                 CENTER PUBLICATIONS
    The Robert A. Taft Sanitary Engineering Center is a national
laboratory of the Public Health Service for research,  training.
and technical consultation in problems of water and waste treat-
ment, milk and food safety,  air pollution control, and radiologi-
cal health. Its technical reports and papers are available without
charge to  professional users in government, education,  and
industry.  Lists of publications in selected fields may be obtained
on request to the Director, Robert A. Taft Sanitary Engineering
Center,  Public Health Service, Cincinnati 26.  Ohio

-------
                      CONTENTS

                                                      Page

FOREWORD  	    v

THE RESIN-IN-PULP URANIUM EXTRACTION PROCESS .    1

    Mines Development Company,
    Edgemont, South Dakota	    1

THE ACID  LEACH-SOLVENT EXTRACTION URANIUM
 REFINING PROCESS	   19

    I. Gunnison Mining Company,
      Gunnison. Colorado	   19

   II. Climax Uranium Company.
      Grand Junction, Colorado	•	   37

THE CARBONATE  LEACH URANIUM EXTRACTION
 PROCESS	   55

    I. Homestake-New Mexico Partners Company,
      Grants. New Mexico  	   55

   II. Homestake-Sapin Partners Company,
      Grants. New Mexico	   73

BIBLIOGRAPHY	   93
                            in

-------
                        FOREWORD

    This report contains the findings of detailed studies of process
and waste flows at five typical uranium mills. The studies were
initiated in 1957 by the Public Health Service for the purpose of
characterizing  the liquid and solid wastes resulting from uranium
milling processes.

    Uranium mills extract the naturally  radioactive  uranium from
its ores and produce a concentrated  product that is subsequently
refined elsewhere for use in nuclear weapons and reactors. The
extraction and recovery processes employed are determined by
the  character of the ore and its uranium  content.  The five mills
reported on here typify the processes normally encountered, i.e. ,
acid or alkaline leaching of the ore,  concentration and purifica-
tion of the leach liquor by ion exchange or solvent extraction,  and
chemical precipitation of the dissolved uranium.

    Although the radioactive waste materials, especially Radium-
226. were of primary interest in these studies, useful information
regarding the chemical characteristics of milling wastes was also
obtained.  The entire body of information thus developed forms an
excellent basis upon which to characterize the waste products
from the industry as a whole. This has resulted in an "Industrial
Waste  Guide for the Uranium Milling Industry", which is pub-
lished  as a separate SEC Technical Report.

    The generous cooperation and assistance of many individuals
and agencies have contributed greatly to  the successful comple-
tion of these studies.  The work was supported in part by funds
made available through the Environmental and Sanitary Engineer-
ing  Branch, Division of Reactor Development. U. S. Atomic
Energy Commission.
               E.G. Tsivoglou. In Charge
               Radiological Pollution Activities
               Field Operations Section
               Technical Services Branch
               Division of  Water Supply and Pollution Control

-------
THE RESIN-IN-PULP URANIUM EXTRACTION PROCESS.
    MINES DEVELOPMENT COMPANY, EDGEMONT,
                      SOUTH DAKOTA.

                        E. C. Tsivoglou *
                          D. C. Kalda
                        J. R. Dearwater

                  Introduction and Background
      During the summer of 1957, the Public Health Service car-
  ried out a field study of liquid wastes resulting from the extrac-
  tion of uranium from its ore in a typical refinery using the resin-
  in-pulp extraction process. The study was performed in coopera-
  tion with the  South Dakota State Department of Health, the Mines
  Development Company of Edgemont, South Dakota, and the United
  States Atomic Energy Commission. It was the first of a series of
  such surveys by the  Public Health Service to develop detailed
  knowledge  of the characteristics of wastes, particularly radio-
  actix'e wastes arising from the extraction of uranium from its
  ores.  Specific objectives  of the studies include detailed analysis
  of the extraction process,  characterization of the resulting liquid
  wastes,  evaluation of their water pollution and public health
  significance, and development of adequate and suitable waste
  control measures. At  the same time, parallel field studies of the
  fate of these  wastes  in the  water environment were carried out. *•

      The uranium refinery at Edgemont,  South Dakota,  is located
  on the banks  of the Cheyenne River about 35 river miles above
  Angustora  Reservoir, a recreational lake  (see Figure 1). This
  refinery of intermediate capacity for ore processing, is a typical
  example of the acid leach-resin-in-pulp process.  At the time of
  the field survey it was  processing slightly more than 500 tons per
  day of ore  that assayed about  0.20 per cent U.,08.  The mill had
  been in operation about 1 year prior to this study.  Virtually all
  liquid wastes were delivered to tailings ponds for storage and
  volume reduction by  evaporation and seepage.  There was a small
  direct liquid  discharge to  Cottonwood Creek, a tributary of the
  Cheyenne.

      The Cheyenne River near Edgemont, South Dakota, is a re-
  latively  shallow stream with a sandy bottom.  Flow is often tur-
  bid, and at the time  of the field survey biological life was rela-
  tively sparse. River flows have been recorded from  1928 to 1933
  and from 1947 to the present by the U. S. Geological Survey at a
  *Respectively, Chief,  Radiological Pollution Activities Unit,
   Division of Water Supply and Pollution Control, Robert A. Taft
   Sanitary Engineering Center,  Cincinnati,  Ohio: Chief, Water
   Pollution Section, South Dakota State Department of Health; and
   Senior Assistant Sanitary Engineer,  Radiological Pollution
   Activities Unit,  (deceased)

-------
                                    RESIN-IN-PULP PROCESS
          Figure 1. Radioactivity monitoring stations, Cheyenne River, 1956.

gaging station located just upstream from Cottonwood Creek.
There are some 7, 143 square miles of drainage area above this
gage. From  1947 to 1955 the average annual flow in the river at
Edgemont ranged from 20.5 to 144 cubic  feet per second (cfs); the
minimum monthly average  flow for this 9 year period ranged as
low as 0.02 cfs and as high as 1.6 cfs.  The flow drops to zero
every year for varying periods; in 1952 there were only 6 days of
zero flow, while in 1950 there were 62 days.  Over the 9-year
period there  were 296 days of zero flow; in other terms, the
records indicate that on the average the river flow at Edgemont
has been zero for 8.1 per cent of  the days.

    During February 1956.  before  the mill went into operation,
a radioactivity background  survey of the  Cheyenne River below
Edgemont was made by the South  Dakota  State Department of Health
and the Public Health Service. At that time  samples of river water,
mud or sand, and aquatic life were collected at four locations and
analyzed for  gross alpha and beta radioactivity.  The sampling
stations are shown in Fibure 1. Biological samples included
plankton,  algae, insects, and minnows.  The water samples con-
tained 10 to 40 micromicrocuries per liter (jt/ic/1) of dissolved
alpha activity and 10 to  120/i^c/l of dissolved beta activity.  Sus-
pended radioactivity was practically nil.  River mud samples con-
tained an average of 10 and 15 micromicrocuries per gram (u/tc/g)
of dry solids  of alpha and beta activity respectively, with no ob-

-------
MINES  DEVELOPMENT COMPANY                          3

servable variation between sampling stations.  The biological
samples showed correspondingly low concentrations of gross
radioactivity. Ground water samples had dissolved alpha and beta
activity in the same range as was indicated by the Cheyenne River
water samples.

     In 1957, a sample of slimes tails before entering the slimes
tailings pond yielded gross alpha and beta activities of 590, 000
and 780, QQQft/ic/1,  respectively.  Virtually all of this activity was
in suspended solids, and the dissolved activities were 980 and 930
/i/ic/l,  respectively.  A liquid sample from  the slimes pond indi-
cated similar dissolved activities, with only slight suspended
radioactivity. Subsequent water samples (June 1957) indicated no
significant radioactivity above background in the Cheyenne River
or in Cottonwood Creek, although the small  direct drainage from
the sands pond to this creek  contained 1, 400 and 1, 800/t/ic/l,
respectively,  of dissolved gross alpha and beta radioactivity. No
radium analyses were made  prior to the mill survey of July 1957.

     In terms of radioactive waste disposal, radium is the most
hazardous radioelement involved in the extraction of uranium from
its ores.  Only the uranium is wanted, and  all of its radioactive
daughters,  including radium, are disposed of as waste products.
Of all of these decay products,  radium has by far the lowest
maximum permissible concentration in water.2  Hence,  the
amount involved, as well as  the course of its passage through the
extraction process,  is of considerable interest. Of special con-
cern is the  question of how much radium becomes dissolved in
the processing of ore, and where this dissolved portion goes. One
of the prime aims of this study, therefore,  was to perform a
radium balance through the  mill,  and in so  doing to answer the
foregoing questions.

                      The Mill Process 3

     Briefly, ore received at the mill is crushed and ground, and
leached with sulfuric acid to dissolve the uranium.  The coarse
sands are separated and discharged to waste,  and  the remaining
slurry, or IX feed,  containing the slimes or fine solids,  proceeds
to ion exchange resin banks. Here the uranium is  extracted from
the feed solution by resin beads, and is in turn stripped from the
beads, precipitated, filtered and dried to form yellowcake,  which
is the final product of the process.  This uranium concentrate is
then shipped to other  facilities for  further refining and process-
ing.  The  IX feed solution,  having been stripped of its uranium,
becomes the slimes tails,  and is neutralized with lime and sent
to the slimes pond.  Figure 2 is a schematic flow diagram of the
entire  process.

-------
  4                                  RESIN-IN-PULP PROCESS

  SAMPLING

      Trucks are used to dump the uranium ore into either of two
  ore bins from the ramp shown at the far left in Figure 2. The ore
  is fed to the sampling plant from these bins by a conveyor belt. A
  magnetic iron separator removes metal scraps, nuts,  bolts,  etc.
  In the sampling plant the ore passes over a 1-1/2 inch vibrating
  grizzly,  then goes to a 1-1/2-inch jaw crusher, and falls on a con-
  veyor belt.  Ten per cent of  the ore is removed from the belt by
  calibrated rotating buckets and 90 per cent goes directly to the
  mill.  The 10 per cent passes over a 3/4-inch  screen,  goes to a
  3/4-inch jaw crusher, and drops to a conveyor belt.  Ten per cent
  of this (one per cent of  the original ore) is removed by calibrated
  rotating buckets, and the remainder leaves for the mill. The
  retained  material (1  per cent) goes to a 1/4- inch screen and a
  1/4-inch jaw crusher, and falls to a conveyor belt from which a
 Vezin sampler  removes 10 per cent. The remainder leaves for
 the mill.  The sample amounts to 0.1 per cent of the original ore
 feed,  or two pounds of ore per ton of ore fed.  This representa-
 tive sample is assayed for its UgOg content. After leaving the
 sampling plant the crushed ore goes to either of two 250-ton ore
 bins or to a 50-ton truck bin from which specification material
 can be taken for blending.

 GRINDING

     The mill bins each have  two bottom discharge hoppers.  One
 means of blending control consists of regulation of the feed rates
 from these four chutes.  Another method of blending involves
 scheduling of various shippers  lots through the mill.  Factors
 considered for primary blending purposes include  the grade of
 ore, particle  size of  sand grains, slimes content,  and oxidizing
 or reducing characteristics of the ore.

     The grinding facility consists of a 4- by 8- foot rod mill in
 open circuit, with a 42-inch spiral classifier.  The feed to this
 grinding section is very  fine,  not only because  the sandstone ores
 are poorly cemented but  also  because crushing and repeated
 handling before  the mill  bins cause a large portion of  the ore to
 be broken down  into individual grains.  The resulting feed is eas-
 ily ground to minus-12-mesh. Water is added at the spiral classi-
 fier to make a slurry. Water for the mill is obtained  from an
 artesian well at a natural temperature of 53 °C.

 LEACHING
    The slurry from the  grinding plant is pumped to a series of
four wood-stave tanks, each measuring 14 feet  in diameter and
 14 feet high, where leaching is carried out.  Each tank is equip-
ped with a 48-inch rubber covered propellor mounted on a 4-1/2-
inch diameter rubber  covered shaft; these tanks are operated in
series.

-------
                       SAIiPLIW AND CRUSHING
                                                                                      GRINDING
                                                                                          AND
                                                                                      LEACHING
                                                                                                                                    SANOS-SLIMES SEPARATION
                                                                                                SCLuTICN

                                                                                         cj   3AN«S
                                                                                                          TANK
 LEGEND:
  II  - ION EXCHANGE
   P  - PREO»A»T  LI l
   3  - BANK
   T  - TAILS
   S  - SUSGf
   E  - EATING
   •  - RESIN
 FCV  - FLO» CONTROL VALVE
 HIP  - KESIN M
                                                                                                                                                                                                                                                        Figure 2.  Flow diagram of resin-in-pulp process. Mines Development,
                                                                                                                                                                                                                                                                  Inc., Edgemont, South Dakota, July 1957.
3FO q13—173—3

-------
MINES DEVELOPMENT COMPANY                           5

    Most of the ore received at this  mill is easily leached.  Most
of the uranium is in an oxidized state,  and hydrocarbons or other
reducing elements are not present.  Excessive sliming is not a
problem,  because the ore is loosely  cemented and contains very
little clay, bentonite. etc.

    Sulfuric acid is automatically metered to the first two leach
tanks from a 30,000-gallon acid storage tank in quantity sufficient
to maintain pH from 0.9 to 1.4, depending upon the type of ore be-
ing processed.  Two continuously recording pH meters with calo-
mel electrodes are used at the leach tanks and are coupled to
controllers that automatically regulate the acid feed.  Leaching is
carried out at about 40°C, with no attempt to regulate heat in the
system.  The retention time in each  leach tank is about 3 hours
and the pH of the  pulp overflow from the fourth tanks is usually
under 1.5.

SAND-SLIME SEPARATION

    The  slurry, or pulp,  from leaching, with the uranium in solu-
tion, undergoes sand-slime separation in  five 30-inch by 17-1/2-
foot spiral classifiers and two 10-inch cyclones.  Three 2-inch
vertical sand pumps and two 3-inch pumps are used in this sec-
tion.  Sands advance from the  first to the fourth classifier, and
from the second cyclone to the fifth classifier, are washed, and
pumped to the sands tailings pond. To facilitate pumping, fresh
water is  added at the  pump.  Water recycled from the sands tail-
ings pond isintroduced at the second  and fifth classifiers to wash
the  leached sands and to reduce pulp density for a sharper sepa-
ration in the cyclone. To minimize losses of uranium-bearing
solution,  the spiral classifiers are set at a relatively steep slope,
which results in a small pool  surface and long drainage deck.
    The  overflow from the first cyclone contains 5 to 10 per cent
solids that are  minus-300-mesh  in size.  After screening to re-
move wood chips and other trash, this overflow  goes to a 21- by
21-foot wooden tank for storage ahead of the ion exchange circuit.
Powdered iron  is usually added at this tank to adjust the solution
EMF to 400.  in order to keep vanadium in its tetravalent state and
thus prevent poisoning of the resin during  ion exchange. The ion
exchange  feed has a pulp density  of about  1.05; entrained solids
are minus-300-mesh,  pH is 1.7 to 1.9, EMF is 400, and the con-
centration of UoOfl in solution is about 1.0  gram  per liter (1,000
parts per million). This pulp, with a small amount of recycle IX
feed from the banks,  is pumped to a small elevated constant head
tank,  from which  it is metered by means  of a weir to the distri-
butor for  the resin tanks.
ION EXCHANGE

    Basically,  the uranium  is absorbed from the IX feed solution
on anion  exchange beads.  The beads are then stripped of their
uranium bv an acidifier nitrate solution, and the uranium is  sub-

-------
 6                                   RESIN-IN-PULP PROCESS
 sequently precipitated from the nitrate solution.
     The resin-in-pulp (RIP) section consists of 14 rubber-lined
 steel tanks,  or banks,  each containing two stainless steel baskets
 which are 4-1/2 by 4-1/2 feet in cross section and 5 feet high and
 have 30-mesh openings.  The baskets hold a 10-inch bed of 20-
 mesh resin beads. The bank is filled with pulp slurry,  and the
 motor-driven baskets oscillate up and down in the tank at 6.3 cy-
 cles per minute, to ensure good contact of bead surfaces with the
 pulp.

     The distribution  of solutions in the RIP section is accom-
 plished by an ingenious device that eliminates all need for expen-
 size valving  in this section. The central unit is a distributor
 wheel through which IX feed solution and eluate enter the banks.
 The wheel is divided  into 14 compartments,  each connected by
 pipe to one bank.  A pH indicator and recorder  continuously gives
 pH of the IX  feed and eluting solution, and meters indicate and
 control the flows of these solutions.  Full flexibility of operation
 is obtained with the system, a detailed description of which   is
 given by Dayton. ^
     The banks are alternated between an adsorption (loading) cy-
 cle, during which bead surfaces are loaded with uranium from the
 pump, and an elution (stripping) cycle, during which the uranium
 is stripped from the beads by a stripping solution. In normal op-
 eration. 7 banks are on adsorption,  5 are on the elution cycle,  one
 between these two cycles is being washed, and  one bank is on
 standby.

     The countercurrent principle is used during loading and strip-
 ping. The  uranium-rich IX feed solution enters the bank contain-
 ing the most  completely loaded resin:  fresh eluate is added to the
bank where the beads have the  least amount of absorbed uranium.
 Pregnant eluate is taken from the bank where the beads are most
 heavily loaded with uranium, and the stripped IX feed leaves from
 the bank containing the least loaded resin. The bank that is taken
 off the loading cycle is the next bank placed in the circuit at the
end of the stripping cycle.  Solution from the seventh bank on the
adsorption cycle  is sent to the  slimes tails neutralization tank.
     Control is obtained by quick fluorimetric uranium analysis of
 samples taken at the first and second banks from the discharge
ends of the adsorption and elution cycles.  For instance, the lead
bank is taken off adsorption and another  bank is added at the end
of the adsorption cycle when the uranium concentration in the IX
effluent to tails builds up to a predetermined level.  Otherwise
uranium would be lost to the tailings.

     The banks are drained  by means of bottom  discharge ports
fitted with  a hollow vertical plunger.  The solution level in the
banks is controlled as desired  via large  funnels connected by
means o£ flexible rubber hose  to the hollow plunger. Excess

-------
MINES DEVELOPMENT COMPANY                           7

solution is thus drained to the bank pumps.  Control of the solu-
tion level provides smooth continuous flow from bank to bank, and
minimizes surging and overloading of any of the banks.

    The beads must be  hosed intermittently to keep them wet and
promote drainage, once a bank has been drained of solution.  The
beads tend to swell and  stick together when dry, but separate
readily if kept wet.  The bank taken off the adsorption cycle is
washed to remove slimes adhering to the beads, in order not to
contaminate the eluting  solution.  After elution. the beads are
washed to remove excess nitrate.  Wash water is kept to a min-
imum, and is  recycled to the IX feed storage tank.

    Eluting solution is made up in three wooden tanks.  At any
time, one tank is delivering fresh eluate to the RIP banks,  one is
receiving filtrate from yellowcake filtration, and the third, full of
filtrate,  is being adjusted with acid and  nitrate. Fresh eluate
entering the RIP  section has about 56 grams per liter  (g/'l) of ni-
trate  ion and is acidified with sulfuric acid to a pH of 1.2.

    Considerable purification and concentration of uranium re-
sults  in the RIP circuit.  The  anion resin beads extract in the
neighborhood of 99.7 per cent of the uranium in solution,  but only
a small fraction of the dissolved iron, vanadium and aluminum.
After elution,  the pregnant eluate  sent to precipitation assays 10
to 20  times the uranium  assay of the IX  feed solution,  and contains
10 to  12 g/1 of uranium,  expressed as U~0n. This uranium-rich
n. expressed as U.,0q.  This
a pregnant eluate holaing ta
solution is pumped to a pregnant eluate holding tank.

URANIUM EXTRACTION

    The pregnant eluate is next filtered and clarified, before
uranium precipitation,  in order to obtain a clean concentrate un-
contaminated by slimes solids carried into the pregnant eluate.
Sufficient milk of  lime  is added to bring the pH up to 3.5 with
subsequent precipitation of calcium  sulfate, or whitecake. which
is returned to the IX feed tank. The pregnant eluate is then clari-
fied by a 38-frame plate and frame filter press.  This prelimi-
nary filtration  also controls  filtrate buildup. Sulfates are con-
trolled also by bleeding off about 10 per cent of the  yellowcake
filtrate  to slimes  tails.
    The clarified pregnant eluate is next cycled through a yellow-
cake dust collector, which strips it  of dust from the yellowcake
dryers.  It then proceeds to one of two 12- by 14-foot precipita-
tion tanks, where magnesium oxide  is added in dry  form to pro-
duce a diuranate  precipitate.  Enough MgO is added to produce a
solution pH of 6.8, The magnesium  oxide  is rather  slow-acting
but produce a large floe as compared to other precipitation agents.
Precipitation required  from  4 to 10  hours.

    All of the chemical reactions involved are indicated on Figure
2, as well as in the process  described by Dayton.

-------
 8                                    RESIN-IN-PULP PROCESS
     The precipitated slurry is pumped to one of two frame filter
 presses for yellowcake filtration. A layer of filter paper and a
 layer of nylon filter fabric are used on the press frames. After
 filtration the cake is washed with water and given an air blow. Be-
 low the filter presses are removable drip pans, which are instal-
 led over a paddle re-pulper that extends the  full length of the  fil-
 ter. The drip pan is removed and the filter opened; the precipi-
 tate is scraped off the frame and drops to the re-pulper,  which
 keeps the cake fluid while it is delivered to an agitator.  The con-
 centrate that forms  on the drum is scraped off and drops to a
 hopper,  from which it is drummed for shipment.

     The drum dryer operates under a slight  vacuum, and the ex-
 haust is pulled through a dust separator fitted with ceramic baf-
 fles.  Clarified eluate is introduced at the top of the separator and
 percolates down, stripping the uranium concentrate from the  dryer
 exhaust.  The scrubbed air is vented to atmosphere.

     The filtrate from the yellowcake presses goes to the eluting
 solution tanks, where it is adjusted with acid and nitrate to  make
 fresh eluant. Ten per cent of the filtrate is delivered to the
 slimes tails to prevent sulfate buildup.

    The slimes tails from the RIP banks are sent to a neutraliza-
 tion tank, where lime is added to bring the pH to about 9.5.  After
 neutralization,  these wastes are  delivered to the slimes tailings
 pond, a large lagoon located near the Cheyenne River.

                      The Mill Survey

    For purposes of analyzing the mill process and characteri-
 zing the  resulting liquid wastes, eight sampling stations were set
 up within the mill (Table 1); the field survey  was carried out from
July 25 through July 30, 1957.  Six other sample types were col-
lected outside the mill.  These included:

    a.    Liquid from the slimes tailings pond.

    b.    Wet solids from the slimes tailings  pond.

    c.    Liquid from the sands tailings pond.

    d.    Dry sand from  the sands tailings pile.

    e-    Silt and wet solids from  the slimes pond outlet to the
         Cheyenne River.

    f-    Direct liquid drainage to Cottonwood Creek.

    Sampling stations 1, 2,  4,  5, and 8 were main mill process
streams, whereas 3, 6. and 7 represented liquid waste streams.

-------
MINES DEVELOPMENT COMPANY

           Table 1. SAMPLING STATIONS WITHIN THE MILL
           Station
           number
            1

            2

            3a

            4
                                 Description
Raw ore slurry before entering leach tanks.

Leached ore slurry before sand-slime separation.

Sand slurry proceeding to sands tailings pond.

IX feed solution.

Loaded eluting solution from RIP circuit.

Slimes tails from IX banks, after absorption or uranium
and before neutralization.

Neutralized slimes tails proceeding to slimes pond.

Yellow cake.
           a Additional dilution water is added to this process stream beyond the
            sampling point, to facilitate pumping.

     Sampling inside and outside the uranium mill was performed
by personnel of the  Public Health Service and the South Dakota
State Department of Health.  Sampling within the mill commenced
on July 25 and was completed on July 30, 1957.  During four days
of this period, samples at each station within the mill were col-
lected hourly for eight hours and composited; in the middle of the
survey the mill samples were collected hourly for an uninterrup-
ted 24-hour cycle and were  composited into 3-hour samples.  This
program resulted in four 8-hour composite and eight 3-hour com-
posite samples at each of the eight sampling stations within the
mill. A full set of samples  was shipped to  the Robert A. Taft
Sanitary Engineering Center of the Public Health Service,  in Cin-
cinnati, Ohio,  for gross radioassay and chemical analysis; iden-
tical samples were  sent to the Occupational Health Field Station,
Public Health Service. Salt  Lake City, Utah, for radium analysis.

     A primary purpose of the study was to make a complete bal-
ance of all radium entering  and leaving the mill in both dissolved
and undissolved form,  and separate radium analyses were per-
formed on the  solid and liquid phases of the samples.  For this
purpose, the hourly samples collected during the 8-hour periods
on July 25 and 26 were composited into a single sample for radium
analysis for each mill  station.  The hourly  samples collected over
the 8-hour periods on July 27 and 30 were similarly composited.
The  hourly samples taken for the 24-hour period during July 28
and 29 were composited into a single sample for each station for
radium analysis.

     Samples of the  direct drainage to Cottonwood Creek were
collected on each of the latter three days of the survey.  A single
representative sample of dry sands was obtained from the sands
pile, and a single representative sample of wet solids was taken
Irom the slimes  pond.  Liquid samples were collected from the
                                                       GPO 813-173-3

-------
 10                                  RESIN-IN-PULP PROCESS

 sands and slimes ponds on each of three days. One representa-
 tive mud sample was collected at the slimes pond outfall to the
 Cheyenne River.  Duplicate sets  of these samples were shipped to
 the Public Health Service  laboratories in Ohio and Utah for the
 analyses noted above.

     A record of the tonnage of ore processed was obtained fre-
 quently during the survey from the weightometer preceding the rod
 mill.  Weir discharge records for flow were kept regularly at the
 weir box preceding the RIP section, and the quantity of eluate
 flow was obtained from plant equipment. Water flows  at the var-
 ious points of addition were obtained from plant personnel, as
 were estimates of chemicals used in the process.   These in-
 cluded ammonium nitrate, magnesium oxide, lime, iron,  and
 sulfuric acid.  Water from Cottonwood Creek was pumped  to the
 sands pond during the survey period in order to provide suf-
 ficient water in the sand-slime separation and for diluting  the
 sand  slurry.  Estimates of solids concentrations,  pulp density, etc.,
 at various points in the mill were obtained from the operating
 personnel.

                    Laboratory Procedures

 RADIOACTIVITY

    Radium was determined generally by coprecipitation with
 barium sulfate. Following pretreatment of the various types of
 samples to put the radium  in solution, the procedure consisted
 essentially of evaporation with sulfuric acid, removal of polonium,
 coprecipitation of  radium with barium  sulfate,  purification, and
 alpha counting  of the precipitate.  Rather extensive pretreatment
 of the undissolved solids of the mill samples was necessary to
 entirely dissolve them. Those samples that contained large quan-
 tities of undissolved solids were centrifuged for separate radium
 analysis of liquid and of solid phases.  The solids were then
 washed with water to wash out the liquid not removed by centri-
 fuging, and this wash water was added to the liquid portion to be
 analyzed for radium.

    Most of the samples had high concentrations of suspended and
 settleable solids.  The gross radioactivity analyses of  the suspend-
 ed and dissolved solids were performed independently  on separate
 representative portions. This procedure eliminated the need for
 a large number of  absolute quantitative transfers for the radio-
 activity determinations.  A known volume of sample was filtered
 through a  membrane filter  and washed.  The filter and filtered
 solids were removed and ashed at 600°C to constant weight.  The
filtrate was evaporated to dryness and ashed at 600°C  to constant
weight.  Solids concentrations were based upon the total sample
volume.

-------
MINES DEVELOPMENT COMPANY                          11

    The general procedure for gross radioactivity determination
is described elsewhere.    Dissolved and undissolved solids were
analyzed separately, and a self-absorption analysis was perform-
ed for a representative sample from each station. This procedure
also is described elsewhere. *

CHEMICAL

    Chemical analysis of the various samples was performed by
standard methods outlined in detail elsewhere.5  Nitrates were
determined by the phenoldisulfonic acid method, and sulfates,
calcium, and magnesium by the gravimetric methods outlined.
Iron was determined by the phenanthroline method,  manganese by
the periodate method, and chlorides by the mercuric nitrate
method.  The results for manganese may have some error due to
interferences caused by  iron and chlorides, and in future work it
is planned to determine manganese colorimetrically by the ammon-
ium persulfate method to correct for these problems.

                      Analytical Results

    During the survey period the average rate of ore processing
was 517 tons per day.  Table 2 shows the average slurry flow at
each mill sampling station, as well as the  specific gravity and
suspended solids content.  Slurry flows were based upon observa-
tions within the plant and on computations that accounted for
specific gravity of the  dry  solids, dry solids flows estimated from
ore processed, specific  gravity of the liquid phase,  and observed
specific gravity of the  slurry.  The slurry  flow at Station 3 is not
the entire flow going to the sands pond, as considerably more
dilution water was added just following the sample collectoin point
at this location.

    Table  3 presents radium concentrations,  dissolved and un-
dissolved,  in the various mill process streams as well as at the
several locations outside the mill.  These are the average re-
sults for the survey period, and are given  in stet of slurry, and
in stet dry suspended solids.  As can be seen, there is good
agreement between the dissolved radium concentrations at Sta-
tions 2, 4, 5,  and 6, and between Station 7 and the  slimes pond
liquid. Similarly,  the  concentrations of radium per gram of dry
solids agree well at Stations 1 and 2, at Stations 4, 5, and 6. at
Station 7 and slimes pond,  and at Station 3 and sands pond (dry
sands).
    Gross alpha and beta radioactivity, dissolved and undis-
solved,  for the mill stations is given in Table 4,  in/i/ic/1.  The
dissolved gross alpha and beta radioactivity shows a large gain as
the result of acid leaching, and, as expected,  slimes neutraliza-
tion results in a major reduction of the dissolved gross activities
in the slimes liquid at Station 6.  The results for Station 1 (Tables
3 and  4) indicate that the radium constitutes about 16 per cent of
the gross alpha activity of the ore.

-------
 12
                              RESIN-IN-PULP PROCESS

Table 2. PROCESS STREAM CHARACTERISTICS
Mill Sampling
Station
1
2
3
4
5
6
7
Slurry flow
gaL min
91
91
91
145
12
142
151
Specific Gravity
of slurry
1.52
1.53
1.45
1.05
1.13
1.05
1.05
Dry suspen
by weight.
60
59
64
7,2
~ 0.07
7.2
8.2
ded solids
per cent







    Table 5 indicates the results of the chemical analyses of sam-
ples of waste flows and pond contents. These results are given in
milligrams per liter (mg/1) of the liquid portion of the samples,
and represent only dissolved chemicals.  Nitrates are expressed
as nitrate nitrogen, rather than nitrate ion.  It should be noted that
the sands pond was also receiving water pumped from Cottonwood
Creek during the survey.

    The  general agreement between Station 7 and the slimes pond
and Station 3 and the sands pond is evident.  Nitrate nitrogen was
      Table 3.  RADIUM  CONCENTRATIONS
Station
1
2
3
4
5
6
7
8
Liquid from
Slimes Pond
Liquid fro-,
Sar.ds Por.d
Drainace to
Co;t-,-r.-.vood Creek
Dry Sar.ds
Solids from
Sliir.es Pond
Siirres Pond
Outlet Ditch
Radium in/i/ii: 1
Dissolved
93
2. 150
114
2.450
2.290
2.450
350
-
270

a

17
-

—

—
Undissolvcd
605.000
640. 000
163.000
273.000
2.050
233.000
250.000
-
--

—

—
-

—

—
Radium in dry
suspended solids, ftfiz g
650
710
170
3.640
~ 2.600
3.760
2.930
ISO
--

--

—
150

2.470

53

-------
 MINES DEVELOPMENT COMPANY

       Table 4. GROSS RADIOACTIVITY.
13
Station

1

2

3
4

5

5
7
Gross Alpha, ppc 1
Undissolved
R
3.66 x 10
R
2.86 x 10
R
1.08 x 10
0.880 x 106
R
0.010 x 10
R
0.810 x 10
0.725 x 106
Dissolved
•3
1.71 X 10
T
473 X 10
T
3.03 x 10
388 x 103
•i
2.930 x 10
T
94.5 x 10
1.03 x 103
Gross Beta, /j/ic, 1
Undissolved
R
4.79 x 10
R
3.96 X 10
R
0 992 x 10
0.962 x 106
f-
0.017 x 10
R
0.866 x 10
0.610 x 106
Dissolved
•3
1.02 x 10J
•3
613 x 10
T
2.05 x 10°
555 x 103
T
5.740 x 10
1
47.5 x 1015
1.23 x 103
also determined for the liquid portions of samples from other mill
stations, and was essentially zero at liquid portions of samples
from other mill stations,  and was essentially zero at Stations 1,
2, and 3. but was 15, 8, 200, and 230 mg/1, respectively, at
Stations 4,  5,  and 6.  Based on the flows in Table 2 (liquid por-
tion), the ore processing rate of 517  tons per day. and a chemical
estimate of the nitrate ion required for exchange with uranium,
the nitrate ion use during the survey was calculated to be about
15 pounds per ton of ore processed.  This is in good agreement
with the mill records.

   Table 5. QUALITY OF WASTE FLOWS AND PONDS CONTENTS.
                        Material dissolved in liquid sarrcle. me 1
Sampling
station
Slimes to Tails
(Station 7)
Liquid [rom
Slimes Pond
Sands to Tails
(Station 3)
Liquid from
Sands Pond
Drainage to
Cottor.'A'ood
Creek
Sulfate
2.330
2. 190
2. 180
1.970
1.090
Chloride
205
200
240
275
170
Calfiurn
730
820
570
440
360
Magnesium
75
SO
120
150
65
Iron
0.14
-0
Tract
~0
~0
Manganese
~0
1.3
7,0
-jo
~0
Nitrate
nitrogen
500
460
-0
-0
Trace
    As has been indicated,  one of the primary purposes of this
survey was to make a radium balance throughout the process,  if
possible.  Table 6, based upon the data in Tables 2 and 3.  pre-
sents the radium balance  for the  mill, and indicates the various
paths by which specific quantities leave the mill. The acid leach-
ing process dissolves a certain amount of radium as well as the
uranium.  Some of the  suspended radium remains tied up with the
sands (about  170 micromicrograms of radium per gram of dry

-------
 14
RESIN-IN-PULP PROCESS
 sands) and is discharged with them to tails at Station 3.  A portion
 of the dissolved radium (150 microgram per day) present at Sta-
 tion 4 goes with the loaded stripping solution at Station 5, although
 the data of Table 3 indicate that dissolved radium is not extracted
 or concentrated by the ion exchange section. As will be seen,  vir-
 tually all of this carried over radium becomes a part of the yellow-
 cake that is shipped out of the mill.  The neutralization of the
 slimes tails with lime results in precipitation of a  substantial
 fraction (about 85 per cent) of the radium dissolved in the liquid
 at Station 6,  as shown by both Tables 3 and 6.

          Table 6. RADIUM BALANCE
Station
1
2
3
4
5
6
7
Radium, mg/day
Undissoived
301
316
81
217
0.1
223
225
Dissolved
0.049
1.06
0.057
1.95
0.150
1.90
0.316
Total
301
317
81
219
0.25
225
225
     As a check on the data in Table 6, the radium content of the
ore  can be estimated within reasonable limits of error.  If radio-
active equilibrium of radium with uranium is assumed on the
basis of  517 tons per day of ore that assays 0.20 per cent U.,0,,,
it has been calculated that 270 milligrams per day of radium
enter the mill with the ore. This is in satisfactory agreement
with the  totals of Table 6.
     Further computations indicate  that virtually all of the radium
present at Station 5 (150/tg/day) becomes a part of the final uran-
ium  concentrate. If  96 per cent over-all recovery of uranium is
assumed and notice is taken that the yellowcake is approximately
75 per cent uranium as U.,0  •*,  calculations indicate that roughly
1.2 tons  of yellowcake are produced per day. If the  radium con-
centration of 150^/tg/g of dry solids in Table 3  is used about  160
micrograms per day ( g/day) of radium leave the mill with the
yellowcake.  This is quite close to the figure of  150 /tg/day for
Station 6  (Table 6).

     From Table 6 it is evident that about 300 milligrams per day
of radium enter the mill.  About 80 milligrams per day leave with
the sands at Station 3, and 220 milligrams continue through the
process to the slimes pond. Some 60 micrograms per day of dis-
solved radium are discharged to the sands pond and about 300
micrograms per day of dissolved radium go to the slimes pond.
The data in Tables 3 and 4 for Station 7 indicate that of the dis-

-------
MINES DEVELOPMENT COMPANY                           15

solved gross alpha activity of 1,080/i/iC/l.,  about 30 to 35 per
cent is due to dissolved radium.

    In similar fashion rough balances of the gross alpha and beta
radioactivity can and have been made.  They need not be reproduced
here, as radium is the specific radioelement of major interest, and
its balance has been shown.  Of interest,'however, is the fact that
the undissolved radium varies between 15 and 38 per cent of the
undissolved gross alpha activity at Stations 1 through 7, while
dissolved radium constitutes from 0.1 to 35 per cent of the  dis-
solved gross alpha activity.  Details of these figures can be ob-
tained by a simple conputation with the data of Tables 3  and 4.

    Although dissolved radium is not concentrated in the ion ex-
change section, dissolved gross radioactivity is.  Table 4 indicates
that the dissolved gross alpha and beta radioactivity was concen-
trated by a factor of 8 to 10 in the RIP  section, or between  Sta-
tions  4 and 5.  This is in good agreement with the estimated con-
centration of uranium by a factor of  10, as quoted by mill per-
sonnel and by  Dayton. ^

                  Summary and Conclusions

    About 17 per cent of the gross alpha activity of the ore pro-
cessed at the uranium refinery under study was due to the pre-
sence of radium. At the ore processing rate of 517 tons per day,
the liquid and  solid wastes from ore  processing contained about
0.6 milligrams of radium per ton of ore processed,  or a total of
approximately 300 milligrams per day.  The bulk of this radium
99.8 per cent, remained in undissolved from throughout the
process, and was effectively retained in the sands and slimes
tailings ponds.

    About 0.2 per cent of the radium that entered the mill with the
ore either left the refinery in the final uranium concentrate or was
delivered in dissolved form to the slimes or sands pond. Specifi-
cally, it has been estimated that some 310/*g/day of dissolved
radium went to the slimes pond, about 60 ^g/day went to the sands
pond,  and about 150/tg/day of radium left the mill in the dried
yellowcake.

    The  radium content of the dried sands was about 150 micro-
micrograms per gram  (/*/ig/g) of sands, while the radium  content
of dried slimes solids was 2, 500 or  more /i/ig/g.  Sands accumu-
lation was in the neighborhood of 440 tons per day, while slimes
accumulated at a rate of about 80 tons per day. Effective reten-
tion and confinement of these solids and of the tailings pond
liquids resulted in practically no radium leaving the plant site
during the survey, except for that contained in the uranium con-
centrate that was drummed and shipped.  The very small amount
of direct drainage to Cottonwood Creek was from the sands pond
and contained little radium or nitrate nitrogen.

-------
 16                                  RESIN-IN-PULP PROCESS

     The data in Table 3 indicate that some of the radium becomes
 dissolved  during the acid leach process (see data for Stations 1
 and 2).  Radium was not concentrated,  however, by the RIP sec-
 tion the concentration of dissolved radium in the loaded stripping
 solution (Station 5) was essentially no different from that in the
 IX feed (Station 4) or the slimes to tails (Station 6).  The opposite
 occurred in regard  to gross alpha and beta radioactivity:  the
 gross alpha activity in the IX feed was 3.88 x 10 puc/l, in the
 loaded stripping solution it was 29.3 x 105/t/tc/l, or greater by a
 factor of 7.5; and in the slimes to tails it was reduced  to 0.95 x
 105/i/jc/l.  or a factor of 4.1 as against the IX feed.  The gross
 beta activity behaved similarly; the loaded stripping solution had
 a concentration 10.3 times that of the IX feed, while the slimes to
 tails showed only 8.6 per cent of the dissolved beta activity of the
 IX feed. Hence, the data indicate that although uranium was con-
 centrated as usual by the RIP  section,  radium was not concentra-
 ted here.

     Slimes neutralization before discharge to the tailings pond
 reduced the dissolved radium concentration from 2,450 to 350
ftfig/l, or by about 85 per cent. Liquid from the slimes pond
 showed a comparable concentration of dissolved radium about 270
j»/jg/l. The  data for Station 7 indicate about one-third of the dis-
 solved gross alpha activity  in the slimes pond liquid was due to
 radium.

     The nitrate nitrogen concentration in slimes pond  liquid was
 460 mg/1.   While this is probably of no public health significance
 in this case, such concentrations could be of considerable im-
 portance at other locations where public water supplies are in-
 volved.  These quantities of nitrogen could also provide sufficient
 nutrient material to result in undesirable blooms of algae and
 other biota in streams and reservoirs.  Shortly after the field
 survey was completed the use  of sodium nitrate in place of am-
 monium nitrate was  instituted  at the refinery, but it is doubtful
 that this would appreciably alter results presented here.

 WASTE DISPOSAL  PRACTICE

    As  has been noted, there has been quite careful control of  .
wastes at the Mines  Development Company refinery at  Edgemont,
South Dakota. Sand and slime solids have been effectively con-
fined at the plant site, and liquid waste releases have been at a
minimum.

    Waste  disposal from this mill and from mills employing a
similar process should continue to be carefully  supervised,  and
should in any instance be based upon available knowledge of down-
stream water uses and of  the fate  of the wastes  in the water en-
vironment.  The sand and  slimes solids should generally be effec-
tively retained and confined,  as they contain considerable radium
and other radioelements.  Their release to a stream would result

-------
MINES DEVELOPMENT COMPANY                          17

in long term contamination of the watercourse, and they could
significantly contaminate equipment at downstream domestic or
industrial water treatment plants.

    Liquid wastes from the slimes ponds at this and similar mills
should be released to surface waters only in accordance with exist-
ing regulations and on the basis of detailed  information as to local
downstream water uses. This liquid, free of suspended solids,
contains considerably more than allowable concentrations of radium
and of nitrate  nitrogen.  At the uranium refinery studied, such con-
trol is especially important in view of the zero flows and extended
low flow periods that occur in the Cheyenne River. As has been
noted, these flows are such that little or no dilution of the effluent
would occur for extended periods of time.  Any susceptible ground
water supplies located near the  slimes tailings ponds of such re-
fineries  should be tested periodically for  nitrate nitrogen and
radioactivity content, as infiltration of the slimes pond liquid into
the ground water may occur at the pond.

    As regards measurement of radioactivity in the effluent from
the slimes tailings pond at Edgemont, South Dakota, it appears
reasonable to  suppose from these studies that the  dissolved radium
content will generally be in the neighborhood of 30 per cent of the
dissolved gross alpha activity. It seems feasible, for routine
measurement  purposes, therefore, to analyze these samples for
dissolved gross alpha activity and apply the factor of 0.30 to com-
pute dissolved radium.  An occasional analysis for radium itself
will then serve as a check,  and refine the percentage figure  to be
used.  In this way, adequate control can be  provided and the
costs  and labor of sample analysis  for routine control can be mini-
mized.  This percentage figure might also prove adequate within
reasonable limits of error for application at other mills using
the same process, but this cannot be known until further studies
are carried out.

    Increases or decreases in the rate of ore processing at  the
Edgemont, South Dakota, refinery, or changes in  the U.,0  con-
tent of the ore, should be proportionately reflected in  many of the
waste quantities. This applies particularly  in the  case of radium.
Thus, if  no important process change is assumed, a capacity
increase to, say, 750 tons of ore per day, at 0.20 per cent U«0
should result in about 440 milligrams per day (mg/day) of
radium entering the  slimes tailings pond.  Any basic change  in
the type of ore received would, of course, affect these figures,
especially the  latter.

    It is not intended here to generalize very much beyond the
local  situation studied.   Each uranium refinery waste disposal
problem  is individual, and must be interpreted in  terms of speci-
fic local  water uses  such as domestic water supply, irrigation of
croplands, recreation,  etc., as well  as in terms of the specific

-------
18                                  RESIN-IN-PULP PROCESS

waste characteristics and mill process.  Available dependable
dilution in the receiving waters is also a critical factor in control
of these wastes. No two problems are likely to be precisely the
same,  and in the dual interest of  radiological safety and economy,
such waste disposal problems should each be carefully analyzed
as individual cases insofar as necessary.  It is hoped that the in-
formation presented herein will provide some  insight into the
waste disposal problems at acid leach resin-in-pulp mills
generally,  but the  data should not be applied to other mills with-
out considerable caution.
                      Acknowledgment
    The generous cooperation and assistance of the following
organizations and individuals is hereby gratefully acknowledged:
Harold Webb,  Plant Superintendent, and personnel of the Miles
Development Company; the U. S. Atomic Energy Commission;
D. E. Rushing  and W. J. Garcia,  Public Health Service,  Salt Lake
City,  Utah; M. Nelson and R. Spicer, South Dakota State Depart-
ment of Health; J. K. Neel, Public Health Service, Kansas City,
missouri; A. F. Bartsch, N. D. Wastler,  S. D. Shearer, J. T.Jones
R. C. Kroner,  H.  L. Faig, and R. J. Lishka, Public Health Service
Cincinnati, Ohio

-------
THE ACID LEACH-SOLVENT EXTRACTION URANIUM
                   REFINING PROCESS.
           I.  GUNNISON MINING COMPANY,
                GUNNISON, COLORADO.
                       S. D. Shearer *
                       C. E. Sponagle
                          J. D. Jones
                       E. C. Tisvoglou

                         Introduction

     This is a report of an in-plant survey of the Gunnison Mining
Company, a uranium refinery utilizing the acid leach solvent ex-
traction process; Part II describes a study performed at the
Climax Uranium Company mill at Grand Junction, Colorado, which
has a similar process.  This study was performed by the  Public
Health Service during August 1958 with the cooperation of the Colo-
rado Department of Public Health, the Gunnison Mining Company,
and the U. S. Atomic Energy Commission.

     The Gunnison Mining Company refinery is located at  Gunnison,
Colorado, on the Gunnison River, about 130 miles upstream from
Grand Junction,  Colorado,  and the confluence of the Gunnison and
Colorado Rivers. At the time of the survey there were no direct
discharges to the Gunnison River. Figure 1 outlines the flow dia-
gram for this refinery and the process is described in detail in the
following sections.

    The Gunnison uranium  refinery began operation about 8 months
prior to this survey. At the time of survey, it was processing an
average of 330 dry tons per day  of ore that assayed from  0.21 to
0.49 per cent U«0fi,  and was producing uranium concentrate at a
rate of about 1, oOl) pounds per day. All wastes from the mill
were retained in a tailings pond,  with no direct discharge to the
Gunnison River.

                      The Mill Process

ORE RECEIVING,  SAMPLING.  AND CRUSHING

     Loaded ore-trucks are emptied into a 50-ton ore hopper,
from which ore is fed on a 30-inch conveyor belt to the sampling
plant.  A magnet removes tramp iron  from the ore as it dis-
charges from this conveyor onto a vibrating grizzly. Ore less

*Respectively, Senior Assistant  Sanitary Engineer. Radiological
 Pollution'Activities Unit; Sanitary Engineering Director,  Colorado
 River Basin Water Quality Control Project. Public  Health Service,
 Denver,  Colorado; Health Physicist (present address:  Radiation
 Control Service, University of Michigan. Ann Arbor:  and Chief,
 Radiological Pollution Activities unit.

                            19

-------
Fllf WCJROML
<7>
SOLHBl
EX me no*
r*«
A-
I«*|C

SOLVfNT
(IIWCTIMI
T*»f



"?/F
                         i J niTt" 1 J^fcitn  j riLTiti i    fFWMCA
              F_«_siuoer.	n  ""(r"i"" rn ""' rfn     r.j

          o-
                                  Figure 1 .  Flow  diagram  for fhe Gunnison  Mining Compony uranium

                                           ,,,HI  ' 	"• •      ^ «    I-  * ,,o,,rt 1 O '-.R

-------
GUNNISON MINING COMPANY                              21

than 3 inches in size drops through the grizzly onto a 24-inch con-
veyor, while larger pieces are crushed in a jaw crusher, and then
rejoin the ore flow.  The 24-inch conveyor feeds a vibrating screen
through which ore less than 3/4-inch in size drops onto another 24-
inch conveyor belt; larger pieces are crushed in a gyratory crush-
er and then fall onto  the belt.  As the ore leaves the end of this
conveyor, a 10 per cent sample is  taken by a chain and bucket
sampler, while the remainder of the ore is conveyed to one of two
250-ton fine-ore bins.

    The sample is screened, the larger pieces are crushed in a
jaw crusher, and a 10 per cent sample is taken by a second chain
and bucket sampler;  the balance of the  original  sample is convey-
ed to the fine-ore bins.
    The remaining sample is again screened and crushed,  and
enters a vezin sampler from which a 5, 10,  15. or 20 per cent
sample  may be taken as required.  The balance of the original
sample  proceeds  to the fine-ore bins.   The final sample collect-
ed amounts to about 0.1 per cent of the original ore fed to the
process, or two pounds per ton. This representative sample  is
assayed for  its U000 content.
                3 o
    An  additional 50-ton fine-ore bin is available for temporary
storage of special ores or excess fine ore.  This  is not ordinarily
used, however, and only a small stockpile is usually maintained.
Blending is not practiced; the ore is processed  immediately upon
receipt  at the plant.

GRINDING

    The fine-ore bins feed a conveyor  belt that transports the ore
to the rod mill.  Feed tonnage  is determined by a weightometer
connected to the belt. The ore  goes to  a 6- x 12-foot rod mill in
series with a 48-inch spiral classifier.  As the  ore enters the rod
mill, it is slurried by the addition of water and a small amount of
sodium  carbonate solution (for corrosion prevention.)

    Slurry from the  classifier  discharges into a sump where  sod-
ium chlorate is added to oxidize ferrous iron to the ferric state and
to maintain an EMF of over 400 in the leach tanks. The slurry is
pumped to the operating floor,  passes through a small cyclone
that returns plus-35-mesh particles to the rod mill feed, and
enters the leach tanks.
LEACHING
    There are four acid leach tanks, each 16 feet in diameter and
16 feet deep, which are arranged for series flow.  Steam is added
to the first tank to maintain a temperature of 85°F.,  and concen-
trated sulphuric acid is added  to the first two tanks to maintain a
pH of 0.8 in the leach liquor leaving the last tank.  Average leach-
ing time is about  17 hours. Constant agitation is  provided by a
propellor-type mechanism in each tank.

-------
 22           ACID LEACH - SOLVENT EXTRACTION PROCESS

 SAND-SLIME SEPARATION

     Slurry leaving the leach tanks is diluted with the overflow
 from the number 3 thickener (see Figure 1), and the  combined
 flow enters a 30-inch classifier.  This is the first of  four class-
 ifiers provided for sand-slime separation, the remaining three
 being 24-inch size.  The sands proceed through the four classi-
 fiers and are discharged to a slurry tank.  In this advance of the
 sands,  they are washed with thickener overflow; number 2 classi-
 fier receives the overflow from number 2 thickener,  and  number
 3 classifier, the overflow from number 1 thickener.  Freshwater
 is used in the number 4 classifier.

     The overflow from each classifier carries the  slimes into
 the thickeners.  Figure 1 shows the manner in which this  is done,
 each classifier discharging its overflow into the thickener that is
 adjacent to it.

     The slimes proceed through each of the four thickeners, are
 washed counter-currently during their travel, and the spent
 slimes are discharged to the slurry tank from the number 1
 thickener.  The washed sands and slimes are combined in the
 slurry tank, repulped with raffinate from the solvent  extraction
 process,  and discharged to tails.  Equipment is available  for
 feeding lime to the slurry tank,  but this is not done.

 ACID LIQUOR STORAGE

     Pregnant acid liquor from  the number four thickener  proceed;
 to a 22-foot diameter by 10-foot deep storage tank.  An EMF ad-
 justing tank and two  filter presses are provided following  this stor
 age tank. These units are incorporated into the plant for the pur-
 poses of (a) reducing ferric iron to ferrous iron,  in order to elim-
 inate interference in the solvent extraction process, and (b) re-
 moving the small amount of slimes remaining in the pregnant acid
 liquor, so as to eliminate difficulties in the solvent extraction
 process.  Usual practice at the time  of the survey,  however, was
 to bypass most of the acid liquor around these units into a second
 storage tank.  This practice at times caused emulsification of the
 solvent due to the presence  of slimes in the acid liquor entering
 the solvent extraction process.

 SOLVENT EXTRACTION
    Pregnant acid liquor from the storage tank is pumped to a cor
 stant head tank above the operating floor,  from which it flows into
the first  of two solvent extraction tanks.  Uranium is  extracted
from the acid liquor  by the organic solvent by alternate cycles of
agitation and quiescence.  There are five such mixing-settling
cycles  in these tanks. Flow of the organic solvent is  counter-
current to that of the acid liquor.  The  raffinate, or barren acid
liquor, is discharged into a holding tank from which a portion is
pumped directly to the tailings pond, while the remainder  is used

-------
GUNNISON MINING COMPANY                               23

to repulp the sand-slime slurry, as previously described. There
is recovery of some solvent that has been carried over, which
rises to the surface of the liquor in the raffinate tank.

SOLVENT  STRIPPING  AND YELLOWCAKE  PRODUCTION

    The pregnant organic  liquor from the solvent extraction pro-
cess goes to a bank of three mixing and settling tanks,  where the
uranium is stripped from it with a counter-current flow of a 10
per cent sodium carbonate solution. The barren organic is re-
turned to a holding  tank for recycling through the solvent extrac-
tion system.

    The prenant sodium carbonate solution is passed through a
filter press for  removal of iron as ferrous  carbonate.  Sludge
from  the press is returned as a slurry to the rod mill, while
the filtrate proceeds to one of two precipitation tanks,  each of
which is 12 feet in  diameter.  Concentrated sulphuric acid is add-
ed to  neutralize the sodium carbonate,  and  the solution is heated
with steam to about 165° F. to drive off any excess carbon dioxide
that may be present.  Magnesium oxide is then added to bring the
pH to about 7. During this process the mixture is constantly stir-
red by a mechanism within the tank.  Uranium is precipitated as a
sodium salt. When precipitation is complete, the tank contents are
filtered and the  yellowcake recovered by use  of a filter press.
Filtrate from the press is returned to the slimes thickeners, and
the yellowcake goes to  a drying furnace.  The dried produce is
drummed,  weighed, and shipped to an Atomic Energy Commission
facility.  Yellowcake is produced only during the day shift.

                       The Mill Survey

    For purposes of analyzing the mill process and  characteri-
zing the resulting liquid waste,  ten sampling stations were selec-
ted.  Samples were collected on an hourly basis from August 7 to
August 11,  1958 and were composited at each  sampling point
over the periods shown:
              Cycle 1 (24 hours): 4 PM.  August 7 to 4 PM, August 8.
              Cycle 2 (35 hours): 4 PM.  August 8 to 4 AM. August 10.
              Cycle 3 (36 hours): 4 AM.  August 10 to 4 PM. August 11.
              The sampling stations selected are described in Table 1.

    All samples were collected inside the mill  building,  with the
exception of those at Stations 6 and 8, which were collected at
the points of discharge  to the tailings pond.

    During the survey  plant flows were obtained at various stations
and locations from  the  operating records of the mill and directly
by the survey party.  Mill records provided frequent data as to the
flow of barren organic  pregnant acid liquor and Na?CO., stripping
solution. Installed  flowmeters gave data as to the acid ilow to the

-------
 24
ACID LEACH - SOLVENT EXTRACTION PROCESS
 first leaching tank and the flow of wash water to the number 4
 classifier.  Flows at Stations 6 (raffinate to tails) and 8 (sand-
 slime slurry to tails) were measured hourly in terms of the a-
 mount of time required to fill a 55-gallon drum.  Hourly readings
 from the weightometer,  together with plant records of the per
 cent moisture in  the belt feed, yielded accurate data regarding
 tonnage of ore processed.  The daily yellow-cake production was
 obtained from plant  records.

           Table 1. GUNMSON MILL SAMPLING STATIONS
          Station Number
              1

              2

              3

              4

              5

              6

              7

              8

              9

              10
                                 Description
            Classifier effluent

            Acid leach tank effluent

            Sands entering slurry tank

            Slimes entering slurry tank

            Pregnant acid liquor to solvent extraction

            Raffinate to tails

            Pregnant organic to stripping circuit

            Sand-slime slurry to tails

            Yellowcake

            Acid leach tank feed
     Plant equipment and records included hourly observations of
 pulp density and per cent solids by weight at Stations 1,  2,  3, 4,
 8, and 10.  Notes were made of any process interruptions during
 the survey.
    Daily composite samples were collected and assayed by plant
 personnel for the U«0  content of the mill heads (from the rod
 mill),  the tails (from the slurry  tank),  the leach tank effluent, the
 pregnant  acid liquor,  and the raffinate,  and these data were made
 available  for this survey.  Records of chemical consumption (acid.
 sodium carbonate, etc.) for the month preceding the survey were
 also made available, and chemical use  figures for the survey
 period were obtained.

    Sampling was performed by personnel of the Colorado De-
 partment  of  Public Health and the Public Health  Service.  The
 yellowcake samples were collected by mill personnel, a small
 vial of about 30 grams being composited from  the day's produc-
 tion.  All  other samples were  collected hourly during each cycle
 indicated.

    All samples collected during the survey were  shipped to the
 U. S. Public  Health Service, Robert A.  Taft Sanitary Engineering
Center, at Cincinnati, Ohio. Portions of selected  samples were
then sent  to a private  laboratory for analysis of  dissolved and un-

-------
GUNNISON MINING COMPANY                               25

dissolved radium.  All other analyses were performed at the Cin-
cinnati laboratory.

    Yellowcake production amounted to 1, 258 pounds, 1, 250
pounds, and 3. 245 pounds during sampling cycles 1, 2, and 3 re-
spectively.  As noted earlier,  yellowcake production was carried
on only during the  day  shift and cycles 1 and 2 each included a
single  day shift, whereas cycle 3 included two day shifts.

                      Analytical Results

    Figure 2 is a schematic flow diagram that indicates the liquid,
solids, and slurry flows at the various process location. At each
station the  slurry flow is given first, with solids and liquid flows
next in order.  These figures were obtained on the basis of obser-
ved specific gravities, tonnages of ore,  and other laboratory and
field data; computations treat the flow as being composed on two
separate streams, liquid and solids. Flows are given to the nearest
gallons per minute (gpm).  These figures and all succeeding  re-
sults represent combined computations for cycles 2 and 3 of the
survey period,  as  the  samples for cycle 1 were not analyzed.
    Late in cycle 2 and extending well into cycle 3 of the survey a
batch of custom ore  was processed.  This ore assayed 0.49 per
cent or more U^O..,  compared to the more usual ore assaying
about 0.22 per cent U  0R. Because of this, and as a result of the
time lags at several points in the process (for instance, the  acid
leach process took about 17 hours),  it was necessary to combine
the data for cycles 2 and 3 in order  to make balancing computa-
tions.

    Table 2 presents certain process stream characteristics for
the stations sampled.  Slurry flows in gallons per  minute,  speci-
fic gravity  of slurry, specific gravity of dry solids, and the  per
cent dry  suspended solids by weight are shown for each sampling
station.  Table  3 indicates the solids balance for the process dur-
ing cycles 2 and 3. Approximately 330 tons per day of ore.were
processed during the survey,  as indicated at Station 10. Acid
leaching  dissolved about 6 per cent of this total tonnage: i.e., 19.8
tons per  day left the leach tanks in the dissolved state. The preg-
nant acid liquor (Station 5) contained about the same amount of
dissolved solids.

    Table 3 shows that the total plant output (Stations 8,9, and 6)
was about 332 tons per day (318.5 +  0.7 + 12.6),  which is in excel-
lent agreement with the 330 tons per day of ore processed. The
total solids into the slurry tank (Stations 3 + 4 + the raffinate
solids going to  the slurry tank) total 318 tons  per day, which
agrees with the output  to tails (Station 8) of 318.5 tons per day.
The dissoK-ed solids into the slurry tank (Stations 3 + 4+  the raf-
finate solids to the slurry tank) total 10.0 tons per day, which is
in agreement with  the  dissolved solids output  at Station 8.  The

-------
                            WASH
                            WATER
                            135 gpm
FROM Cli/
PREGNANT
ORGANIC -
99 gpm
21
78

ACID
LEACH

SO
EXTF
i
BARRtM
"W29.6 gpm


S
CAR
EXT
LV
AC
i
:NT
TION
(

10IUM
BONATE
(ACTION
9 ,
97 gpm
21
76
PREGNANT (T)
ACID LIQUOR V'
176 gpm '
0
175
RAFFINATE (BARREN AC
PREGNANT
SODIUM
CARBONATE
CLASSIFIERS

SLIMES
f
THICKENERS
D LIQUOR)


Tfl THrrrrKFfli _rf


URANIUM
PRECIPI-
TATION
WASHED SANDS
WASHED SLIMES
63 gpm *
13 f
50


©-•
150 gpm
8
SLURRY .
TANK -4—*" TO TAILS
rr
V

22
I'm
RAFFINATE
53 gpm
0
£~
122 gum
0
122
,YELLOWCAKE
fi)


1500 LB./DAY
KEV:

                   EXAMPLE'

    SLURRY,  gpm       97
    SOLIDS,  gpm       21
    LIQUID,  gpm       76

    'STATION-
Figure 2.  Schematic flow diagram, Gunnison Mining Co., uranium mill,
          August 1 958.

-------
GUNNISON MINING COMPANY
27
total dissolved solids output (Stations 6 plus 8) of 22.6 tons per
day  isin good agreement with the dissolved solids after acid leach
and  in the pregnant acid liquor.
           Table 2.  PROCESS STREAM CHARACTERISTICS
Station
1
2
3
4
5
6
7
8
9
10
Slurry Flow,
gals, min
-
97
50
63
175
122
29.5
165
3°
99
Specific
Gravity of
slurry
1.52
1.36
1.33
1.30
1.01
1.01
(2)
1.19
3C
1.35
Dry Susp.
Solids by-
Weight, ^c
55.4
39.6
38.7
38.1
~ 0.005
~ 0.004
,b
26.0
3°
41.2
Specific
Gravity of
dry solids
2.61
2.52
2.64
2.49
-
-
2b
2.37
n.d.
2.61
            Average of cycles 2 and 3.
            Liquid (negligible suspended solids).
           c Solid sample.
    Table 4 indicates the concentrations of radium 226 in dis-
solved and undissolved form at the several stations.  Undissolved
radium is that portion retained on a millipore filter, while dis-
solved radium represents that contained in the filtrate.  The data
are representative of cycles 2 and 3, and are given as micromi-
           Table 3. SOLIDS BALANCE a
           Station
                                      Tons per Day

10
2
3
4
5
6
8
9
Raffinate to
slurry tank
Suspended
323. 0
309.5
120.8
187.5
-0
-o
308.5
0.7
-0
Dissolved
1.6
19.8
3.3
1.2
19.9
12.6
10.0
0.0
5.5
Total
329.6
329.3
124.1
183.7
19.9
12.6
318.5
0.7
5.5
           ' During cycles 2 and 3.

-------
 28
ACID LEACH - SOLVENT EXTRACTION PROCESS
        Table 4. RADIUM CONCENTRATIONS

Station
10
2
3
4
5
6
7
8
9
Radium 228 in total sample, /i>ig/l

Undissolved
271,000
345.000
94. 500
338,000
50
130
70
156.000
a
Dissolved
125
270
180
110
490
480
3
155
a.

Radium 226 in dry
undissolved solids.
PPS e
490
640
235
680
905
3,500
130
505
35
        a  Solid sample
crograms of radium-226 per liter of slurry, as well as per gram
of dry suspended solids.

     Portions of the samples were also assayed for gross alpha
and beta radioactivity;  these results are given in Table 5.

     The dissolved alpha activities show  clearly the  effects of the
various steps in the process.  A sharp increase in activity result-
ed from the acid leach  (to 250, OOO^u/tc/1 at Station 2), and the
bulk of this activity was contained in the pregnant acid liquor from
the thickeners  (Station  5). Most of it was removed by solvent ex-
traction (see Station 6, barren acid liquor), and appeared finally
in the yellowcake as uranium (Station 9,  306, 000/i/ic/g of dry sus-
   Table 5.  GROSS RADIOACTIVITY CONCENTRATIONS.
Station
10
2
3
4
5
6
7
8
9
Activity in total sample ftpc,- 1
Undissolved
Alpha
2.560,000
1.980.000
304.000
1. 810.000
1.400
730
n.d.
973.000
a
Beta
3.000.000
2.740.000
494.000
1.840.000
1.550
660
n.d
1.052.000
a
Dissolved
Alpha
5.900
250.000
17.400
4.400
272.000
5,900
n.d.
14. 030
a
Beta
15,300
830.000
50.600
33.200
475.000
11,500
n.d.
33.300
a
ActivHv in
undissolved solids.
we g
Alpha
4.633
3.6SO
760
3. 640
25.800
20.600
n.d.
3. 140
306.000
Beta
5.440
5.090
1.230
3,700
28.000
18.000
n.d.
3. 400
445. 000
    Solid sample

-------
GUNNISON MINING COMPANY

        Table 6. CHEMICAL QUALITY OF MILL EFFLUENTS
29
Determination
Total acidity as CaCO~
Mineral acidity as CaCO,
Hardness as CaCOj
Sulfate
Chloride
Iron
Manganese
Copper
Seler.ium
Sodium
Fluoride
Beryllium
Vanadium
Arsenic
Concentration, mg. 1
Station 6
Kaffinate
10.000
9.000
1.850
12.600
ISO
48
17
0.2
0
1.400
12
2.4
0.06
11
Station 8
Sands-slimes slurry
5.400
3.500
1.550
10.000
275
82
7
0.2
0
830
13
2.1
0.03
17
pended solids). Although the suspended solids in mg/'l of the preg-
nant and barren acid liquors were low. with correspondingly
slight alpha activity on a per liter basis, the suspended solids
that were present contained relatively high alpha activity on a per
gram basis.

    Table 4 clearly shows that little radium-226 was dissolved by
the acid leach (see Stations 10 and 2,  dissolved radium), and rad-
ium stayed mainly in the undissolved  state through the entire mill
process.  The uranium concentrate (Station 9) contained only 35
micromicrograms of radium-226 per gram, dry weight. While
the pregnant and barren acid liquors (Stations 5 and 6) contained
relatively high-radium suspended solids,  in small quantities,  the
dissolved radium content of these liquors was also relatively
high.  This indicates that very little radium was extracted from
the pregnant acid liquor, and this is also borne out by the low
radium content of the yellowcake.

    The radium-226 and gross alpha  balances that follow clarify
these conclusions.
    Samples of the barren acid liquor, or raffinate. (Station 6)
and of the sands-slimes slurry going  to the tailings pond (Station
8) were analyzed for various chemical constituents of interest.
The results are shown in Table 6. Only the liquid portions of the
samples were analyzed, and the results, in mg/1, represent
only  dissolved chemicals.  The pH of  the raffinate was 1.3.. and
that at Station 8 was 2.1.

-------
30
 ACID LEACH - SOLVENT EXTRACTION PROCESS

Table 7. CHEMICAL CONSUMPTION
Chemical
H2S04
Mt-O
Kerosene
DEHPA
TBP
Separan
Na2CO3
NaC103
Decinol
Amount used per ton of ore
processed, Ibs.
70.2
0.85
0.225a
0.0054
0,0054
0.354
15.6
2.89
0.016
              Gallons.
    Chemicals used in the mill process are indicated in Table 7.
together with average consumptions per ton of ore processed.
These data are averages for the month preceding the survey, and
were supplied by company officials.

    During the mill survey, about 36,000 Ibs H?SO,,  about 900
Ibs of NaC103, and an average of about 5, 000 Ibs of Na^CO- were
used per day.  MgO and Separan were used at rates of about 310
and 97 Ibs per day, respectively.

    One of the primary purposes of this  survey was to determine
the amounts of radium-226 in the process at various locations,
and the amounts in suspended and dissolved form in the effluents.
To that end, a radium balance for the process has been carried
out (Table 8).

         Table 8. RADIUM BALANCE
         Station
                                 Radium - 226. me/day

10
2
3
4
5
6
7
8
9
naltinate to
s!-irrv tank
Undissolved
146
180
26
116
0.0-13
0.037
0.012
142
0.025
0.033
Dissolved
0.063
0.142
0.049
0.038
0.467
0.319
~0
0.141
-
0.139
Total
146
180
26
116
0.52
0.41
0.0!
142
0.025
0.18

-------
GUNNISON MINING COMPANY                               31

    The values in Table 8 were computed directly from the
radium concentrations of Table  4 and the slurry flows of Table 2;
a value of 1, 500 Ibs per day of concentrate for Station 9 was used.
    An initial test of  the results was made on the basis  of the
uranium assays,  yellowcake production, and the  assumption of
radioactive  equilibrium between uranium and radium in  the raw
ore. The yellowcake  production rate during the survey was 1, 500
Ibs/day, containing about 1.0 per cent moisture and 82 per cent
tLOp,.  If radium-226  were in equilibrium with this  much uranium
there would be 156 mg/day of radium-226 involved  in the process.
The uranium assays of  raw  ore  ("heads") and tailings indicated an
over-all efficiency of about  90 per cent for uranium recovery;
hence it is estimated  that about 173 mg/day of radium-226 en-
tered the mill with the 330 tons per day of ore.  This is  in good
agreement with the data of Table 8.

    A  separate computation, based on the assumption of 0.22 per
cent U«0_ in the raw  ore (because of the time lag between raw ore
and concentrate in the mill), the tonnage processed,  and radio-
active  equilibrium between uranium and radium,  indicates about
180 milligrams of radium-226 enter the mill daily.
    The radium balance is in generally good agreement through-
out the process with the exception of some discrepancy between
the undissolved radium  results for Stations 2 and 10, before and
after the acid  leach.  Due to the presence part of the time of the
smaller batch of higher grade ore. and to the large time delay
during  the acid leach  process  step, the result for Station 2 may
not be fully  representative for the survey period  and may be some-
what high.  Other than this,  the  undissolved radium data indicate
146 mg/day  entering from the cyclone,  142 mg/day at Stations 3
and 4 combined (separate washed sands and slimes slurries) and
142 mg/day  at Station 8 (the combined slurries).  Total mill input
and output therefore agree adequately so far as undissolved
radium is concerned.
    The dissolved radium data also indicate good balances.  The
46 7/jg/day in  the pregnant acid liquor,  together with the outputs
of 49 and 38/ig/day in the washed  sands and slimes, respectively.
yield 554^g/day. This is accounted for adequately by the 319
/tg, day in the raffinate to tails (Station 6). the 25 fig/day in the
yellowcake.  and the 141 /tg/day at Station 8  a total of 485 /tg/day.
The dissolved radium entering and leaving the slurry tank totals
226 /ig, day (Stations 3.  4. and the portion of raffinate to the slurry
tank) as against 141 /ig/day  at Station 8. It appears possible here
that some of the radium initially dissolved in the raffinate preci-
pitated on mixing with the washed sands and slimes slurry in the
slurry  tank.
    It  is also  of interest to  note that the dissolved  radium leaving
the acid leach tanks (142/ig, day) does not account for the 554/ig/
day at  Stations 3. 4.  and 5.  This, together with  the observed con-

-------
 32
ACID LEACH - SOLVENT EXTRACTION PROCESS
         Table 9. GROSS ALPHA BALANCE
Station
10
2
3
4
5
6
7
8
9
Raffinate to
slurry tank
Gross alpha Radioactivity, me, day
Undissolved
1.380
1.040
85
620
1.4
0.5
n.d.
680
210
1
Dissolved
3.2
130
•4.7
2.9
260
3.9
n.d.
13
a
1.7
Total
1.380
1.170
90
620
260
. S. Geological Survey almost adjacent to the mi
property.

    Discharge records for this station are available from Octobe
1911 through September 1958: continuous daily discharge measur
ments are available  for the period October 1,  1945,  through Sep-
tember 30, 1958.  Only these continuous records were analyzed.
In order  to obtain reasonable estimates of  flow frequencies ex-
pected,  arithmetic-probability and Gumbel type 8 curve fitting
techniques were employed. In general, the two types of analyses

-------
GUNNISON MINING COMPANY
                                 33
were in agreement.  Figure 3 presents the arithmetic-probability
analysis of the minimum daily and minimum monthly average
flows for the continuous years of record.  This curve shows that
half the time the minimum daily flow has been equal to or greater
than 135 cfs,  while the minimum monthly average flow has been
equal to or greater than 150  cfs; the minimum daily flows ranged
from 96 to 200 cfs, with an average of 136 cfs, and the minimum
monthly flows ranged from 111 to 252 cfs,  with an average of 160
cfs. During the survey period. August 7 through August 11, 1958,
the flow ranged from 790 to 825 cfs,  with an average of 803 cfs.
A hydrograph of daily flows for this station for the period Octo-
ber 1.  1957,  through September 30,  1958, is shown in Fugure 4.
The discharge records  from  which this hydrograph was platted
show that the  average flow for the water year was 838 cfs,  while
the average flow for the six-month period October 1957 through
March 1958 was 355 cfs.  This is a typical hydrograph of rivers
in the western United States,  in which the spring snow-melts pro-
duce high runoffs during the three spring months and the  flows for
the remainder of the year remain relatively steady.
         300
         250 -\—
         200
      en
      u.
      o
MINIMUM MONTHLY AVERAGE
      (1945-58)
         50
                    20       40       60       80       100
                    % OF TIME  *THAN STATED VALUE
         Figure 3. Occurrence of minimum daily and monthly average flows,
               Gunnison River near Gunnison, Colorado.
    It must be pointed out that  the probability methods utilized
are statistical ones and are subject  to variations such as length

-------
                                                           11 TTITTTTTTTTTTTT'TTTTTI
  --uJ          JVN" AI|


LLLLLLLLL J.I LI 1	|..1_1 U.LJJJ LI..I ill
.'* ''*,''' \  * '° '^ ?''
                 Figure 4.  Daily flows for Gunnison River near  Gunnison, Colorado.

                         Water year 1958.

-------
GUNNISON MINING COMPANY                               35

of record,  stream regulation, amount of upstream irrigation, etc.
For this particular gage the years of continuous record include
only a 14-year period.  Since September 1937,  the flow at this
station has been partly  regulated by Taylor Park Reservoir about
37 miles upstream.  There are also about 22.000 acres of irriga-
ted land above the station.

    So far as can be ascertained, no public water supplies are
taken from the Gunnison River below Gunnison, Colorado.  The
extent of other uses of  the stream is not known in detail, but ex-
tensive downstream diversion for irrigation purposes occurs, with
the possibility that individual ranch families use the river water
in its raw state for domestic purposes.

WASTE DISPOSAL

    During the mill survey a large unused tailings pond area was
available, and all wastes from the mill process were retained in-
definitely in this  pond area.

    About 12 per cent of the gross alpha activity of the ore  that
was being processed at  the uranium refinery under study was due
to the presence of radium. The liquid and solid wastes from ore
processed at the  rate of 330 tons per day contained about 0.6 milli-
grams of radium  per ton of ore processed,  or  a total of approxi-
mateily 150 mg/day.  Some 99.6 per cent of this radium remained
in undissolved form throughout the process and was effectively
retained in the tailings  ponds.
    About 0.4 per cent  of the radium that entered the mill with the
ore either left the refinery in the final uranium concentrate or was
delivered in dissolved form to the pond. Specifically,  an estimated
480 micrograms of dissolved radium per day went to the tailings
pond, and about 25  micrograms of radium  per  day left the mill in
the dried yellowcake.

    These studies indicate that two waste constituents in particular
are present  in potentially hazardous quantities: radium-226 and
arsenic (see Tables 6 and 8).
    As indicated  earlier, the effluents from the Gunnison Mining
Company uranium mill  contained per day about 500 micrograms
of dissolved radium-226. The  records of flow  of the Gunnison
River near the  refinery indicate that if all of this dissolved rad-
ium were released  routinely to the river, the dissolved radium
content of the river water would show an increase from essenti-
ally zero at high or flood flows to about 1.2^/ig/l at average or
usual flows, and to about 2.0/i/ig/l at low flows, which are  rela-
tively rare.

    Thus, by itself, the quantity of dissolved radium regularly
produced as waste from this refinery, while detectable in the
river,  would not constitute a major hazard in terms of existing
standards/ In practice, of course.it is not usually released to the

-------
  36           ACID LEACH - SOLVENT EXTRACTION PROCESS

  river directly, although the extent of possible seepage from the
  tailings pond is not known.

      The undissolved radium-226 wasted daily from the mill to the
  tailings pond constitutes a much more significant source of po-
  tential environmental contamination.  As has been shown else-
  where.  10 the initially undissolved spent ore solids can result in
  a relatively high  degree of water pollution if discharged to a
  river and permitted to accumulate on the  stream bed.  It is also
  true here (see Table 4), as at other mills, that the lighter sus-
  pended solids that are carried by the  effluents contain relatively
  high radium concentrations.  As a result of these considerations,
  it  is quite important that tailings or spent ore solids should be re-
  tained at the  plant site effectively and regularly,  and should not
  be released to the Gunnison River in any regular or significant
  quantity.

     The problem of  arsenic is somewhat different.  From the data
 of  Table 6 and Figure 2 (liquid flow rates) it can be shown that the
 two main effluent streams. Stations 6 and 8-, carry about 20,000
 grams of dissolved arsenic per day to the tailings pond.  No esti-
 mate of the arsenic content of undissolved tailings solids has been
 made.
     If this quantity of arsenic were released routinely and the
 available dilution in the Gunnison River, was considered the ar-
 senic content of the river would, for a considerable portion of the
 year,  equal or slightly exceed the  allowable concentration based
 upon the Public Health Service Drinking Water Standard. ^  For
 instance,  the minimum monthly average flow, which would be ex-
 ceeded only 50 per cent of the years,  is 150 cfs (see Figure 3). At
 this flow, the arsenic concentration would be about 0.057 mg/1 or
 slightly more than the allowable 0.05 mg/1.
     From the standpoint of chemical pollution the mill effluents
 clearly should continue  to be retained indefinitely, as has been the
 practice in the past.   This retention, however, raises the question
 of the potential accumulation of arsenic at the rate of 20, 000  g/dav
 in the soil near and surrounding the tailings pond.  The possibility
 that the accumulating quantities of arsenic may leach to the river
 in increasing amounts, or to nearby well supplies (such as that
 used by the Gunnison Mining Company) should not be ignored.  For
 a time, a minimal amount of monitoring of the river and any  near-
 by  wells for arsenic  appears to be desirable.

                       Acknowledgment

    The generous cooperation and  assistance of the following are
gratefully acknowledged:  personnel of the  Gunnison Mining Com-
pany; the U. S. Atomic Energy Commission; Stan May,  Colorado
Department of Public Health; Bill  Fixen, Public Health Service
Region VTH, Denver;  and E. A. Pash, G. Harlow, Carl Shadix and
Carl Hirth,  Public Health Service,  Cincinnati, Ohio.

-------
      THE ACID LEACH-SOLVENT EXTRACTION
              URANIUM REFINING PROCESS
            II. CLIMAX URANIUM COMPANY,
             GRAND JUNCTION, COLORADO
                        J. B. Cohen *
                        C. E. Sponagle
                          R. M. Shaw
                          J. D. Jones
                         S. D. Shearer

                        Introduction

     An in-plant survey of the Climax Uranium Company uranium
 refinery at Grand Junction, Colorado,  was performed by the Public
 Health Service during August 1958 with the cooperation  cf the Colo-
 rado Department of Public Health, the Climax Uranium Company,
 and the U. S. Atomic Energy Commission.

     The Climax Refinery is located on the  Colorado River,  about
 30 miles upstream from the Utah-Colorado State Line.  It provides
 an example of the acid-leach solvent extraction process for uranium
 recover. Vanadium is also recovered at this plant.  Figures 1 and
 2 present the process flow diagram for this refinery and the waste
 pond arrangement  respectively, which are  discussed in detail in
 the following sections.

                        Mill Process

 ORE RECEIVING, SAMPLING AND CRUSHING

     Ore delivered to the plant by truck is weighed and unloaded
 into truck bins. Each load is run separately through a jaw crush-
 er, which reduces it to a maximum size of two inches.  A sample,
 varying in size from two pounds per ton on large lots (20 tons or
 more) to about four pounds  per ton on small lots (8 to 10 tons),  is
 taken automatically during crushing.  High grade ore is hand sam-
 pled or specially sampled,  according to the size of the  lot.  Crush-
 ed and sampled ore goes either to the stockpile or to fine-ore
 storage bins.  From these bins it proceeds to the rod mill where
 it is ground to less than 14-mesh size  (2 to 4 per cent retained on
 14-mesh screen).  During this process water is added to the ore
 so that the finely ground material leaves as a slurry for the next
 phase.
^Respectively,  Sanitary Engineer, Radiological Pollution Activities
 Unit; Sanitary Engineering Director, Colorado River Basin Water
 Quality Contro. Project, Public Health Service, Denver, Colorado;
 Captain, U. S.  Army, Fort McPherson, Atlanta, Georgia; Health
 Physicist: and Senior Assistant Sanitary  Engineer, Radiological
 Pollution Activities Unit.
                             37

-------
Figure 1.  Flow diagram for the Climax Uranium Company, Grand
         Junction, Colorado, Augusf 1958.

-------
CLIMAX URANIUM COMPANY
39
                                 	OEBQTE! OLD TAILINGS POM (ROT III USE)

                                 Q	 DEMOTE! SiWtlKG STATIC*

                                 HOTt: USUAL OPERATIC OF POMS I. 2, «nd 3
                                       DESCRIBED 111 THT
         Figure 2. Mill area and pond arrangement of Climax Uranium Co.,
                Grand Junction, Colorado, August 15-18, 1958.

CONDITIONING AND  CLASSIFICATION

    Slurry from the rod mill is pumped to acid-conditioning tanks,
of which there are six arranged in series.  A strong H?SO.  solu-
tion (2-1/2 to 7 per cent acid) is added in the first tank: most of
this solution is recirculated from  storage tanks following the sand
leach. A more dilute  solution of HC1 and H?SO, is also added, this
being recirculated from the  roaster gas  scrubbing unit.  The pH in
the first three conditioning tanks is about 1.0 to 1.5, rising  in the
last three conditioning tanks in the range of 1.5 to 4.0. Some raf-
finate from  the solvent extraction process  is also returned to the
No. 1 conditioning tank along with the acid  liquor from the sand
leach.
    The purpose of this conditioning is to destroy the lime in the
ore, and change CaCO., to insoluble CaSO4. 13  This prevents
formation of the water-insoluble calcium vanadate during the
roasting process; instead, the water-soluble sodium vanadate is
formed.  Acid conditioning takes about 1-1/2 hours.

    Upon leaving the acid conditioning tanks, the slurry is sub-
jected to a second conditioning with ammonia to neutralize the re-
maining acid and to raise the pH to about 6.5.  The principal pur-
poses of this neutralization are  (a) to avoid corrosion, and (b) to
precipitate any uranium and vanadium dissolved during acid con-
ditioning or entering in the recirculated  acid solutions.  Precipi-
tation will occur in the pH range of 5.3 to 7.5.
    Upon completion of conditioning the  slurry is partially de-
watered in a cyclone  separator before proceeding to a hydraulic
sizer.  Liquor from the cyclone goes to the thickeners. In the
sizer, sands and slimes are separated hydraulically.  An attempt
is made to maintain the slimes coming out of this unit in such a

-------
 40          ACID LEACH - SOLVENT EXTRACTION PROCESS

 manner that 85 per cent of the  solids are less than 200-mesh in
 size. All larger particles are  removed as sands.
 SAND LEACH

     Sands from the sizer go to a spiral classifier for dewatering.
 Overflow from the classifier goes to the thickeners,  while the
 underflow is discharged to one of ten acid leach tanks where the
 uranium and some vanadium are leached from the sands. As the
 sands leave the classifier, concentrated H?SO.  is added at a rate
 of about 105 pounds H?SO. per ton of dry solids.  The acid leach
 process is conducted as a catch operation; a tank is filled with the
 sands-acid slurry and then handled as a unif
     The leach tanks are constructed with false bottoms so that the
 sands will be retained, while the liquor content of the tanks can be
 drained off as underflow. When a  tank is filled, the underflow is
 recycled through the  sands for about 2 hours. At the start of the
 recycling operation,  NaC103 is added to the liquor in an amount
 equal to 3-1/2 pounds per ton of dry sands in the tank.  At the  sarr.e
 time the liquor is heated to 90  degrees Fahrenheit by a heat exchan-
 ger, using steam as the source of heat.  At the end of 2 hours, re-
 cycling is  stopped, and the tank contents are allowed to "cure"
 for about 8 hours.  At the end of the curing  period the acid liquor
 is drained into a storage tank.  Fresh water is then flushed
 through the sands and run into  the same  storage tank until the  pH
 is in the range of 1.8 to 2.3.  The water flow into the  leach tank
 is then  stopped and the liquor remaining in  the tank is diverted
 into a second small (4, 500 gal.) storage  tank to be used for roas-
 ter gas scrubbing.  The spent sands are  then removed from the
 leach tank by repulping with water and discharged to  the tailings
 pond.
     Acid liquor from the first  storage tank, which contains uran-
 ium  and vanadium leached from the sands, is returned to the No.l
acid conditioning tank as described in the previous  section.

THICKENING AND ROASTING

     The components  comprising the thickener feed are slimes
 from the classifier receiving sands from the sizing operation,  and
 filtrate  from the disc filters following the thickeners. As this  cm:;
 bined flow enters the  thickening tanks. Separan (200 pounds per da;
 is added to improve sedimentation.  Overflow from the thickeners
 is discharged to Settling Pond No. 1,  while  the underflow, about
 30 per cent solids, is filtered on disc filters. The filtrate is re-
 turned to the thickener feed as mentioned above.
     The filter cake drops into  a screw conveyor and  proceeds  to
a mixer where Nad is added at the rate of  15 per cent by weight
of dry solids. This material with about 40 per cent water is still
too wet  to serve as dryer feed, so it  is mixed in a pug mill with
the previously dried filter cake-NaCl mixture in an amount suf-
ficient to reduce the moisture content to about 25 per cent.  Thi?

-------
CLIMAX URANIUM COMPANY                              41

material is fed to a gas-fired rotary dryer, which reduces the
moisture content to about 13 per cent.  The dried material passes
through a 50-ton storage bin and thence to a gas-fired 10-hearth
roaster, operating at  1400°F. to 1600°F.

    The roasting process converts insoluble vanadium compounds
in the ore to water-soluble  sodium vanadates.  This reaction can
be written
                V0O..+  2NaCl -f H0O-^2NaVO0+ 2HC1.
                  25            2           3
The hydrochloric acid gas is recovered for use in the acid condi-
tioners. by passing the roaster gas through the scrubbing units.
The uranium compounds  remain insoluble in water.  Calcines are
-split into two "streams"  on leaving the roaster.  One "stream" is
carried on a vibrating conveyor to a Baker cooler in  which cool-
ing takes place by  heat transfer into cooling water circulated a-
round the  outside of the mechanism.  In this process  there is no
contact between calcines  and water.  The second "stream" is car-
ried on a vibrating conveyor to a quench tank where it is quenched
with filtrate and wash water from an Oliver filter that follows the
thickeners.

    Cooled calcines from the Baker cooler are reground in a ball
mill: quenched calcines go to a spiral classifier. Sands from this
classifier join the  cooled calcines entering the ball mill.  The
classifier overflow is thickened in two 32-foot thickeners.  Thick-
ener overflow is processed for vanadium, underflow  for uranium.

VANADIUM  EXTRACTION

    The thickener overflow feeds five vanadium precipitation tanks
of 7, 100-gallon capacity  each.  Precipitation is carried out as a
batch process: the typical operation in any tank is as follows:
    The tank is filled with thickener overflow,  agitation begun,
and the contents brought  almost to the boiling point by injection
of steam.  Concentrated  H?SO. is then added in an amount suf-
ficient to  lower the pH to below 4.0.   The tank contents are agi-
tated until complete precipitation of vanadium as sodium poly-
vanadate (redcake) is effected. The time required for this may
vary from 1/2 to 3 hours, depending upon the "grade" and charac-
teristics of the liquor. The tank contents are then run onto a fil-
ter tray and the liquor is drained to Pond No. 3. The redcake is
washed with water to leach sulfates from the cake.  The amount of
water used is roughly equal to the volume of the tank, e.c.,  about
7, 500 gallons.  Wash water is also drained to Pond No. 3.  Washed
redcake is fused at 650 3C in a fusion  furnace, and emerges  as
V^O  product, which is then cooled and drummed for shipment.

URANIUM EXTRACTION
    Underflow from the  thickener is filtered on an Oliver filter
and the filtrate, together with water used to wash the filter cake.

-------
 42          ACID  LEACH - SOLVENT EXTRACTION  PROCESS

 is returned to a storage tank from which it is used for quenching
 roaster calcines. Washed filter cake is repulped with water and
 goes to three acid conditioning tanks, in series. Here concentra-
 ted HpSO. is added  at about 230 pounds per ton of dry solids.  The
 uranium in the cake, together with vanadium that was not water
 soluble, is dissolved in the acid during the contact time (about one
 hour) in these  tanks.

     Upon leaving the last of these tanks the  slurry goes through
 three additional cycles of conditioning, filtration and repulping to
 extract the maximum amount of uranium.  Cake from  the last fil-
 ter is repulped with water and pumped as "slime leach tails" to the
 tailings pond.  Filtrate from the  last filter is returned to the
 process.

     The pregnant acid  liquor filtrate from the first and second fil-
 ters is settled for removal of slimes. The liquor then goes to onu
 of three 7500-gallon storage tanks and then to the solvent extrac-
 tion tanks.  Here, uranium is stripped from the acid liquor with
 an organic solvent.  The raffinate is returned to the No. 1 condi-
 tioning tank, while the  pregnant organic goes to a storage tank,
 then into a circuit where the uranium  is stripped from the solvent
 with a 10 per cent Na~CO,, solution.  Barren organic is returned
 to the solvent extraction circuit, while the loaded Na^CO,, proceed;
 to one of two precipitation tanks, operated as a parallel batch pro-
 cess.  When one of these tanks is filled, about 800 pounds of con-
 centrated H?SO. is  added to neutralize the NapCO,,, and the con-
 tents are brougnt to a boil with steam to drive off excess carbon
 dioxide. Ammonia is added to precipitate the uranium as an am-
 monium di-uranate.  When precipitation is complete,  the tank
 contents go to a filter press.  Filtrate is returned to the No. 4
 conditioning tank. Filter cake is fed to a gas-fired dryer operatin;
 at a temperature of  900°F. where drying takes place,  and the feed
 is converted to U«0R (yellowcake), which is then drummed and
 shipped to  an Atomic Energy Commission facility.

 OPERATION OF SETTLING  PONDS NOS. 1, 2, and 3

     As shown in Figure 2, in addition to a large tailings pond, thiv
 smaller settling ponds, each approximately 150 feet square, are
 provided for settling thickener overflow and  vanadium tank filtrate
 and wash water prior to their discharge to the river.  Ordinarily
 thickener overflow enters Pond No. 1. This  pond discharges to
 Pond No. 2, and thence to Pond No. 3. where the waste vanadiun;
 liquors enter.  The only discharge from this series of ponds is
from Pond  No.  3 to the  river.

     During Cycle I of the mill survey, Pond No. 2 was being
cleaned out.  During Cycle  II, Pond No. 2 was still out of opera-
tion and cleaning operations were underway in Pond No. 3.  Be-
cause of this situation,  the contents of Pond No. 1 were dischar-
ged directly to  the river, as were those  of Pond No. 3. Conse-
                                                     GFO 813—1"'-

-------
CLIMAX URANIUM COMPANY
                                           43
quently, the effluent samples collected from these ponds did not
reflect the discharge which would have occurred if the ponds were
being operated in their normal manner.

     Since the survey in 1958, an additional tailings pond has been
constructed.  The large tailings pond, shown discharging to the
drainage ditch on Figure 2, now discharges to the new pond.  No
direct discharge to the river is anticipated from this new pond.
In addition, the  settling pond arrangement has been changed so
that the only  discharge to the river is from the vanadium extrac-
tion circuit.  All other discharges  now enter the tailings ponds.

                           Mill  Survey

     During the survey period,  August 15 through August 18,  1958,
nineteen sampling stations were established for the purpose of ob-
taining representative samples from the mill process for analysis.
Table  2 describes the stations sampled;  Figure 1 gives the loca-
tions.  Sampling Stations 1, 2, 3, 4, 8,9,10,13,14,16,  and 17 were
main mill process streams,  whereas Stations 5,6, 7,11,12,15,18,
and 19 were representative of waste streams.
     Sampling was performed by personnel of the Public  Health
Service and the Colorado Department of Public Health.  Two sepa-

             Table 1. CLIMAX MILL SAMPLING STATIONS
             Station
               1

               2

               3

               4

               5

               6

               7

               8

               9

              10

              11

              12

              13

              14

              15

              16

              17

              18

              19
                                     Description
Classifier discharge to conditioning tanks

Conditioning tank effluent to hydraulic classifier

Thickener feed (65-foot tanks)

Sands to acid leach

Spent sands to tails

Influent to Pond No. 1

Effluent from Pond No. 1

Roaster feed

Roaster calcines

Thickener feed - vanadium  circuit

Influent to Pond No. 3

Effluent from Pond No. 3

Vanadium product

Filter cake to solvent-extraction circuit

Slime leach tails

Filtrate to solvent extraction tanks

Uranium  product

Combined effluent from Ponds Nos. 1 and 3

Sand-slimes tailings pond discharge

-------
 44          ACID  LEACH - SOLVENT EXTRACTION PROCESS

 rate sampling cycles were selected during which samples were
 collected continuously.  The sampling periods were:

 Cycle I  - 3:00 PM, August 15 - 11:00 PM,  August 16, 1958
                                               (32 hours)

 Cycle H - 3:00 PM, August 17 - 3:00 PM,  August 18, 1958
                                               (24 hours)

 Compositing of samples during these periods varied with the type
 and duration of operation. Adequate sampling was  difficult at
 several locations due to the batch type operations described pre-
 viously. Also,  flow at  several other stations was not continuous
 over the entire length of the  sampling cycles. Detailed records
 were maintained for each  sampling  station in order that data de-
 rived from laboratory analyses could  be adequately interpreted.

     All samples collected during the survey were shipped to the
 U. S. Public Health Service Robert A. Taft Sanitary Engineering
 Center at Cincinnati, Ohio.  Portions  of all samples were sent to
 a private laboratory for determinations of dissolved and undis-
 solved  radium; all other determinations were performed  at the
 Cincinnati laboratory.

                     Analyses and Results

     To adequately determine the operating characteristics of the
 individual units of the mill process, a flow balance was calculated
 for the  mill. This was  accomplished by combining the laboratory
 analyses with the  flow measurements taken on the effluent streams.
 To estimate the quantities of liquid, suspended solids,  and dis-
 solved solids at the  various stages of the process during  each cy-
 cle a solids balance throughout  the plant was assumed.  Because
 of this assumption the solids for each cycle necessarily balanced.
 The  input to the plant on the basis of a balance was calculated as
 540 and 583 tons of ore per day for Cycles I and H, respectively.
 Results from the two cycles as  presented in  this section of the re-
 port have been combined and adjusted to represent the mill pro-
 cessing 540 tons of ore per day; the solids flow at each station is
 presented in Table 2.
    Figure  3 illustrates the average flow for the two cycles in
terms of gallons per minute of liquid and  solids  at the stations sam-
pled  during  the process. The  flow at Stations 4, 5, 7,11, 12.13, 15.
16, and 17 were measured during the survey while  the flows at the
remaining stations were calculated from the  solids balance.

    Table 3 presents process stream characteristics, both mea-
sured and calculated. These  and all other process  stream char-
acteristics are reported on the basis of uninterrupted operation
at all stations and steady flow conditions for batch operations  such
as the sand  leach and the vanadium precipitation operations.

-------
CLIMAX URANIUM COMPANY
45
          Table 2.  SOLIDS IN PROCESS STREAM
Station
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Tons per day
Suspended
537
448
120
394
390
1.7
0.3
123
123
97
0.04
0.04
7.9
97
88
0.1
2.0
0.2
0
Dissolved
3.0
72
71
1.0
1.6
56
50
a
a
74
6.2
6.3
a
12
0.9
20
a
49
0.2
Total
540
560
191
395
392
57
50
123
123
171
6.2
6.3
7.9
109
89
20
2.0
49
0.2
          •* Solid Sample

    From Figures  1 and 3 comparisons between stations can be
made of the flows entering and leaving various sections of the mill.
For the over-all plant balance, the input to Station 1 can  be com-
pared with the total output of the  spent sands, Station 5; the in-
fluent to Pond No.  1.  Station 6: the influent to Pond No. 3. Station
11; the vanadium product.  Station 13; the slime leach tails. Sta-
tion 15, and the yellowcake product, Station 17.  The effluent from
the ammonia conditioners, Station 2, should be accounted for at
the feed to the acid leach tanks. Station  4; Pond No. 1,  Station 6;
and the  roaster feed.  Station 8. The roaster calcines at Station 9
should divide between the filter cake at Station 14. and Stations
11 and 13.  The roaster feed at Station 8 should compare  with the
calcines at Station 9.  Lastly, the filter  cake at Station 14 should
be measurable in the  filtrate to the solvent extraction tanks.
Station 16, and in the slime leach tails,  Station 15.  These com-
parisons can also be used for estimating the quantities of water
added between sampling  stations.

    Table 4 shows  the gross alpha and beta radioactivity concen-
trations.  The figures are an average for the two  sampling cycles.

-------
 46
ACID LEACH - SOLVENT EXTRACTION PROCESS
4—
K
• a

ORE
\
CLASSIFIER
36
61
— <
CC»niTIC»-
C«S
31
JIL
262
— C
DOWrLr-F
'I8

SIZE'
5,
\ 	
S»I»AL
CLASSIFIER
SO.^.
103 .


~>
— »
_^
*
^

— CD
A: o
LEACH
TA.»«S
i
F
Si»r
, TO TA
-
TANKS


"3
-C
rfj
•v— ..-
(5)«!
26




SCR'J33ERS
P
MA
J
PCI
V£»
TFR
»3
-^1. 1 	 ^
N.
~^l

«' Fl(^
©! ~c
9C
7) S°
~ C
•|J2L^
THIC.E.FR-;


PISC
FILTERS

^*-x
ROASTER


,_>

-*•
'«^_

)f F r :
sours FU
till! IS FL
SLL'^'r FU
MLi»"ILL FILT"
CLASSIFIER
•-0 Hf
TBn.EBEI1 ™ERFL»

)W -
V -
V -
WLINS
PE

1C gp«"
CC go
-------
 CLIMAX URANIUM COMPANY

            Table 3, PROCESS STREAM CHARACTERISTICS3
47
Station
1
2
3
4
5
6
7
a
9
10
11
12b
13
14
15
16
17
18b
19
Slurry
flow, gpm
97
282
562
58
82
469
408
c
c
133
90
145
c
35
41
50
c
545
8
Specific
gravity
of slurry
1.56
1.23
1.02
J.70b
1.49
1.01
1.01
-
-
1.14
1.01
1.01
-
1.33
1.22
1.08
-
1.01
1.00
Dry suspended
solids by
weight, %
59.2
23.9
35.1
67.3b
53.9
d
d
- 87
-100
11.0
d
d
- 100
35.0
29.5
d
- 100
d
d
Specific
gravity of
dry solids
2.47
2.63
2.63
2.56
2.63
-
-
2.13
2.30
2.21
-
-
2.73
2.60
2.66
-
5.81
-
-
              Average for two cycles

              One cycle only
              Liquid (negligible suspended solids)

cines at Station 8, 1, 626 me/day, can be equated to the 1, 499
(16 +  5 + 1, 478) me/day calculated for Station n, the influent to
Pond  No. 3; Station 13, the vanadium  product; and Station  14. the
filter cake going to the uranium extraction process. In turn the
activity at Station  14,  1478 me/day,  agrees well with the 1472
(1066 + 406) me/day encountered at Station 15, the slime leach
tails;  and Station 16,  the filtrate to the solvent extraction tanks.
For the over-all plant balance, 1778 me/day enter the plant as
compared with 1990 (273 + 31+16+5+ 1066 + 599) me/day that
leave at Stations 5, 6,  11, 13. 15, and 17. It should be noted here
that the concentrations of alpha and beta activity present in Table
4 are  laboratory determinations and as such could be different
from  the actual activity that might be measured at the plant,  if
this had been possible. Material that was not in radioactive equil-
ibrium during the  process would tend to return to  equilibrium in

-------
 48
ACID LEACH - SOLVENT EXTRACTION PROCESS
    Table 4. GROSS RADIOACTIVITY CONCENTRATIONS
Station
1
2
3
4
5
6
7
8
9
10
11
12 a
13
14
15
16
17
18 a
19
Gross radioactivity of total sarr.ple (slurry), >i/ic/l
Undissolved
Alpha
3,350.000
1, 180.000
533.000
1.500.000
631.000
11.900
3,270
b
b
1.780.000
19.800
2.560
b
7.900,000
4.800.000
15.300
b
452
235
Beta
4.150.000
1.640.000
753.000
1. 500,000
923.000
20. 700
3.830
b
b
1,660.000
29.900
1.320
b
7.960,000
3.520.000
8.320
b
1.030
335
Dissolved
Alpha
329
3.030
1,880
418 a
I,i70
260
4.130
b
b
5.380
10. 700
5. 100
b
3.250
725
1.520,000
b
4.300
320
Beta
905
7.320
2.760
1.5C03
5.770
700
5. 460
b
b
14.700
48,200
26.800
b
8.600
1.090
2.970.000
b
10. 150
330
Gross radioactivity of dry
undissolved solids. >*^ic, g
Alpha
3.640
4.050
15.300
1, 180
770
22.600
27.700
16.000
14.700
14,300
291.000
42. 100
690
16,800
13,300
33.000
329,000
70,500
3, 190
Beta
4.530
5.670
21,600
1.330
1. 150
35.000
33,200
16,300
13, 100
13,400
451,000
21.800
808
16,800
9.750
18.000
436, 000
168. 000
3.730
     Single cycle.
   jj
     Solid sample.

the sample before being assayed for radioactivity.
     Radium-226  analyses were performed by a private laboratory
after sample preparation at the Taft Center.  Solids were separated
from the liquid by means of a membrane filter and then ground t-~>
less than 100 mesh. The solids and liquid portions were then ana-
lyzed;  results were reported in terms of/i/ig/1 of liquid sample and
/*Ag/g of solid sample. These results were  then converted to the
units presented in Table 6.  As in  the case of the alpha activity,
the sand-slime separation apparently tends  to concentrate the un-
dissolved radium in the slimes. The effect  of Pond No. 1 is  also
evident from Table 6.  Here, the undissolved radium is settled,
going from 1370 to 142npg/l of slurry, while the dissolved activ-
ity increases from 125 to 475^/ig/l of slurry.  This is not as evi-
dent in Pond No.  3 where river water is returned from the Baker
cooler.

-------
CLIMAX URANIUM COMPANY

         Table 5. ALPHA ACTIVITY IN PROCESS STREAMS
49
Station
1
22
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Undissolved
1,778
1.790
1.653
423
272
30
7
1.778
1.626
1 257
10
2
5
1.477
1.066
4
599
1
< 1
Dissolved
< 1
4
6
< 1
1
1
9
a
a
4
6
4
a
1
< 1
402
a
13
< I
Total
1.778
1.794
1.659
423
273
31
16
1.778
1.626
1.251
16
6
5
1.478
1.066
406
599
14
< 1
     The radium concentrations in Table 6 have been combined with
the flows in Table 3 (by cycle) to estimate the quantity of radium-
226 in the process stream (Table 7 and Figure 4).  Determination
of this radium distribution was one of the main  objectives of the
mill survey. In balancing the radium  input against output at the
locations indicated previously, we see that the 256 me/day at
Station 2, the conditioner tank effluent,  compare favorably with
the 270 (59 + 4 + 207) me/day at Station 4. the sand to the acid
leach; Station 6,  the influent to Pond No.  1; and Station 8. the
roaster feed, the 219 me/day of radium in the roaster calcines.
Station 9, are accounted for in the 184 (2.1 + 182) me/day at
Station 13, the vanadium  product, and Station 14, the filter cake
from the Oliver filter.  The  182 me/day at Station  14 are in
agreement with the 176 me/day at Station 15. the slime leach
tails.  The over-all plant balance equates 331 me/day at Station 1
to 232 (50 + 4 + 2 + 176) me/day, leaving the plant at Stations 5,
6, 13, and 15,  respectively.  An examination of the results from
Stations 1 and 2 in  Tables 6  and 7,  and consideration of the man-
ner in which Station 2 balances when compared with other stations

-------
50          ACID LEACH - SOLVENT EXTRACTION PROCESS

        Table 6. RADIUM - 226 CONCENTRATIONS.
Station
1
2
3
4
5
6
7
8
9
10
11
12a
13
14
15
16
17
18a
19
Concentration in slurry, Itpg/'l
Undissolved
625,000
168,000
75.800
194,000a
112.000
1.370
142
b
b
305.000
127
7.3
b
963.000
739.000
600
b
35
645
Dissolved
187
660
345
72a
138
125
475
b
b
4.950
56
5.8
b
2.710
65
1.250
b
490
74
Concentration in dry
undissolved solids,
Wg/g
680
575
2,100
165
140
2,450
1,200
1,850
1,950
2,550
1,950
120
300
2.050
2.200
1,300
26
550
690
       a  Single cycle.
       b  Solid sample
tend to indicate that the laboratory analysis for the undissolved
radium was higher than would normally be expected at Station 1.
On the basis of these comparisons,  the radium-226 passing
Station 1 should more probably be about 240 me/day in order to
produce more over-all agreement throughout the process.
    Chemical analyses were performed on several of the effluent
samples collected during the survey; Table 8  presents these re-
sults. No reason is apparent for the large variations between the
analyses from Cycles I and n at Station 19.
    Chemical utilization was reported as follows:

      HpSO.     -   105 Ib. per ton of sand, for sand leach

                 -   250 Ib. per ton of slime, uranium extraction

                 -   400 Ib. per tank, vanadium precipitation

                 -   800 Ib. per tank, uranium  precipitation

-------
 CLIMAX URANIUM COMPANY
                                       51
          NH,
         NaC103

WASTE DISPOSAL
-  548,000 Ib. per month

-  3.5 Ib. per ton of sands
    Results of the survey indicated that Radium-226 was entering
the river at Stations 18 and 19 at rates of approximately 0.16 and
0.005 mg/day, respectively.  Of this quantity 0.11 mg was sus-
pended radium and the remainder, dissolved.  These quantities may
have since decreased, due to tailings ponds rearrangement pre-
viously mentioned.  A survey of the Colorado River in the vicinity
of Grand Junction was conducted during August 1960 for the Colo-
rado River Basin Water Quality Control Project. *4  At that time
samples of river water and sediment were collected about 0.3
miles above the Climax Mill (Station G-l at the lower end of a
diversion  channel at the mill (Station G-M), and below this di-
version channel, about one-fourth mile above the mouth of the
Gunnison River (Station G-2).  The results of this survey are pre-
sented in Table 9.  Each of the two consecutive cycles was of

           Table 7. RADIUM - 226 IN PROCESS STREAM
Station
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Racium - 226 mg/day
Undissolved
331
255
230
59
50
3.5
0.32
207
219
220
0.06
0.01
2.1
181
176
0.16
0.05
0.11
0.01
Dissolved
0.10
0.92
1.07
0.02
0.06
0.33
1.05
a
a
3.5
0.02
< 0.01
a
0.50
0.01
0.33
a
1.45
< 0.01
Total
331
256
231
53
50
3.8
1.4
207
219
224
O.OB
0.01
2.1
162
176
0.49
0.05
1.6
<0.01
            Solid sample.

-------
 52
ACID LEACH - SOLVENT EXTRACTION PROCESS
                                   KEY: UK3ISSOLVED - 3CC 
-------
CLIMAX URANIUM COMPANY

    Table 8. CHEMICAL ANALYSES OF EFFLUENT SAMPLES
53
Determination
Total acidity, as CaCO,
Mineral acidity as CaCOj
Salfate
Hardness, as CaCOj
Chloride
Sodium
Fluoride
Vanadium
Arsenic
PH
Concentration mg/1 station - Cycle
12-1
500
250
2.400
2.500
1.340
1.800
12
.03
0
2.7
15-1
250
20
1.300
2.100
140
135
13
.03
0.2
3.2
19-1
310
10
7.600
3.600
2.280
1,350
13
.OS
3
3.8
i9-n
150
0
1.300
1,300
415
420
6
.007
0
6.9
would be from seepage,  pond overflow if this should occur and
spills due to washout of pond walls. The latter is not very likely
although not without precedent in the industry.

    The increase in the alpha activity activity and radium and
uranium concentrations at Station G-M is clearly detectable.  In
terms of the maximum permissible concentrations (MFC) of radio-
nuclides in water outside of a controlled area,  as specified by
NBS Handbook 69 ^ the  uranium concentration is well below the
allowable 20 mg/1 while the  radium-226 is  definitely above the
allowable 3.3/i^c/l; however,  due  to the limited access to the
diversion channel before reaching  the main body of the river
where the radium concentration is below the MFC,  this cannot
be considered of important public health significance.

Table 9. ANALYSES OF RADIOACTIVITY IN MUD AND WATER SAMPLES
     OF THE COLORADO RIVER - SURVEY RESULTS. AUGUST I960 a





G-] Cycle I
Cycle II
G-M Cycle I
Cycle II
G-2 Cycle I
Cycle II
Mud Sa.rr.ple b

Alpha
activitv.
we f
14.6
26.4
386
653
249
343
Beta
activity
Itlic e
39.6
58.9
478
615
509
546

Ra-226
A^c g
3.2
3.4
13
19
3.7
4.8
Water Sarr.ple l

Alpha
activity
W*c 1

4.4

10.7

7 0
Beta
activity.
,11/lC 1

13.1

-

31.6

Ra-226
we 1
0.3
1.6
3.9
4.8
1.1
1.5

Uranium
MS 1
12
14
44
35
22
21

-------
54          ACID LEACH - SOLVENT EXTRACTION PROCESS

                     Acknowledgment

    The generous cooperation and assistance of the following are
gratefully acknowledged:  Ralph Musgrave,  Chief Metallurgist;
and other personnel of the Climax Uranium Company; the U. S.
Atomic Energy Commission; S. May,  Colorado Department of
Public Health; and G.  Harlow and E. A. Pash, U. S. Public Health
Service.

-------
 THE CARBONATE LEACH URANIUM EXTRACTION
                         PROCESS
L HOMESTAKE-NEW MEXICO PARTNERS  COMPANY,
                 GRANTS, NEW MEXICO

                        J. B.  Cohen*
                        H. R. Pahren
                      M. W. Lammering

                        Introduction

     This report presents the results of an inplant survey of a uran-
 ium refinery that utili/es the carbonate leach extraction process for
 the separation of uranium from  its ore. The study was performed
 by the Public Health Service during September 1959 with the co-
 operation of the New Mexico Department of Public Health, the
 Homestake-New Mexico Partners  Company, and the U. S. Atomic
 Energy Commission.
     The Homestake-New Mexico mill is located about 10 miles
 northeast of Grants, New Mexico.  Operation began during 1958.
 The mill is rated at 750 tons per day: although between Septem-
 ber 22 and September 28, 1959, the  dates of the survey, about
 900 tons of ore per day were processed. The ore assayed from
 0.173 to 0.195 per cent U.,0  and yeilded about 4,000  pounds per
 day of yellowcake. Wastes Trom the mill were discharged to a
 tailings pond,  and the liquid  portion  not lost by seepage and eva-
 poration was recycled for use as process water. There was no
 surface water in the vicinity likely to receive any of these wastes.

     Figure 1 is a flow diagram  of the process, which is descri-
 bed in detail in the following sections.

                     Process  Description

 ORE  PREPARATION
     Ore is brought from the nearby  uranium mines by truck and
 stored outdoors until used.  The ore is transferred by means of a
 bulldozer to a feed hopper, from which  the ore is conveyed to a
 jaw crusher.  Here the lumps are  crushed to less than 3/4 inch.

     A 10 per cent sample of the crushed ore is obtained by moving
 buckets as they pass through the stream of ore falling from a con-
 veyor into a hopper.  A second 10  per cent sample is obtained from
 the original 10 per cent,  cut in a. similar manner,  to give a one
 per cent portion of the original ore.  The stream representing one
 per cent of the ore is further crushed and then impinges on a  ro-

*Sanitary Engineer,  Radiological Pollution Activities  Unit: Sanitary
 Engineer. Colorado River Basin vVater  Quality Control Project.
 Public Health Service, Denver.  Colorado: and Senior Assistant
 Sanitary Engineer. Radiological Pollution Activities  Unit.
                              55

-------
56
CARBONATE LEACH PROCESS
                                            Q	 S»MPLING POINTS
                                               •SS"
         Figure 1. Flow diagram of Homestake-New Mexico Partners uranium
                mill, Grants, New Mexico, September 1959.

 tating disk containing slots.  The number of open slots determines
 whether the disk takes a 2.5,  5.0, 7.5, or a 10 per cent sample of
 the ore stream. Thus,  the final sample may contain from 0.025
 to 0.10 per cent of the ore passing through the plant.  The remain-
 ing ore is stored in a fine ore storage tank of 3, 000 ton capacity.

 GRINDING
     After storage there are parallel circuits for the grinding,
 classification and cyclone separator steps.  The remainder of the
 process is essentially a series  circuit.
     Ore is conveyed from the fine ore storage bin to a ball mill
 where it is ground to less than 65 mesh.  Carbonate solutions will

-------
HOMESTAKE - NEW MEXICO PARTNERS COMPANY        57

not react with many ore components other than the uranium mine-
rals, and it is necessary to grind the ore to this size to provide
the necessary amount of surface area for efficient leaching.  The
necessary particle size is determined by the type of ore processed.
Limestone  ores in which the uranium is finely dispersed throughout
the matrix  must be ground finer  than sandstone type ores in which
the uranium material is part of the bond between the sand grains.
    Mill solution  is added to the  ball mill with the ore.  A spiral
classifier separates any oversize particles and returns them to
the ball mill.  The specific gravity of the slurry from the classi-
fier is  automatically controlled at 1.20 by means of an automatic
valve.  The flow of additional mill solution added to the classifier
is either increased or decreased, depending on the amount neces-
sary to maintain the desired specific gravity in the classifier
overflow.  This effluent is pumped to a small cyclone separator
where the coarse  solids are separated  from the slime.

    The cyclone overflow is sent to a 75-foot thickener where  the
solids are further concentrated.  About 20 pounds per day of Sepa-
ran are added here to aid in solids separation.  Overflow from the
thickener is recycled to the mill solution storage  tank.  The thick-
ener underflow is combined with the underflow from the two cy-
clones  in an agitated tank.  Combined effluent from the  agitated
tank flows to a sump,  from which it is  pumped to  the leaching
section.

LEACHING

    Leaching of the uranium from the solid particles is accom-
plished with a sodium carbonate-bicarbonate liquor in six Pachuca
tanks,  19 feet in diameter and 48 feet high, operating in series.
There is a  7-hour retention time in each tank or 42 hours total
retention in the leaching circuit.  The Pachuca tanks are operated
at 179°F. and at atmospheric pressure.

    The insoluble quadrivalent uranium must be converted to the
soluble hexavalent form.  Soluble uranyl tricarbonate then forms
in the carbonate solution, under  the leaching conditions.  The solu-
bilization of the hexavalent form, such as the uranium mineral
carnotite in order to produce the uranyl tricarbonate ion may be
represented as follows: ^- 17
       + H00
          z
    The  Pachuca tanks are aerated to provide for the oxidation of
reduced uranium  compounds.  Copper sulfate and ammonia are nor-
mally added to the ore slurry to catalyze the oxidation reaction.
During the period of the survey however, a test run was made
with cyanide instead of the copper sulfate and ammonia.  Crude
cyanide that contained about 50 per cent NaCN equivalent was

-------
 58                            CARBONATE  LEACH PROCESS
 added to the ore at the rate of 0.8  pounds per ton of ore as it was
 conveyed to the ball mills.  The cyanide reacts with the iron balls
 in the ball mills to form the complex ferricyanide ion, a strong
 oxidizing agent for uranium.

 PREGNANT LIQUOR PREPARATION

     The slurry from the final leaching tank flows to the first-
 stage rotary filters, where the solids are separated from the
 liquid containing the uranium or the pregnant solution. There are
 three filters in each stage of filtration.  The filter cake is washed
 with recarbonated barren solution and this wash water becomes a
 part of the filtrate.
     The pregnant  solution is then  pumped to an aerated flotation
 tank where  any hydrocarbons in the  solution are removed. If these
 hydrocarbons are  not removed they  interfere with the next step by
 preventing the complete precipitation of uranium with the  caustic.
 A small amount of organic chemical is  continuously added to pro-
 mote frothing and  to aid in removing chemical is continuously adde:
 to promote frothing and to aid in removing the hydrocarbons.  The
 froth overflows to the floor sump,  from which it goes to the pri-
 mary thickener. The flotation tank underflow is sent to a  second
 75-foot settling tank for further clarification.  Overflow from this
 settling tank flows to the pregnant  liquor storage tank while the
 underflow is pumped to the primary thickener.

 PRECIPITATION AND PREPARATION OF YELLOWCAKE

     Pregnant solution is pumped to the first of three 20-foot dia-
 meter precipitation tanks operating in series.  A 50 per cent
 sodium hydroxide  solution is added at a rate of about 5,000
 pounds per day to precipitate the sodium diuranate  yellowcake.
 The  reaction is as follows:

               3 +  6NaOH—*Na2U207+  6Na2C03 + 3H20

 The precipitated yellowcake is concentrated in a 12-foot diameter
thickener.  Underflow from the thickener passes through a Burwell
filter press that removes the solids.  After the yellowcake is re-
moved from the filter, it is dried  in a gas fired drier, ground in
a hammermill, and drummed for shipment to the U. S. Atomic
Energy Commission.

    Overflow from the yellowcake thickener and the filtrate from
the Burwell presses are combined.  This barren solution  is padded
through a Sperry filter press to remove any remaining yellowcake
particles before it is pumped to the recarbonation tower where the
boiler plant flue gas is passed countercurrent to the barren solu-
tion.  Carbon dioxide in the flue gas neutralizes the excess caustic
alkalinity thus forming additional carbonate and bicarbonate. The
recarbonated barren solution is then reintroduced into the process
at the first stage filters,  following the leach process.
                                                    GFO 813—17'-

-------
HOMESTAKE - NEW  MEXICO  PARTNERS COMPANY        59

FILTRATION  OF  TAILINGS

    The solids that are separated from the pie gnant solution by the
first stage filters are washed by a portion of the recarbonated
barren solution. The  remaining recarbonated  solution is used to
repulp the first-stage filter cake. This repulped slurry is then
filtered by the second stage filters.  The water used to wash the
filter cake from the second stage filters is return water decanted
from the tailings pond.  Filtrate from the second stage filters is
returned to the primary thickener.  About 50 pounds per day of a
flocculating agent, guar gum solution,  is added to the filter feed
to aid in filtration.

    The filter cake from the second stage filters is repulped so
that it may be  pumped to the tailings pond.  Part of the repulping
water consists of fresh water from  the plant well.  The balance of
the water is return water from  the tailings pond.

    Normally  the water returned from the  tailings pond passes
through an ion exchange system to recover any dissolved uranium,
and a flotation unit to  remove any hydrocarbons before the water
is used.  During the period of the survey,  however, the ion ex-
change system was not in operation and the return tailings water
by-passed this step.  The flotation unit, however,  remained in
operation. A gas flame was used to burn off any hydrocarbons in
the froth from  the flotation unit, before the froth flowed to the
floor sump and the primary thickener.

TAILINGS POND

    The tailings pond contained approximately 40 acres within the
dikes,  of which about  10 acres were occupied by clear water and
about 10 acres by moist tailings.  The remaining area was dry. At
one end, an area is partially separated from the remainder of the
pond. Water overflows from the main portion  of the pond into this
area, and from here is returned to  the mill process.

                      The Mill Survey

    For the  purpose of analyzing the individual components that
make up the  entire extraction process, samples were obtained
during two sampling cycles of 72  hours each.  The sampling per-
iods  were as follows:

Cycle 1-7 AM. Sept. 22,  1959. to  7 AM, Sept. 25, 1959.

Cycle 2-7 AM, Sept. 25,  1959. to  7 AM, Sept. 28, 1959.

    Table 1  gives  a list of the sampling stations, together with a
brief description of each..

    A sample  of the ore being fed to each ball mill was collected
by plant personnel every hour and composited over a 24-hour

-------
 60
                   CARBONATE LEACH  PROCESS
Table 1. SAMPLING STATIONS
              Station number
                  1

                  2

                 34

                  5

                  6
                 9

                 10

                 11

                 12

                 13

                 14

                 IS

                 17
                                  Description
            Raw ore feed to ball mill

            Mill solution

            Combined overflow from classifiers

            Slurry to leach taris

            Leach tank effluent

            Waste to tailings pond

            Pregnant filtrate from first stage filters

            Pregnant solution to precipitation tanks

            Yellowcake product

            Barren solution from Burwell Press

            Filtrate from second stage filters

            Raw well water

            Return water from tailings pond

            Recarbonated barren solution

            Combined barren solution to Sperry Press
period.  The samples from the two ball mills were combined, pul-
verized, blended and resampled.  Portions of the daily composite
samples were weighted according to tonnage of ore fed to process
and combined for each cycle (Station 1).

     Equal portions were composited from each drum  of yellow-
cake  packaged during each three-day sampling period to give the
yellowcake sample representative of Station 10.

     Every 2 hours,  samples were collected at both classifier
overflows (Stations 3 and 4); combined slurries were pumped to the
leach tanks (Station 5), and from the final leach tank (Station 6).
These samples were composited for each cycle.
    During Cycle  1 samples were collected from  the drinking
water fountain every two hours and composited (the plant raw wa-
ter sample,  Station 13). For the second cycle,  one grab sample
of water was obtained to determine if there was any difference  in
the two methods of collection, such as possible contamination of
the composite sample by plant dust.
    All other plant samples were collected by  means  of automatic
samplers, and daily samples were composited for each 3-day
period of the cycle.
    The samples from  this survey were  sent to the Public Health
Service's Robert A. Taft Sanitary Engineering Center, at Cincinnat:
Ohio, where all physical, chemical,  and radiological  determination
except radium analyses, were conducted. Duplicate portions of the
liquid and solid samples were sent to a private laboratory for
determination of dissolved and undissolved radium.

-------
HOMESTAKE - NEW MEXICO PARTNERS COMPANY
61
                     Analyses and  Results

    The results presented here are a combination of field data,
laboratory analyses and flow balances; together they present an
over-all picture of the units composing the carbonate-leach pro-
cess. With physical data from the field and laboratory, a flow
balance for the mill was obtained.  This was done through the use
of a solids balance and a total weight balance between the various
mill process units.  These results enabled the preparation of  Fig-
ure 2,  a schematic flow diagram giving the estimated slurry flow,
separated in  terms of solids and liquid,  at each of the sampling
points.  Table 2 presents the physical characteristics of the process
stream  at the various stations.  The data, including all  balances,
were measured or computed separately for each cycle and then
averaged because of their close agreement.  As previously men-
tioned,  the 42-hour time lag between Stations 5 and 6 (the leach-
ing process)  tends to make a mill balance difficult, but  averaging
the two  72-hour cycles minimized this problem.
    In Table 2 the data in colums (3) through (6) are laboratory
determinations, while the slurry flows in column (2) were Calcu-
   lable 2. PROCESS STREAM CHARACTERISTICS a
Station
(1)
1
2
34
5
6
7
8
9
10
11
12
13
14
16
17
Slurry flow.
cpnl
h)
b
508
580
183
178
285
117
80
b
d
73
70
150
82
81
Specific
gravity cf
slurrv
(3) "
-
1.09
1.23
1.48
1.54
1.32
1.10
1. 10
-
1.10
1.07
1.00
1.01
1.03
1. 10
Dry suspended
solids by
weicht. ^~
(4)
94.1
c
25.4
54.6
52. B
39.1
c
c
99.4
c
c
c
c
c
c
Specific
gravity of
dry solids
' (5)
-
-
2.64
2.59
2.66
2.72
-
-
-
-
-
-
-
-
-
PH
(6)
-
10.1
10.1
10. 1
10.0
9.6
10.1
10.1
-
12.0
10.2
-
9.8
10.3
12.0
     Average of cycles 1 and 2

     Solid sample.

     Liquid (neirlitrible suspended solids).

     Flows not calculated.

-------
 62
CARBONATE  LEACH PROCESS
 lated. Although the fresh water flow, used for repulping the final
 tails, at Station 13 was calculated as 70 gpm,  additional fresh
 water entered the process at other points,  including water for
 chemical feed preparation, water used for plant housekeeping,  and
 makeup water needed because of evaporation and other losses.

     Table 3 presents the average solids flow (average of cycles
 1 and 2).  Because of the selectivity  of the leaching process for
 the uranium compounds and the low concentration of uranium in
 the ore, there was little detectable change through the mill as  re-
 gards tonnages of undissolved ore solids.
          Table 3. SOLIDS IN PROCESS STREAMS
Station
1
2
34
5
6
7
8
9
10
12
13
14
16
17
Tons per day
Suspended
331
0.8
834
633
663
633
0.6
0.2
2.0
0.9
"0
-0
0.2
0.1
Dissolved
-
302
299
83
111
12
64
57
-
35
0.5
12
47
55
Total
881
303
1183
9?6
979
895
85
57
2.0
36
0.5
12
47
55
     Figure 2 shows that there are locations throughout the mill
where quantities entering and leaving sections of the process can
be directly balanced. The mill solution. Station 2, and the ore
from storage. Station 1, entering the ball  mill and classifier
should be quantitatively accounted for at the exit from the classi-
fiers.  Station 34.  Quantities present in the pregnant liquor at
Station 9 should be measurable in the yellowcake product at Sta-
tion  10 and in the recarbonated barren  solution at Station 16.  For
the over-all plant balance, the ore entering at Station 1,  the fresh
water at Station 13,  and makeup water  at Station 14 should approxi-
mate the yellowcake produced  at Station 10 and the slurry flow  to
the tailings pond at Station 7.  A balance at the primary thickeners
would equate the input from the classifiers at Station 34 and the
primary and secondary filtrates at Stations 8 and 12 to the output to

-------
HOMESTAKE - NEW MEXICO PARTNERS COMPANY
63
         Figure 2. Schematic flow diagram of Homestake-New Mexico Partners
                mill, September 1959.

leaching at Station 5, the mill solution at Station 2 and the preg
nant liquor storage at Station 9 respectively.
    Results for the stations mentioned above show that  the
solids flows of Table 3 are  fairly good. At the ball mill and
classifier,  1184 (881 + 303) tons of solids  per day  enter  as  raw
ore (Station 1) and mill  solution (Station 2) compared to the  1183
tons per day  leaving  these units. In the yellowcake  precipitation
stage  of the process,  57 tons per day of solids enter in the pregnant
liquor (Station 9) compared to the 49  (2 + 47) tons per day leaving as
rellowcake (Station 10) and recarbonated barren solution (Station 16). In
terms of theover-all  mill balance 894 (881  + 12 + 0.5)  tons per day
enter the plant as raw ore (Station 1),  return water (Station 14)
and well water (Station 13)  compared to the 897 (895 + 2) leaving
the plant via the tailings pond (Station 7) and as yellowcake pro-
duct (Station 10).  A balance at the thickeners shows 1304 (1183 +
85 * 36) tons per day entering from the classifiers (Station 34)
and the  primary and secondary filters (Stations 8 and  12).  This is
in good  agreement with the 1326 (966 +  57  +303) tons comprising

-------
                                CARBONATE LEACH PROCESS
 the slurry to leach tanks (Station 5), the pregnant liquor (Station 9)
 and the mill solution (Station 2) in that order.

     The gross radioactivity concentrations determined in the lab-
 oratory are  listed in Table 4.  These data demonstrate how the
 leaching process affects the activity of the sample.  Between
 Stations 5 and 6  (the leaching process) the  undissolved activity
 decreases while the dissolved activity increases.  The effect of
 particle size is also shown in this table. At those stations where
 the large solids  had already been filtered and only the fine par-
 ticles remained,  the activity was several times greater than that
 of  the unfiltered solids when considered  in terms of ppc/g of dry
 undissolved  solids.
  Table 4. GROSS RADIOACTIVITY CONCENTRATION'S
Station
1
2
34
5
6
7
8
9
10
11
12
13
14
16
17
Total Sample (Slurry) Radioactivity, ppc 1
Undissolved
Alpha

2.580
662.000
2.300.000
2.200,000
1. 100.000
4,830
3. 630
-
b
16. 500
b
b
4.420
25.200
Beta
-
5.100
676.100
2,810.000
2.660.000
l.OBO. 000
9.940
7.690
-
b
26. 500
-
-
8,430
66.100
Dissolved
Alpha
i
393,000
456.000
386.000
554.000
5.720
902.000
637.000
a
12.600
41.400
2.6
11.600
27.700
5.430
Beta
d
I, 110.000
1. 190.000
1.030.000
1.760.000
17,500
2.120.000
1,730.000
•\
28,400
122.000
41
41.600
98.400
23.400
Radioactivitv in dry
undissolved solids, ^/*c-g
Alpha
3.300
9.910
2.630
2.870
2.700
2.150
7.780
10.500
164.000
-
8. EDO
-
-
10.200
112.000
Beta
3.300
19.400
2.660
3.520
3.230
2.100
15,700
21,400
374,000
-
14,200
-
-
19.300
281,000
  a Solid Sample.
  b Liquid Sample
    Gross alpha quantities are presented in Table 5.  The flows
computed in Table 2 have been combined with the assay of Table  4
to trace the alpha activity through the various mill processes.
The figures for the alpha balance are a function of flow and as
such would be expected to vary with production.

    The balance at the ball mill  was good with  3727 (2630-i- 1097)
me/day entering as raw ore (Station 1) and mill solution (Station  2)
compared to the 3500 me/day at  Station 34.  The effect of the
leaching process is illustrated between Stations 5 and 6 where the

-------
HOMESTAKE - NEW MEXICO  PARTNERS COMPANY
65
           Table 5. ALPHA ACTIVITY IN PROCESS STREAMS
Station
1
2
34
5
6
7
8
9
20
12
13
14
16
17
Gross Alpha Radioactivity, me, day
Undissolved
2.630
7
2.060
2.310
2. 140
1.730
3
2
300

-
-
2
12
Dissolved
-
1 090
1.440
450
550
11
600
273
-
17
- 0
9
12
2
Total
2.630
1.097
3.500
2,760
2.690
1.741
603
280
300
24
0
9
14
14
dissolved alpha activity increased as the undissolved activity en-
tered solution.  A comparison of these two stations shows 2760
me/day entering the leaching stage  and 2690 me/day leaving at
Station 6. Table 5 also shows that 280 me/day entered the uran-
ium extraction stage of the process while 314  (300 + 14) me/day
were accounted for in the yellowcake product and the barren solu
tion.  A total of 4127 (3500 + 603 + 24)  me/day entered the thick-
ening stages from  the classifier (Station 34), and the filters
(Stations 8 and 12), which is in excellent agreement with the 4137
(2760 +  1097  + 280) me/day leaving for the leaching process(Sta-
tion 5) as mill solution (Station 2) and as the pregnant liquor
(Station 9).

    The over-all plant balance required further study since the
2639 (2630 + 9) me/day entering with the raw  ore and the  return
water was in relatively poor agreement with the 2041 (1741 •*•
300) me/day in the slurry to the tailings pond  (Station 7) and in the
yellowcake product (Station 10). A  close investigation of the pro-
cedures used in collecting and analyzing the samples revealed that
the discrepance could have been due to the manner in which the
samples are collected at Station 7.  At this station a portion of the
flow is collected in a large drum.  To obtain a sample for analysis
the contents are mixed and the tank is completely drained from
below.  A sample bottle is passed through the stream every few
seconds until the drum is emptied.  Because the heavier particles
would tend to settle first at a higher flow, the sample would con-
tain a disproportionately greater quantity of large solids  than the

-------
 66
CARBONATE  LEACH  PROCESS
          Table 6. RADIUM - 226 CONCENTRATIONS
Station
1
2
34
5
6
7
8
9
10
11
12
13
14
16
17
Radium - 226 in slurry ,u>ig, I
Undissolved
a
890
124.000
427.000
463.000
225.000
1.410
270
a
2.800
7.900
b
b
170
530
Dissolved
a
8, 160
9.820
5.620
7.830
100
19,600
15.200
a
2.4
2.400
0.4
160
990
5.4
Radium - 226 in dry
undissolved solids.
WS £
530
3.450
490
540
570
490
2.150
740
3.490
26.400
4.250
-
-
390
2.260
           Solid Sample.
         b .Vecligible Solids Content.

actual flow.  The larger particles usually contain less activity per
gram than the smaller particles, and we would,  therefore, expect
to find a discrepancy of the type involved.  This was evident pre-
viously in Table 4. Thus,  while the activity at Station 7 was cal-
culated at 1741 me/day,  it was more probably in the  neighborhood
of 2200 me/day.  Table 5 also illustrates that the majority of the
alpha activity remained undissolved through the process, and was
discharged to the tailings pond with the solids.

    Samples to be analyzed for Ra-226 were prepared at the Taft
Center and forwarded to a private laboratory for these determina-
tions.  Preparation entailed separating the solids from the liquid
with membrane filters, and grinding the solids portion to less than
100-mesh. The undissolved activity is in the solids remaining on
the membrane filter: the dissolved activity  is in  the liquid portion
of the sample passing through the filter. Table 6 presents the
results of the laboratory analyses for radium per liter of  slurry
and per gram of dry undissolved  solids. The finer particles ap-
parently have a greater activity than the coarser particles.
    When the laboratory results  of Table 6 are combined  with the
flow balance  of Table 2 a radium balance throughout  the mill may
be computed.  This is shown in Table 7 and Figure 3. A balance

-------
HOMESTAKE - NEW MEXICO PARTNERS  COMPANY
67
OPE
STOCK
PILF


CffUSHIKG
AND
SAMPLING
          Figure 3. Schematic flow diagram for Radium-226 at the Homestake-
                  New Mexico Partners miJI, September 1959.
           Table 7. RADIUM - 226 IN PROCESS STREAM
Statira
1
2
34
5
fi
7
S
9
10
12
13
14
16
17
Ridi'-m-226 rng day
Undissclvcd
420
2.5
394
432
449
395
0.90
0.12
6.4
3.2
-
-
0.08
0.23
Dissolved
a
22.5
31
6.0
7.7
0.16
12.5
6.6
a
- 1.0
0
0.18
0.44
- 0
Total
423
25
425
438
457
335
13
6,7
6.4
4.2
- 0
0.18
0.52
0.23
              S.-.lid San-pie

-------
 68                             CARBONATE LEACH  PROCESS

             Table 8. CHEMICAL CONSUMPTION
Chemical
Cyanide (50^ CN~)
Caustic (100^ NaOH)
Soda Ash
Jaguar
Separan
Pounds added per ton
of ore processed
0.82
2.84
2.65
0.052
0.023
was calculated for each cycle and then averaged at each station.
Thus the figures represent the quantities of radium in the process
while the mill was processing an average of 880 tons of ore per day.
     Table 7 and Figure 3 indicate the radium balance between
stations is good in most cases.  The 445 (420 + 25) mg/day of
radium entering the ball mill with the raw ore and the mill solu-
tion agrees well with the 425 mg/day at Station 34. The uranium
precipitation stage only contains a small  amount of Ra 226- the
6.7 mg/day entering at Station 9 are accounted for as  6.9 (6.4 +
0.5) mg/day in the yellowcake and the recarbonated barren solu-
tion.  The 457 mg/day of Ra 22^ leaving the leaching  stage at Sta-
tion 6 is slightly higher than expected though still in line with the
other stations having high  solids content.   A total of 442 (425 + 13
-t- 4.2) mg/day enter the thickener units from the  classifiers and
the filters as compared to the 470 (438 +  25 -t- 6.7) mg/day leaving
these units for the leaching stage (Station 5) as mill solution
(Station 2) and pregnant liquor solution (Station 9). The over-all
plant balance shows 420 mg/day in the raw ore (Station 1) as a-
gainst 402 (396 -1-6.4) mg/day in the slurry to tails (Station 7) and
the yellowcake product (Station 10).  Because the Ra  22^ at Stations
13 and 14 is negligible for balancing purposes,  again it would
appear that Station 7 has less activity than would actually be ex-
pected. In all cases mentioned above  the discrepancies  in the
radium balance were less  than 10 per cent of the quantity being
considered.
    Most of the radium remains undissolved through the mill, al-
though dissolved radium concentrations build up at several process
locations, due to recycling of mill solution.  If the ore entering
the process at Station 1 is  considered to be in radioactive equili-
brium,  the amount of radium entering with the ore each day can
be computed from the assay for UgO   On the basis of the average
assay of 0.186 per cent V^o during the survey period, the rad-
ium entering was calculatedas 420 me/day,  in exact agreement
with the value calculated from the radium analysis. On this same
basis,  if it could be  assumed that all of the daughters of U 238 are
in radioactive equilibrium  with the parent isotope, we would expect
the gross alpha activity entering the process to be about 3, 300 me
day. Radon-222t however, is a gas; hence a portion of it will be
lost to the atmosphere in the ore body during mining  and handling

-------
HOMESTAKE - NEW MEXICO PARTNERS COMPANY        69

and milling,  thus partially breaking the chain. As a result, the
observed value of 2,630 me/day in Table 5 is within the range
that would be expected.

    Of the 420 mg/day of Ra ^26 entering the mill in the raw  ore,
6.4 mg/day,  or 1.5 per cent, leave with the yellowcake. This is
substantially more than has been found in mills using the acid
leach process.

    Chemicals added to the process at the time of the survey are
listed in Table 8.  On a yearly basis, the plant uses caustic at a
rate  of 17 to 22 pounds per ton of ore processed and soda ash at
a rate of 0 to 5 pounds per ton, with an average of 20 pounds  and
1 pound per ton. respectively.  In addition, the use in the recar-
bonization tower of carbon dioxide from the flue gas must be
recognized as part of the  chemical consumption.   Analysis of  the
filtrate  at Station 7 provided the following data in  terms of mg/1
of filtrate:
               Arsenic          -     0.20 mg, I
               Chlorjda         -     275 mg 1 Cl"
               Phenol Aklalinity    -     900 mg 1 as CaCO.
               Total Alkalinity     -     4050 rcg; 1 as CaCOj
               Sodium           -     2950 rr.g/1  Na+
               pH             -     9.6

    These figures, with the exception of pH,  could be expressed
in terms of slurry flow by reducing the concentrations by 20 per
cent.
                       Waste Disposal

    As the tabular data indicate, 285 gpm of slurry entered the
tailings  ponds during the survey period.  Of this,  54 gpm were
undissolved  solids and the remaining 231 gpm were liquid.  The
solid portion contributed an estimated 396 milligrams of Ra 226
per day  to the pond and the liquid phase only 0.16  mg/day.  The
return water from the pond (Station 14) was slightly higher  in
Ra-226 concentration,  probably due to leaching and evaporation
in the pond.  As the flow diagrams indicate, the liquid returned at
Station 7 is mainly a combination of repulping water from Station
14 and fresh water from Station 13.
    The wastes are  contained within the tailings pond and there
are no surface waters within the vicinity: hence,  surface water
pollution is not a problem. The possibility of ground water pollu-
tion, however, cannot be  ignored.
    A number of prixrate wells are in use in the area and ground
water is used for domestic consumption,  for  livestock,  and for ir-
rigation. In addition, the towns of Milan  and Grants, a few miles
south of the  mills, have municipal supply wells.
    Three aquifers are present in the area:  Silt,  sand and gravel
of the alluvium: interbedded clay, siltstone and sandstone beds in

-------
 70                             CARBONATE  LEACH  PROCESS  j

 the Chinle formation: and the San Andres limestone.  Domestic
 and stock wells usually obtain sufficient water from the  alluvium or
 from the Chinle formation. Water levels in the  general  area indi-
 cate that the water in the alluvium occurs under water table condi-
 tions and moves southwestward from the mill sites into  the valley
 of the Rio San Jose. It then moves down the valley in a southeast-
 ward direction.  The alluvium is in contact with the San  Andres
 limestone along the southwest edge of the valley, west of Milan.
 and at these points water in the alluvium can enter the limestone
 formation.  Water in the San Andres limestone is under  artesian
 pressure and moves in a  general easterly direction from the mills.
 The level and movement of water in the  Chinle formation has not
 been determined definitely. At the mills the depth to water in the
 alluvium is thought to be  about 80 feet.  The depth to the San Andres
 limestone is 600 feet but  the artesian pressure  raises the water
 level in  this formation to 130  feet below the land surface. ^

     Samples from test  wells  in the vicinity of the mill were analy- ;
 zed for radium-226. The results  (0.8,  1.8, 0.7, 4.5. 9.5, and 3.1  <
ftpc/l) are shown on Figure 4, a map of  the area.  Radioassays     •
 on the solids in the samples from the north and  the east  test wells  <
 indicated radium 226 concentrations of 551 and  1, 685/i/ig/g of      j
 solids, respectively. These concentrations are higher than in the  j
 ore feed to  the mill. Additional samples were analyzed  from wells j
 several  miles to the west of the mills  and from  wells between      i
 Grants and  San Rafael.  Results of these analyses indicate a radiuir
 content of from  0.1 to 0.4/i/zc/l.  or the usually expected natural
 concentrations.                                                  •

     If monitoring of the test wells at the mill should indicate a
 buildup or spread of the radium-226 consideration may have to be
 given to  ways and means of sealing the ponds and preventing fur-
 ther buildup.  This  is suggested in view of the facts that the radiur
 concentration of the discharge to the pond (Station 7) was measured
 as 124/i^ug/l of liquid and, at  the time the  wells were sampled, the
 mill had been in operation less than two years.

                       Acknowledgment
     The  generous cooperation and technical assistance of the fol-
lowing are gratefully acknowledged: John Hernandez,  New Mexico
Department  of Public Health: personnel of the Homestake-New
Mexico Partners Mill: the U. S. Atomic Energy  Commission:
Charles  E. Sponagle. Public Health Service,  Region VIII, Denver.
Colorado: E. A.  Pash, Carl Hirth,  Carl  Shadix.  and H. D. Nash.
Public  Health Service,  Cincinnati,  Ohio: and D.  E. Rushing. Publit
Health Service. Salt Lake City,  Utah.  The technical advice and
guidance of  Dr. E. C. Tsivoglou.  Public  Health Service,  Cincinnati.
Ohio, was greatly appreciated. This study was  supported by funds
made available through the Environmental and Sanitary Engineer-
ing Branch,  Division of Reactor Development, U. S. Atomic Energy
Commission.

-------
HOMESTAKE - NEW MEXICO  PARTNERS COMPANY
                                     71
      -- *,-«
                             	NOSI"_::-,V.t^4- —, - iSr
                             •••  • "-'^i!L^i:-r£\  >   <
                                    v-^7  ^^^ -m v-'^, /»
                                            'tr.l'.^W^'"

                                             5ai --S^^ 'as
WflTE"

*ELL
                                                 ,-\'xi^;,/
                                                  ""' "^ * •''
                                    TA-L'iGs OCND   ' '*•;   »//

                                            ^"*.j
                                                   • ,G*rc

                                                   L^ ^O

                                                    BALL fw*v
                                           HOMESTflKE-PARTNERS

                                               MILL-SITE
                                         ezc  ADC  c
         Figure 4.  Radiym-226 in water samples from test wells at Homesfake-

                 Partners mill-site.

-------
 THE CARBONATE  LEACH  URANIUM EXTRACTION
                          PROCESS
      II.  HOMESTAKE-SAPIN PARTNERS,  GRANTS,
                      NEW MEXICO

                        H. R. Pahren*
                       M.  W. Lam me ring
                        J. Hernandez

                         Introduction

     This report presents the results of a study conducted at a
 second uranium refinery employing the "carbonate leach extrac-
 tion process.  The study was performed by the U. S.  Public Health
 Service and the New Mexico Department of Public Health during
 September 1959, with the cooperation of personnel of the Home-
 stake-Sapin Partners mill and the U. S. Atomic Energy Commis-
 sion.

     The Homestake-Sapin mill with a design capacity of  1500
 tons  of ore per day was placed in  operation during the latter half
 of 1953.   Located about  ten miles  northeast of Grants. New
 Mexico,  (adjacent to the Homestake-New Mexico Partners mill)
 it lies in an arid area with  no surface water in the near vicinity.
 At the time of the  survey, the mill was processing about 1640
 tons  per day of dry ore. which assayed from 0.143 per cent to
 0.235 per cent IL08.  Recovery of U^O,, averaged 90 per  cent
 with  a yield of 5uOO pounds per day of yellowcake. The plant  ef-
 fluents were discharged to  a tailings pond, where all but a small
 liquid portion (recycled as  process water) was retained for dis-
 sipation  by evaporation  and possibly  seepage. Figure 1 is a de-
 tailed flow diagram of the refinery.  The units are described  in
 detail in the following section.

                     Process  Description
 ORE  PREPARATION

     Ore  is brought from the mines by truck and is initially stored
 outdoors on a storage pad.  A bulldozer transfers the ore to a
 feeder from which it is carried by belt conveyor to a jaw crusher.
 If the ore contains more than about 10 per cent moisture it is
 passed through  a gas fired rotary kiln dryer: ores with less  than
 10 per cent moisture by-pass the  dryer.
"Respectively.  Sanitary Engineer.  Colorado River Basin Water
 Quality Control Project.  Public Health Service. Denver, Colorado:
 Senior Assistant Sanitary Engineer, Radiological Pollution Activi-
 ties Unit: and  Associate Engineer. Environmental Sanitation Ser-
 vices. New Mexico Department of Public Health. Santa Fe.
                              73

-------
 74
                                 CARBONATE  LEACH PROCESS


          Figure 1. Flow diagram of Horrtesrake-Sapin Partners uranium mill.
                 Grants, New Mexico, September 1959.

     Next, the ore is sampled to ascertain the uranium content for
 each lot. The ore passes through a sample cutter which diverts
 10 per cent of the ore stream.  The 10 per  cent sample  then passe-
 through  a second cutter which takes another 10 per  cent sample.
 After four such cuts,  a sample of 0.2 pounds per ton of  ore pro-
 cessed is obtained. The balance of the ore  is approximately
 equally distributed in four fine-ore bins  of  1500 tons capacity
 each.
     In the process following the fine ore storage there is a dupli-
 cate circuit through the plant.  This duplication of facilities con-
 tinues until the pregnant solution streams from each circuit are
 combined.

     Two of the fine ore bins feed to each circuit.  Ore fed to
process  is first weighed by an automatic weightometer and then
is added to a ball mill along with mill solution to make a relatively
thick slurry.  Following this, the slurry flows to a spiral classi-
fier where oversize material is separated and returned  to the
ball mill. Additional mill solution is added  to the classifier  to
give an effluent with a specific  gravity of 1.20. Approximately 45
per cent  of the solids in the  classifier effluent are minus 200 mesh
and 12 per cent are plus  65 mesh.

     The  classifier effluent is pumped to a 100-foot diameter pri-
mary thickener where the solids are concentrated in the bottom
and the overflow is returned to mill solution storage. Separan is
added  in  a one per cent solution to facilitate settling.
                                                      GFO 813—I 73-"

-------
HOMESTAKE - SAPIN PARTNERS                           75

LEACHING

    The underflow from  the primary thickener is pumped by a
diaphragm pump to the first of eight leach tanks in series. The
retention time in each leach tank is about one hour.  Under leach-
ing conditions of approximately 225° F. and 60 pounds pressure,
most of the uranium remaining in the solids is leached out.

    Leaching solutions contain carbonate and bicarbonate ions,
and under the process conditions, the soluble uranyl tricarbonate
ion, UOr.(COo)Q   , is formed.
       ^    o o
    After leaching, the slurry flows to a 100-foot diameter
secondary thickener where additional separan is added to"aid in
settling the suspended solids.  The secondary thickener overflow,
known as pregnant solution, from each circuit is combined. The
pregnant solution is clarified by means of a rotary filter,  and then
is stored in the clarified-pregnant-solution storage tank.  The
small amount of solids removed by the filter is slurried and re-
turned to the  secondary thickener. Underflow the secondary
thickeners is pumped to the tailings filtration area.

PRECIPITATION AND PREPARATION OF YELLOWCAKE

    Clarified pregnant solution is pumped to the first of seven
agitated precipitation tanks operated in series.  Here,  sodium
hydroxide is added as a 50 per cent solution until the pregnant
solution contains 8 grams of NaOH per liter. With the increase of
pH, the uranium precipitates as sodium diuranate or yellowcake,
as follows: ^

2 Na4U02(C03)3+ 6 NaOH—>Na2U20?+ 6 NagCOg-s- 3 HgO

    The  yellowcake slurry is transferred to a thickener to which
a solution of locust bean  gum, a natural polysaccharide, is added
to aid in settling the yellowcake.  Overflow from the thickener is
filtered through a plate and frame filter press to remove the small
amount of yellowcake that is carried over,  and the clarified barren
solution istthen passed through a recarbonation tower.   Flue gas
from the boiler plant is passed countercurrent to the barren solu-
tion in the tower, and the carbon dioxide in the flue gas neutralizes
the caustic alkalinity and forms  additional carbonate. The recar- ,
bonated barren solution is then available for re-use in the process.

    Underflow from the thickener contains the yellowcake, which
is filtered out with a rotary filter. The yellowcake is repulped
with fresh, softened water and refiltered, and the filtrate  is added
to the yellowcake thickener. The resulting filter cake is washed,
dried, pulverized and then drummed for shipment to plants of the
U. S. Atomic  Energy Commission. Yellowcake is filtered  and
packaged only during the  day shift.

-------
 76                             CARBONATE LEACH PROCESS

 FILTRATION OF TAILINGS

     The tailings filtration area consists of three stages of five
 rotary filters each.  The combined underflow from the two secon-
 dary thickeners passes through the three filtration stages and the
 pregnant solution that carries over is separated from the solids
 before the solids are discharged to the tailings pond.  Filtrate
 from the first stage filters is added to the secondary thickener,
 while filtrate from the second stage filters is pumped to the mill
 solution storage tank.

     Filter cake from the first stage filters is washed by part of
 the third stage filtrate.  The balance of the third stage filtrate is
 used to repulp the filter  cake prior to  the second stage filtration.
 About one-third of the recarbonated barren solution is used to
 wash the filter cake on the second stage filters. The remaining
 recarbonated barren solution is used for repulping between second
 and  third stage filtration.  Third stage filter cake  is washed by
 reclaimed water from the tailings pond and, after removal from
 the filter drum, is repuloed with unsoftened raw water.  This
 slurry passes through an automatic sampler. After passing throus;.
 the sampler, additional water is added before the  slurry is pumpen
 to the tailings pond.   Mill personnel indicated that the additional
 water consisted of approximately 10 gpm from the reagent build-
 ing,  40 gpm of cooling water from the vacuum pumps  and 25 gpm
 of excess  return water from the tailings pond.

                         Tailings  Pond
     The tailings pond is circular in shape and has 60  acres within
 the dikes.  Water covered only 39 acres of this area at the time of
 the survey. In addition to the main pond there is a smaller pond
 used as a  surge pond for the tailings pond  water that is returned
 to process. Return water is decanted  from the main pond.  In
 addition to the process water, boiler plant blowdown water and the
 plant domestic  sewage are added to the tailings pond.  The sewage
 is first treated in a septic tank.

                     Sampling  Procedures

 Samples were obtained during two 72-hour cycles as follows:

     Cycle I  -  8 AM Sept. 17, 1959, to 8 AM Sept. 20. 1959:
     Cycle II -  8 AM Sept. 20, 1959, to 8 AM Sept. 23, 1959
     At all but three  sampling stations, a sample was  obtained for
each cycle by compositing equal volumes every 2 hours for the  dur-
ation of the cycle. Thus, 36 portions were used to make up the
single composite sample.

     The wastes going to  the tailings pond were sampled by an
automatic  sampler operated by mill personnel.  The Public Health
Service obtained a portion of the  sample collected by this auto-

-------
HOMESTAKE - SAPIN PARTNERS                          77

matic sampler each day and composited the three equal portions
for each cycle.

     A 1-day lead time was used for the raw ore samples. A por-
tion of the official sample from each ore lot was obtained from
that collected by the plant sampling equipment before the ore was
discharged to the fine ore storage bins. To make the composite
sample for each cycle, the portion taken from each ore lot pro-
cessed was weighted according to the tonnage of the lot.  For the
ore samples,  the period from 8 AM Sept.  16 to 8 AM Sept. 19 was
Cycle I and the period from 8 AM Sept. 19 to 8 AM Sept. 22 was
Cycle II.

     Composite yellowcake samples were obtained for each of two
consecutive  3-day periods beginning September 18.  thus giving a
1-day lag period.  Representative samples of each day's produc-
tion were weighted according to tonnage of production for the day
and composited for the cycle.  Yellowcake was not packaged on
September 20, the third day of Cycle I. Thus, the production on
September 21 was equally divided between both cycles to compen-
sate for this  variation.

     As mentioned in the process description, there are duplicate
circuits in the plant from  the beginning of the process through the
secondary thickener step.  During the survey individual samples
were obtained at significant points for  both the north and south
circuit.  Samples from the same point  in each circuit vvere then
combined. A list of all the sampling stations is  shown in Table 1.

                      Sample Processing

     The samples collected during the survey were shipped to the
Public Health Service's Robert A. Taft Sanitary  Engineering
Center, where all chemical and physical analyses were performed,
except the analysis for radium-226.

    Although the  results presented in the  following section are
average values for Cycles I and II,  analyses for dissolved and un-
dissolved gross alpha and beta activity, dissolved and undissolved
radium-226,  sodium (Na+),  and pH were performed on individual
samples from each cycle.  Thus, each station is associated with
two results for each applicable analysis.  Supplementary analyses
for arsenic,  chloride,  phenolphthalein alkalinity, and total alka-
linity were performed on the liquid portions of Station 11 and 13
samples.

    The radium analyses  were performed by a private laboratory
under contract to the Public Health Service. Pretreatment of the
samples by the Taft Center laboratory consisted of liquid-solid
separation by filtration through a membrane filter,  and grinding
the dried  suspended material to less than  100 mesh.

-------
78
         CARBONATE  LEACH PROCESS
           Table 1. SAMPLING STATIONS
           Station Number
               9

              10

              11

              12

              13

              14
                              Description
Raw ore

Mill solution

Overflow from classifiers

Overflow from primary thickeners

Underflow from primary thickeners

Effluent from digesters

Underflow from secondary thickeners

Pregnant solution to precipitation tanks

Recarbonated barren solutions

Yellowcake product

Repulped third stage filter cake to tailings pond

Return water from tailings pond

Unsoftened well water

Softened water
                      Discussion of Results

    Table 2 through 8 show average values obtained from the re-
sults  of Cycle I and II. This technique of averaging over the two
72-hour periods was used because of the close agreement between
analytical results for the two cycles, and to further minimize
possible errors associated with representative sample collection
and retention times, such as the 8-hour retention across the
leaching circuit.  The material balances were developed from a
combination of field data (ore tonnage,  yellowcake production,
flows, etc.) extracted from mill records and laboratory analyses.
The balances characterize each unit in the carbonate-leach pro-
cess  and  form the basis for comparison between varying ore  feed
rates and ore quality.
     Throughout the survey period, the mill operated somewhat in
excess of design capacity.  The ore was fed at the rate of 1590 dry
tons  per  day during Cycle I, and 5130 pounds per day of yellowcake
was packaged.  Corresponding figures for Cycle n were  1686 dry
tons  per  day and 4250 pounds, respective. The general physical
characteristics and the concentration of sodium ion for the in-plant
process streams, waste effluent, and raw water supply are pre-
sented in Table 2.  Concentrations of dissolved  and  suspended solids
are presented in  Table 3.
     In Table 2, the data in Columns (3) through (7)  are laboratory
determinations.  The slurry flows (Column 2) except for Stations
8,  12, and 13,  are calculated values.  The flow at Station 9 was
metered  at an average value of 232 gpm for the survey period,
                                                       GPO 813-173-8

-------
HOMESTAKE  - SAPIN  PARTNERS
 Table 2. PROCESS STREAM CHARACTERISTICS a
79
Station
(1)
1
2
3
4
5
5
7
j,
9
10
II
12
13
14
Slurry flo*.
gpm
(2)
b
1501
1665
1291
357
361
316
236
241
b
625
eo
399
3.5
Slurry specific
gravity
(3)
_
1.12
1.19
1.12
1.50
1.50
1.56
1.11
1.13
-
1.26 d
1. 01
1.00
1.00
Dry suspended
solids by weight ^
(4)
_
c
13.9
c
51.2
50.2
54.6
c
c
-
34.6 d
C
c
c
Specific gravi
of dry solids
(5)
.
-
2.60
-
2.46
2.59
2.50
-
-
-
2.40
-
-
-
ty / Na
/ rr.g; 1 in slurrv
' (6)
_
50.3 x I03
39.2 x 103
44.5 x 103
24.2 x 103
25.1 x 103
23.6 x 103
51.5 x 103
53.3 x 103
-
4.45 x 103
5.15 x I03
3. TO x I03
4.45 x 103
pH
(7)
-
10.5
10.4
10.3
10.3
10.2
10.1
10.1
10.6
-
10.1
9,9
7.5
6.9
   Average "f Cycles I and II.
   S">hd sample.
   Liquid sarrple (negligible solids).
   Based on Cycle I only.
           Table 3. CONCENTRATION OF SLURRY SOLIDS
Station
(1>
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Dissolved solids.
mg 1 slurry
(2)
b
138.000
12S.OOO
137.000
87.900
94.500
91.300
159.000
162.000
b
12.200
14.500
1.860
1.700
Suspended solids.
rr.g 1 slurry
(3)
b
850
165.000
7SO
765.000
752.000
854.000
145
50
b
435.000
193
0
56

-------
 80
CARBONATE LEACH PROCESS
 which is in good agreement with the calculated value of 241 gpm.
 The field notes indicated, however, that the average metered
 flow would probably be low  because of marked fluctuations in line
 flow for a period of about 8 hours during Cycle II. In the  same
 8-hour period softened water was added to the recarbonation
 barren tank to maintain suction.  Fresh water was substituted
 for recycled water (Station  12) during the last half of Cycle II.
                                                       10 gfl
                                                      no llm
                                                     10.- O
               TO TAILUGS PCND
           Figure 2. Schematic flow diagram of the Homestake-Sapin mi!!,
                  September 1959.

     Figure 2 illustrates the basic flow diagram of the mill in
schematic form. The small differences in estimated slurry flows
across process units are considered to be insignificant. The
relatively small discrepancies in solids flow across a unit pro-
bably resulted from the limits of sensitivity of the dry-solids
specific gravity analysis.  Theoretically,  if a constant feed rate
and a steady state process throughout the mill are assumed the
solids flow would be expected to remain unchanged throughout  the
mill because the decrease in undissolved  solid tonnage is slight.
Discrepancies in slurry flow are the result of the experimental
errors that entered into the  total weight and solid balances in the
laboratory determinations.   As will be seen,  these discrepancies
are not of major importance in this study.

-------
HOMESTAKE - SAPIN PARTNERS
81
     Figure 2 also indicates that material balances can be made
across grouped process units. Proceeding from the ore storage
bin, (Figure 2) the first balance quantitatively equates the input to
the ball mill and classifier circuits of ore (Station 1) and mill
solution (Station 2) with the output from the classifiers (Station 3).
The output from the classifiers should also be accounted for in the
overflow (Station 4) and in the underflow (Station 5) from the pri
mary thickeners.  Quantities present in the underflow from the
primary thickener (Station 5)  should be measurable in the effluent
from the leaching  circuit (Station 6).  The recarbonated barren
solution (Station 9) and yellowcake  production (Station 10) should
approximate the pregnant solution to the precipitation tanks
(Station 8) and the caustic stream.   As an over-all plant balance,
the ore feed (Station 1), caustic stream, return water at the third-
stage filters (Station 12),  and untreated well  water for tails re-
pulping (Station 13) should approximate yellowcake production
(Station 10) and the effluent to the tailings pond (Station 11).
      Table 4. SOLIDS QUANTITIES a
Station b
(1)
1
2
3
4
5
6
7
e
9
10
11
12
13
14
Dissolved solids.
tons dav
(2)
-
1.250
1.260
1.050
189
204
174
224
235
-
46
5.8
5.0
- 0
Suspended solids.
tons dav
(3)
1.643
8.0
1.640
6.5
1,640
1.620
1.620
-0
-o
2.5
1.620
-0
-0
-0
Total solids.
tons, day
(4)
1.640
1,253
2.900
1.067
1.829
1.S24
1.794
224
235
2.5
1.666
5.8
5.0
-0
        Average of Cycles I and II.
        Approximately 18 tons per day of dissolved solids are added to the precipitation
        tanks (caustic stream)

    Table 4  presents the average solids quantities of both cycles.
If the balance points listed above are used, it is apparent that the
materials balance (Figure  3) is quite good.  The raw ore feed
(Station 1) and mill  solution (Station 2) comprise  an input of 2. 898
(1. 258 + 1, 640) tons per day to the classifier circuit as compared

-------
 82
CARBONATE LEACH PROCESS
                                          DISSOLVED SOLID? - 1C
                                          SISPEKPED SOLUS - IOC
                                          TOTAL SOLIDS   - IIC
                                          SAMPLING STATION »0. -
                      TO»S'P»Y
                      TINS'OY
         Figure 3.  Schematic flow diagram of solids balance, Homestake-Sapin
                 mill, September 1959.

to the measured output  of 2, 900 tons per day (Station 3).  The out-
put of 2,900 ton's per day from the classifiers is also balanced by
2,896 (1,067+ 1,829) tons per day from  the primary thickener
overflow (Station 4) and underflow (Station 5).

     In good agreement with the 1, 829 tons of solids per day in the
slurry to the digesters  (Station 5) is the 1, 824 tons per day leaving
the leaching circuits (Station 6).   Further examination of the data
shows an increase in dissolved solids of  15 tons per day across
the leaching circuit and a corresponding  decrease in suspended or
undissolved solids of 20 tons per day. Pregnant solution (Station
8) and caustic at the precipitation tanks yield 242 (224 -f- 18) tons
of solids per day as compared to 238 (235 +  2.5) tons per day in
the recarbonated barren liquor (Station 9) and yellowcake product
(Station  10).  For the process as a whole. 1,669 (1,640 -f- 18+  5.8
+ 5.0) tons of solids per day enter the refinery as raw ore (Sta-
tion 1),  caustic,  return water (Station 12),  and well water (Station
13) in that  order. This  is in exact agreement with the 1, 669 +
(1,666 + 2.5) tons per day leaving the  refinery as waste to the
tailings pond (Station 11) and yellowcake product (Station 10).

-------
HOMESTAKE - SAPIN  PARTNERS
 Table 5 GROSS RADIOACTIVITY CONCENTRATIONS a
83

Station
(I)
1
2
3
4
5
R
7
3
9
10
11
12
13
H
Suspended
Alpha
(2)
-
3.813
423.000
7.550
1.843,000
1.710.000
1.640.000
850
b
-
7 13. GOO
430
0 5
1.3
Beta
(3}
-
6.410
545.000
9.240
2.050.000
1.700.000
1.710.000
2.150
b
-
642.000
740
0.7
2.5
Dissolved
Alpha.
(4)
-
191.000
227.000
217.000
111.000
225.000
327.000
353.000
53.400
-
7,900
8.500
5.S
4.5
Beta
(5)
-
575.000
562,000
555.000
341.000
763.000
775 . 000
1,030,000
153.000
-
22. 100
23. 100
30
20
Dry suspended solids
Alpha
>>MC g
(6)
3.070
4.430
2.590
9.6JO
2,400
2.280
1.923
5. 670 c
b
317.000
1.650
2.120
b
23 C
Beta
•MfC- 8
(7)
3.340
7,540
3,310
11.600
2.670
2.260
2.000
14. 300 c
b
420.000
1.930
3.720°
b
45 r
   Average rf Cycles I and II.
    Gross radioactivity concentrations are presented in Table 5.
These data show a consistent decrease in the gross alpha activity
of one gram of undissolved ore solids as it moves from the ore
storage bins through the digesters and ultimately to the tailings
pond.  Leaching of the uranium from the ore solids at the digesters
is illustrated by the twofold increase in dissolved alpha activity
between Stations 5 and 6 and the  small decrease in the suspended
solids radioactivity.  The comparison of the alpha activity (u/ic/g)
of the heavy settleable solids (Stations 3, 5, 6, 7, and 11) with
that of the smaller  size suspended particles that are characteris-
tic of thickener overflows  (Stations 2, 4, and 8) was in agreement
with observations from previous mill surveys; that is, the alpha
activity of the smaller particles  (slimes) exceeded that of the
sands by factors of about 2 to 4.
    If radioactive equilibrium in the ore among uranium-238,
radium-226 and the  other daughter products is assumed, the alpha
activity of the dry suspended  solids should  remain constant across
the ball mill and classifier circuit. The observed decrease  at
Station 3 compared to Station 1 may be due in part to leaching and
a resulting loss of the third member in the decay chain, radon-
222,  to the atmosphere as a gas.  That further leaching occurs as
far as the secondary thickener is illustrated by the  increase  in the
dissolved alpha activity of the pregnant liquor (Station 8) in com-
parison to the dissolved alpha activity of the digester effluent
(Station 6)

-------
 84
CARBONATE LEACH PROCESS
     The gross alpha concentrations of Table 5 and the balanced
 slurry flows of Table 2 form the basis for the gross alpha quanti-
 ties presented in Table 6.  If previously mentioned balance points
 are used, the gross alpha balance appears to be adequate. The
 6,170 (4, 570+1, 600) me/day input to the classifiers from the ore
 (Station 1) and mill solution (Station 2) is in  good agreement with
                  Table 6. GROSS ALPHA QUANTITIES3
Station
(1)
1
2
3
4
5
6
7
a
9
10
n
12
13
14
Gross Alpha Radioactivity, me, day
Dissolved
(2)
b
1 570
2.070
1.550
216
445
562
456
73
b
27
3
d
d
Suspended
(3)
4.570
30
3. 860
53
3.570
3.340
2.830
- 0
c
631
2.430
~0
d
d
Total
(4)
4. 570
1.600
5.930
1,603
3. 786
3.785
3.392
456
73
681
2.457
3
d
d
                    Average of Cycles I and 11.
                   Solid sample.
                   Not determined.
                   Negligible.
the 5.930 me/day leaving the classifiers (Station 3).  The output of
5, 389 (1, 603 + 3, 736) me/day from the primary thickeners is
split between the overflow (Station 4) and underflow (Station 5).
This agrees sufficiently well with the 5,930 me/day input to the
thickeners (Station 3),  although a definite loss of alpha activity is
noted.  The balance across the digesters is excellent: 3, 786 me/
day enter at Station 5 and the same amount leave at Station 6.
Gross alpha activity in the recarbonated barren solution  (Station 9)
that leaves  the process as yellowcake product (Station 10) totaled
759 (78 + 681) me/day. This does not agree well with the 456 me/
day in the pregnant  liquor (Station 8) and the negligible amount
contributed by the caustic.  The explanation for this discrepancy
is not readily apparent, especially in view of the fact that the
balances for solids and slurry flows at these stations were very
good.

-------
HOMESTAKE - SAPIN  PARTNERS
         Table 7. RADIUM-226 CONCENTRATION'S '
85
Station
(1)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Radium-226 in slurry, pug, I
Dissolved
(2)
-
9.560
Suspended
(3)
-
1.220
13.200 (8. 300)° 83.900
8.790
5,160
6.620
5.280
18.900
59
-
34
35
5.4
4.2
b
323.000
370,000
380,000
286
5.4
-
222,000
bb
b
b
Radium-226 in
dry suspended
solids. nn« g
(41
497
1.440
508
b
429 (500) c
492
445
1.970
107
7,190
510
b
b
b
           Average of Cycles I and n.
           Not determined (negligible solids).
         c  The parenthetical result is most probable. See discussion in text.

    The over-all plant balance also showed considerable disagree-
ment; 4, 573 (4, 570 + 3) me/day entered the process as ore (Sta-
tion 1) and recycled process water (Station 12), whereas the effluent
to the tailings pond (Station 11) and the yellowcake product (Sta-
tion 10) accounted for only 69 per cent of the input,  or 3,138 me/
day.  The over-all plant balance in the similar Homestake-New
Mexico Partners Company mill showed a 77 per cent accounting.
In that case it was suspected that the semiautomatic sampling de-
vice for collecting a tailings sample actually collected a non-
representative excess quantity of large versus fine  solids, there-
by lowering the gross alpha activity; this may in part  explain the
discrepancy noted above.  It appears probable that the alpha acti-
vity of representative dry suspended solids at Station  11 was with-
in the range of 1,900 to 2, 300M/*c/g (Stations 7 and 6) as compared
to the 1,650/i/ic/g determined for the  sample obtained.  Based on
2,100/t/jc/g. the alpha activity at Station 11 would be  increased to
about 3,100 me/day and the over-all recovery to about 83 per
cent.
    Concentrations of radium-226 per liter of slurry and per  gram
of suspended matter are presented in  Table 7.  The dissolved  rad-
ium concentrations reflect those solids passing through a membrane
filter,  and the suspended concentrations, that portion retained on

-------
 86
CARBONATE LEACH  PROCESS
 the filter.  From the data for dry suspended solids,  it is apparent
 that most of the radium remained in the undissolved form through-
 out the process; 497/i/ig/g entered in the ore feed and about the
 same concentration left in the spent ore solids (Station 11). At
 intermediate steps in the process (Stations 3,  5,  6,  and 7) the
 range  of observed concentrations in the ore solids varied between
 429 and 508/t/tg/g.  As was the case with gross alpha activity, the
 fine-grained particles exhibited the highest activity.
            Table 8. RADIUM-226 QUANTITIES a
Station
(D
1
2

3
4

C
6
7
8
9
10
11
12
13
14
Radium-226 mg/ day
Dissolved
(2)
b
78
e
130 (81)
61

14
18
13
25
- 0
b
-0
c
c
c
Suspended
(3)
739
10

755
9
e
639 (744)
724
655
-0
- 0
16
750
d
d
d
Total
(4)
739
88
e
885 (836)
70
£
653 (748)
742
668
25
- 0
16
750
c
c
c
              Average of Cycles I and II.
              Solid sample.
              Negligible.
              Not determined.
           6  The parenthetical result is more probable. See discussion in text.

     Table 7 also indicates certain discrepancies that are  impor-
tant to interpretation.  The concentration of dissoh'ed radium-226
at Station 3 (overflow from the classifiers) is clearly not consis-
tent with the dissolved radium-226 data for Station 2 (mill solution
added at the classifiers) or with the data for Stations 4 and 5,
while are locations immediately following Station 3.  Specifically,
the average dissolved radium was 9, 560/*x*g/l of liquid in mill
solution entering the classifiers (Station 2), 8, 800/i/ig/l in the
overflow from the  primary thickeners (Station 4) that follow the
classifiers,  and 10, 100/i/ig/l in the liquid portion of the underflow

-------
 HOMESTAKE - SAPIN PARTNERS
87
                                         < E Y:
                                       DISSOLVED PAOU'M-226  - 10 mg/d»
                                       SUSPEKKD R»DIW<-22«  -ICO »a'da»
                TO T»IU»GJ PC«D
         Figure 4.  Schematic flow diagram of Radium-226, Homestake-Sapin
                 mill, September 1959.

slurry leaving the primary thickeners (Station 5). In contrast, a
value was  reported of 15,300 micromicrograms of radium-226 per
liter of liquid in the slurry leaving the classifiers (Station 3) prior
to entering the primary thickeners.  This latter value is clearly
inconsistent with the  surrounding data; between 8. 800 and 10, 100
A/*g/l of liquid are required for Station 3. As will be seen, a mean
value of 9, 600/
-------
88
CARBONATE LEACH PROCESS
    Table 9.  CHEMICAL CHARACTERISTICS OF THE FILTRATE AT
          STATIONS 11 AND 13

Station
11
13

Arsenic
mg/1
0.49
-

Chlorides^
mg/1 Cl
286
250
Phenolphthalein
alkalinity
mg/1 as CaCO
1,720
5
Total
alkalinity
mg/1 as CaCO,
3,560
690
somewhat low, although this is less directly interpretable because
of the side^-routes between Station 6 and 7 (see Figure 2). As a re-
sult, the radium-226 concentration of undissolved solids at Station
5 was probably about 500^/ig/g,  rather than the reported value of
429/i/ug/g.  This would also perfect the radium-226 quantity bal-
ance, as noted below.

    No explanation for the foregoing two discrepancies  can be of-
fered at this time,  although they appear to be quite definite.

    The assumption of equilibrium in the ore between uranium-238
and its  daughter, radium-226, allows a separate check  to be made on
the concentration in the raw  ore.  If the average ore assay of 0.167
per cent UoCL for the survey period is used, a concentration of 474
micromicrograms of radium-226 per gram of ore is derived.  This
agrees  quite favorably with the analyses  average of 497^i/tg/g. If
the possibility of Radon-222  loss did not  exist,  it could  be assumed
that all the daughters of uranium-238 were in equilibrium with the
parent during the survey period.
          Table 10. CHEMICAL CONSUMPTION
            Product
           Guar Gum

           NaOH (100%)

           Separan
                                       Quantity
   25 Ibs/each three days

   30.82 Ibs/ton of ore processed

   0.05 Ibs/ton of ore processed
    Table 8 presents the radium quantities at the various process
stations at an ore processing rate of 1, 640 tons per day.  The
values in Table 8 were obtained with the concentrations of Table 7
and the slurry flows of Table 2.  As indicated earlier,  the reported
dissolved radium-226  concentration at Station 3 and the undis-
solved radium-226  concentration for Station 5 are probably erron-
eous.  The more  probable results are given as parenthetical fig-
ures in Tables 7  and 8.  If these figures are used,  the  radium
balance throughout  the process (Figure 4) is generally quite good,
as follows:
    Mill solution combines with the raw ore feed for 827  (88 + 739)
mg/day of radium-226 entering the ball mill and classifiers.  This
is in good agreement with the estimated 836 mg/day at Station 3.
Overflow (Station 4) and underflow (Station 5) from the primary
thickeners account  for 828 (70+  758) mg/day. The balance across

-------
HOMESTAKE - SAPIN PARTNERS                           89

the leaching circuit consists of an estimated 758 mg/day entering
at Station 5 and 742 mg/day leaving at Station 6.  In comparison to
the 836 mg/day at Station 3 is the 812 (70 + 742) mg/day from the
primary thickener overflow (Station 4) and the digester effluent
(Station 6).  Pregnant liquor carries 25 mg/day  to the precipita-
tion circuit.  This is in acceptable agreement with the 16 (16 + 0)
mg/day accounted for as yellowcake product (Station 10) and re-
carbonated barren solution (Station 9).

    The total plant balance shows 739 mg/day entering as  raw ore
(Station 1) and 766 leaving via the waste effluent to the tailings pond
(Station 11) and as yellowcake product (Station 10): plant input is
in good agreement with plant output.

    Based on the 739 mg/day of radium-226 entering the mill in
the raw ore,  16 mg/day,  or 2.2 per cent leave with the yellowcake
product.  This percentage  is in general agreement with the 1.5 per
cent for the Homestake-New Mexico mill but in  substantially higher
than that found in mills using the acid leach process and other  ores.

    To delineate further the characteristics of the main plant
effluent to the tailings pond and the raw well water,  a series of
five chemical tests was run on the filtrate portions from Stations
11 and 13. The average values based on 1 liter  of slurry,  are
tabulated in Table 9.

    The rate  of consumption of chemical additives as reported by
mill personnel is presented in Table  10.  Carbon dioxide,  which
was used to partially neutralize the caustic alkalinity in the recar-
bonation tower, was not listed because it was generated within the
process as flue gas.  Oxidizing agents were not added during the
survey  and had not been added to the  digesters for 3 months pre-
ceding the survey.  In addition to the  products listed in Table 10,
lime,  a small amount of soda ash, and sulphuric acid were used
in the treatment of the mill's raw water supply.

                       Waste  Disposal

    Slurry flow to the tailings pond,  as shown in Figure 2, aver-
aged about 700 gpm, of which 112 gpm was solid flow, and 588 gpm,
liquid.  The spent ore solids are  retained in the tailings pond,  but
the liquid phase,  in addition to being  retained for concentration
by evaporation, may be recycled as return filter wash water or
may drain by  seepage. The opinion of mill personnel was  that the
soil at this location provided an excellent seal,  thereby minimiz-
ing the problem of seepage into the ground water.

    The 0.12  mg/day of dissolved radium entering the tailings pond
at a concentration of 35/i/tc/l does not constitute the only  source
of possible ground water contamination.  The insoluble radium ac-
cumulating at the rate of 750 mg/day provides a radium reservoir
for leaching which, if coupled with seepage,  could also produce a

-------
 90
CARBONATE  LEACH PROCESS
 significant ground water contamination problem.  That leaching of
 radium from spent ore solids does occur has been discussed in the
 reports on the study of the Animas River by the U. S. Public Health
 Service. 10> 12
        Table 11.
              RADIUM CONCENTRATIONS IN SAMPLES FROM TEST WELLS
              ON HOMESTAKE - SAFIN" MILL PROPERTY.
Depth of well. ft.
70
80
95
-
60
Description
Cased
Cased
Cased
Observation Well
Cased
Dissolved Radium
1959
0.8
1.8
0.7
0.2
1
- 225. fipK 1
1961
0.22
0.13
0.24
-
1.79
     The Homestake-New Mexico Partners report points out the
need for considering ground water contamination inasmuch as
ranchers in the general area of the mills use well supplies for
domestic consumption,  for watering livestock, and for irrigation.
In addition, the communities of Milan and Grants, located a few
miles south of the mills, take their domestic supplies from wells.
The Homestake-New Mexico  Partners report also describes the
water-bearing strata in the area, and the direction of ground
water movement,  and summarized the radium results for num-
erous well samples.  Table 11 presents the radium concentrations
found in samples from test wells located around the periphery of
the tailings pond  on the  Homestake-Sapin mill property.  These
were collected in 1959 and 1961.

     These results may  be compared to the radium  content of wells
located several miles to the west of the mills and that of wells be-
tween Grants and San Rafael.  The concentration  range for these
wells is 0.1 to 0.4/t/ic/l. a natural background concentration range.
     An evaluation of the Homestake-Sapin  pond must  take into
account the presence of the Homestake-New Mexico Partners
pond and their combined effect.  Therefore, it seems advisable,
in view of the reservoir of radium in both tailings ponds and the
variability of results to date, that the monitoring of well supplies,
particularly in the near-vicinity of the ponds,  be continued until a
firm conclusion on the presence or absence of seepage can be
reached.  If a significant buildup of radium in the test wells is ob-
served over a period of time,  remedial measures may be neces-
sary. The need for such measures should be based on the obser-
ved rate of buildup, the  possibility of peaking at a given concen-
tration, and the human exposure potential in accordance with
applicable radiation protection criteria.

-------
HOMESTAKE - SAPIN  PARTNERS                          91

                      Acknowledgment

    The generous cooperation and technical assistance of the
following are gratefully acknowledged:  Personnel of the Homestake-
Sapin Partners Mill; the U. S. Atomic Energy Commission; Charles E.
E. Sponagle,  Public Health Service, Region VIII, Denver, Colo-
rado; E. A. Pash, Carl Hirth, Carl Shadix and H. D. Nash,
Public Health Service.  Cincinnati, Ohio; and D. E. Rushing,
Public Health Service,  Salt Lake City,  Utah.  The technical advice
and guidance of Dr. E.  C. Tsivoglou,  Public Health Service,
Cincinnati, Ohio, was greatly appreciate.  This study was suppor-
ted by funds made available through the Environmental and Sani-
tary Engineering Branch,  Division of Reactor Development, U. S.
Atomic Energy Commission.

-------
                           Bibliography

 l.Tsivoglou,  E. C., Bartsch, A. F., Rushing, D. E.,  and Holaday,
       D. A.  Effects of Uranium Ore Refinery Wastes on Reciving
       Waters. Proc.  1958 Nuclear Science and Engrg. Conf.,  Re-
       print 40, Am. Inst. Chem. Engrs., New York, N.Y.
       (March 17, 1958).

 2. National Bureau of Standards.  Maximum Permissible Amounts of
       Radioisotopes in the Human Body and Maximum Permissible
       Concentrations  in Air and vVater.  Handbook 52. U. S. Gov't.
       Printing Office,  Washington, D. C. (1953).

 S.Dayton, S.  UgOq Processing Made More Efficient by Plant-Con-
       trols at Edgemont Mill.  Mining World, 19.  No. 1, 36-43
       (January 1957).

 4. Public Health Service.  Determination of Gross Radioactivity in
       Water. Silt,  and Biological Materials.  Robert  A. Taft Sani-
       tary Engineering Center, Cincinnati, Ohio. (1957).

 5. Standard Methods for the  Examination of vVater. Sewage and Indus-
       trial Wastes. Am. Publ.  Hlth. Assn., New York,  N.Y.,  522 p.,
       10th Ed. (1955).

 6. Part 20 - Standards for Protection Against Radiation.  U. S.  At.
       Energy Comm.,  Federal Register.  22.  No. 19. (Tuesday,
       Jan. 29. 1957).

 7. Public Health Service.  Report on an Investigation of Ground Water
       Pollution.  Grants-Bluewater. New Mexico.  Robert A. Taft
       Sanitary Engineering Center, Cincinnati,  Ohio.  (Aug. 1957).

 S.Velz.  C. J., Graphical Approach to Statistics.  Water and Sewage
       Works, 97.  (1950).

 9.Recommendations of the International Commission on Radiological
       Protection.  Main Commission Report,  Sept. 9, 1958: Perga-
       mon Press,  London,  England.  (1959).

lO.Tsivoglou.  E. C. Shearer. S.  D.. Shaw.  R.M.. Jr.,  Jones, J. D.,
       Anderson, J. B., Sponagle, C. E., and Clark, D. A.,  Survey
       of Interstate Pollution of the Animas River. Colorado-New
       Mexico. U. S.  Pablic Health Service. Robt. A.  Taft San.
       Engrg. Ctr., Cincinnati,  Ohio.  (May 1959).

H.Public Health Service Drinking Water Standards. 1946. Public
  '     Health Reports.  61, No. 11, 371-384: Reprint No. 2697.
       (March 15, 1946: revised March 1956).

                                93

-------
 12.Tsivoglou, E. C., Shearer, S. D., Jones, J. D. Sponagle, C. E.,
        Pahren,  H. R., Anderson, J. B., and Clark, D. A. Survey of
        Interstate Pollution of the Animas River. Colorado-New
        Mexico.  II. 1959 Surveys. U. S. Public Health Service, Robt.
        A. Taft San. Engrg. Ctr., Cincinnati, Ohio.  (Jan.  1960).

 IS.Knott, Woodrow, The Processing of Uranium Ores of the Colorado
        Plateau by the Climax Uranium Company. The Mines Maga-
        zine, p. 29. (January 1955).

 14.Survey of the Colorado and Gunnison Rivers in the Vicinity of Grand
        Junction.  Colorado.  Technical Report  No. 2,  Colorado River
        Basin Water Quality Control Project, U. S.  Public Health
        Service.  (May 1961).

 15 National  Bureau of Standards.  Maximum Permissible  Body Bur-
        dens and Maximum Permissible Concentrations of Radio-
        nuclides in Air and Water for Occupational Exposure.
        Handbook 69, U. S. Gov't. Printing Office, Washington, D.C.
       (June 5  1959).

 16.McClaine, L. A., Bullwinkel, E. P., and Huggins, J. C.  The
        Carbonate Chemistry of Uranium: Theory and Applications.
       Internat.  Conf. of Peaseful Uses of  Atomic Energy, A/CONF.
       8/P/525, USA.  (July 18,  1955).

17-Clegg, J. W.. and Foley, D.  D. Uranium Ore Processing. Atoms
       for Peace,  USA. Geneva 1958.  Prepared for USAEC.  Addison-
       Wesley Publ. Co., Reading. Mass.  (Sept. 1958).

lg Review of Waste Disposal Problem at Homestake-Partners Mill-
       site (N.M.) Communication from the New Mexico Department
       of Public Health.
                                94

-------