REMOVAL OF URANIUM FROM DRINKING
WATER BY ION EXCHANGE AND
CHEMICAL CLARIFICATION
by
Steven W. Hanson
Donald B. Wilson
Naren N. Gunaji
New Mexico State University
Las Cruces, New Mexico 88003
Cooperative Agreement
CR-810453-01-0
Project Officer
Steven W. Hathaway/Richard P. Lauch
Drinking Water Research Division
Water Engineering Research Laboratory
Cincinnati, Ohio 45268
WATER ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
The information in this document has been funded wholly or in part by
the United States Environmental Protection Agency under assistance agreement
number CR-810453-01-0 to New Mexico State University. It has been subject
to the Agency's peer and administrative review, and it has been approved for
publication as an EPA document. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with pro-
tecting the Nation's land, air, and water systems. Under a mandate of
national environmental laws, the agency strives to formulate and implement
actions leading to a compatible balance between human activities and the
ability of natural systems to support and nurture life. The Clean Water Act,
the Safe Drinking Water Act, and the Toxic Substances Control Act are three
of the major congressional laws that provide the framework for restoring and
maintaining the integrity of our Nation's water, for preserving and enhancing
the water we drink, and for protecting the environment from toxic substances.
These laws direct EPA to perform research to define our environmental prob-
lems, measure the impacts, and search for solutions.
The Water Engineering Research Laboratory is that component of EPA's
Research and Development program concerned with preventing, treating, and
managing municipal and industrial wastewater discharges; establishing prac-
tices to control and remove contaminants from drinking water and to prevent
its deterioration during storage and distribution; and assessing the nature
and controllability of releases of toxic substances to the air, water, and
land from manufacturing processes and subsequent product uses. This publica-
tion is one of the products of that research and provides a vital communica-
tion link between the researcher and the user community.
This report details the demonstration of ion exchange and chemical clari-
fication technology for the removal of uranium (as U02 ) from drinking water.
These technologies were shown to be applicable for reduction of uranium con-
centrations to below 10 pCi/L. This report further describes a strategy for
the ultimate disposal of the recovered uranium in a safe manner.
Francis T. Mayo, Director
Water Engineering Research Laboratory
-------
ABSTRACT
A research project was conducted to remove uranium solute from drink-
ing water with ion exchange and chemical clarification. Three different
ion exchange resins were used: DOWEX SBRP, DOWEX 21K, and IONAC A641. Four
ion exchange columns were constructed, each containing 2 ft^ of resin.
Three columns were operated in the conventional down flow mode; the fourth
column was operated with upward flow of the feed water. Pretreatment con-
sisted of particulate filtering only. Regeneration was by chloride ion.
Resin capacity was represented by bed volumes between 12,000 and 20,000.
Four cycles of the resin were completed, processing approximately 4 million
gallons of feed containing an average uranium concentration of 300 ug/L.
A small, 1-gpm chemical clarification unit was built consisting of a
rapid-mix vessel and a pre-coat rotary vacuum filter. This system was
operated continuously over a period of 3 months using various pH values
ranging from 6 to 10.0 and various ferric chloride concentrations ranging
from 15 to 40 mg/L. Better than 99 percent uranium removal was achieved
by operating at a 30 mg/L ferric chloride concentration and pH 10.0. The
diatomaceous earth precoat filter achieved complete solid-liquid separation.
Additional work included a review of current drinking water operations
for uranium removal, geohydrology studies on the origin of the raw water
uranium for this study, fluidization characteristics of upflow ion exchange
columns, and ultimate disposal problems associated with the separated
•uranium.
This report was submitted in fulfillment of Contract No. CR-810453-01-0
by New Mexzico State University under the sponsorship of the U.S. Environ-
mental Agency. This report covers the period of October 1982 to July 1986,
and work was completed as of July 1986.
iv
-------
CONTENTS
Page
Disclaimer n
Foreward iii
Abstract iv
Contents v
Figures yi
Tables yii
Acknowledgments viii
1. Introduction 1
A. Objectives of the Study 1
B. Uranium Effects and Chemistry in Aqueous Solution . 1
C. Chemical Clarification 6
D. Description of Ion Exchange 7
E. Description of Current Data on Uranium Removal ... 9
F. Ultimate Waste Disposal from Uranium Removal from
Drinking Water 12
2. Summary and Conclusions 14
A. Ion Exchange 14
B. Chemical Clarification 16
C. Survey of Current Operations . : . . 17
. D. Waste Disposal 18
3. Recommendations 20
4. Experimental Equipment and Procedures . 21
A. Ion Exchange System and Operation 21
B. Start-up 26
C. Chemical Clarification System 28
D. Up-Flow Ion Exchange System 28
E. Analytical Procedures 28
5. Results and Discussion 32
A. Ion Exchange System 32
B. Chemical Coagulation 40
C. Monitoring Program 45
D. Waste Disposal 48
E. Design Analysis 51
References 56
Appendix. Method of Uranium Analysis 58
-------
FIGURES
Number Page
1. Pourbaix Diagrams for Uranium 4
2. Chemical Clarification 8
3. Typical Single Column Ion Exchange Flow Diagram ...... 10
4. Ion Exchange Column Operation 10
5. Percentage of Uranium Removed From Pond Water by Ferric Sul-
fate 11
6. Percentage of Uranium Removed from Pond Water by Aluminum Sul-
fate 11
7. Percentage of Uranium Removed as a Function of pH 11
8. Uranium Breakthrough From an Anion Exchange Column 11
9. Demonstration Van 23
10. IX Equipment Overview in Trailer 23
11. Organization of Van System 24
12. Pre-Filter Equipment Ion Exchange System. 25
13. IX Columns • . 25
14. Schematic Diagram of a Flotation/Filtration Unit 29
15. Chemical Clarification System as Installed 30
16. Up-Flow Ion Exchange System 30
17. Summary of Operation of Ion Exchange Columns 34
18. Breakthrough Curves - Unit 1 (DOWEX 21K) 35
19. Breakthrough Curves - Unit 2 (DOWEX SBR-P) 36
20. Breakthrough Curves - Unit 3 (IONAC A641-Downflow) 37
21. Breakthrough Curves - Unit 4 (IONAC A641-Upflow) 38
22. Chemical Clarification Operation 42
23. Combined Distribution of Species in Chemical Clarification .
System 44
24. Cost of Ion Exchange (1985) 54
25. Cost of Flotation (1985) • 54
26. Cost of Filtration (1985) 55
vi
-------
TABLES
Number Page
1. Uranium Concentrations in Raw and Treated Water Samples
Taken from Selected Municipal Water Treatment Plants ... 3
2. Uranium Ions in Aqueous Solutions 5
3. Drinking Water Solutes 6
4. Mine Water Ion Exchange Tests 12
5. Demonstration Ion Exchange Units 15
6. Ion Exchange Resin Capacity 16
7. Survey of Current Operation on Uranium-Containing Waters . . 18
8. Ion Exchange Equipment List (Spring 1983) 22
9. Chemical Composition of Well Water Used in Demonstration ... 31
10. Summary of Operation of Ion Exchange System 33
11. Resin Capacity for Each Unit for Cycle . . . .• 39
12. Up-Flow Column Experiment 40
13. Summary of the Results for Uranium Removal Using 30 mg/L of
Ferric Chloride 41
14. The Effect of Ferric Chloride Concentration on Uranium Removal
Efficiency in a Solution of pH 6 and pH 10 43
15. State Survey for Uranium-Enriched Waters 46
16. Results of Harrisburg, S.D., Operation Analysis 47
17. Ultimate Disposal Values for Uranium 48
18. Ion Exchange System Design 53
19. Chemical Clarification System Design 55
vn
-------
ACKNOWLEDGMENTS
We thank Mr. Steven Hathaway, Project Officer for his interest and
patience in this project. We would also like to acknowledge the willingness
of the Kerr-McGee Corporation to volunteer a water supply and operating loca-
tion for the project, although it was not used. In addition we acknowledge
the communities of Harrisburg, South Dakota, and Arvada and Denver, Colorado,
for providing operating data on their drinking water treatment plants.
Projects conducted at universities generally depend on students as
research assistants. We were fortunate to have students from the Water
Utility Operators Training Program at the Dona Ana Branch of New Mexico State
University assist in the construction and operation of the ion exchange sys-
tem. Ms. Trina Rankin (B.S.C.E.) wrote an undergraduate research paper on
disposal of uranium wastes.
Graduate students involved were Ms. Li Huey Wu (M.S.Ch.E.), whose thesis
was entitled, "Chemical Clarification for Uranium Removal;" Mr. Chi Dong Ho
(M.S.Ch.E.) whose thesis was entitled "Up-Flow Ion Exchange;" Mr. John
Cochran (M.S.C.E.), whose thesis was entitled "Investigation of the Elevated
Uranium Concentrations in the Groundwaters of the Las Cruces, New Mexico,
Area Based on Uranium Disequilibrium." Mr. Cochran currently works for the
U.S. Environmental Protection Agency in the Dallas Regional office.
vm
-------
SECTION 1
INTRODUCTION
A. Objectives of the Study
The overall study objective.was to demonstrate the operation of the
equivalent of commercially available ion exchange equipment and to develop a
pilot-scale chemical clarification unit for the removal of uranium from
drinking water supplies. Specific objectives were:
1. To demonstrate extended operation of four ion exchange columns (3
to 5 gpm capacity each) using three different resins with one
column operated in an up-flow mode. These columns were to be the
equivalent of commercially available units and were to include
automatic regeneration capability.
2. To develop a pilot-scale chemical clarification unit (5 gpm maximum
capacity) and to test selected coagulants and/or flocculating rea-
gents. The system includes continuous solid-liquid separation.
3. To assemble the necessary ion exchange equipment, chemical clarifi-
cation equipment, monitoring instrumentation, and analytical
instrumentation in a self-contained trailer unit.
4. To develop and conduct a monitoring program for several currently
operating conventional drinking water treatment plants that have
uranium in their feedwater supplies.
5. To prepare a general evaluation of radioactive waste disposal tech-
nology that would be appropriate for the selected processes of ion
exchange and chemical clarification.
6. To prepare the necessary engineering data for process selection,
process design, and cost evaluation for removing uranium from
individual community water supplies.
7. To publish the project results. In addition to this final report,
research related to this project will provide thesis or disserta-
tion material for graduate students, and appropriate material will
be published in the technical literature.
B. Uranium Effects and Chemistry in Aqueous Solution
At present, the maximum contaminant level (MCL) for radioactivity in
drinking water does not include regulation of naturally-occurring uranium. [1]
The MCL for alpha-emitting radionuclides specifies 5 pCi/L for Ra226 and 228
combined and gross alpha activity of 15 pCi/L excluding the alpha contribu-
tion from radon and uranium. Establishing a safe unit for human consumption
-1-
-------
is the primary goal of the MCL, but it is also based on technology for
removal and cost of the technology.
Although the effects of ionizing radiation on humans is much better
known than those of many other environmental pollutants, it cannot accurately
and definitely be predicted from known animal effects. Our bodies may be
exposed to both external and internal radioactivity. For exposures to drink-
ing water, the internal exposures are the most important. Once the radioiso-
tope enters the body through ingestion, it will move to locations determined
by the body's metabolism and chemistry. Uranium is considered to concentrate
in the kidney rather than in bone marrow; therefore it may have a short bio-
logical half-life (e.g., the time duration for the body to eliminate one-half
the original concentration). [2]
A recent study of the occurrence of uranium in drinking water in the
United States estimates the average concentration as 2 pCi/L. [3] By con-
suming 2 liters of drinking water per day, the average person would ingest
1440 pCi in 1 year. The average annual ingestion in food is about 240 pCi/
year. Current model estimates indicate that the ingestion of 10 pCi/day of
radium (i.e., 2 liters of water containing 5 pCi/L) produces a dose of 150
mrem/year to the skeletal bone. [4] Corresponding models_|or uranium have
estimated the risk from ingested uranium to produce 3 x 10 excess cancers
per lifetime from ingestion of water with a uranium concentration of 10 pCi/L.
Therefore current considerations are for limiting uranium concentrations in
drinking water to 10 pCi/L. This is the level of activity that must be
achieved by current available technology.
Public water supplies are subjected to a variety of treatments, depend-
ing on the condition of the raw water and the needs of the consuming com-
munity. Several different drinking water treatment processes are available
for possible removal of naturally occurring uranium. These processes are
classified into conventional water treatment processes and non-conventional
water treatment process. Conventional water treatment processes for removal
of inorganic contaminants are (1) coagulation (alum or iron) followed by set-
tling and filtration; (2) coagulation (alum or iron) followed by filtration;
(3) lime softening (with or without recarbonation). Conventional treatment
plant operations on uranium-containing water supplies were surveyed. [5]
The results of this survey are summarized in Table 1. [6] The survey
indicated that further research and development of chemical clarification and
ion exchange should be undertaken.
Before considering the technology of uranium removal a brief review of
the aqueous-phase chemistry of uranium is appropriate. Pourbaix [7] gives
two potential pH diagrams for uranium-water systems. Tlji|se are summarized
as Figure 1. These diagrams show the region of the UO- ion that is the
predominant form of the uranium ion in drinking water. Cotton and Wilkinson
[8] summarize the aqueous chemistry of uranium as follows:
"Uranium ions in aqueous solution can give very complex species
because, in addition to the four oxidation states, complexing
reactions with all ions other than CIO. as well as hydrolytic
-2-
-------
TABLE 1. URANIUM CONCENTRATIONS IN RAW AND TREATED SAMPLES TAKEN
FROM SELECTED MUNICIPAL WATER TREATMENT PLANTS
I
CO
I
Water Treatment
Coagulation
NaOH
Laotian and Hirst Water Iron Activated for pH
at Treatment Flint Source Aim S«lt limt Polymer Stlici Carbon Lime Control
Arizona (Phoenix)
Squaw Peak
Val Vlit.
Verde
California (Lot An|elei)
Hawthorne
Jensen
Long Beach
Veynouth
California (San Diegeo)
Alvarado
Eacondldo-VI«ta
Clay
.Sweetwater
Colorado (Denver)
Haraion
Moffat
Michigan (Midland)
rUliouri (Kanaat City)
Nebraska (Lincoln)
Texai (Houaton)
Utah (Salt Lake City)
Big Cotlonvood
City Creek
Little Cottonvood
Si X X
Si X X XX
Si X X
G: X
Si X
c x
s
Si XX
Si X X
Si X X
Si X
Si X X
Si X XX
Si X
Si XXX
C: X
si x x x
Si X
Si X
Si X X
Uranium
Raw
nitration Chlorination Water
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
x
X
X
X
X
X
X
X
X
x
X
X
X
1.70
4.45,4.30
4.30.4.10
0.74,0. 29, <0. 10
0.30,0.27
0.32,0.30
6.61,6.10
1.68,1.90
5. 43, 5. 10, 8. IS
1.05,1.00
4.07,3.30
1.60,1.60
15. 911.58
0.2710.03
5.3310.23
7.2910.47
0.3210.14
0.90
I. 00, 0.30
1.70
Treated
Water
1.80
4.05
4.20
O.IO,<.IO
0.28
1.5913.03
6.06,6.60
2.31,1.80
6.25,5.60
2.10,2.30
3.30
1.50
4.00
0.35
4.0710.30
7.39
0.28
0.80
0.90,1.00
0.90,1.00
For replica analyaea (n>2), valuea are mean 11ttandard deviation.
n = 3
n = 4
-------
-2024 6 8 10 12 1\pH -202
6
b]
8 10 12 14 pH
Figure I. Pourbaix Diagrams for Uranium
-------
reactions leading to polymeric ions occur under appropriate con-
ditions."
Table 2 lists the simple ions and their properties plus the pH dependency of
the uranyl complex ions in the presence of carbonate.
TABLE 2. URANIUM IONS IN AQUEOUS SOLUTIONS [8]
Ion
Color
Preparation
Stability
,3+
Red-Brown Na or Zn/Hg on
U Green
U02 ?
Air or 02 on U
Transient species
U022+ Yellow Oxidize U4+ with HNOg, etc.
Slowly oxidized by d-0,
rapidly by air to U
Stable; slowlv+oxidized
by air to U02
Stability greatest at pH
2j4; disproportionates to
U* and U02
Very stable; difficult to
reduce
Adjusted pH [6]
(In the presence of carbonate)
PH
10
Uranyl Species U02
2+
U02(C03)2~ (U02)3(OH)*
Aqueous solutions of uranium salts haye ajji acid reaction due t hydroly-
sis, which increases in the order U
-------
TABLE 3. DRINKING WATER SOLUTES
Metals Ions Water Parameter
u
Fe
Ca
Na
As
Ra
Mg
Si
cr
F"
so4=
N03"
co3=
HCO^-
Pfl *~
A
P&M Alkalinity
Hardness
SP Conductance
PH
TDS
Temperature
species depending on the pH. As shown in Table 2 carbonate complexes can be
either neutral or negative in the pH range characteristic of drinking water
supplies2+ Phosphate complexes can be either positive, such as UCLH^PO. and
U07H-PO, ,9or negative at higher concentrations. Finally, silica readily
adSofbs^UO^ and IT at low pH's [9].
Sorg and Logsdon [10] have published a series of articles summarizing
existing treatment technology to meet the inorganic National Interim Primary
Drinking Water Regulations. Part 5 of their series covers barium and "radio-.
nuclides" which are treated as radium 226 and radium 228. Uranium and other
members of the actinides are not discussed. Recent work by Oak Ridge National
Laboratories has provided the best available data for uranium removal. [11]
The Oak Ridge studies addressed the technologies of chemical clarification
and ion exchange, supporting the decision that it is these technologies which
should be examined further.
C. Chemical Clarification
In general, chemical clarification is made up of three operations:
(1) coagulation, (2) flocculation, and (3) sedimentation [12]. The litera-
ture does not make a clear distinction between these operations and in some
work 'coagulation1 and 'flocculation1 are used interchangeably. The descrip-
tion of the process becomes even less clear when chemical reactions occur and
filtration is substituted for sedimentation. It is defined in this work that
"chemical clarification" will mean "...the addition of coagulants, flocculants
or oxidants to cause the precipitation of material which removes inorganic
compounds or which adsorbs inorganic compounds (ions) and hence removes them
from solution."
-6-
-------
Coagulation takes place in rapid mix, or flash mix basins. Some systems
use in-line mixers. The primary coagulants that have been used in wastewater
treatment are: (1) lime, (2) alum, and (3) iron salts such as ferric chlo-
ride, ferric sulfate and ferrous sulfate. While the mechanism of uranium
removal via coagulation has not been demonstrated conclusively, it most
likely occurs through adsorption of the uranyl ion complex by the coagulant
precipitate. Subsequent filtration removes the coagulant precipitate and the
adsorbed uranyl ion complex. Depending on the filter media there may also be
adsorption of the uranyl ion complex on the filter media. Successful opera-
tion requires the uranyl ion complex to remain adsorbed i.e., should not
desorb during filtration.
Conventional water treating systems which use coagulants or flocculating
agents most often use sedimentation basins for separating the solid-precipi-
tate, agglomerated solids from the water. These systems are selected on the
basis that the feed water contains suspended solids. In applications where
chemical clarification is used to remove metal ions from water in the absence
of suspended solids the resulting precipitates are usually so small that sedi-
mentation alone is not adequate and flocculation and filtration are required.
Filtration is the key process in production of high quality effluents
from wastewater. In solid-liquid separation filtration is generally through
a very thin layer of porous material deposited by flow on a support septum.
[13] While these types of filters have special applications in water trea-
tment, most frequently filtration in water treatment is understood to mean
the removal of impurities from water by passage through a relatively deep
(2-3 ft) bed of granular material. The major difference in the operation of
the two types of filtration is that the mechanical-straining type filter pro-
duces a product cake having from 5-8 percent water and the deep bed filter
produces a slurry (as a result of back washing to regenerate the bed). Both
types of filters.will need to be evaluated for ultimate product disposal when
separating radioactive nuclides.
A conventional chemical clarification system is shown in Figure 2.
D. Description of Ion Exchange
Ion exchange is a separation process in which ions held by electrostatic
forces to charged functional groups on the surface of an insoluble solid are
replaced by ions of like charge from the solution. Unlike simple physical
adsorption phenomena, ion exchange is a stoichiometric process in which every
ion removed from solution is replaced by an electrical equivalent amount of
another ionic species of the same sign from the solid. Ion exchange is, in
general, a reversible process and is selective in the removal of dissolved
ionic species. Although many naturally occurring materials exhibit ion
exchange properties, synthetic ion exchange resins having a wide range of
properties for specific applications have been developed. [14,15]
The characteristic properties of ion exchange materials are due primarily
to their structure. These materials consist of a solid matrix held together
by chemical bonds. Attached to this framework are soluble ionic functional
-7-
-------
Chemical
Reagents
(as
I
oo
I
Feed
Water
— ^
Pressure Filtration
NJ
Coagulation
Regent
V
^
V
^
r^
^j
Contact Tank
Flocculation Tank
Backwash
Continuous
Vacuum
Filtration
Product
Filtrate
Backwash
Slurry
Product
Filtrate
Filter
Figure 2. Chemical Clarification
-------
groups containing ions which are relatively free to move and exchange with
similarly charged ions in solution. Ion exchange materials must possess the
following characteristics:
1. Ion-active sites throughout the entire structure e.g., very uni-
form distribution of activity;
2. High total capacity, that is, a high degree of ion substitution or
low equivalent weight;
3. Good degree of selectivity for ionic species but capable of being
regenerated;
4. Extremely low solubility;
5. Good structural chemical stability;
6. Good structural physical stability;
and 7. Costs competitive with other processes.
Ion exchangers are classified by the type of ionic functional group
attached to the structure and the charge sign of the exchanging ion. Five
major classes of ion exchange resins, categorized according to functional
group, are (1) strongly acidic cationic; (2) weakly acidic cationic;
(3) strongly basic anionic; (4) weakly basic anionic; and (5) a broad miscel-
laneous category of ion-specific structures. In addition to these classes,
there exist some intermediate strength acid and base resins.
Normal ion exchange is operated in a column system as shown in Figure 3.
There are four distinct steps to one cycle of operation: 1) service period;
2) backwash; 3) regenerate; and 4) rinse.
The primary operating parameters in ion exchange units are shown in
Figure 4 e.g., the exchange zone and breakthrough. These parameters provide
the necessary operation period before regeneration. The third operating par-
ameter can only be obtained on pilot-scale experiments using the actual water
supply and this is the number of generation/regeneration cycles a given
charge of ion exchange resin can undergo. Bottle tests can give some indica-
tion of the percentage regeneration (aging) of the resin but are not conclu-
sive.
E. Description of Current Data on Uranium Removal
As a result of the literature review of removing uranium from drinking
water, Oak Ridge National Laboratory undertook bench-scale testing on uranium
removal from a natural water. [11] Their work is summarized in Figures 5
through 8 indicating that conventional coagulents i.e., ferric sulfate or
alumni urn sulfate remove more than 85 percent of dissolved uranium (83 ug/L)
when an optimum pH and dosage were provided. At pH 10 a dosage of greater
than 12 mg/L for both coagulants gave maximum removal. In addition, it was
verified that a strong base anion-exchange column is a recommended option for
the treatment of private well waters containing uranium at higher than
desired levels; although the published data did not represent column break-
through.
Further bench tests on ion exchange were initiated at the Drinking Water
Research Laboratory of the Environmental Protection Agency. [1] In addition
McClanahan examined ion exchange for removal of uranium from mine waters.
[16] These results are summarized in Table 4.
-9-
-------
Feed water.
CO-
Backwash discharge
~
X
\
v
_Pfoduct water,
W—
Reeenerate
Water rinse
•*-»
Backwash watpr
Regenerate waste,
Water rinse waste
Figure 3. Typical Single-Column Ion Exchanger Flow Diagram
H
Z
u
o
§
u
0.95 X,
< u
u ~>.
H
3 ~]
-3 **
u- Z
0.05 X,
V, EFFLUENT VOLUME CUBIC CENTIMETERS T
Figure 4. Ion Exchange Column Operation
-10-
-------
^_»
e
§
o
a.
•a
a>
o
a>
Od
E
'c
u
100 .
80.
60.
40.
20.
0.
/ '•'•- — •
. ^ •" *"
• -
/ X
•'
• ^
' • ^ . — • *" _
x^'x'^>*>vl^ ^" "^
••"-'-
«^
e
o
o
Cu
•o
>
o
OS
e
3
1
100 .
80.
60.
40.
20.
A
. ^^
* ^^
r ^
/
/
/
/ /
• / /
: / _s
.' / ^^^
i — --^^^^
0. 4. 8.12.16.20.24.28.
Ferric Sulfate Dosage mg/l
Figure 5. Percentage of Uranium
Removed From Pond Water
by Ferric Sulfate.
_ pH 4; — pH 6;
__ pH 8; ... pH 10.
=> 0. 4. 8.12.16.20.24.28.
Aluminum Sulfate mg/1
Figure 6. Percentage of Uranium
Removed From Pond Water
by Aluminum Sulfate.
_ ph 4; — pH 6;
_ _ pH 8; ... pH 10.
c
o>
o
u
T3
0)
E
'c
u
S
100.
80.
60.
40.
20.
0.
o>
ex
E
3
'S
2. 4. 6. 8. 10. 12. 14.
PH
Figure 7. Percentage of Uranium
Removed as a Function
of pH (Using Ferric Sul-
fate. —; no coagulant
—; Aluminum sulfate
Figure 8.
30.
25.
20.
15.
10.
5.
0.
0. 500. 1000. 1500. 2000.
Col Vol of Effluent ml
Uranium Breakthrough From
An Anion Exchange Column.
(DOWEX 1-X2) Influent
Contained 23.8 mg U/L and
960 mg (NH4)2 C03/L.
-11-
-------
TABLE 4. MINE-WATER ION EXCHANGE TESTS
Ion Exchange Material Uranium Removal, % Comment Ref.
DOWEX SBR
DOWEX 50W-X8
Clinoptiloite
DOWEX 21K
DOWEX 1-X2
DOWEX SBR-P
IONAC A-641
99+
35.
20.
97.
99* (Max)
99
99+
Mine Water
Mine Water
Mine Water
Mine Water
Parameter Dependent
Column Test
Column Test
16
16
16
16
1
1
1
The majority of applications of ion exchange and chemical clarification
for recovery of uranium (and other radioactive materials) has taken place at
uranium ore processing operations or at national laboratories. While ion
exchange is a primary unit operation for the recovery of uranium from aqueous
solution in ore processing the concentrations considered are several orders
of magnitude greater than the equivalent concentration anticipated in drink-
ing water supplies. In fact the IX column effluent from ore processing is
usually greater than drinking water supply raw feed.
Chemical clarification and ion exchange for treating radioactive waste
solutions have been used extensively at Los Alamos National Laboratory (LANL).
[17] Liquid wastes originate at many LANL facilities and are treated at two
separate plants in conventional water purification equipment. The treatment
removes most of the radioactivity in the wastes by controlled precipitation
of ferric hydroxide, followed by ion exchange if necessary. Settling tanks
and filtration (precoat) are used to separate the precipitate (called sludge)
from the treated water. The piping system permits recycling of the liquid
through the plant when additional contamination removal is necessary. This
process removes more than 99.9 percent of the original transuranic radio-
activity.
F. Ultimate Waste Disposal From Uranium Removal From Drinking Water
This topic has not been discussed in the literature. The subject of
radioactive waste disposal is a major concern that does not have a univer-
sally acceptable solution at present. There are regulations for handling
various levels of radioactivity in wastes. The two that are pertinent to
this proposed technology are maximum allowable uranium discharge to surface
water i.e., the ion exchange regeneration solution and the level of radio-
activity in solid waste.
-12-
-------
New Mexico restricts the discharge to surface waters of uranium to con-
centrations of less than 5 mg/L [18] Solid waste having less than 100 nano-
curies per gram can be packaged for ground burial at an approved site. Solid
waste having greater than 100 nanocuries per gram must be package and even-
tually transported to an approved waste repository.
Those communities currently using conventional waste treatment operations
(Table 1) on uranium containing waters dispose of their sludge similar
to non-radioactive sludge disposal practices.
-13-
-------
SECTION 2
SUMMARY AND CONCLUSIONS
A. Ion Exchange
Water treatment using ion exchange technology is a well established oper-
ation. Much of the growth, however, has been relatively recent. This current
popularity can be attributed to the increasing awareness of declining water
quality and the concurrent requirements for meeting water standards. Manufac-
turers have responded by providing an increasing variety of ion exchange
resins to handle these demands. Three resins were selected to demonstrate
the effective removal of uranium from a potential drinking water supply.
The resins selected and their operational configuration are shown in
Table 5. As shown, three of the four units were operated in the conventional
downward mode, while the fourth was operated in an upflow mode. The effi-
ciency (as characterized by the breakthrough curve) of each system was compar-
able, although the upflow unit processed the most water.
Several physical and chemical characteristics of the resins need to be
considered in the selection of ion exchange as a water treatment technology.
Physical bead breakage may occur in some applications. For down-flow.systems,
this results in increased pressure drop and reduced efficiency. In an upflow
unit, breakage could result in subsequent downstream problems in addition to
the loss of exchange capacity through loss of resin.
Thermal stability of the resins is a second consideration in selecting
ion exchange. Weak-base resins are usable up to 212 °F followed by Type I
strong bases to 122 °F (the resins of this demonstration). Temperatures in
the van housing the units ranged from a winter low of 35 °F and a summer high
of 115 °F. The average inlet water temperature was 65 °F.
Although chemical degradation is not normally a problem in most water
treatment applications, strong oxidizing agents can rapidly degrade the
polymer matrix and should be avoided. Slower degradation with oxygen may be
catalytically induced, so ionic iron, manganese, and copper should be mini-
mized. Both equipment manufacturers and resin producers recommend techniques
for managing these situations.
Capacity of ion exchange resin is the measure of ionic attraction per
volume, and is expressed in a number of.ways. Total capacity is the theore-
tical measure of the total number of exchange sites available and is normally
-14-
-------
TABLE 5. DEMONSTRATION ION EXCHANGE UNITS
Resin
Type
DOWEX 21K
DOVES SBR-P
IONAC A-641
IONAC A-641
Flow
Condition
Down
Down
Down
UP
Total
Gallons
Processed
797,540
890,600
853,870
903,450
Number
of
Cycles
4
4
4
4
Average
Bed-Volumes
Treated/ Cycle
12920
14925
16038
15370
Maximum
Removal
Efficiency
99%
99%
99%
99%
Inlet water concentration ranged between 200 and 400 ug/1 uranyl complex
(as uranium).
calculated in three different ways: Dry weight capacity, (mini-equivalents/
dry gram), wet weight capacity (milli-equivalents/wet gram), and wet volume
capacity (milli-equivalents/milliliter). Regardless of how expressed, operat-
ing capacity is the most realistic performance measure for ion exchange resins
and in water treatment is usually expressed as kilograins CaCO, per cubic
foot of resin.
A final consideration in the selection of an ion exchange resin is the
cost of regeneration. Small differences in efficiency will be magnified over
the life of the system. Perhaps of equal importance are the potential envir-
onmental problems encountered in disposing of waste regenerant. These prob-
lems will be addressed after a discussion of the chemical clarification
results. '"
NaCl in a 10-percent by weight concentration was selected as the regeneration
solution.
Specific Conclusions
1. When operated under prescribed conditions, all resins performed
satisfactorily to produce effluent water that would meet the sug-
gested standard of less than 10 pCi/L (approximately 14 jjg/0- The
IONAC A-641 resin had the greater capacity under the operating
conditions used.
2. All four systems operated well. No unusual mechanical problems were
experienced.
3. The three units operated with downflow showed no resin breakage as
determined by pressure drop.
4. The upflow unit exhibited no resin breakage based on no change in
pressure drop and examination of the resin.
5. No noticeable change in operation occurred as a result of the
temperature of the operation over the range 35°F (winter) to 115 °F
(summer).
-15-
-------
Feed water iron and manganese concentrations (4 mg/L and 1.1 mg/L,
respectively) were higher than recommended for operation, however
there was no quantitative effect on the measured resin capacity over
the four cycles, although the measured capacity was less than the
manufacturer's listed capacity
Table 6 gives the measured resin capacity for the resins under the
operating conditions used.
Table 6. Ion Exchange Resin Capacity
Resin
DOWEX 21K
DOWEX SBRP
IONAC A641
IONAC A641
Uranium
Capacity*
Kilograins/ft3
2.95
3.15
3.67
3.54
Uranium
Wet Volume
Capacity"*"
(meq/mL)
0.03
0.032
0.038
0.037
Total
Wet Volume
Capacity^
(meq/mL)
1.20
1.20
1.16
1.16
(upflow)
*As U0
+Average over three cycles of operation (i.e., cycles 1,2, and 4).
o
^Manufacture's value: 1.2 meq/mL is 26 Kilograins/ft as CaCO,.
B. Chemical Clarification
A bench-scale flotation cell was used as the flash mixer and adsorption
vessel. Entrained air only was used i.e., no forced air mixing, and no sur-
factants to enhance bubble stability or flocculation were used. This vessel
had a one-minute residence time at 1 gpm feed rate. The effluent was chan-
neled, by gravity flow to the filter vessel, where a continuous rotary vacuum
filter separated the ferric hydroxide precipitate.
The filter membrane was 0.45 micron polymer mesh and when operated with-
out precoating allowed precipitate to bleed through into the filtrate. A
one-quarter inch pre-coat of diatomaceous earth was used and gave complete
solid-liquid separation as measured by our analytical procedures. Additional
diatomaceous earth (12% by wgt) was added to the filter feed and mixed by a
mechanical agitator with the feed slurry. A stationary knife removed the
cake build up. The residual precoat showed (visual inspection), no blinding
by the ferric hydroxide precipitate. Sampling the precoat material after a
run for uranium showed that all uranium removed was adsorbed on the ferric
hydroxide and not on the diatomaceous earth.
-16-
-------
Bed filtration was considered and previous work in the Sanitary Engineer-
ing program had demonstrated the suitability of mixed media bed filtration
for separating ferric hydroxide. Bed filtration was rejected for two reasons:
first, successful use required coagulation chemicals and second, back-flushing,
which again produces a slurry, is required to regenerate the bed. Ultimate
disposal of the uranium suggests that the disposal product should be solid
and of minimum volume, hence the selection of pre-coat continuous vacuum fil-
tration. Further discussion of ultimate disposal will occur in Section D.
Specific Conclusions
1. This work verified the prior work of Oak Ridge, Los Alamos, and EPA
that conventional chemical clarification would successfully remove
uranium from drinking water.
2. Ferric chloride at a concentration of 30 mg/L removed more than 98
percent of the uranium at concentrations of 300 and 400 ug/L at
pH 10.
3. At pH 6, and an average uranium concentration of 300 ug/L, 30 mg/L
ferric chloride removed 82 percent of the uranium.
C. Survey of Current Operations
A detailed search was conducted to locate currently operating conven-
tional water treatment facilities with uranium in their feed supplies. For
this search, a minimum level of 15 ug/L of uranium was arbitrarily selected
and conventional water treatment facilities were defined as any type of treat-
ment facility more "complex" than sand separation and chlorination. A total
of 34 municipal systems and an additional 21 municipal wells were located in
a six state area (Table 7). Of these 55 possible study sites, only 4 provide
treatment above and beyond sand separation and chlorination.
Three of these cities are located in Colorado (Denver, Arvada and North
Table Mountain) and they all draw their water from the uranium contaminated
Ralston Reservoir. The fourth city is a small city in South Dakota
(Harrisburg). These communities were contacted and with the exception of
North Table Mountain they agreed to help with this study.
The city of Arvada treats their water using a microfloc system.
The system employs alum and Separan (a polyelectrolyte) to create the micro-
floc. The water is then passed through mixed media filters. A 125 cc water
sample is collected daily from the raw water and the treated water to form
monthly composite samples. For the past couple of years, these monthly samples
have been tested for uranium. The raw water from Ralston Reservoir contains
from less than 1 pg uranium/L to 36 ug uranium/L with an average of 14.7 ± 9.6
(SD) ug/L of uranium.
The Moffat Treatment Facility (Denver) also draws water from Ralston
Reservoir and has been keeping uranium records for about two years. Unfor-
tunately, Denver's monthly samples are "grab samples" and correlation between
raw and treated waters are less meaningful.
-17-
-------
TABLE 7. SURVEY OF CURRENT OPERATIONS ON URANIUM-CONTAINING WATERS
Colorado, Water Quality Control Division of Colorado Dept. of Health, 3 com-
munity composites, <1 - 24 pCi/L, conventional alum systems.
New Mexico, Water Supply Section of New Mexico Environmental Improvement
Division, 2 community composites and 15 municipal water sources, 10 -
110 pCi/L, sand filters and chlorination.
Oklahoma, Association of Central Oklahoma Governments, 22 communities (about
200 wells) near Oklahoma City, 10 - 190 pCi/L, sand removal and chlorina-
tion.
South Dakota, S.D. Dept. of Water and Natural Resources, 1 community, 12
pCi/L, iron and manganese removal, chlorination.
Texas, Texas Water Hygiene Division, 5 community composites and 6 individual
water sources, 10 - 55 pCi/L, sand removal and chlorination.
Wyoming, EPA Drinking Water Branch, Denver, 1 community (Cheyean), 29 pCi/L,
this well water is chlorinated, they also use surface water which is
treated with conventional alum.
NOTE: This survey is representative of existing state and federal records and
may not be complete.
Harrisburg, South Dakota treats about 100,000 gpd using aeration,
KMnO/rgreensand filters, chlorination and fluoridation. No previous data
existed on the uranium removal efficiency of this facility.
Specific Conclusions.-
1. The Arvada facility removed 18 to 90 percent of uranium in
the feed with an average efficiency of 67 percent ± 15 percent.
2. The Denver facility data were inconclusive (probably because their
sampling technique was the use of grab samples).
3. Two sets of samples were taken from the Harrisburg facility. Before
and after treatment had the same uranium concentration (19 ug/L).
D. Waste Disposal
As previously described, uranium contaminated drinking water is a common
problem, particularly in the western United States. If regulations governing
the concentration of uranium in drinking water are accepted and enforced, many
communities will be required to remove uranium from their drinking water sup-
plies. Removal of uranium produces a new problem for these communities—the
ultimate disposal of a radioactive waste. Three disposal alternatives were
considered in this study: dilution/release, reuse or resale, and burial.
-18-
-------
Specific Conclusions
Environmental considerations aside, dilution and release is
easily the least expensive alternative.
Depending on the treatment used for removal, a concentrated solution
of uranium could be shipped to a standard uranium mill for process-
ing.
Solid material containing the uranium could be packaged and shipped
for burial at an approved repository.
-19-
-------
SECTION 3
RECOMMENDATIONS
1. The recommended process is a combination of ion exchange for primary
removal of uranium from the drinking water supply and chemical clar-
ification (with filtration) for recovery of the uranium from the
regeneration solution. This is similar to current practice at Los
Alamos National Laboratories.
2. During part of the operation, the uranium concentrations in the well
reached 450 ug/L. Uranium concentrations in the regeneration solu-
tion were measured as greater than 25 mg/L Research needs to be
conducted to establish the operating conditions for successful
chemical clarification at these loadings of uranium and solution
conditions.
3. Clarification of existing federal (and state) regulations on licens-
ing of uranium processing facilities would be required should
recovery of uranium be acceptable practice.
4. Because ultimate disposal of the uranium through an ore-processing
facility is an optimal solution, research should be conducted to
establish the conditions, (i.e. pH, temperature, etc.) for desorption
of the uranyl complex from the ferric hydroxide precipitated.
5. Permissible water content of the filter cake for inter- (or.intra-)
state shipping should be established, and the necessary research
should be supported to demonstrate appropriate technology should the
5 to 8 percent residual moisture of vacuum filtration be unaccept-
able.
6. Conventional ion exchange operation on waters containing iron suggest
prefiltering for iron removal. Green sand filtration is the recom-
mended treatment. The survey data obtained from Harrisburg, South
Dakota, showed that green sand filters would pass the uranyl complex
through. This should be confirmed by further testing.
7. Alternative quantitative analysis procedures that could be performed
by community water treatment personnel would be required. If not
available, they, (e.g., equipment and or procedures) need to be
developed.
8. This research did not examine the leaching of organic materials from
the ion exchange resins used. If this information is not available,
it should be obtained before such resins are introduced into water
treatment for uranium removal.
9. Kinetic rates and mechanisms of uranyl complex adsorption on
precipitating ferric hydroxide compounds should be researched.
-20-
-------
SECTION 4
EXPERIMENTAL EQUIPMENT AND PROCEDURES
The demonstration objectives of the project were met through the design
and construction of three experimental systems:
(a) Four-column, semi-commerical-scale ion exchange system.
(b) Pilot-scale chemical clarification unit, and
(c) Pilot-scale upflow ion exchange unit.
The major ion exchange system was assembled and operated by personnel from
the Water Utilities Operator Training program at the Dona Ana Branch of
New Mexico State University. The chemical clarification unit and upflow ion
exchange were assembled and operated by graduate students in the chemical
engineering program, and present treatment and waste disposal studies were con-
ducted by civil engineering students at NMSU.
A. Ion Exchange System and Operation
The ion exchange system was shop-fabricated onsite under the supervision
of Mr. Steven Hanson. Table 8 is a summary of the equipment used and the cost
of the major items including the resins. Figures 9 through 13 show the
components of the system and their organization in the trailer.
The basic design criteria for the ion exchange units were as follows:
Downflow columns 2.75 gpm/ft2
Upflow column 1.83 gpm/ft2
The following is a description of the basic operation of the ion exchange
system.
Operational Steps
There are five steps in the operation of an anion exchange column, the
service cycle and four-step regeneration.
1. Service—The raw water enters the anion columns and passes either up
or down through the resin bed and is piped to the product storage
tank. The service cycle continues until either: 1) the seven day
timer indicates regeneration, or 2) the water quality falls to an
unacceptable level which is indicated by performing a water analysis
for uranium. (Initially these were daily until experience suggested
a weekly sample). Sample frequency increased when breakthrough
occurred.
2. Backwash—Flow through the resin bed is reversed - this is true only
for the three down-flow units, the flow for the up-flow unit will be
-21-
-------
Table 8. Ion Exchange Equipment List (Spring 1983)
Company Name
Amount
Item
CERAC Inc.
Fisher Scientific
Sargent-Welch
McMaster Carr
Meter Service Supply
Control Devices
Automatic Switch
Plastic Systems Inc.
L. A. Piping System Inc.
Ronnigen-Potter Corp.
Electrical Supply Co.
Cashway
Great Lakes Inst.
McMaster Carr
Great Lakes Inst.
Raven Industries
Aquamatic
Metro Harrington Plastics
Unistrut of El Paso
Southwest Piping Supplies
Southwest Piping Supplies
Southwest Piping Supplies
TOTAL
27.13
10.38
87.50
130.44
326.93
553.70
2232.50
5815.00
3310.00
5269.00
120.00
365.50
880.00
332.93
972.50
931.00
2481.00
760.00
989.32
454.52
913.04
984.00
$ 28,706.39
Sodium Sulfide
Hydrocloric Acid
D. Earth
Filter Regulator
Pressure Gauges
Level Switches
Solenoid Valves
Diaphram Valve
Centrifugal Pumps
Multiplex Filter
Slotted Angle
B-Board Insulation
pH Analyzer
PVC Welding Kit
Conductivity Analyzer.
Fiberglass Tanks
Programmable Controller
PVC Pressure Regulators
Unistrut Channel
PVC Valves
PVC Fittings
PVC Pipe
Ion Exchange Resins
IONAC A641
DOWEX SBRP
DOWEX 21K
4 cubic feet
2 cubic feet
2 cubic feet
$ 664.00
288.00
295.00
-22-
-------
fflf
i
Figure 9. Demonstration Van
Figure 10. IX Equipment Overview in Trailer
-23-
-------
1. Water Supply
2. Feed Pump
3. Pre-Filter
4. Chemical Clarification System
5. Four Unit Ion Exchange Columns
6. Ion Exchange Control Sytem
7. Product Pump
8. Product Tank
9. Product Discharge
10. Regeneration Solution
11. Regeneration Perm
12. Waste Discharge
13. Pilot-Scale Up Flow Unit
i
ho
H iTl M
"
\^
^^
^s
^s
. 1 . X_y (9) ^^
Figure 11. Organization of Van System
-------
I,
Figure 12. Pre-Filter Equipment Ion
Exchange System
Figure 13. IX Columns
-25-
-------
the same as the service flow - allowing water to pass upward loosen-
ing the particles and flushing away accumulated resin "fines" and
sediment.
3. Regeneration—The resin is now regenerated by allowing the brine
solution to be pumped from the brine tanks and passed UD_ through
the resin bed. This generating procedure displace the ions picked
up during the preceding service cycle and exchange chloride ions for
them. The anion resin is once again converted to its functional
form.
4. Slow Rinse—After the regenerant chemical has passed through the
resin, a displacement rinse continues at the same rate and in the
same direction through the resin gradually forcing any remaining
regenerant out to drain. The step helps insure good chemical utili-
zation.
5. Fast Rinse—The regenerated resin is now rinsed at a flow rate near
the service rate until all the residual chemicals are removed. This
rinse continues until the water going to drain meets the required
effluent quality as determined by analysis.
Operating Conditions
1. Physical Conditions—The equipment is designed to operate at inlet
line pressures of 45 to 75 psig. However, variations in line pres-
sure within these limits cannot be greater than ±5 psi. The con-
stant pressure requirement is necessary to insure that the regenera-
tion steps are not upset by erratic flow rates. A pressure regula-
tor has been installed to insure that a constant pressure is kept.
The temperature of the incoming water should not fluctuate much
during the year, however, the system should not be operated at tem-
perature higher than 105 °F due to the fact that plastic piping is
. subject to deformation at that temperature and strong base resin
- should not be operated much above that temprature.
2. Chemical Conditions—Due to the use of the 5-10 micron filter, the
turbidity of the incoming water should not be a problem. The chemi-
cal composition of the raw water should be checked periodically for
dissolved iron and manganese as these ions can and will cause foul-
ing of the anion resin should they exist above 0.3 mg/L.
B. Start-up
The start-up of the deionizers should proceed smoothly. At any time the
system is restarted, however, there may occur minor problems which must be
corrected to insure proper operation. Most problems can be easily handled by
using good common sense and frequent referral to this procedure.
1. Start-up Preparation
a. Close all column valves.
b. Water connections should be checked for tightness. Line pres-
sure to the system should be brought to 50 psig.
c. Drain lines should be clear of any blockage.
-26-
-------
d. This completes the Start-Up Preparation. The sequence should
be repeated for all units.
2. System Fill and Pressurization
a. Open main water valve and bottom feed valve and allow the vessel
to fill from the bottom with air escaping from the vent. When
the vessel is full as evidenced by water flowing from the vent
line, close bottom feed valve.
b. Now open column valves on the down-flow units - shutting off
feed valve. For the up-flow unit, open column valve while leav-
ing feed valve open. The entire system is now pressurized.
Carefully go over the entire piping, checking for any leaks and
tighten connections as necessary.
c. Close all valves. This completes the fill and pressurization for
the system.
Regenerant Chemical Make-Up
1. Salt (NaCl)—Fill the regenerant tank with 10 percent by weight
sodium chloride, NaCl.
Initial Start-Up
The following steps will be performed prior to start-up. During these
steps, the flows will be set, backwash rates checked and chemical regenerant
concentrations verified.
1. Backwash
a. Open main valve and column backwash valve on down-flow units.
Open bottom valve and column backwash valve on up-flow unit.
b. Regulate flow by means of stopwatch and alien wrench, until a
flow of 3 gpm is reached. This is the normal backwash rate for
the anion units of 60 °F. If the water is cooler, this rate
will have to be reduced, while if the water is warmer, the rate
must be increased.
c. Take a beaker or other suitable container and take a sample
every minute or so at the drain and inspect to be sure that no
whole beads are being carried over. Resin "fines," small, dust-
like particles are expected to be in the sample. If whole beads
appear, the backwash rate should be reduced at column backwash
valve until the carryover stops.
2. Slow Rinse
a. Close main feed valve and column backwash valve on the down-flow
units. Close backwash valve on the up-flow unit.
b. Open column feed valve on the down-flow units. Open column feed
valve on the up-flow unit.
c. Adjust the flow rate until a flow of 1 gpm is reached at each
unit.
d. Allow this step to continue for 5 minutes, then close column feed
valve on the down-flow units and column feed valve on the up-flow
unit.
-27-
-------
3. Fast Rinse
a. Now open feed valve on the down-flow units and bottom feed valve
on the up-flow unit.
b. Adjust the flow rate until a flow of 6 gpm is reached on each
unit.
c. Allow this step to continue for 3 minutes. This is the end of
the rinse step.
4. Service
a. Now adjust flow on all columns to desired operating flow. Oper-
ation should now begin.
C. Chemical Clarification System
The study of the removal of uranium ions with ferric chloride was per-
formed using a model D-12 Laboratory Sub-A Flotation Machine attached to a
rotary-drum vacuum filter. The flotation machine is manufactured by Joy
Manufacturing Company, Denver Equipment Division, Denver, Colorado. The
rotary-drum vacuum filter design was adapted from that of an Oliver filter
manufactured by Dorr-Oliver, Inc. The vacuum filter system was built in the
shop of the Chemical Engineering Department. A schematic showing the main
components of the flotation/filtration system is shown in Figure 14. Figure
15 is a photograph of the system. The filtered well water was combined with
an appropriate amount of ferric chloride solution and the pH of the influent
adjusted to a desired level, then fed directly to the flotation cell. The
mixed feed was agitated using the impeller in a cell [20]. Air was sucked
into the cell near the impeller zone. The air bubbles dispersed by the
impeller, attached to particles which were precipitated as coagulant, rise
to the cell top and then flow into the filtration vat. The slurry in the vat
was agitated using mechanical stirring to keep the suspension uniform. The
drum speed, degree of vacuum and precoat thickness have been adjusted to a
suitable value to maintain about 40 percent of the drum surface submerged in
the slurry. Continuous operation was achieved by having an automatic level
control on the filtrate receiver. At a high level, the pump discharged fil-
trate to drain, shutting off at the low level. Uranium activity remaining in
the filtrate was determined by a fluorometric method using a Turner Filter
Fluorometer.
D. Up-Flow Ion Exchange System
The pilot-scale column for the up-flow ion exchange experiments was con-
structed in the Chemical Engineering Department. Figure 16 shows the system
assembled for operation in the van. This unit was operated as an M.S. Thesis
project. [20]
E. Analytical Procedures
The water used in this experiment was taken from a university well
located at the Physical Plant of New Mexico State University, Well Number 8.
The chemical composition of the well water is shown in Table 9. The pH and
concentration of carbonate and other liquid species in the well water indicate
that the dominant uranium species in well water would be uranyl carbonates
-28-
-------
i
NJ
VD
I
Rapid-Mix
Tank
Raw
Water
Rotary
Vacuum
Filter
Filtrate
Receiver
Vacuum
/pjSystem
-t
t *~-
L
-tr
1
-------
UJ
O
Figure 15.. Chemical Clarification
System as Installed
Figure 16. Up-Flow Ion Echange System
-------
Component
Table 9. Chemical Composition of Well
Water Used in Demonstration
Concentration
(mg/L)
Component
Concentration
(mg/L)
u
Na+
.
K
Ca2+
Mg2+
Cl"
9-
co32
HCO,
H
SO/
n
"t
TDS
AS5+
9+
Ba2
0.300 (1)
17.9
18.4
375.2
69.0
555.2
0
96.4
400.0
2152
0.005
0.12
Cd2+
Cr3+
9+
Pb^
Hg2+
Se2'
Ag"
^
N03
F"
3+
9+
MM
Hardness (CaCO,)
«5
Alkalinity
<0.005
<0.01
<0.005
0.0004
0.005
<0.05
0.01
0.54
1.69
0.95
1220
79
Notes:
1. Uranium concentration in well waters varied and ranged from
180 pg/L to 450 pg/L.
2. pH of well water was 7.62.
3. Electric conductivity of well water was 2.36 m mhos/cm.
,2-
([U02(CO,)2] and [UOpCCO-),] ). The well water was passed through a pre-
filter to remove suspended impurities.
Two standard methods of uranium analysis were used by the student assis-
tants. These are given in the Appendix.
-31-
-------
SECTION 5
RESULTS AND DISCUSSION
The overall project operation went smoothly, although the well pump did
require replacing on two occasions. The student personnel from the Dona Ana
Water Utilities Training Program used the ion exchange system operation as
part of their technical education. The three graduate students involved
wrote Master of Science Thesis on work related to the project and one senior
did an independent study project related to the project. All the students
stated they gain a great deal from involvement in the project.
A. Ion Exchange System
The ion exchange system processed nearly four million gallons of well
water which contained on the average 300 ug/L uranium. The ion exchange
systems where constructed and operated similar to commercially available
equipment. Readily available commercial anion exchange resins were used.
Table 10 summarizes the operation (throughput) of the four columns which made
up the system. The table shows differential volume treated per period. Units
1, 2, and 3 were operated in a down-flow mode and Unit 4 was operated in an
up-flow mode. Although the operation was intended to try and maintain uniform
volume flow through all four units, this was not achieved.
Figure 17 gives the overall operation summary for the four units. Fig-
ures 18 through 21 show the breakthrough curves for each cycle of each unit.
These curves show the total volume of water processed during that cycle by
each unit. As seen in these figures, breakthrough to exhaustion did not occur
for each column during each cycle. (This is shown by the dashed portion of
the curve.) The columns were all regenerated in the same sequence at the same
time. The consistency of the four units operation can be seen by the similar
behavior as reflected in the breakthrough curves.
All four units operated well. Each resin tested was successful in
removing uranium from the raw water supply. Each resin stood up well under
the conditions of operation, as shown by the resin capacity calculations
summarized in Table 11. There was no degredation of the resin through foul-
ing or decrepitation during the project. As seen in Figure 17 and reflected
in Table 11, the results for Cycle 3 were not included. Examination of the
effluent sample analysis indicated that regeneration was not complete and
there was a period when the influent concentration varied widely, e.g. <100
ug/L to >500 ug/L. Also several effluent samples tested at greater concentra-
tions than the corresponding feed.
-32-
-------
TABLE 10. SUMMARY OF OPERATION OF ION EXCHANGE SYSTEM
This table shows the operational data taken on the ion-
exchange columns as operated between August 1983 and
October 1984.
Date
Gal. Treated
Per Period
Unit # Comments
9/15/83
11/10/83
12/12/83
1/15/84
3/6/84
4/25/84
5/10/84
5/15/84
5/16/84
24,580
78,360
88,840
84,790
42,390
116,060
128,780
102,680
67,020
128,910
185,070
165,780
119,970
189,280
236,570
216,730
130,670
301,620
333,090
369,850
152,670
343,620
369,090
417,850
159,700
350,540
374,940
426,120
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
Well pump keeps
shutting off, not
enough flow to run
IX and Vac. filter
Still not enough
Water to run
System
Need new well pump
Perhaps frozen
New pump installed
Higher Capacity -
Needs new wiring
-33-
-------
1.0
Unit 1
(DOMEX 21k)
0.0
1.0
Unit 2
(DOWEX SBR-P)
0.0
E
F
F
L
U
E
N
T
C
0
N
C
E
N
T
R
A
T
I
0
N
1.0
c/co Unit 4
(IONAC A641)
(Upflow) n n
1.0
Unit 3
(IONAC A641)
0.0
0
rio
10
10
10
20
1
20
20
30
30
30
40 50 60
Bed Volumes x 10~3
5060
Bed Volumes x 10~3
40
50 60
Bed Volumes x 10~3
20 30 40 50 60
Bed Volumes x 10"3
Figure 17. Summary of Operation of Ion Exchange Columns.
-------
Cycle 1
Cycle 2
en
1.
0>
^
O
O
CM
II
O
O
O
O
O
0.
1.
en
o
O
CO
o
o
o
o
0.
10 20
Bed Volume x 10'3
10.
Bed Volume x 10"
1.
O
in
o
o
o
o
0.
Cycle 3
~^T-
Cycle 4
0.
O
o
CO
o
o
o
o
u
0.
10 20
Bed Volume x 10-3
Figure 18.
10 f 20
Bed Volume x 10-3
Breakthrough Curves-Unit 1 (DOWEX 21k).
Bed Volumes Processed Measured for Each Cycle Separately.
-------
Cycle 1
Cycle 2
O>
3.
o
o
CM
en
3.
O
O
CO
o
I
0.
0.
10
Bed Volumes x 10"3
20
0.
3D
Bed Volumes x
en
3.
O
in
ro
o
o
o
o
Cycle 3
Cycle 4
1.
en
a.
o
o
CO
o
o
o
u
o
0.
10 20
Bed Volume x 10-3
2TJ
Bed Volume x 10-3
Figure 19. Breakthrough Curves-Unit 2 (DOWEX SBR-P).
Bed Volumes Processed Measured for Each Cycle Separately.
-------
Cycle 1
Cycle 2
0>
O
o
CM
o
o
o
o
1.
0.
Bed Volumes x 10
20
-3
o>
3.
O
o
CO
o
o
o
u
I tO 30
Bed Volumes x 10"-3
Cycle 3
o
in
ro
o
o
o
o
0.
0.
10
Bed Volumes x
en
o
o
CO
o
u
o
u
0
Cycle 4
10
Bed Volumes
Figure 20. Breakthrough Curves - Unit 3 (IONAC A641-Downflow).
Bed Volumes Processed Measured for Each Cycle Separately.
TO
10-3
-------
Cycle 1
Cycle 2
O
o
CM
0>
O
CD
CO
O
o
o
o
(J
o
o
o
o
o
CO
CO
I
0.
10 20
Bed Volumes x 10'3
10.
Bed Volumes x 10"
Cycle 3
O
ID
CO
O
O
O
O
0.
0.
10 20
Bed Volumes x 10'3
o>
o
CD
CO
o
(J
o
o
0.
Cycle 4
0.
10 20
Bed Volumes x 10~3
Figure 21. Breakthrough Curves - Unit 4 (IONAC A641-Upflow).
Bed Volumes Processed Measured for Each Cycle Separately.
-------
Table 11. Resin Capacity for Each Unit Per Cycle
Unit 1
Cycle 1
Cycle 2
Cycle 4
Unit 2
Cycle 1
Cycle 2
Cycle 4
Unit 3
Cycle 1
Cycle 2
Cycle 4
Unit 4
Cycle 1
Cycle 2
Cycle 4
C0, M9/1
200
300
300
200
300
300
200
300
300
200
300
300
VT Gallons*
12 x 104
26 x 104
20 x 104
Average
28 x 104
24 x 104
15 x 104
Average
15 x 104
27 x 104
30 x 104
Average
13 x 104
36 x 104
20 x 104
Average
meg
ml
0.013
0.044
0.034
O35
0.031
0.040
0.025
O32
0.017
0.046
0.051
O35
0.015
0.061
0.034
OTU37
Bed- Volumes
Treated at
Breakthrough
8020
17375
13365
12520"
18712
16038
10024
I4~525
10024
18043
20048
15015
8687
24057
13365
15370
*The volume VT is the differential volume processed
during the cycle.
While industrial ion exchange columns have been used in an up-flow mode
for recovering uranium from mine and mill waters, very few drinking water
treating operations have used up-flow in their operation. As seen in Table
10 the up-flow unit, #4, processed a greater volume of water during several
cycles than the down-flow units although the operation procedure was to try
and balance the load on all units.
To better understand the operation of the up-flow mode, a small bench-
scale unit pilot-scale units were constructed and operated on a side stream
from the project feed. Table 12 summarizes these experiments. The resin was
IONAC A-641, the same as used in the large column. As seen in Table 12, the
beds were operated in an expanded mode, e.g. bed porosity fr°, but not in a
complete fluidized state (as seen by visual inspection). Tnese runs were
operated until at least 50 percent of breakthrough had been achieved and the
data analyzed for conventional mass transfer/adsorption parameters. The com-
plete analysis is available in Reference 20. These analyses confirmed that
-39-
-------
the Exchange Zone method (Figure 4) gives satisfactory analysis for up-flow
ion exchange.
TABLE 12. UP-FLOW COLUMN EXPERIMENT
Run
Resin Mesh Size 16-50 16-50 16-50 16-50
Resin Weights, g 29.11 58.4 45.2 1965.
Height of Packed Bed
(cm) 2.12 4.32 3.02 10.12
Bed Diameter 5.08 5.08 5.08 19.05
Height of Expanded
bed (cm)
CQ (ug/ml)
V (cmVmin)
U (cm/sec)
fr, dimensionless
4.02
422.
80.
0.0658
0.63
8.02
443.
86.
0.0707
0.62
6.32
444.
90.
0.0740
0.67
22.2
336.
3785.
0.2213
0.68
B. Chemical Coagulation
Effect of pH on Coagulation/Flotation Process
Different percentages of uranium removal were obtained with different pH
of solutions in chemical treatment for the samples which have the same dosage
of the ferric chloride solution. The results in Table 13 and Figure 22 show
the difference observed when pH of solutions were adjusted from 4 to 10 using
0.1 N NaOH and HC1 solutions with 30 mg/L
The results in Table 13 show that among these five pH solutions the
highest uranium removal efficiency by ferric chloride was found at pH 10
about 98 percent removal was obtained. The same dosage of coagulant was less
effective at pH 6 (31 percent of uranium was removed). However, at pH 6 the
removal was much higher than at pH 4 and 8. Only 31 percent and 18 percent
removals were obtained in pH 4 and pH 8, respectively, with the same dosage
of ferric chloride. No data are shown for similar experiments examining the
pH dependencies for other dosage of ferric chloride.
-40-
-------
TABLE 13. SUMMARY OF THE RESULTS FOR URANIUM
REMOVAL USING 30 mg/L OF FERRIC
CHLORIDE SOLUTION AT DIFFERENT pH
OF WATER SAMPLES.
Adjusted pH Uranium Remaining in Percent of
the Tested Water (ug/L) Uranium Removed
4.0
4.9
5.3
5.8
8.2
9.4
10.0
309
251
180
86
369
6
<2
31.3
44.2
60.0
80.9
18.0
98.7
99.0
The Effect of Variation of Ferric Chloride Concentration
In order to study the effects of variations in the different dosages of
ferric chloride solution, the best conditions of pH dependencies test have
been chosen i.e., the solution which had pH 6 and 10.
The results shown in Table 14 indicate that the uranium removal percent-
age increased with increased dosage in the 30 to 90 mg/L range on the effect
of the ferric chloride as a coagulant at the pH 6. The results show the
uranium removal efficiency in this range was about 81 percent to 87 percent.
The same experiments also have been done at pH 10 with varying amount of
ferric chloride concentration. The results show that greater than 99 per-
cent of the uranium was removed at a ferric chloride dosage of 30 mg/L and
above (Table 14).
Experimental data for the effect of dosage of the ferric chloride and
the equilibration pH of the solution, show that pH is a major controlling
factor in the removal of uranium from drinking water via the coagulation/
filtration process. Proper choice of pH is a requirement for effective
-41-
-------
K)
0>
>
O
6
0)
C
«J
t-
o>
O
(-1
Q)
P-,
100.
80.
60.
40.
20.
0.
3.
4.
5.
6. 7.
PH
8.
9.
10
O-DATA OF LI WU
A-DATA OF Y. Y. CHEN
Figure 22. Chemical Clarification Operation
-------
TABLE 14. THE EFFECT OF FERRIC CHLORIDE
CONCENTRATION ON URANIUM REMOVAL
EFFICIENCY IN A SOLUTION OF pH 6
and pH 10.
Uranium Remaining in Percent of
FeCl3 Dosage (tng/L) the Tested Water (ug/L) Uranium Removed
At Ph 6
30
60
90
86
60
59
80.9
86.6
86.9
At Ph 10 30 <2 >99.0
60 <0 100.0
90 <0 100.0
chemical coagulant treatment. The results of the pH dependence for uranium
removal with iron salt can be interpreted by the stability and charge char-
acteristics of uranyl species and metal hydroxide precipitates at the
adjusted pH of the solution. No data are shown for similar experiments
examining the pH dependencies for other dosages of ferric chloride. However,
the similar results can be expected from the physicochemical properties of
the metal hydroxide formed-during the coagulation process at a given pH and
the dominant uranyl species in the solution. The role of ferric hydroxide as
a coagulant formed from ferric chloride in aqueous solution as well known.
The stability, solubility and reactivity (adsorption) of the hydroxides are
pH dependent. The pH dependence of the distribution of uranyl species in
natural water is also known. Dominant uranyl species and charge characteris-
tics of iron hydroxide floe at pH 4, 6, 8 and 10 are shown in Figure 23_. At
a low pH, say, pH < 5, ferric hydroxide has a positive charge, Fe(OH)- , and
the uranyj carbonate complex also dissociates to a positively charged uranyl
ion (UOp , UOjOH ). Therefore, at pH 4, less adsorption can occur due to the
strong repulsion between ferric hydroxide and uranyl ion. QAs the pH was
increased to 6, the dominant uranyl complex would be UOpCO- with a small posi-
tive charge of uranyl ions, and the mixed charge of negative and neutral
hyrolyzed ferric iron would be negative. The possible reason is that at
least there is no repulsion between the two charge particles, UO?CO_ and
Fe(OH).. At pH 8, the same phenomenon was observed as at pH 4 except both
interacting species would have negative charges. Therefore, a lower removal
efficiency was obtained in both conditions (pH 4 and 8). When the+pH exceeds
9.5, the U02(C03)?~ species are known to be stable, but (U02)-(OH)t- would be
the dominant species in carbonate-depleted water. The carbonate in water
-43-
-------
100
Figure 23. Combined Distribution of Species in
Chemical Clarification System
-44-
-------
could be depleted by CaCO, precipitation during the coagulant treatment pro-
cess. The reaction is shfiwn as follows:
3 U09(CO,)J~ + 15 Caz + 2 Fed, + 15 OH%
^(UOp3(OH)5 + 12 CaC03(s) * 2 Fe(OH)g (s) + CaCl£
Since the charge of the ferric hydroxide is pH-dependent, it should
adsorb on its active surface sites the stable uranyl complex through electro-
static attraction. Therefore, minimum uranium removal was observed when the
charge of the uranyl species was the same as the charge of the floes, and
maximum removal occurred when the charges were opposite or neutral.
The experiments in this work examining the effects of pH of the solutions
and variation of coagulant concentration in the coagulation/filtration con-
tinuous process with ferric chloride were similar to previous investigations
using ferric sulfate and aluminum sulfate as coagulants in batch tests. The
optimum dosage is related to the type of coagulant and the amount of uranium
contained in the water. However, for the optimum effectiveness of uranium
removal with different iron salts (ferric sulfate and ferric chloride), the
results are essentially the same.
C. Monitoring Program
A detailed search was conducted to locate currently operating conven-
tional water treatment facilities with uranium in their feed supplies. For
this search a minimum level of 15 ug/L of uranium was arbitrarily selected
and conventional water treatment facilities were defined as any type of
treatment facility more "complex" than sand separation and chlorination.
The wide variety of state and U.S. government agencies responsible for
the radiological monitoring of water supplies made the search very time con-
suming. A total of 34 municipal systems and an additional 21 municipal we-lls
were located in a 6 state area (Table 15). Of these 55 possible study site's,
only 4 provide treatment above and beyond sand separation and chlorination.
Three of these cities are located in Colorado (Denver, Arvada, and North
Table Mountain) and they all draw their water from the uranium contaminated
Ralston Reservoir. The fourth city is a small city in South Dakota
(Harrisburg). These communities were contacted and with the exception of
North Table Mountain they agreed to help us with this study.
The City of Arvada treats their water using a microfloc system. The
system employs alum and Separan (a polyelectrolyte) to create the microfloc.
The water is then passed through mixed media filters. A 125 cc water sample
is collected daily from the raw water and the treated water to form monthly
composite samples. For the past couple of years, these monthly samples have
been tested for uranium. The raw water from Ralston Reservoir contains from
less than 1 ug/L to 36 ug/L with an average of 14.7 ±9.6 (SD) ug/L of urani-
um. The Arvada facility removed from 18 to 90 percent of this incoming urani-
um with a mean efficiency of 67 ± 15 percent. The meticulous nature of these
records and the time span of the monitoring clearly indicates that conven-
-45-
-------
TABLE 15. STATE SURVEY FOR URANIUM ENRICHED WATERS
Colorado, Water Quality Control Division of Colorado Dept. of Health, 3
community composites, 1-24 pCi/L, conventional alum systems.
New Mexico, Water Supply Section of New Mexico Environmental Improvement
Division, 2 community composites and 15 municipal water sources, 10 - 110
pCi/L, sand filters and chlorination.
Oklahoma, Association of Central Oklahoma Governments, 22 communities (about
200 wells) near Oklahoma City, 10-190 pCi/L, sand removal and chlorination.
South Dakota, S.D. Dept. of Water and Natural Resources, 1 community, 12
pCi/L, iron and manganese removal, chlorination.
Texas, Texas Water Hygiene Division, 5 community composites and 6 individual
water sources, 10-55 pCi/L, sand removal and chlorination.
Wyoming, EPA Drinking Water Branch, Denver, 1 community (Cheyenne), 29 pCi/L,
this well water is chlorinated, they also use surface water which is
treated with conventional alum.
NOTE: This survey is representative of existing state and federal records
and may not be complete.
tional water treatment facilities can greatly reduce the uranium content of
natural waters.
The Moffat Treatment Facility (Denver) also draws water from Ralston
Reservoir and has been keeping uranium records for about two years. Unfor-
tunately, Denver's monthly samples are "grab samples" and correlation between
raw and treated waters are less meaningful. Using their records, uranium
removal efficiency was 78 ± 190 percent.
Harrisburg, S. D. treats about 100,000 gpd using aeration, KMnO. green-
sand filters, chlorination and fluoridation. No previous data exists on the
uranium removal efficiency of this facility.
Harrisburg, South Dakota, is a town of about 500 people located near
Sioux Falls. Because the Harrisburg water supply contains some 3 parts per
million (ppm) of iron and quite a bit of manganese, a small water treatment
facility was constructed in 1970. In late 1980, the Water Quality Division
of the S.D. Department of Water and Natural Resources determined that the
Harrisburg water also contained about 20 ug/L uranium.
Harrisburg appeared to meet the two criteria of our study; an elevated
concentration of uranium in the raw water and a water treatment facility more
involved than sand separation and chlorination. To determine if the
Harrisburg system still contained uranium, we requested Ross Abbott of
Harrisburg to ship us a raw and a treated water sample. We had these samples
-46-
-------
tested fluorometerically in June of 1983. Both samples contained about 19
ug/L of uranium.
A visit to Harrisburg was conducted in July of 1983. In brief, the water
treatment process consists of permanganate, counter-flow aerator-degassifier,
chlorine, dual green sand filters (66 inches in diameter and 24 inches deep)
and fluoridation. This facility treats 0.07 and 0.1 millions gallons per day.
A sample of the raw, treated and backwash water was collected.
A standard fluorometeric test was run on the raw and treated water sam-
ples. Because of the tremendous iron concentration in the backwash sample
an isotopic uranium analysis was run on this sample. The isotopic tests
includes a U-232 spike, ion exchange column, electrodeposition on a planchet,
alpha spectrometer and a multichannel pulse height analyzer.
The test results show that the water contained about 13 ug/L of uranium
before treatment and about 15 ug/L after treatment. The increased uranium
in the treated water is attributed to the fickle and statistical nature of
the fluorometeric test. The backwash sample contained 9.6 ± 1.1 picocuries
per liter of alpha activity. This radiation is the sum from U-234, U-235
and U-238. The backwash water contains about 15 ug/L of uranium (assuming
1 pCi/L equals 1.5 ug/1). Therefore, the backwash water from the green sand
filters contains the same concentration of uranium as the raw water. Since
the backwash water is the only possible outlet for the uranium in the water,
we can conclude that a permanganate/aeration/green sand filter system does
not remove uranium from drinking water. These results are shown in Table 16.
TABLE 16. RESULTS OF HARRISBURG, S.D., OPERATION ANALYSIS
Sample
Feed
Product
Feed
Product
Backwash
Date
Collected
6/83
6/83
7/83
7/83
7/83
Type of
Analysis*
Total Uranium
Total Uranium
Total Uranium
Total Uranium
Total Uranium
u/1
19
19
13
15
9.eti.i"
Backwash was accomplished with product water; measurement pCi/L
-47-
-------
The results of these tests are commensurate with theoretical considera-
tions. Dissolved iron in water is in the ferrous state, Fe (a cation).
This iroll's removed by aeration which changes the ferrous iron to ferric
iron, Fe . The ferric iron combines with hydroxide to form a filterable
solid, Fe(OH)~. In waters of a neutral_pH, the uranium_exists as uranyl car-
bonate complex ions (anions), U02(C03)2 and UCLCCO,), . In an oxidizing
environment, uranium always changes to the very soluble hexavalent state. In
fact, if an insoluble uranium precipitate, U0? were introduced into the
aerator it would quickly dissolve into the water as uranyl carbonate complex-
ions.
In summary, a permanganate/aeration/green sand filter system does not
remove any uranium under normal operating conditions. This conclusion is
born-out by both theoretical considerations and as the result of a testing
program of an operating greensand filter system in South Dakota.
D. Waste Disposal
As previously described, uranium contaminated drinking water is a common
problem, particularly in the western United States. If regulations are
accepted and enforced, many communities will be required to remove the uran-
ium from their drinking water supplies. Removal of this radioactive element
produces a new problem—that is, a problem of radioactive waste disposal.
The physical form of this waste and the longevity of uranium cause difficul-
ties in formulating a waste disposal plan. Three disposal alternatives were
considered in this study. These include: dilution/release, reuse or resale,
and burial; however, the choice must also be based on environmental acceptabi-
lity. Each participating community will have unique drinking water and waste
characteristics. For this reason, it is not possible to prescribe one solu-
tion to the problem. Each community must consider their situation and choose
the optimum plan on that basis. Two cases are analyzed: (1) 150,000 gallons
per day with 30 ug/L uranium, and (2) 150,000 gallons per day with 200 ug/L
uranium. Table 17 summarizes the quantity of uranium to be disposed.
TABLE 17. Ultimate Disposal Values for Uranium
System Capacity: 150,000 gallons/day
Case 1: Concentration 30 ug/L Total Uranium = 1.79 g/day
(a) Ion Exchange Treatment - Regeneration Solution
1500 gallons @0.3 mg/L
(b) Chemical Clarification - Filter Cake
17.03 kg Cake
Case 2: Concentration 200 ug/L Total Uranium = 11.35 g/day
(a) Ion Exchange Treatment - Regeneration Solution
1500 gallons @2.0 mg/L
(b) Chemical Clarification - Filter Cake
17.03 kg Cake
-48-
-------
Dilution and Release
The first alternative considered is to dilute, if necessary, and return
the uranium to surface waters. In New Mexico, the New Mexico Water Quality
Control Commission regulates such releases. They have established a maximum
allowed concentration of uranium of 5 mg/L to be returned to surface waters.
In case 1 (Table 15) the ion exchange regeneration solution could be dis-
charged directly to the surface, i.e. 0.3 mg uranium/L concentration. In
Case 2, the average regeneration solution could be discharged directly to sur-
face waters, i.e. 2.0 mg uranium/L. Note that this is the average concentra-
tion of the regeneration solution. During operation of the ion exchange
system, samples of regeneration solution were obtained which exceed 25 mg/L
uranium. These samples were taken shortly after starting regeneration and
were highly concentrated in uranium.
If the treatment is chemical clarification, the waste product is a fil-
ter cake. For Case 1, this cake would contain 0.057 nanocuries uranium per
gram and in Case 2 the cake would contain 0.38 nanocuries per gram. (These
calculations are based on 0.67 picocuries = 1 ug uranium). In both cases
these quantities are well below the 100 nano curie per gram definition of
low-level waste. (In analyzing the currently operating systems in Section C,
the practice was to send the flocculated sludge to sludge beds).
Reuse and Resale
A second alternative would be to ship liquid waste to a uranium mill,
where it could be processed along with the incoming ores. It could be pos-
sible to co-dispose sludge waste with the uranium mill tailings. Reuse or
sale of the waste would involve several steps. It would involve collection,
intermediate storage, shipment and possible packaging of the waste.
Collection does not pose any major problem and should be a relatively
inexpensive part of this process. Intermediate storage, however, presents
difficulties. A storage facility is necessary to hold waste until a large
enough volume is collected for a shipment. This facility would be some type
of holding tank from which trucks could be filled. It would require licens-
ing and a permit from the State Hazardous Waste Management division. This
is a new situation and would require an indepth study. Since uranium is an
alpha emitter, the tank construction material is not a critical factor for
preventing radiation exposure under normal operation.
Assume that the regeneration solutions from ion exchange is 1500 gallons
per day per 1000 people. Assume, also, that approximately 10,000 gallons of
waste will constitute a shipment. The storage period would be 6.5 days. Cost
involved in storage will be the initial cost of construction of the tank and
any maintenance required thereafter. A current rule of thumb for estimating
construction costs of steel tanks is about twenty-five cents per gallon of
volume. For a 10,000 gallon holding tank, the construction cost would be
about $2,500.
-49-
-------
The next stage of this alternative is shipment. Two mills are currently
operating in the U.S., both of which are located in northern New Mexico.
Shipping distances would thus be fairly short and feasible for most of New
Mexico, Colorado, and Arizona. As distances to the mills increase, however,
costs and disadvantages of this plan also increase. The risk associated with
accidents during shipping and storing the waste must be considered. Licens-
ing by the State Hazardous Waste Management Division would be reported for
shipment on public roads and highways. The following shipping cost estimates
are based on the 1983 Book Rental Rates for a 5000 gallon diesel powered
truck. The hourly cost for shipment by this means is about $10.25. This
does not include pay for the driver. Actual cost per mile would be about
$1.75 per loaded mile. Obviously this becomes very expensive over any large
distance.
Disposal Through Burial
The third choice available for either liquid or a sludge waste is
storage or burial. Burial of the waste at a commerical facility would involve
several processes similar to those involved in shipment of the waste to mills.
In addition to collection, intermediate storage, and shipment, a solidifica-
tion and/or volume reduction process would have to be included before shipment
could take place.
Collection again would be a relatively simple process. Intermediate
storage will be necessary as volumes are accumulated. This also would
require the type of structure discussed for use in shipment of the waste to
uranium mills. Either solidification or volume reduction is necessary prior
to shipment. A solidification process could be implemented for either a
liquid or sludge waste. Solidification can be achieved by several methods.
In-drum solidification is one process currently being used by the Department
of Energy. The waste (either liquid or sludge) is put into drums. Portland
cement is added, the drums are sealed and then tumbled for mixing. Vermicu-
lite is also used for solidification of sludge or liquid waste. Waste is
added to vermiculite filled drums. The vermiculite will expand and incor-
porate the waste in its matrix. A method which will solidify sludge waste as
well as reduce its volume, is vacuum filtration. Sludge cakes will result
from the vacuum filtration system. The filter cake containing 5-8 percent
moisture could be further dried or package with a cement binder as above.
It is difficult to predict costs for these processes because of their
recent development and difficulties in predicting sludge volumes. Cost of
shipment of solid waste would be comparable to that of shipping the liquid
waste. The solid form presents a much smaller shipping risk as compared to
the liquid waste, thus licensing would be easier to obtain.
In using ion exchange, the uranium is removed through the regeneration
solution. Evaporation to dryness would give a greatly reduced volume of
solid, i.e. 1.7 g/day Case 1 and 11.35 g/day Case 2, but would be impractical
from an energy standpoint. The alternative is to follow the ion exchange
process with chemical clarification. Since the volume to be treated is the
regeneration solution, the scale of the chemical clarification unit will be
-50-
-------
greatly reduced from that of the ion exchange system. Although it was not
tested by this project, the quantity of filter cake would be comparable to
that previously discussed.
E. Design Analysis
The main purpose of this section is to present cost analysis for the
application of ion exchange/chemical clarification for removal of uranium
from drinking water supplies. Complete cost estimation has been covered by
a recent EPA publication "Estimation of Small System Water Treatment Costs."
[19] This publication treats ion exchange both for centralized treatment
systems and for point-of-use treatment systems. There are currently other
projects examining ion exchange point-of-use treatment for removing uranium
and this is not included in this section. It should be noted that the excep-
tional capacity for uranium of standard ion exchange resins makes point-of-
use applications feasible on a one-cycle charging of resin, then collection
by a service company for regeneration as is practiced in any point-of-use
water-softening applications.
Ion Exchange
An ion-exchange system usually consists of the exchange resin (cation or
anion), with provisions made for regeneration and rinsing. Prior to applica-
tion to the ion-exchange bed, wastewater may be subjected to pretreatment to
remove certain contaminants which may hinder the performance of the exchange
bed.
Input Data
(1) Wastewater Flow.
(a-) Average flow, mgd.
(b)~ Minimum and maximum flows, mgd.
(2) Cation and anion concentrations, mg/1.
(3) Allowable effluent concentrations, mg/1.
Design Parameters
(1) Type of resin.
(2) Resin exchange capacity, lb/ft3 (manufacturer's specifications).
(3) Regenerant dosage, lb/ft3 (consult resin manufacturer's specifica-
tions).
(4) Flow rates.
(a) Treatment flow rate (2-5 gpm/ft3).
(b) Regenerant flow rate (1-2 gpm/ft3).
(c) Rinsing flow rate (0.5-1.5 gpm/ft3).
(5) Amount of rinse water (30-120 gal/ft3).
(6) Column depth (24-30 in. minimum).
(7) Operation per day, hr.
(8) Amount of backwash water, gal/ft3.
(9) Regenerant level, lb/ft3.
-51-
-------
(10) Regenerant concentration, percent.
(11) Regenerant specific gravity.
(12) Backwash water rate, gpm/ft3.
One case study has been summarized in Table 18. Capital equipment cost is
given in Figure 24.
Chemical Clarification
The chemical clarification system consisted of a flotation cell for fer-
ric hydroxide colloids precipitation and adsorption of the uranyl ion complex
and a vacuum filter for sol id-liquid separation. The flotation design and
filter design are outlined as follows:
Capital equipment for the two systems of this study are shown in Figures
25 and 26.
Input Data
(1) Wastewater flow, mg.
(2) Solute solids concentration in the feed, mg/1.
(a) Average concentration.
(b) Variation in concentraiton.
Design Parameters. From laboratory or pilot plant studies.
(1) Air-to-solid ratio (A/S).
(2) Air pressure (P), psig.
(3) Detention time in flotation tan (DTFT), hr.
(4) Solids loading (ML, Ib/ft2/day).
(5) Hydraulic loading (HL), gpm/ft2.
(6) Detention time in pressure tank (DTPT), min.
(7) Float concentration (Cp), percent.
Filtration
Input Data
(1) Volume of slurry to be dewatered, gpd.
(2) Initial moisture content of slurry, percent
Design Parameters
(1) Final moisture content or slurry, percent.
(2) Specific resistance, secVgm (Buchner funnel test).
(3) Applied vacuum, psi.
(4) Fraction of cycle time for cake formation (formation time/cycle
time), depends on degree of submergence.
(5) Cycle time, min (usually 1.5 to 5 min).
(6) Filtrate viscosity, centipoises.
(7) Chemical dose, percent of dry weight in solids fed to filter.
(8) Operation per week, days.
(9) Operation per day, hr.
-52-
-------
TABLE 18. ION EXCHANGE SYSTEM DESIGN
Based on Avg. Influent Cone, of:
84 (jg/L as (U02)2 (COg) = 50 ug/L as uranium
Using Avg. Capacity of 3.3 KGR./Ft3 Grains as U02
600 S|l x .084 ppm x 0.0547 9ra^9a1 x 24 hrs x 60 ^ = 3960 grains
3960 grains feed. (1 day basis)
Resin Needed for 1 Days Operation:
3.96KGR=120Ft3
3.3 KGr/Ft3
Size of Tank:
Assume 10 GPM/Ft3
600 GPM = 6Q Ft2
10 GPM/Ft* DU ri
Use 9J_ diam. Tank = 63.6 Ft2
Use 3' Bed Depth
Resin Quantity = 63.6' x 3' = 191' Ft3
Using 100% Freeboard for B/W and/or expansion
60" side shell tank
Tank Size:
9' <|) x 51 SS.
Approximate Running Time Between Regenerations:
191 Ft3
1.20FWday
Salt Needed/Regeneration
15#/Ft3 x 191 Ft3 = 2?865#
Estimated Capital Equipment Cost (Figure 24) $125,000.
(10) Loading rate, Ib/ftVhr.
(11) Number of units.
One case study for chemical clarification is summarized in Table 19.
The cost of curves for the three unit operations considered were updated
in 1985 costs using the CE plant, cost index data published in Chemical
-53-
-------
I0N EXCHANGE
o:
s
a
en
a
z
<
'Si
s
10
<*•
CO
c:
Q
01
a
"Z.
c/i
s
Capacity gpm
Figure 24. Cost of Ion Exchange (1985)
FL2TATIBN
Capacity cu ft
Figure 25. Cost of Flotation (1985)
-54-
-------
VACUUM FILTRATI0N
en
s
D
<
<
S
I
1C
10 o
10°
riLTER SURFACE AREA.SQ
Figure 26. Cost of Filtration (1985)
TABLE 19. CHEMICAL CLARIFICATION SYSTEM DESIGN
Based on Avg. Influent Cone, of: 0.5 mg/1
Feed Rate: 60 gpm
Joy Manufacturing Flotation Cell 16 cu. ft.
Approximately two minutes residence time
Ferric Chloride Addition 25 mg/1
35.7 Ibs/day
pH Adjustment Assuming Feed at pH 7 -»• pH 10
2 liters Cone (37%) HC1 per day
Continuous Rotary Vacuum (Precoat) Filter
6 ft2 Total Surface Area
Filter-aid Addition 2x weight of Ferric Chloride (Diatomaceous earth)
70 Ibs/day
Total Filter production per day (dry weight and wet weight 8% moisture)
106 Ibs dry weight
115 Ibs wet weight
Estimated Capital Equipment Cost (Figures 25 and 26) $44,000
Engineering. The index used was 375.2 (July 1985) based on 1957 - 59 as 100.
These cost curves were scaled for the specific processing of uranium.
-55-
-------
REFERENCES
1. Reid, G. W., P. Lassovszky, S. W. Hathaway, Proceedings of National
Workshop on Radioactivity in Drinking Water, Washington, D.C., 1983.
2. "Radioactivity in Drinking Water", Health Effects Branch, Office of
Drinking Water, EPA 570/9-81-002, Washington, D.C., January 1981.
3. Water Quality Criteria 1972 EPA R3-73-033 Com. on Water Quality Criteria,
Washington, D.C., March 1973.
4. National Interim Primary Drinking Water Regulations, EPA-570/9-76-003,
Washington, D.C., September 1976.
5. Drury, J. S., D. Michelson and J. T. Enswinger, "Methods for Removing
Uranium from Drinking Water", EPA-570/9-82-002, September 1982.
6. White, S. K. and E. A. Bondietti, "Removing Uranium by Current Municipal
Water Treatment Processes", JAWWA, pp. 374-380, July 1983.
7. Pourbaix, M., Atlas of Electrochemical Equilibria in Aqueous Solutions,
Trans, by J. A. Franklin, Pergamon Press, New York, p. 203, 1966.
8. Cotton, F. A. and G. Wilkinson, Advanced Inorganic Chemistry, Inter
science Publishers, (John Wiley & Sons) New York, pp. 1089-1096, 1972.
9. Her, R. K., The Chemistry of Silica, John Wiley and Sons, New York,
p. 595, 1979.
10. Sorg, T. J. and G. S. Hogsdon, "Treatment Technology to Meet the Interim
Primary Drinking Water Regulations for Inorganics," JAWWA, Vol. 72 (7),
pp. 411-422, July 1982.
11. Lee, S. Y. and E. A. Bondietti, "Removing Uranium from Drinking Water by
Metal Hydroxides and Anion-Exchange Resin," JAWWA, pp. 536-540, October
1983.
12. Culp, R. L., G. M. Wesner and G. L. Culp, "Handbook of Advanced Waste-
water Treatment," Van Nostrand Reinhold Company, New York, pp. 22-28,
89-94, 1978.
13. "Process Design Manual for Suspended Solids Removal," U.S. Environmental
Protection Agency, EPA 625/l-75-003a, Cincinnati, Ohio, January 1975.
-56-
-------
14. Helfferich, F., Ion Exchange, McGraw-Hill Book Co., New York, 1962.
15. Paterson, R., An Introduction to Ion Exchange. Heyden and Sons, Ltd.,
London, 1970.
16. McClanahan R., "Demonstration of Three Ion Exchange Materials for Treat-
ing Uranium Mine Water," Master of Science Thesis, Civil Engineering,
New Mexico State University, 1980.
17. Emelity, L A., J. R. Buchholz and P. E. McGinnis, "Review of Radioac-
tive Liquid Waste Management at Los Alamos", LA-UR-77-1195 and "Liquid
Waste Management Systems at Los Alamos", IAEA-SM-246/33, LANL, Los
Alamos, New Mexico, 1980.
18. "Water Quality Control Commission Regulations," State of New Mexico,
WQCC 81-2, Santa Fe, New Mexico, July 1981.
19. Gumerman, R. C., B. E. Burn's and S. P. Hansen, "Estimation of Small
System Water Treatment Costs", NTIS PB85-1610644, Springfield, Virginia,
1985.
20. Ho, C. D., "Up-Flow Ion Exchange", Master of Science Thesis, New Mexico
State University, Las Cruces, NM, 1984.
-57-
-------
APPENDIX A. METHOD OF URANIUM ANALYSIS
The first method needs a blank pellet, a sample pellet, and a sample
with an uranium spike pellet to analyze a sample. The detailed steps are
described as below.
Reagents
All chemicals were of analytical grade ("Baker Analyzed" Reagent, J. T.
Baker Chemical Company, Phillipsurg, New Jersey). Aqueous reagents were pre-
pared in deionized water. The following chemicals were used in the uranium
analysis:
1. Flux mixture: Mix together 9 parts of sodium fluoride (NaF), 45.5
parts of sodium carbonate (Na^CO,) and 45.5 parts of potassium carbonate
(I^CO,) by weight in a ball mill until it becomes homogeneous.
2. NTtric acid (1+1): Mix one volume of nitric acid HNO,(sp. gr. 1.42)
with one volume of water.
3. Nitric acid (1+9): Mix one volume of nitric acid HN03 (sp. gr. 1.42)
with nine volumes of water.
4. Nitric acid (1+99): Mix one volume of nitric acid HN03 (sp.
gr. 1.42) with ninety nine volumes of water.
5. Potassium pyrosulfate (K^S^Oy): For cleaning the platinum dishes.
6. Uranyl nitrate (UOptNOOp): For preparation of uranium standard
stock solution.
Uranium Standard Stock Solution
1. Prepare 1000 mg/1 of uranium solution as follows:
a. Weight 1.0549 grams of uranyl nitrate and dissolve it in 20 ml
of HN03 (1+1) solution.
b. Slowly evaporate to near dryness.
c. Dissolve residue with 10 ml of HNO, (1+9) solutions.
d. Transfer to 500-ml volumetric flasR.
e. Dilute to 500 ml with HN03 (1+99) solution.
f. Mix solution and transfer to a clean polyethylene bottle.
2. Prepare 50 mg/1 of uranium solution as follows:
a. Pipet 25 ml of 1000 mg/1 of uranium standard solution into a
500-ml volumetric flask.
b. Dilute to 500 ml with HN03 (1+99).
c. Mix well and transfer to a clean dry polyethylene bottle.
-58-
-------
3. Prepare more dilute uranium solution as follows:
a. Dilute appropriate volumes of 50 mg/1 of uranium solution with
HN03 (1+99) solution.
b. Consecutive tenfold dilutions of the 10 and 5 mg/1 uranium
standards are used for improving accuracy in preparing the more
dilute standards.
c. Mix each standard well and transfer to a clean, dry polyethylene
container.
Procedure of Uranium Assay
0. The procedure is for samples greater than 20 ug/1.
1. Transfer two 100-ul aliquots of the filtered samples to each of two
platinum dishes and evaporate to dryness in a drying oven at 103 °C.
Drying takes about ten minutes.
2. To one of the platinum dishes add 100 ug/1 of uranium standard solu-
tion.
3. Evaporate to dryness in a drying oven at 103 °C.
4. Weigh out 400 ±4 mg flux into each of the two platinum dishes
where the flux is the mixture of 9 parts of NaF, 45.5 parts of
Na2C03 and 45.5 parts of K2C03 by weight.
5. Prepare a blank flux sample by weighing out 400 ±4 mg flux into a
clean platinum dish.
6. Place the three platinum dishes on a stainless steel plate and put
into a preheated muffle furnace at 625 °C for 15 minutes.
7. Remove from furnace and cool in a desiccator for 30 minutes.
8. Read fluorescence in a fluorometer.
Cleaning the Platinum Dishes
1. Remove the disk from the platinum dish.
2. Wash the dish in hot water.
3. Fuse each dish with potassium pyrosulfate (I^SpOy).
4. Wait until it cools down.
5. Dissolve the residue in hot water.
6. Store the dishes in dilute HNO, (1+9) solution until needed.
7. Rinse in water prior to use next time.
Fluorometeric Determination
1. Choose the appropriate lamp and filters (excitation and emission
filter) for the analysis to be performed.
2. Select the desired range and pull out the "Range Selector" knob.
The numbers IX, 10X, and 30X indicate the approximate increase in
sensitivity which is obtained in order to select the appropriate
setting for emission energy.
3. Insert the blank disk containing a reagent blank into the "Uranium
Pellet Holder Door" (which is designed to accept pellets fused from
0.5 g of flux in platinum dishes).
4. Close the door and turn on the "Power" rocker switch.
-59-
-------
5. Momentarily depress the "Lamp Start" button and release. Confirm
lamp ignition by observing the lamp indicator.
6. Allow 15 minutes for warming up.
7. Set the End Point or Kinetic Chemistry Mode switch to the appropri-
ate position. In End Point Mode, the QA lamp indicates when to take
a reading. In Kinetid Mode, the QA lamp is deactivated and readings
are taken as required for the procedure being used.
8. Using the "Coard Blank Control", adjust the display to read any
value from -01.0 to +01.0. Then adjust the Fine Zero Control to
0.000 +0.02.
9. Remove the blank sample.
10. Insert the sample into the "Uranium Pellet Holder Door".
11. Repeat steps 4 and 5 and record the output.
12. Note that the pellets should be handled with tweezers.
Calculation
Calculate the uranium concentration in micrograms per liter as follows:
Uranium, ug/1 = (((R-Rb)/(T -R ))*a)/V (3-1)
where
R = Reading of the blank
R? = Reading of the blank
R ° = Reading of the spiked sample
a = Mass of the uranium spike, ug
V = Initial sample size in liters.
Precision
The standard deviation, S, is calculated from the equation:
S - ((I(s-) - (Ix.)2/N)/(N-l))'5 (3-2)
where n 1
2
Z(x-) = summation of the squares of the individual results
(Zx.™ = square of the summation of the individual results
N = number of results
The coefficient of variation, CV, is calculated from the equation:
CV = 100S/X (3-3)
where
S = standard deviation from the equation (3-2)
X = mean value of the individual results.
Second Method of Uranium Analysis
With the second method, the data can be graphed and the unknown concen-
tration can be read from the graph. The second method resembles the first
method except the reagents and procedure of the uranium assay is as below.
-60-
-------
Reagents
Reagents are the same as in the first method except the flux. The flux
mixture contains 98 parts of sodium fluoride (NaF) and 2 parts of lithium
fluoride (LiF) by weight.
Procedure of Uranium Assay
1. Pipet 0.1 ml aliquots of blank reagent solution, 0.01, 0.05, 0.75
and 0.1 mg/1 uranium standards into five of the platinum dishes.
2. Evaporate blank sample and standard to dryness in a drying oven at
103 °C.
3. Add one scoop of flux into each dish.
4. Put into a preheated muffle furnace of 980 °C for 15 minutes.
5. Remove from furnace and cool in a desiccator.
6. Read fluorescence in a fluorometer.
7. Plot fluorescence versus uranium concentration as a calibration
curve.
8. Repeat steps 2, 3, 4 and 5 to make one blank (for setting a zero
point) and sample pellets.
9. Repeat step 6.
10. Read concentration from calibration curve.
-61-
------- |