\
V
600/0-^89/020
REMOVAL OF URANIUM FROM DRINKING HATER
BY CONVENTIONAL TREATMENT METHODS
by
Thomas J. Sorg
Prinking Water Research Division
Risk Reduction Engineerng Laboratory
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Drinking Water Research Division
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
January 1989
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse
ment or recommendation for use.
ii
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before compter" nBnil »».,«.¦»»»«
t, REPORT NO. 2.
600/D-89/020
4. TITLE AND SUBTITLE
REMOVAL OF URANIUM FROM DRINKING WATER RY
CONVENTIONAL TREATMENT METHODS
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7, AUTHOfUS)
THOMAS J. SORfi
S. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND AOORE5S
DRINKING WATER RESEARCH DIVISION
RISK REDUCTION ENGINEERING; LABORATORY
CINCINNATI, OHIO 45268
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
DRINKING WATER RESEARCH DIVISION
RISK REDUCTION ENGINEERING LABORATORY
CINCINNATI, OHIO 45268
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The HSEPA currently does not regulate uranium in drinking water but will be
revising the radionuclide regulations during 1989 and will propose a maximum
contaminant level for uranium. This paper presents treatment technology infor-
mation on the effectiveness of conventional methods to removal uranium from
drinking water. Treatment information based primarily on laboratory and pilot
pi ant studies is presented on conventional coagulation/fil tration, ion exchange,
lime softening, and reverse osmosis. Ion exchange treatment has been applied
successfully on ground waters by small systems.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. cosati Field/Group
18. DISTRIBUTION STATEMENT
Release to the public
19. SECURITY CLASS (ThisReportj
unclassified
21. NO. OF PAGES
2g'
20. SECURITY CLASS (This page)
unclassified
22. PRICE
EPA Form 2220-1 (R»v. 4-77) previous coition is obsolete
-------
-------
REMOVAL OF URANIUM FROM DRINKING WATER
BY CONVENTIONAL TREATMENT METHODS
Thomas J. Sorg
Drinking Water Research Division
Risk Reduction Enrigineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
INTRODUCTION
The U.S. Environmental Protection Agency's (USEPA) drinking water
regulations currently do not contain a maximum contaminant level (MCL)
for uranium. However, sometime during 1989, the Agency will he propos-
ing new regulations for radionuclides in drinking water and the Agency
has indicated that these regulations will contain a proposed MCL for
uranium. Presently, the Agency is reviewing the health effects data
and depending on the outcome of this review, the MCL could be as low as
10 pCi/L or as high as 40 to 50 pC1/L.l Recently, the State of Califor-
nia has proposed a MCL of 20 pCi/L.
Uranium occurrence data reported by the IJSEPA Eastern Radiation
Facility, Montgomery, AL^ in 1983 and by Longtin^ in 1987 indicates
that a number of water supplies will exceed the MCL even if it is set
as high as 50 pCi/L. Treatment to remove the uranium will be one
alternative for consideration by the water utilities in non-compliance
with the uranium MCL.
Because uranium has not been regulated, treatment methods have
not been extensively investigated nor has full scale treatment been
widely applied to water supplies. With the anticipation that a MCL
would likely be established in the future, EPA sponsored several labor-
atory and pilot plant studies during the past 4-5 years. Because
uranium is found in surface and ground waters, these studies investi-
gated a variety of methods including conventional coagulation, lime
softening, ion exchange and reverse osmosis. Of these methods, ion
exchange treatment has received the most attention and several small
full-scale ion exchange systems have been constructed.
Uranium Chemistry
Uranium (atomic weight - 23R) is a radioactive element having
four oxidation states usually represented by U , U , U0^,+( +5) and
UOg (+6), Of these, U+3 and U02+(+5) are unstable. Uranium is a very
reactive element readily combining with many elements to form a variety
of complexes. The uranium containing oxygen compounds and the common
uranyl (UO^) ion can combine easily with CI", NO,, SO^"^ and CO3 .
In aerated aqueous solutions at pH _^2.5, the uranyl ion is very
stable. Near pH 7, the uranyl ion form stable complexes with phosphate
and carbonate, the latter being the most significant for drinking water
supplies. In natural waters of pH 7-10, the soluble carbonate complexes
of U02 are the predominant anion species—UO^COj^ and U02(C03)3"4.
1
-------
As shown in Figure 1, these two complexes exist in various ratios
depending on the pH of the water.4 Above pH in and in carhonate-free
waters, hydroxide complexes (cation) are formed, UO^OH"1", (HOo)p(0H)?+z
and (UO2),(0H)J. Knowing which uranium ion species is actually present
in the water supply is critical for the selection of the treatment
process and successful operation.
TREATMENT METHODS
Conventional Coagulation/Filtration Treatment
During the past several years, the USEPA has sponsored two research
studies on the removal of uranium from drinking water at Oak Ridge Na-
tional Laboratoryand New Mexico State University®, lee and
Rondletti6 conducted jar test studies for the removal of uranium (83
ug/L) from pond water (low level radioactive waste settling basin)
using both iron (ferric and ferrous sulfate) and aluminum (alum) coagu-
lants. To simplify the analytical procedure, the pond water that
contained 1)238 (alpha emitter) was spiked with a tracer, n?37, a gamma
emitter, and the U"7 used to measure the effectiveness of the treat-
ment procedures. In these experiments, coagulant dosage ((1.5 - 30 mg/L)
and pH (4, 6, 8 and 10) were varied.
The results (Figures 2-5) of the laboratory experiments showed
that the removal efficiency was dependent on the coagulant dosage and
that the final pH of the test solutions. With ferric sulfate, the
highest uranium removals achieved were around 8n percent with a dose of
10 mg/L or more at pH 10, and with 20 mg/L or more at pH 6. At pH a
and 8, the maximum removals attainable were 18 percent and 44 percent
respectively with 25 mg/L of coagulant (Figure 2).
Ferrous sulfate results show a pattern similar to ferric sulfate
(Figure 3). Best removals (92 - 93 percent) were achieved with 20-25
mg/L at pH 10. For 25 mg/L doses, lower removals were obtained at the
other three pH values with pH 6 having better results (44 percent) than
either pH 4 (33 percent) or pH 8 (20 percent).
The test results with alum as the coagulant were similar to those
of the two iron coagulants (Figure 4). About 95 percent removals were
achieved at pH 10 with alum dosages of 10 mg/L or more. At pH 6, re-
movals did not increase as fast with increasing doses, but above 80
percent removals were obtained with 20 mg/L or more. At pH 4 and 8,
and dosages of 25 mg/L, removals were much less at 21 percent and 48
percent, respectively.
In bar graph Figure 5, uranium removal for each coagulant (25 mg/L
dose) are compared at pH 4, 6, 8 and 10. This figure clearly shows the
effect of pH and the difference between coagulants at the least effec-
tive pHs of 4, 6, and 8. At pH 10, 1itt1e difference in removals exist
between the three coagulants.
Lee and Bondiettifi suggested that the pH dependency of uranium re-
moval was related to the stability and charge characteristics of the
2
-------
uranyl species and the metal hydroxide precipitates at the adjusted pH
of the solution. For example, at pH 6 the predominant uranium complex
would be the neutral 1102003° (Figure 1) and the charge of the hydroxide
precipitates would also be near zero. In the pH 7-9 range, the princi-
pal uranium complexes would be the negatively charged tKWCOg).?"2 and
U0j,(C03)3"4 and the metal hydroxides would also be negatively charged.
In this pH range, therefore, lower uranium removal would be expected
because the interacting species would both have the same negative
charge. As the pH exceeds 9.5, the positively charged uranium hydrox-
ide (UOg^fQH)! complex is formed that might then coagulate with the
negatively charged metal hydroxide. Alum, however, is not recommended
to be used above pH 10 because the solubility of aluminum hydroxide
increases. At the low pH of 4, a similar situation exist as occurs at
pH 7-9. Both the uranium species IJO? and the metal hydroxides have
the same positive charge that would discourage removal.
To evaluate the role of carbonate, Lee and Bondietti6, conducted
experiments with and without aluminum sulfate and ferric sulfate in car-
bonate depleted solutions. The test results showed that in the pH
range of 5.5 to 9.0 more than 90 percent of the uranium was removed
with ferric and aluminum sulfate. Even in the absence of the coagulants,
up to 60 percent of the uranium was removed. These data suggest that
the hydrolyzed uranyl complexes are probably the positively charged
uranium hydroxides rather than the negatively charged carbonate com-
plexes that would form in the carbonate solutions.
Hansen et ala conducted a pilot plant study for uranium removal
using a chemical clarification system (air flotation and rotary vacuum
filter) with ferric chloride (30 mg/L) as the coagulant. The pH was
varied from 4 to 10 and the results were very similar to the labor-
atory studies of Lee et al as shown in Figure 6. The highest removals
were achieved at pH 6 and 10. Lower removals were obtained at pH 4,
and 7-9. Hansen et al also conducted several tests to evaluate coagu-
1 ant dosage (30-90 mg/L) on uranium removal at pH 6 and 10. Results
showed that increasing ferric chloride dosage above 30 mg/L produced
very little benefit at either of these pH values (Figure 7).
White and Rondletti^ and Hansen et al conducted surveys of full-
scale treatment plants for uranium removal. White and Bondietti report-
ed on data (grab samples) from 20 full scale treatment plants, but only
one had feed water uranium above 10 ug/L (the Moffat treatment pi ant in
Henver, CO with 15 mg U/L). Data from the plants with less than in
ug/L of uranium in the raw water showed 1ittle or no removal of uranium.
The Moffat plant (alum, lime, polymer) on the other hand, showed a re-
moval of 75 percent with the pH of raw water samples at 7.5. Hansen et
al also reported uranium data collected by the same pi ant. These data
showed a wide variation in removals, 78 *1 90 percent. Because the
data was based on grab samples collected at different times, correla-
tion between raw and treated water was difficult. Hansen et al report-
ed that an alum plant (microfloc system) at Arvada, CO which draws its
water from the same source as the Moffat pi ant had removals of 18 to 90
percent with an average efficiency of 67 *15 percent. Data from these
two treatment plants, although rather sketchy, did indicate that alum
3
-------
coagulation could potentially reduce uranium rather significantly. On
the other hand, White and Rondietti concluded, considering all of the
data collected fron the ?0 plant survey, that conventional coagulation
treatment was not effective in removing uranium from raw waters when
the uranium concentrations were in the range of 0.O1-16 ug/L.
Lime-Softening Treatment
Lee and Bondietti6 reported on laboratory jar test experiments to
evaluate lime softening for uranium removal using the same pond water
(83 ug/L U) that was used for the coagulation tests. With the addition
of lime (50-250 mg/L) alone to raise the pH of 10.6 to 11.5, 85 to 90
percent of the uranium was removed (Figure 8). As shown, very little
difference in removals occurred in the pH range of 10.6 to 11.5. The
same investigators also conducted tests with lime and magnesium carbon-
ate to determine whether HgC03 could improve removal. Varying amounts,
of MgCOj and Ca(0H)2 were added to test solution and the pH and uraninum
content of the solution measured. The results (Table 1) indicate that
at low pH (9.8 to 10.6), the MgC03 additions reduced the effectiveness of
lime for uranium removal and at higher pH (>10.6), uranium removals in-
creased with the increase in MgC03 doses. The data also show that the
critical pH appears to be near pH 10.6. Above pH 10.6, the lime and
MgC03 combination increased uranium removals to 93-99 percent. The
investigators suggest that at the high pH, the water becomes depleted
of carbonate and the chemical state of the uranium species in the
natural pond water is converted to uranyl hydroxide (1102)3 (0H)|.
To investigate the significance of magnesium and to gain a better
understanding of the removal mechanism, Lee and Bondietti conducted
several experiments with solutions of magnesium chloride, magnesium
bicarbonate, calcium chloride and calcium biocarbonate using uranium
test water in place of the pond water. The pH of the solutions were
also varied from about 9 to 11.5. The results showed the high pH
magensium solutions to achieve much higher removals (80-99 percent)
than the calcium solutions (10-30 percent). The authors concluded from
these facts that at high pH (10.7-11.3) Mg(0H)2 precipitates plays a
major role in uranium removal. Also, in the pH range of 8.5-10.6, the
calcium and magnesium carbonate precipitates can remove some uranium,
but the efficiencies are low (10-30 percent). And, finally, they
concluded that magnesium is an essential ingredient for uranium removal
by lime treatment and that the natural magnesium in the pond water was
probably the major reason why high removals were achieved in the initial
experiments with only lime addition.
The Drinking Water Research Division (DWRP), USEPA conducted several
lime softening experiments for uranium removal on a lake water having a
uranium concentration of near 60 ug/L. The test results conf1rmed the
work of Lee and Bondietti showing a pH effect. As the pH increased
from 9 to 11.2, removals increased from 16 to 97 percent (Figure 9).
Reverse Osmosis
Pilot plant studies for the removal of a number of inorganic con-
4
-------
tami nants and uranium by reverse osmosis (RO) have been reported by
Huxstep and Sorg.9 The study showed that natural uranium in a Florida
ground water (300 ug/L) was highly removed (99%) by four different
reverse osmosis membranes: Filmtec BW30-4021, a Dow 5K, a Dupont R-9,
Model 0440-042, and a Hydranautics P/N 4040 LSY-IFCI. Because removal
of ions in water by RO is a function of the size and charge of the ion,
uranium should be very effectively removed by most all RO systems.
For household water supplies, a point-of-use reverse osmosis system
has also been shown by Fox and Sorg10 to be very effective for uranium
removal. These investigators conducted a laboratory study using an Aqua
Clear H-82, Culligan International Home RO System on Cincinnati tap
water spiked with uranium concentrations of 69.2 ug/L and 182.5 ug/L.
Both test runs showed uranium removal to be greater than 99 percent.
Ion Exchange Treatment
Ion exchange resins, particularly anion exchange resins, have been
used to recover uranium from uranium mine water for many years.^ To
investigate this application for water treatment, the Oak Ridge National
Laboratory study6»' Included pilot column studies for uranium removal
using both cation and anion resins. The cation resin (Dowex 50-X8)
tests used the resin in the H+, Na+ and Ca+2 forms. With the H+ form
resin, effluent samples (pH 3.5) showed removals of 93-97 percent re-
moval of the uranium. The results suggested that the uranyl carbonates
in pond water were changed to uranyl cations in the acid resin bed. Re-
moval of the uranium by the Na and Ca form resins on the pond water (pH
>8.3) was low. Low removals would be expected if the uranium was in
the carbonate (anion specie) form.
Follow-up tests were conducted with the cation resin in the Ca and
Na form at pH 8.2, 7.0, 5.6 and 4.0 by lowering the solution pH in the
order of high to low. At pH 8.2, neither form removed the uranium
(Figure 10). At pH 7, the Ca form resin did not remove uranium, but
the Na form removed about 85 percent. The Ca form resin started to
remove uranium at pH 5.6 and achieved about 60-65 percent removal at pH
4.0. The Na form resin continuously removed 70 percent at pH 5.6 and
4.0. The pH dependence demonstrated the importance of the uranium
complex in the feed water and the difference 1n efficiency between the
Ca and Na forms suggest a difference in selectivity order of the three
ions. The overall results of the cation resin tests indicate that this
process would probably not be practical for drinking water supplies.
Studies on the use of anion resins for uranium removal have been
reported by several authors."-8*12-17 tee and Bond1ett1fi conducted
1aboratory column tests using Dowex 1-X2 anion resin. The initial test
with pond water (83 ug/L) showed 99 percent removal after passing 50
1 iters through a small column containing 5.5 mL of resin. Assuming an
exchange capacity of 0.7 meq/mL of resin, and assuming the dominant
uranium species in the pond water was U02 (C03)o 5 (50 percent
dicarbonate and 50 percent tricarbonate), the theoretical capacity of
the resin was calculated to be about 55 mg/mL of resin. Follow-up
tests with much higher uranium feed water concentration (23.8 mg ll/L)
5
-------
and a smaller column (0.5 ml resin) showed a gradual Increase in the
uranium level in the effluent with time. When the resin reached 65
percent of the calculated capacity, the uranium concentration in the
effluent was 10 percent of the influent concentration (90 percent
removal). At 81 percent of the resin capacity, the column was removing
72 percent of the uranium. The hreakthrough curve was, therefore, a
smooth exponential curve (y = 0.79e^'n^4x; where y = % U in effluent;
x = BV treated in milliequivalents) and not a sharp breakthrough common-
ly experienced in ion exchange water softening treatment.
Hansen et als conducted pilot plant tests using three differ-
ent anion resins, Dowex SBR-P, Dowex 21K, and lonac A641. The results
of these tests using 2 cu ft columns on a well water containing 200-300
ug/L of uranium are shown in Table 2. The test runs consisted of four
cycles each, hut cycle number 3 was rejected because the regeneration
step (10 percent salt solution) prior to this run was incomplete and
uranium leakage through the columns was very high. During cycles 1, 2
and 4, the effluent had a uranium concentration of less than 15 ug/L up
to breakthrough.
Studies on the use of anion exchange to remove uranium from drinking
water have been conducted by DWRD, USEPA, Cincinnati and reported by
Hathaway12 and Sorg.l® The research conducted consisted of two phases;
an inhouse bench-scale study and a field study at ten locations In
Colorado and New Mexico. The bench scale study consisted of column
tests using three different anion resins Dowex SP8-P, Dowex 21K and
lonac A641, the same resins used by Hansen et al.® The test water was
a natural Florida ground water containing around 300 ug/L of uranium.
Because several tank trailer loads of ground water were used, the
quality varied during the test period. The uranium ranged from 175 to
300 ug/L, sulfate from 40 to 131 mg/L, Tns from 376 to 1300 mg/L, and
pH 7.4 to 7.7.
The first series of tests were conducted with the Dowex SPR-P
anion resin in the chloride form. Two parallel columns each containing
55 mL of resin were used. One column was put through four cycles of
column loading and regeneration and the second column through three
cycles. The results show very good removal with greater than 90 per-
cent uranium removal for test runs lasting from 1400 to 7500 bed vol-
umes. The test runs varied significantly because of variations in
regeneration procedures but more importantly because of flow restriction
(media clogging). In other words, several test runs were terminated
before uranium breakthrough because of flow problems.
These initial tests agreed with the results of the Oak Ridge
studies6*? that when uranium begins to breakthrough the column, it does
so on a gradual basis. Moreover, sulfate, which is also removed, broke
through long before (600 BV) uranium and the sulfate breakthrough curve
1s a very sharp peak as normally occurs.
Following the column tests with the Dowex SBR-P resin, column
tests were run with Dowex 21K (55 mL) and lonac A-641 (55 mL). Each
resin had an estimated capacity of 1.2 meq/mL and in order to minimize
6
-------
the flow problems, a ?5 micrometer particulate filter was installed on
the influent line to the columns. This filter helped, but did not
totally eliminate the flow problem because of the very long test runs
(60+ days).
Each resin was run through multiple cycles; six with Dowex ?1K
resin and seven with the Ionac A-641 resin (Table 3). The results of
the first test cycle with the Dowex ?1K. resin in the chloride form as
supplied by the manufacturer is shown in Figure 11. The column test
was run for about 60 days and treated a total of 17,432 bed volumes of
water before the test was terminated at an effluent uranium concentra-
tion of 12 ug/L (95.2 percent removal). Based upon the average influent
concentration, the calculated loading on the 55 mL of resin was 275,000
ug of uranium. The column was regenerated with one liter of 1 percent
NaOH followed by one liter of IN HC1. An analysis of the regneration
solution Indicated that only 51,000 ug/L of uranium was removed (18
percent). Because of the low recovery, different regeneration procedures
were used for the remaining tests. An analysis of the regenerate
solutions for the subsequent tests showed incomplete regeneration up to
the sixth test run. In other words, although recoveries varied from
about 5 percent to 86 percent because of the different regeneration
procedures, the resin column continued to accumulate uranium on the
column through six regenerations.
The bench-scale test showed that uranium could be removed by all
resins tested down to the 5-10 ug/L range for many thousands of bed
volumes of water treated. The regeneration information suggested that
a combination of 4 percent NaOH and 1 N HC1 was more effective than 10
percent NaCl although neither process was totally effective in remov-
ing all of the uranium on the resin.
An evaluation of regenerants for the removal of uranium from anion
exchange resins was conducted by Utrecht using the uranium loaded resins
of the 0WR0.17 Utrecht employed batch and column regeneration tests
using sodium chloride, sodium hydroxide, sodium bicarbonate and hydro-
chloric acid as regenerants. Based upon the tests results, Utrecht
concluded that hydrochlor1c acid was the most effective and sodium
hydroxide the least effective. In general, the studies showed that
uranium was difficult to remove by all regenerants and large quanti-
ties of regenera.ts would be required if 100 percent removal of uranium
was necessary. Repetitive column tests were not conducted, therefore,
the regenerants were not evaluated on their long term application.
The second phase of the DWRD research consist of 11 (initially 12)
small 1/4 cu ft anion exchange systmes (Ionac-A641) installed in homes,
schools and on community water systems at ten sites in Colorado and New
Mexico (Table 4). The systems were equipped with inlet and outlet
screens, a timer, solenoid valve and an in-line flow regulator. Five
of the systems were operated continuously (a sixth was shut down early)
and the other six operated intermittently to simulate home use. The
continuous flow systems treated 3fi0 gpd and intermittent flow units
treated about 37 gpd. At all locations, water meters were installed to
measure the total amount of water treated. Two systems (numbers 4 and
7
-------
5) were Installed at one site; one operated continuously and the other
intermittently.
Water samples were collected for a two year period at least once a
month, although some units were sampled weekly. Some gaps in sampling
occurred at remote sites when the units had to be shut down because of
a water shortage.
The raw waters at the ten sites varied in quality. The uranium
concentration ranged from 22 ug/L to 104 ug/L, TPS from 166 mg/L to
1200 ntg/L and sulfate from less than 5 to 408 mg/L.
The results of the study showed only three (continuous flow) of
the ten systems to achieve uranium breakthrough within the two years of
monitoring. All the other systems continually produced treated water
with less than 1 ug/L of uranium, through the entire two years. A
summary of the results are shown in Table 4.
The amount of raw water treated by the three systems that achieved
breakthrough were about 8000 RV (System No. 7) 14,000 BV (System No.
5), and 10,000 RV (System No. 3). Of the other systems that did not
achieve uranium breakthrough, one system (28 ug/L U raw water) had
treated over 62,000 RV of water and another (64 ug/L U raw water) had
treated over 21,000 BV of water. Overal1, the field study showed anion
resins have a large capacity for uranium and treated a variety of qual-
ity of raw waters for long periods of time.
Extensive ion exchange pilot plant studies in New Mexico and Ari-
zona have also been reported by Jel 1 nek and Correll J3 Using 2-inch
diameter, 24-inch depth PVC columns in parallel, three anion resins
(Dowex 21K, Ionac Afi4l, and Dowex SRR-P) were tested at flow rates of
5 gpm/ft3 and 2.5 gpm/ft3. At the New Mexico site, the Oowex 21K resin
consistantly removed 99 percent of the influent uranium (measured as
86-120 pCi/L gross alpha). Rreakthrough (<1 pCi/L) had not occurred at
64,000 BV with a loading rate of 5 gpm/ft3, and at 33,000 BV with a
loading rate of 2,5 gpm/ft3. Results of the Ionac A641 resin were
similar while uranium breakthrough with the Dowex S8R-P resin occurred
after approximately 3,000 BV.
Test results at the Arizona site that had a uranium activity of 13
to 38 pCi/L were consistant with the New Mexico results. The Howex 12K
and Ionac A641 resins consistantly removed the uranium to the 1 pCi/L
level while treating more than 60,000 BV at both flow rates. Also,
early breakthrough occurred with the Oowex SRR-P resin.
Based upon pilot plant studies, small scale anion exchange systems
have been installed at several schools in Jefferson County, CO.I4>Is
Recause the well yields and uranium concentrates are similar, the systems
were designed for essentially the same capacity. Each of the treatment
systems consist of two spiral-wound cartridge prefilters in parallel,
two commercial water softener tanks in series, a brine tank for batch
regeneration and facilities to store or transfer spent regenerant.
P
-------
Each ion exchange tank is 16-in (40.6 cm) in diameter and 52-in
(13? cm) in height and contains 3 cu ft (0.085 M3) of Sybron Ionac
A642 resin.
One of the uranium removal systems was placed on line at the Coal
Creek Elementary School in March 1987.14 Between March 1987 and February
198R, the system was regenerated only twice--in July 1987 and February
1988. A summary of the results of operation of the system is shown in
Table 5.
A typical regeneration sequence consists of 1 PV of backwash, ap-
proximately 5 RV of saturated NaCl solution regenerant and approximately
5 BV of slow rinse. The total volume of wastewater produced during a
regeneration is approximately 220 gallons. The characteristics of the
first regeneration solution is shown in Table 6. These data show that
the uranium concentration in the brine water was 16,502 ug/L. Comparing
the amount of uranium recovery during the regeneration and the average
amount removed indicates that 97 percent of the uranium loaded onto
the resin was removed during regeneration. Currently the waste brine
is hauled away for disposal to a school owned and operated secondary
wastewater treatment plant. Limited data suggest that the uranium
concentrates in the sludge of the wastewater treatment plant,
Jelinek and Correll13 also reported that a full-scale ion exchange
system consisting of two ion exchange beds in series have been designed
for the Arizona well site mentioned above. Based upon a system flow
rate of 11,000 gpd, the system has been designed to operate for one
year before regeneration.
SUMMARY
Because uranium can occur in both surface waters and ground waters,
treatment technologies have been evaluated to handle both types of
waters. Conventional treatment technologies have been evaluated on
either a laboratory or field-scale level and a summary of their effec-
tiveness is shown in Table 7. This information shows that coagulation/
filtration, lime softening, anion exchange, and reverse osmosis are
capable of removing uranium down to the 1-5 ug/L range.
Although 1aboratory and pilot pi ant studies have been conducted to
evaluate a variety of treatment methods, few full-scale systems have
been built to remove uranium primarily because an uranium regulation
has not been formally established. Data from full-scale conventional
coagulation/f1ltration has shown uranium to he effectively removed, but
the data 1s very 1imited and influent concentrations are generally very
low. Several small anion exchange systems have been installed at
schools and these systems are performing very effectively.
9
-------
REFERENCES
1. Cothern, C.R. "Regulations and Health Effects of Radionuclides
in Drinking Water," in Proceedings of AWWA Seminar on Radionuclides
1n Drinking Water, (Denver, CO: AWWA, 1987), pp 3-10.
2. Horton, T.R. "Methods and Results of EPA's Study of Radon in Drink-
ing Water," Office of Radiation Programs, U.S.EPA Report 520/5-83/027
{1983)
3. Longtin, J.P. "Radon, Radium, and Uranium Occurrence in Drinking
Water from Groundwater Sources," in Proceedings of AWWA Seminar
on Radionuclides in Drinking Water, (Denver, CO: AWWA, T987),
pp. 11-33.
4. Langnulr, D. "Uranium Solution Mineral Equilibria at Low Tempera-
tures with Application to Sedimentory Ore DepositsGeochim.
Cosmochin. Acta. 42:547 (1978).
5. White, S.K., and E. A. Bondietti. "Removing Uranium by Current
Municipal Water Treatment Processes," Jour. AWWA. 75(7):374-380
(1983).
6. Lee, S.Y., and E. A. Bondietti. "Removal Uranium from Drinking
Water by Metal Hydroxides and Anion Exchange Resin," Jour. AWWA.
(75(10):536-540 (1983).
7. Lee, S.Y., S.K. Hall and E.A. Bondietti. "Methods of Removing
Uranium from Drinking Water: II. Present Municipal Water Treat-
ment and Potential Removal Methods," Office of Drinking Water,
U.S.EPA report EPA 570/9-82/003 (1982).
8. Hanson, S.W., d.r. Wilson, and N.N. Gunaji. "Removal of Uranium
from Drinking Water by Ion Exchange and Chemical Clarification,"
U.S.EPA report EPA/600/52-87/076 (December 1987).
9. Huxstep, M.R. and T. J. Sorg. "Reverse Osmosis Treatment to Re-
move Inorganic Contaminants from Drinking Water," U.S.EPA Report
EPA/600/52-87/109 (March 1987).
10. Fox, K.R. and T. J. Sorg. "Controlling Arsenic, Fluoride, and
Uranium by Point-of-Use Treatment.," Jour. AWWA. 79(10) :8l -84 (1987).
11. Ross, J.R. and D.R. George. "Recovery of Uranium from National Mine
Waters by Countercurrent Ion Exchange," U.S.Bureau of Mines Report
R1-7471(1971).
12. Hathaway, S.W. "Uranium Removal Processes," Paper presented at
AWWA Annual Conference, Las Vegas, NE, June 7, 1983.
13. Jelinek, R.T. and R. J. Correl1. "Operation of Small Scale Uranium
Removal Systems," in Proceedings AWWA Seminar on Radionuclides in
Drinking Hater (Denver, CO: AWWA, 1987), pp. 99-117.
10
-------
14. ilelinek, R,T. and T.J.Sorg, "Operating a Small Full-Scale Ion
Exchange System for Uranium Removal," Jour AWWA. 80(7);79-B3(1988).
15. Varani, F.T., R.T. Jelinek, and R..J. Correl 1. "Occurrence and
Treatment of Uranium in Point-of-llse Systems in Colorado," Radon in
Around Water. (Chelsea, MI: Lewis Publishers, Inc. (19R7) pp. 535-546.
16. Sorg, T..1. "Methods for Removing Uranium from Prinking Water,"
Jour AWWA. 80(7):105-111 (1988).
17. titrecht, P.M. "Evaluation of Regenerants for the Removal of Uranyl
Carbonates from Strong-Base Anion Exchange Resin," Masters Thesis,
Environmental Engineering, University of Cincinnati (1985).
NOTE: Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
11
-------
TABLE 1
REMOVAL OF URANIUM FROM POND WATER BY COMBINED
Ca(OH)^ and MgCflj TREATMENT
MgCHj dose
mg/L
50
Ca(OH)? dose
100 150
- mg/L
znn
250
in
Percent IJ
removal
32
90
90
RR
89
Final pH
10.6
11.0
11.1
11.3
11.3
40
Percent U
removal
9
95
95
94
94
Final pH
9.8
10.9
10.8
11.1
11.4
80
Percent U
removal
24
98
93
98
98
Final pH
10.3
10.8
10.7
10.9
11.2
120
Percent M
removal
15
99
99
99
99
Final pH
9.9
10.6
10.8
11.0
11.2
w o
_L^
-------
TABLE 2
URANIUM REMOVAL WITH ANION EXCHANGE TREATMENT*
Resin Influent II Red Volumes
Test Concentration Treated at
Cycle ug/L H Breakthrough
HOWEX 21k
1 200 8,020
2 300 17,375'
4 300 13,365
Average (12,920)
DOWEX SBR-P
1 200 18,712
2 300 16,038
4 300 10,02*
Average (14,925)
IONAC A641
j 2nn 10,024
2 300 18,043
4 300 20,048
Average (16,038)
13
-------
TABLE 3
URANIUM REMOVAL WITH ANION EXCHANGE TREATMENT
Resi n
Test Run
DOWEX 21k
1
2
3
4
5
6
7
IONAC
1
2
3
4
A-641
Influent U
Concentration
ug/L
300
300
30n
300
300
350
432
300
300
300
300
350
432
Regeneration
Solutions
1% NaOH/l N HC1
IN HCL/10% NaCl
2% NaOH/l N HC1
10% NaCl /2% NaOH
4% NaOH/l N HC1
4% NaOH/l N HC1
None
10% NaCl
2% NaOH/l N HC1
1% NaCl/?.% NaOH
$% MaOH/1 N HC1
4% NAOH/1 N HC1
None
BV Treated
n Cone termi-
nation (ug/L)
17,432 (12)
12,460 (1.3)
MOO (10)
34,500 ( 9)
5,980 ( 7)
23,980 flO)
23,320 (10)
21,110 (9.5)
6,280 (0,6)
28,880 (10)
Not Effective
20,130 (9)
18,260 (10)
-------
TABLF 4
SUMMARY OF URANIUM REMOVAL WITH ANION EXCHANGE TREATMENT
.. .. AT_LEN FIEL0 SIT£S
Influent U Influent Bed Volumes Percent
System Concentration Sulfate Concen- Treated at Uranium
Unit No* Flowt ug/L tration mg/L Termination Removed
(total)
1
I
22
< 5
9,400
99.8
2
I
30
320
25,000
99.8
3
C
23
15
24,700
78.1
4
I
52
390
2,800
>99.5
5
C
52
390
34,500
72,1
6
C
64
124
21,200
>99.9
7
C
35
408
11,900
29.8
8
C
28
3
62,900
99.6
9
I
100
< 5
13,900
98.6
10
I
40
< 5
20,000
99.9
11
I
104
9
7,900
99.9
* - System No 4 and No 5 at same location
tl - Intermittent Flow,
C - Continuous Flow
15
-------
TABLE 5
OPERATIONAL RESULTS FROM A (IRANIIH ION EXCHANGE SYSTEM14
= si * 933'S's***** *.*.•* arm-*- *- a *3 stm *..j®us»s a * •* at ¦% at at twe ». * « at-ar-ar «• «-at «* s* 4 a a a -sp-si a as-
Amount Removal
of Water Red Volumes Uranium Concentration—11 g/L Across
Sample
Treated
After
Raw
After
After
Column 1
Date
gallons
Regeneration
Water
Column 1
Column 2
percent
7-?-R7*
66,000
2,940
56.6
<0.1
0.6
99.8
9-2-87
116,740
2,260
45.2
0.1
0.1
99.8
10-6-87
165,220
4,420
39.7
0.1
0.1
99.8
11-18-87
218,620
6,800
47.0
<0.1
0.1
98.8
2-3-88*
284,790
9,750
110.0
0.3
0.2
99.7
*The first ion exchange column was regenerated on these dates.
1.6
-------
TARLE 6
CHARACTERISTICS OF REGENERATION WASTEWATER
Parameter Value
Uranium -ug/L lfi,502
Gross alpha particiles-pCi/L 6,154 t 580
Chloride - mg/l 18,480
pH 8.75
TPS - ng/L 39,600 .
Volume - gal 220
17
-------
TABLE 7
SUMMARY OF EFFECTIVENESS OF TREATMENT METHODS
FOR URANIUM REMOVAL FROM DRINKING HATER
e •* a* a
Treatment Percent Special Conditions/
Method Removal Notes
tmat at*;'*'* am. * a
Iron Coagulation 80-85 pH fi and 10
Alum Coagulation 90-95 pH 10
80-85 pH P
Lime Softening 99 pH 10.6+ Mg ion
Cation Exchange
H+ form 90-95 pH 3,5
Ca+ form 70 pH 4
Na+ form 70-85 pH *7
Anion Exchange 99 10,000-50,000 BV
1.8
-------
100
D
1
Figure 1. Distribution of uranyf hydroxy and carbonate complexes versus pH.4
I
coagulant dosage - mg/L
Figure 2. Uranium removal by ferric sulfate coagulation.8
19
-------
0 5 10 15 20 25 30
coagulant dosage - mg/L
Figure 3. Uranium removal by ferrous sulfate coagulation.8
coagulant dosage - mg/L
Figure 4. Uranium removal by alum coagulation.8
20
-------
Alum
Ferric Sulfate
Ferrous Sulfate
Figure S. Effect of pH ami coagulant ori uranium removal by coagulation
(25 mg/L) treatment*
pH - units
Figure 8. Effect of pH on uranium removal by Iron coagulation.6-*
21
-------
1
s
E
3
e
2
s
5
a.
100
90
SO
70
60
50
40
30
20
10
0
pH 10
a
pH 6
U - 450 mg/L
•
20 40 60 80 100
coagulant dosaga - mg/l
Figure 7. Effect of coagulant dosage on uranium removal by iron coagulation.1
E
1
3
100
90
80
70
60
SO
40
30
20
10
0
—
U - 83 ug/L
Final pH (10.6 • 11.5)
50
100
150
200
250
300
lime dosaga - mg/L
Figure 8. Removal of uranium by lime softening.*
22
-------
pH • units
Figure 9. Effect of pH on uranium removal by lime softening
100
90
80
70
60
50
40
30
20
10
0
3.5
resin
form
¦H ^
Ca
• Ma'
8.2
4 5.6 7
pH • units
Figure 10. Removal of uranium by cation exchange treatment.7
VI
-------
a
3
E
5000
I 1 i1 i
10000
15000
20000
bad voiuims
Figure 11. Removal of uranium by Dowex 21 resin in chloride term.
------- |