E P A/540 /R-93/515
METALS TREATMENT AT SUPERFUND SITES BY ADSORPTIVE FILTRATION
Mark M. Benjamin and Ronald S. Slctten
Environmental Engineering and Sciences
University of Washington, Mail Stop FX-10
Seattle, WA 98195
Contract No, CR 81S784-QI-Q
Project Officer:
Nonna Lewis
Emerging Technology Section, SDEB
Risk Reduction Engineering Laboratory
Cincinnati, OH 45268
RISK REDUCTION LABORATORY
OFFICE Or RESEARCH AND DEVELOPMENT
U.S. KtfVmONMENTAL PROTECTIOH AC20ICY
I. OH 45268
-------
FOREWORD
The Superfund Innovative Technology Evaluation (SITE) Program was authorized in the
1966 Super-fund Amendments. The purpose of the Program is to assist the development of
hazardous waste treatment technologies necessary to implement new cleanup standards which
require greater reliance on permanent remedies. A key part of EPA's effort is its research into our
environmental problems to find new and innovative solutions.
The Risk Reduction Engineering Laboratory (RREL) is responsible for planning,
implementing, and managing research, development, and demonstration programs to provide an
authoritative, defensible engineering basis in support of the policies, programs, and regulations of
the EPA with respect to drinking water, wastewater, pesticides, toxic substances ,solid and
hazardous wastes, and Superfund-related activities. This publication is one of the products of that
research and provides a vital communication link between the researcher and the user community,
The SITE Program is part of EPA's research into cleanup methods for hazardous waste sites
around the nation. Through cooperative, agreements with developers, alternative or innovative
technologies are refined at the bench-and pilot-scale level then demonstrated at actual sites. EPA
collects and evaluates extensive performance data on each technology to use in remediation
decision-making for hazardous waste sites.
This report documents the University of Washington's studies of the use of adsorptive
filtration for removing inorganic contaminants from the liquid phase. The research project assessed
the ability of coated sand media to remove soluble and particulate metals simultaneously and the
ease with which the media could be regenerated.
Copies of this report can be purchased from the National Technical Information Service,
Ravensworth Bldg., Springfield, VA, 22161,703-487-4600. You can also call the SITE Clearinghouse
hotline at 1-800-424-9346 or 202-382-3000 in Washington, D.C. to inquire about the availability of
other reports.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
in
-------
TABLE OF CONTENTS
NOTICE «ii
FOREWORD iii
ABSTRACT iv
PLATES and FIGURES vi
TABLES viii
INTRODUCTION 1
TECHNICAL DESCRIPTION OF THE TECHNOLOGY 1
METHODOLOGY 5
Preparation and Characterization of the Coated Media 6
Composition of Test Solutions 6
General Procedures for Column Tests 7
RESULTS AND DISCUSSION 8
Organization 8
Terminology 9
System Performance in Runs with Solutions Containing Only One Metal.. 9
Media Capacity 9
Adsorption Kinetics 12
Effect of Solution pH 12
Regeneration Kinetics.. 13
System Performance in Runs with Cu, Cd, and Pb Present in Influent 15
Systems with 0.5 mg/L Cu, Cd, and Pb 15
Systems with 5 mg/L Cu, Cd, and Pb 16
Backwashing 22
Regeneration 23
Anion Removal 25
Effects of Other Contaminants on Removal of Cu, Pb, and Cd 27
Tests with complexed metals 27
Effects of organic surfactants 29
Effects of biofilms on the adsorptive filtration media so
Effects of oily substances on adsorptive filtration 31
Effects of antifreeze 32
Adsorbent Longevity 32
Evaluation of Options for Residuals Management 33
Treatment of Real Superfund Solutions 33
CONCLUSIONS 39
ABBREVIATIONS 41
REFERENCES 42
-------
Figure 12. Headless buildup during Run 11 and 17 as a function of influent volume treated. Run
conditions: Pb-Cd-Cd-5mg/L; Run 11 EBDT=2 minutes; Run 17 EBDT=I minute.
Figure 13. Column regeneration data for Run 15. The upper graph shows metal concentrations in the
first regenerant solution, which was maintained at pH-2 and continuously recycled through the
column. The lower graph is for the second regenerant solution, which was adjusted to pH 2 and
passed through the column only once.
Figure 14. Breakthrough curve for arsenate and selenite in Runs 18 and 19. EBDT=2 min, pH=3.5.
Figure 15. Breakthrough curve for selenate with and without sulfate added. EBDT=2 min, pH=3.5.
Figure 16. Effluent breakthrough curves for ammonia wmplexed metals. The dip in concentration in the
middle of the graph occurred when switching from the first to second batch of influent. Influent
conditions: Cd=Cu=5 mg/L, NH3=0.04M, pH=IO, EBDT=2 min.
Figure 17. Effluent breakthrough curves for EDTA-complexed metals. Runs were completed at pH 4.5
and pH 10, as indicated. Influent conditions: Cd=Cu=Pb=S rngL, EDTA: metal ratio = 1.25:1
O^KT'MEDTA), EBDT=2 min.
Figure 18. Effluent breakthrough curves for ammoniawmplexed metals in the presence of sodium
dodecyl suffate (SDS). Influent conditions: Cd=Cu=5 mg/L,NH3=0.04M, pH=IO, EBDT=2 min.
Figure 19. Breakthrough curves for ammonia-complexed metals in a column containing iron oxide
coated sand onto which a biofilm had grown. Influent conditions: Cd=Cu=5 mg/L, NH3=0.04 M,
EBDT=2 min.
Figure 20. The effect of oil contamination of the media, followed by a cleaning step using isopropyl
alcohol on breakthrough curves for ammoniawmplexed metals. Influent conditions: Cd=Cu=5
mg/L, NH3-0.04 M, EBDT=2 min.
Figure 21. Schematic diagram of setup used for automatic treatment runs using Superfund site
solutions.
Figure 22. Effluent zinc wncentration for Western Processing Run 2 as a function of bed volumes
treated.
Figure 23. Effluent zinc wncentration for Western Processing Run 3 as a function of bed volumes
treated.
Figure 24. Effluent zinc concentration for Western Processing Run 4 as a function of bed volumes
treated.
Figure 25. Regeneration kinetics after Western Processing Run 3. The upper curve is for a solution that
was recirculated continuously through the column and maintained at pH 4. The lower curve is for
the second stage of regeneration, showing the metal concentration in a solution that has passed
through the column only once.
Figure 26. Headless buildup during Western Processing Run 2.
vn
-------
1HTRODOCTION
This project evaluated an inexpensive, innovative approach for removing inorganic contaminants from the
liquid phase at Superfund sites. In the process, called adsorptive filtration, metals are collected by
attachment to a thin layer of ferrihydrite (iron oxide) that has been immobilized on the surface of sand
grains. Since iron oxide is known to be a good adsorbent for heavy metals, it was hoped that the
modification of the surface of the sand would allow the grains to adsorb soluble heavy metals as they
passed through a column packed with the media. At the same time, it was anticipated that the grains
would perform comparably to plain sand as a filter media for collecting particulate metals. Finally, for the
system to be effective, it would have to be easily regenerated. That is, once the media was saturated
with metals, it would have to be possible to release and concentrate the metals for further processing
and, possibly, recovery.
The research project characterized various aspects of the adsorptive filtration process. Specifically, the
goals of the research were to assess the ability of the coated sand media to remove soluble and
particulate metals simultaneously as water containing those species passed through a column packed
with iron oxide coated sand, and to assess the ease with which the media could be regenerated. In this
report, the technology is described in a general way, and then the results of the experimental, pilot-scale
program to test the technology's applicability Siiperfund remediation efforts are presented.
TECHNICAL DESCRIPTION OF THE TECHNOLOGY
Conventional technology for removing metals from solution involves precipitation of the metals as oxides,
hydroxides, or sulfides, and then separation of the particulate metals by settling, usually aided by a
coagulant such as iron hydroxide. This approach has several practical limitations, some of which are
exacerbated when the metals are present in complex matrices such as commonly is the case for
solutions at Superfund sites. Among these limitations are that precipitation is ineffective if the metals are
complexed or if they are present as anions (e.g., Cr04,SeO3, AsOJ. and that those metals which do
precipitate may form small particles that do not settle readily. As a result, large settling basins may be
required to collect the metals, usually followed by a sludge thickening operation. Additionally, polymeric
coagulant aids are frequently required for solids separation to be effective.
Adsorption is the binding of chemical species on the surface of suspended particles. If the adsorbent
(solid surface) is chosen carefully and the solution chemistry is adjusted appropriately, adsorption-based
-------
Despite the high affinity of ferrihydrite for metals, its use as an ion exchange medium has not been
successfully developed because its physical properties (it is a bulky, flocculant material with extremely
low hydraulic conductivity) are not conducive to such a process. To overcome these problems, a means
was developed by which the ferrihydrite could be coated onto the surface of sand.* Simply precipitating
ferrihydrite in the presence of sand generates a partial coating, but this coating is very thin and patchy,
and It is relatively easily lost by abrasion or acid attack. However, a much more concentrated and
resilient ferrihydrite coating can be produced by mixing sand with an acidic, iron-containing solution, and
heating the mixture to complete dryness. Photographs of the uncoated sand used in the current study
and of a typical batch of coated sand at various magnifications are shown in Plates 1 through 3. Once
the coated sand grains are rinsed, the coating is stable to the mild abrasion that is typically encountered
during fitter backwashing, and it is very resistant to acids.
'Plate 1. Visual appearance of the uncoated and coated sand.
-------
When the coated sand is placed in a column and a solution containing metals is passed through it, the
ferrihydrite can adsorb the dissolved metals rapidly and efficiently. That is, the ion exchange properties
of the iron oxide remain viable during the coating process, though the ion exchange capacity of coated
ferrihydrite is reduced compared to that of ferrihydrite in a freely suspended state. The loss of this ion
exchange capacity Is at least partially offset by the fact that, at the same time that the surface of the
coated sand is acting as an ion exchanger, the bulk media can perform the function of a normal granular
media sand filter, i.e., it can remove precipitated metals and other particulate matter.
After some period of time, either the coating reaches its maximum capacity to remove metals or the filter
requires backwashing. At this time, an acidic backwashing solution can be applied to recover the metals
from the column and regenerate the column for further use. Because the ferrihydrite is trapped on the
sand particles, only the contaminant metals and not the ferrihydrite are released. Thus, the need to
dispose of large amounts of iron oxide with the metal sludge, one of the main drawbacks of a
conventional treatment process, is eliminated. Furthermore, the regeneration solution could potentially
be used repeatedly, so that over time the metals concentrations in it build up to quite high values, making
subsequent metal recovery an economic possibility.
In sum, there are six essential aspects of adsorptive filtration that combine to make it a potentially
valuable and widely applicable technology:
» Ferrihydrite is a strong metal adsorbent that can be regenerated by changing pH.
* Ferrihydrite can be coated onto sand, retaining much of its ion exchange activity.
* Ferrihydrite can adsorb some metal complexes that are not removed from solution by conventional
precipitation.
* Many metal oxyanions that cannot be treated by conventional precipiitbn can adsorb onto
ferrihydrite.
• A column of coated sand acts as a fitter as well as an adsorbent.
» The technology appears to be applicable over a wide range of contaminant concentrations.
METHODOLOGY
The project consisted primarily of pilot-scale testing of the ferrihydrite-coated sand. The experiments
performed addressed the kinetics of adsorption and regeneration, the adsorption capacity of the media
for both soluble and particulate metals, the effects of various organic contaminants on metals removal by
the media, the long-term stability of the adsorbent, and the metal concentrations achievable In the
-------
Finally, as noted above, a few tests were run with a solution collected from a Superfund site where
conventional treatment is currently being applied.
General Procedures for Column Tests
The central component of the test apparatus a set of Plexiglas columns filled with iron oxide coated sand.
The columns were 1.8 cm ID by 120 cm high. They were packed to a depth of 88 cm, so that each
contained 250 ml (bulk volume) of the coated sand. The rest of the pilot-scale set-up consisted of
automated instruments for maintaining pH of the influent water, pumping water through the column, and
regenerating the column at fixed intervals. A schematic diagram of the experimental setup is shown in
Figure 1. Samples were collected automatically at frequent intervals and were analyzed for metal
content. In some cases, soluble and particulate metals were distinguished, based on filtration through a
membrane fitter with 0.45 \un pores (Millipore type HA). Headless across the bed was also monitored.
Metal
Stock
solution
Figure 1. Schematic of setup used to determine breakthrough curves for the synthetic waste runs.
7
-------
Terminology
For the purposes of this report, a single 'run' is defined as the period between any two regeneration
steps. The columns were sometimes, but not always, backwashed in the middle or at the end of a run.
In those cases where the column was backwashed one or more times during a single run, the influent
solution that entered the column between backwashing steps is referred to as a 'batch' of influent. Thus,
in some runs, only one batch of influent was treated (the column was backwashed and regenerated at the
end of one treatment cycle), and, in others, several batches of influent were treated (the column was
backwashed after each treatment cycle, but was not regenerated until several treatment cycles had been
completed).
System Performance in Runs with Solutions Containing Only One Metal
The packed columns were used in several adsorption tests with influents containing onty a single metal
(Cd, Cu, or Pb) at a concentration of 0.5 m^L, which was almost all soluble in all cases.
Media Capacity
The results from several runs investigating the effect of solution pH on sorption capacity are summarized
in Table 2, and typical breakthrough curves for influents with 0.5 mgfl. of a single metal are presented in
Figures 2 through 4. The total amounts of metal adsorbed at the points of 50% breakthrough and at the
end of the run are included in Table 2 and are typically several hundred to a few thousand mg of metal
per liter of packed bed. The gross density of IOCS in the packed bed was 1670 g/L, so 1000 mg of metal
per liter of packed bed corresponds to 600 mg of metal per kilogram of IOCS. The adsorption density
can then be computed based on the iron content of the media, given in Table 1; 1000 mg of metal per
liter of packed bed corresponds to 28 mg/g of Fe for Media I and 19 mg/g Fe for Media II.
-------
0.50
0.40 •
0.30 -
0.20 •
0.10
0.00
— -
p«
p«
pi
pi
pi
pi
pi
pi
<
7
7
7
1
•
*
t
.0
.0
.0
.s
.0
.s
.0
.»
10
JO
JO
10
10
10
10
IS
rr/
w/
VI
w/
VI
•v/
rr/
•v/
ft
•r
•r
•r
•r
«r
200 400 600 800 1000
B«d volume*
1200
1400
1800
Figure 2. Breakthrough curves for 0.5 mg/L copper influent showing total metal in effluent as a function
of bed volumes of water treated at various pH values. Several flow rates were used as indicated in
leaend
o.so
—
pi
pi
pi
pi
pi
pi
pi
pi
«.s
7.3
7.5
(.0
I.S
t.o
I.S
• .s
10
10
10
10
10
10
IS
IS
Wfl
w/
•T/
rr/
wn
1V/
•v/
Ir
•r
•r
Br
•r
Ir
0.00
200 400 800 800 1000
B*dvolum««
1200
1400
1800
Figure 3. Breakthrough curves for 0.5 mg/L cadmium influent showing total metal in effluent as a function
of bed volumes of water treated at various pH values. Several flow rates were used as indicated in
legend.
11
-------
steadily from pH 6.5 to pH 8.0, and then decreased again when pH was increased to 6.5 and 9.0. For
Cd, capacity increased steadily to pH 6.5. When the pH was increased to 9.0, adsorption of Cd onto the
walls of the plastic influent container prevented any significant amount of Cd from entering the column, so
pH 6.5 was chosen as the highest value to study in this phase of testing. Sorption of Pb was quite good
at pH 7.0, and no systematic variation of pH was conducted for the Pb runs.
Regeneration Kinetics
Several tests were conducted in association with these runs to evaluate regeneration kinetics. In these
tests, regeneration consisted of recirculating 500 ml (2 empty bed volumes) of a pH 2.0 solution through
the column for 30 minutes and analyzing the metal concentration In the solution every 5 minutes. This
was followed by passing another 500 ml of pH 2.0 solution through the column in a single pass. After
each 100 ml of this solution passed through the column, a subsample was collected and analyzed. The
results for one such run for each metal are shown in Figures 5 and 6.
The metal concentration in the recirculation fluid increased rapidly at first and then only slowly thereafter.
Based on these results, it appears that, if time were critical, a recirculating period as short as 10 minutes
would release a large fraction of the available metal. Similarly, any releasable metal remaining on the
media after the first step was released to the first 100 ml of the next batch of pH 2.0 fluid passing
through the column.
Although the performance of the system during this set of experiments indicated that the media had
substantial capacity to remove soluble metals, the capacity was somewhat smaller than hoped for, and,
as noted above, a different batch of media (Media II in Table 1) was used in the remainder of the testing.
13
-------
System Performance in Runs with Cu, Cd, and Pb Present in Influent
System s with 0.5 mg/l Cu. Cd. and P_b
In the first runs with Cu, Cd, and Pb present simultaneously in the influent, all at a concentration of 0.5
mg/L, the influent was adjusted to pH 8.5 and was fed at a rate corresponding to a 2-min empty bed
detention time. However, the effluent pH fluctuated significantly and often dropped below 8.0 due to
exchange of H+ ions for the adsorbing metal ions. When the influent pH was adjusted to 9.0, the effluent
pH was stable and near that of the influent, and 7000 bed volumes of influent were treated effectively
(Figure 7). The metal concentrations in the effluent were increased gradually during the run, with virtually
no short-term fluctuations. Headless was usually under 5 psi at the beginning of a run and increased
gradually thereafter. When the headloss reached around 10-13 psi, the column was backwashed to
removed particulate matter that had been trapped. This process was successful, in that the pressure
drop through the column was reduced after backwashing and the metal removal efficiency was at least as
good, and often better, after backwashing compared to before.
1 T
o.a • •
o.e • •
0.4 > •
0.2 • •
1000 2000 3ooo 4ooo
B«d volumes
5000
8000
7ooa
Figure 7. Breakthrough curves for Pb, Cd, and Cu using Media II. The influent concentration was 0.5
mg/L for each metal. EBDT=2 minutes, pH=9.0.
15
-------
table 3. Conditions for Runs 14 to17. These runs were performed using batches containing all three
metals (Pb, Cd, and Cu) each at 5 mg/L concentration.
Run
14
15
18
17
PH
8.6
8.9
8.7
9.3
Batches
3
3
1
1
EBOT
2 min.
4 min.
4 min.
1 min.
BV Treated
831
772
199
178
7 -r
6 • •
5 • •
3 • •
2 • •
1 +•
Lead
Cadmiun
Copp«r
loo
200
300 400 too
Bad Volumes
600
700
Figure 8. Breakthrough curve for Run 14. Influent contained Cd, Cu, and Pb at 5 mg/L each. EDBT=2
min, pH=9.0.
17
-------
0.25 T
•*— LEAD
-•— CADMUM
•»- COPPER
80 100
B«d volume*
120
140
100
180
Figure 11. Breakthrough curve for Run 17. Influent contained Cd, Cu, and Pb at 5 mg/L each. EDBT=2
imin, pH=9.0.
During each run, effluent was sampled until the pressure drop across the column reached a pre-
determined value, usually either 20 or 25 psi. At that point, an automatic shut-off switch was activated,
and flow to the column was terminated. In Runs 14 and 15, the column was then backwashed with pH
9.0 water, and flow was re-initiated. As noted earlier, influent to the column at that point was identified as
Batch 2.' A similar sequence occurred when the pressure reached the target value again, and Batch 3
was treated. When the pressure reached the maximum allowable value after treatment of Batch 3, the
column was backwashed and then regenerated with water adjusted to pH 2.0.
Since the total influent concentration during Runs 14-17 was considerably greater than the soluble
influent concentration, it is apparent that a significant fraction of the metal in the influent precipitated prior
to entering the column. The total concentrations (corrected for background) of all the metals in the
effluent were well below 0.1 mg/L until about 150 to 200 bed volumes had been treated (a 6-to 12-hour
period), at which point particulate metals began breaking through the column. Removal of soluble metal
was always significant throughout these runs. Typical removal efficiencies for soluble metals in the
influent were 80% for Cu, 90% for Pb, and 98% for Cd, and typical overall removal
(comparing total effluent and total influent) were 99% or greater for all three metals (Table 4).
19
-------
25
20
10
Run 11
20
200
300
—I h-
400 500
B*d volum**
800
700
800
9oo
30
25
Run 17
10
20
40
60
80 100
B«d volume*
140
160
I
190
IFigure 12. Headless buildup during Run 11 and 17 as a function of influent volume treated. Run
[conditions: Pb=Cd=Cd=5mg/L; Run 11 EBDT=2 minutes; Run 17 EBDT=1 minute.
21
-------
Table 5. Recovery of metals from the column after various treatments after Run 15. All values
are mg/L unless indicated otherwise.
Metal, Batch #
Pb, 1
Pb, 2
Pb,3
Cd, 1
Cd,2
Cd,3
Cu, 1
Cu, 2
Cu, 3
Cumulative
Paniculate
Metal Removed
80,7
63.0
163.9
307.6
119,0
41.4
158.8
319.2
109.3
85.1
201.4
395.8
Metal
Recovered by
Back-washing
35.4
67.3
150.1
252.7
29.0
54.3
135.9
219.2
46.2
84.7
164.0
294.9
Cumulative
Soluble Me
Removed
29
36
32
102
47
85
2Z
209
5.5
12.3
ilfl
28.8
Metal
Recovered by
Regeneration
161.1
234.1
59.8
Total Metal
Recovery
Efficiency (%)
101
86
84
Regeneration
The maximum metal concentration attainable in the regenerant solution is a function of pH and is
important because it controls what can ultimately be done with the regenerant. The higher the
concentration, the more likely it is that the metal values can be recovered, turning the process into a
resource recovery technology, rather than just a treatment technology. The upper limit to this
concentration is imposed by the solubility of the metals and the fact that the regeneration efficiency
decreases when the regenerant becomes highly concentrated.
The regeneration protocol was typically to circulate 2 bed volumes of water adjusted to pH 2.0 through
the column for two hours, although in some cases a larger volume of solution was used. After two hours,
another, equal volume of acidified water was passed through the column and was not recirculated. The
metal concentration in the recirculation fluid increased rapidly at first and then only slowty thereafter. The
kinetics of regeneration after Run 15 are shown in Figure 13. Based on these results, it appears that a
recirculating period as short as 10 minutes Would release a large fraction of the available metal. Metal
concentrations in the first and second regenerant solutions were as high as 3000 and 500 mg/L after the
5 mg/L runs.
23
-------
An ion Removal
Five runs were conducted to evaluate the removal of As and Se from model solutions by the coated sand.
In this case, the influent was adjusted to pH 3.5, but conditions were otherwise similar to those for
removal of cationic metals. The run conditions are summarized in Table 6, and the results of several
runs are shown in Figures 14 and 15. In runs 18 through 20, oxyanionic forms of both As and Se were
present in the influent simultaneously, and both were removed quite efficiently from solution. System
efficiency was characterized by a classical, S-shaped breakthrough curve, with significant Se or As
appearing in the effluent after about 200 to 300 bed volumes of solution had been treated. The removal
pattern was remarkably consistent, regardless of the metal (As or Se) being treated and its oxidation
state (+3 or+5 for As; +4 or+6 for Se) (Figure 14). The latter result was particularly surprising, since
selenate (Se042")is generally much moredifficult to remove from solution than is selenit^Se032~). To
confirm this result and to minimize the possibility of any oxidation-reduction reaction involving selenium,
Run 21 was conducted with selenate as the only metal anion in the influent. Consistent with the prior
results, about 1000 mg Se was removed from solution per liter of packed bed prior to any significant
breakthrough. However, when 100 rngL sulfate was added to the influent, it apparently competed
effectively with the selenate for binding sites on the iron oxide coating, and breakthrough occurred much
sooner (Figure 15).
'able 6. Conditions for Run 18-22 using As and/or Se influent at 5 mq/L each. Media II was used for
hese runs.
Run # Metals
18 Arsenate, Selenite
19 Arsenate, Selenite
20 Arsenite, Selenate
21 Selenate only
22 Selenate, 100 mg/l S04
PH
3.5
3.5
3.5
3.5
3.5
EBDT
2 min.
2 min.
2 min.
2 min.
2 min.
Bed volumes
treated.
1289
773
750
438
454
25
-------
Effects of Other Contaminants on Removal of Cu, Pb, and Cd
Tests with complexed metals
Figures 16 and 17 show the breakthrough cures for the metals when complexed by ammonia and
EDTA, respectively. Only Cd and Cu were tested with ammonia present as a complexing agent, since
the ammonia did not maintain Pb in a soluble form. All three metals were present in the influent when
EDTA was used as the chelating agent.
Substantial amounts of ammonia-complexed metal can be sorbed by the iron-coated sand. The
breakthrough curves for Cd and Cu are almost identical. About 1500 mg of each metal can sorb per liter
of media before the effluent concentration exceeds a few tenths of a mg/L, and about 4000 mg of each
metal can sorb per liter of bed before the effluent concentration reaches 4 mg/L. Regeneration of this
column using a total of 16 bed volumes of regenerant solution at pH 2.0 recovered 93% of the sorbed Cd
and 100% of the Cu.
When the metals were complexed with EDTA, they broke through the column almost immediately. The
capacity of the media to remove metals under these conditions is not significant either at pH 10.0 or at pH
4.5, and the adsorptive filtration process would not be applicable for treatment of waters containing
EOTA-complexed metals.
In situations where complexed metals need to be treated, tests investigating the behavior of the
complex will be required.
27
-------
7 -.
5 «•
4 . .
*
a ••
4-
10
20
30
40 SO
B*d voiurrws
60
70
Figure 17. Effluent breakthrough curves for EDTAcomplexed metals. Runs were completed at pH 4.5
andpH 10, as indicated. Influent conditions: Cd=Cu=Pb=5 mg/L.EDTA:metal ratio = 1.25:1 (1 .84*10"^
E:DTA), EBDT=2 min.
Effects of organic surfactants.
Sodium lauryl sulfonate is a surfactant that might interfere with the adsorptive filtration process by
interacting either with the metals or the surface of the media. Figure 18 shows the column performance
in the presence of 0, 15, and 30 mg/L of this surfactant, along with enough ammonia to keep the metals
in solution. Tests with higher concentrations of the surfactant could not be conducted due to its limited
solubility. However, 30 mg/L is a higher concentration of surfactant than one would expect in most
wastewaters where the adsorptive filtration process might be used, and Figure 18 indicates that this
concentration had no noticeable effect on metal sorption. Computed regeneration efficiencies were 117%
forCd and 97% for Cu.
29
-------
process as a result of biogrowth is unlikely to be a severe problem. Regeneration efficiencies were 94%
forCd and 105% for Cu.
9 T
200
300 400
Bad volume
500
600
700
figure 19. Breakthrough curves for ammoniacomplexed metals in a column containing iron oxide coated
s>and onto which a biofilm had grown. Influent conditions: Cd=Cu=5 mg/L, NH^=0.04 M, EBDT=2 min. j
** " mm-if
Effects of oily substances on absorptive filtration.
One test was conducted in which the media was exposed to motor oil, in order to simulate a situation
where, by accident, a large amount of some oily substance entered an adsorptive filtration column.
Normally, such substances would be removed upstream of the column. To investigate a worst-case
scenario, a damp sample of the coated sand was exposed to a 10% by volume mixture of SAE 30 motor
oil in water and was then packed into a column. Oil remained attached to the media, and air bubbles
were trapped in the column. When the influent was applied, severe channeling was observed, and
breakthrough occurred almost immediately. The coated sand was then removed from the column and
cleaned by rinsing it twice in isopropanol. The cleaned sand was returned to the column, and the
standard ammonia-complexed influent was applied. As shown in Figure 21, the media performed
reasonably well, but a little more than half of the sorption capacity was lost. It is not known whether this
loss was due to residual oil on the media, which might be removed by more strenuous cleaning efforts, or
whether it reflects a permanent loss of capacity.
31
-------
Evaluation of Options for Residuals Management
One of the major factors affecting the value of this technology is the ultimate disposition of the regenerant
solution. An experimental evaluation of that issue was not conducted as part of this project. However,
based on the high metal concentrations achieved in the regenerant with relatively little effort, it is likely
that some recovery and reuse will be feasible in most cases, even if only by treating the sludge as an ore.
Treatment of Real Superfund Solutions
Once reasonable operating parameters for the technology were established, a real waste from a
Superfund site was collected and treated in one of the columns. The untreated water at this site contains
tens of mg/L of ferrous iron in addition to a few mg/L Zn and less thanl mg/L of several other metals. It
is treated by conventional precipitation/ coagulation at pH around 8.0 at the site, and, since the ferrous
iron is oxidized and precipitated in the process, the metals are exposed to a large amount of iron oxide in
the process. As a result, a significant fraction of the metals that can adsorb onto iron oxide do so in the
treatment process. The treatment process at the site is slightly unconventional, in that conventional
precipitation/ coagulation is usually carried out at pH 10 or greater. However, operation at that pH would
be problematic with this particular water, since it contains high concentrations of Ca and alkalinity,
causing massive amounts of CaCO, to precipitate when the pH is raised to 9.0. Fortunately, the
combination of metal precipitation and (probably most importantly) adsorption onto iron oxide removes
most of the metals from the soluble phase, so treatment at pH 8.0 is adequate.
We chose to treat the effluent from the precipitation/coagulation process being used on site. The upper
practical limit imposed on the treatment pH in the on-site process also applied to the adsorptive filtration
process. That is, we were constrained to operate the test system at pH near 8.0, well below the pH that
would have been optimal for metal removal. On the other hand, the use of our process on this water
represents a realistic test of whether adsorptive filtration can remove a substantial amount of metal from
a solution that has been treated optimally by more conventional methods.
The water was therefore treated as received, at pH 8.0 with Zn as the only metal present in significant
quantities. The total and soluble Zn concentrations in the samples collected were in the ranges 0.6 to 4.0
and 0.3 to 0.6 mg/L, respectively. The treatment set-up for these tests was more automated than that for
previous tests, as shown by the schematic in Figure 21. As shown in Figures 22 through 24, typical total
and soluble Zn concentrations in the effluent were around 0.2 and <0.1mg/L, respectively. Regeneration
kinetics for one of these runs is shown in Figure 25. The build-up of headloss during treatment of this
solution followed the same pattern as was observed for treatment of the synthetic solution (Figure 26).
33
-------
Column
(250 ml)
50% Expansion
Effluent Sample
Solenoid Switch
IBM AT with A./D and
control board using
Labtech Notebook
Software.
Rownwtor
JFigure 21. Schematic diagram of setup used for automatic treatment runs using Superfund site solution;
35
-------
0.4 T-
0.35 «•
N
200
400
600 800
B*d volumes
1000
1200
1400
Figure 24. Effluent zinc concentration for Western Processing Run 4 as a function of bed volumes;
treated,
37
-------
25 -r
400
600
800 1000 1200
B*d volum**
1400
1600
1800
2000
JRgure 26. Headless buildup during Western Processing Run 2.
CONCLUSIONS
In conclusion, simultaneous sorption and filtration of Cu, Cd, and Pb are feasible using iron oxide-coated
sand under reasonable engineering conditions. Total and soluble effluent concentrations of less than 100
\ig/L, and sometimes considerably less, are achievable. The media can be backwashed to recover most
of the particulate metals, and regenerated by exposure to an acid solution to recover the remaining
particulates and most of the adsorbed metals. The regenerant solution typically contains metal
concentrations a few hundred times as concentrated as the influent. In tests with 5 mg/L of each of three
metals in the influent, filtration limited process perfomance more than sorption did. It should be
recognized that this outcome is not generalizable: the limiting factor in any application would depend on
the specific operating conditions and chemical composition of the influent solution. Modifications to the
influent, such as adjusting solution pH or adding a polymeric filter aid, and modifications to the operation,
such as adjusting the hydraulic loading rate, would certainly affect the relative importance of headless
and effluent concentration limits. However, there was no indication that such factors could prevent
adsorptive filtration from serving as a viable treatment technology.
The solutions investigated contained one to three heavy metals, at total heavy metal concentrations
ranging from 0.5 to 15 mg/L. These solutions were treated successfully at empty bed contact times
.TO
-------
ABBREVIATIONS
BV Bed Volume
EBDT Empty Bed Detention Time
EDTA Ethylene Diamine Tetra-acetic Acid
IOCS Iron-Oxide-Coated Sand
SDS Sodium Dodecyl Sulfonate (also called Sodium Laury! Sulfonate)
41
------- |